https://chalkbrain.com Sketch-style boards, lessons, videos and interactive labs on how software systems work. en https://chalkbrain.com/lessons/from-scratch/container-everything/ https://chalkbrain.com/lessons/from-scratch/container-everything/ Before you start. Lessons 0651-container-fork-and-exec to 0656-container-network-namespace-and-veth built six walls one at a time: a child process, a private… https://chalkbrain.com/lessons/from-scratch/container-network-namespace-and-veth/ https://chalkbrain.com/lessons/from-scratch/container-network-namespace-and-veth/ Before you start. 0655-container-cgroups-limits finished the box's ceilings. This stage gives it a network of its own. A network device is something a machine… https://chalkbrain.com/lessons/from-scratch/container-cgroups-limits/ https://chalkbrain.com/lessons/from-scratch/container-cgroups-limits/ Before you start. 0654-container-proc-and-dev finished the walls that change what a box can see. This stage adds the one that changes how much it can use. You… https://chalkbrain.com/lessons/from-scratch/container-proc-and-dev/ https://chalkbrain.com/lessons/from-scratch/container-proc-and-dev/ Before you start. 0653-container-mount-namespace-and-pivot-root gave the box Alpine's files and found that ps printed only its header. This stage fixes that.… https://chalkbrain.com/lessons/from-scratch/container-mount-namespace-and-pivot-root/ https://chalkbrain.com/lessons/from-scratch/container-mount-namespace-and-pivot-root/ Before you start. 0652-container-uts-and-pid-namespaces gave the box its own hostname and its own process numbers. This stage gives it its own files. A file… https://chalkbrain.com/lessons/from-scratch/container-uts-and-pid-namespaces/ https://chalkbrain.com/lessons/from-scratch/container-uts-and-pid-namespaces/ Before you start. 0651-container-fork-and-exec started a command as a child and found that the child sees the host's hostname and the host's process list. This… https://chalkbrain.com/lessons/from-scratch/container-fork-and-exec/ https://chalkbrain.com/lessons/from-scratch/container-fork-and-exec/ Before you start. 0650-container-runtime-from-scratch set out the build: a container is an ordinary process with walls around it. This stage builds the process… https://chalkbrain.com/lessons/from-scratch/container-runtime-from-scratch/ https://chalkbrain.com/lessons/from-scratch/container-runtime-from-scratch/ Before you start. This series builds a small copy of the part of Docker that starts a container. A container is a program running with a limited view of the… https://chalkbrain.com/lessons/mind-blown/a-rounding-error-that-rewrote-the-weather/ https://chalkbrain.com/lessons/mind-blown/a-rounding-error-that-rewrote-the-weather/ Before you start. Nothing technical is needed for the story. The second half uses a one-line rule and a short Python program. 0227-no-one-is-in-charge shows… https://chalkbrain.com/lessons/mind-blown/the-hunters-and-the-hunted-take-turns/ https://chalkbrain.com/lessons/mind-blown/the-hunters-and-the-hunted-take-turns/ Before you start. Nothing technical is needed for the story. The second half uses simple arithmetic and one short Python program. 0227-no-one-is-in-charge is a… https://chalkbrain.com/lessons/mind-blown/the-person-who-changed-mid-sentence/ https://chalkbrain.com/lessons/mind-blown/the-person-who-changed-mid-sentence/ Before you start. Nothing technical is needed. If you have read 0230-you-have-a-hole-in-your-vision you have met the idea that your picture of the world has… https://chalkbrain.com/lessons/security/the-thing-you-let-in/ https://chalkbrain.com/lessons/security/the-thing-you-let-in/ Before you start. 0072-application-hardening said to check what knocks, in one line that included "a file upload". 0070-web-vulnerabilities showed what happens… https://chalkbrain.com/lessons/database-internals/make-this-postgres-fast-the-fixes-and-the-write-up/ https://chalkbrain.com/lessons/database-internals/make-this-postgres-fast-the-fixes-and-the-write-up/ Before you start. 0638-make-this-postgres-fast-the-brief built the shop, the stopwatch and the list of what is slow. If you did its Try it, keep your written… https://chalkbrain.com/lessons/database-internals/make-this-postgres-fast-the-brief/ https://chalkbrain.com/lessons/database-internals/make-this-postgres-fast-the-brief/ Before you start. This is the capstone of the track, and it uses the lessons before it: reading a plan (0259-the-query-planner, 0579-the-plan-running),… https://chalkbrain.com/lessons/database-internals/build-a-lock-manager-and-deadlock-detector/ https://chalkbrain.com/lessons/database-internals/build-a-lock-manager-and-deadlock-detector/ Before you start. 0260-locks-latches-and-transactions taught strict two-phase locking, the wait-for graph and a forced deadlock on a real Postgres.… https://chalkbrain.com/lessons/database-internals/build-three-ways-to-join/ https://chalkbrain.com/lessons/database-internals/build-three-ways-to-join/ Before you start. 0259-the-query-planner showed the three join methods in a real plan and the planner's bet on which to use.… https://chalkbrain.com/lessons/database-internals/build-sort-and-aggregate-with-a-memory-limit/ https://chalkbrain.com/lessons/database-internals/build-sort-and-aggregate-with-a-memory-limit/ Before you start. 0579-the-plan-running watched Postgres sort and group with workmem as the limit, and counted temp-file blocks as the sign of a spill.… https://chalkbrain.com/lessons/database-internals/build-a-bounded-buffer-pool/ https://chalkbrain.com/lessons/database-internals/build-a-bounded-buffer-pool/ Before you start. 0255-the-buffer-pool explained the room: frames, pinned and dirty pages, the clock sweep, and the small ring a big scan gets. 0288-db-pages… https://chalkbrain.com/lessons/networking/build-read-the-capture/ https://chalkbrain.com/lessons/networking/build-read-the-capture/ Before you start. You need the street address (IP) of 0004-how-packets-find-their-way, the corridor address (MAC) of 0005-inside-your-building and… https://chalkbrain.com/lessons/networking/build-ping-and-the-dying-parcel/ https://chalkbrain.com/lessons/networking/build-ping-and-the-dying-parcel/ Before you start. You need 0004-how-packets-find-their-way, which did traceroute by hand: depots, hops, and the hop limit that stops a parcel circling forever.… https://chalkbrain.com/lessons/networking/build-the-concierge/ https://chalkbrain.com/lessons/networking/build-the-concierge/ Before you start. You need 0001-dns-resolution (the concierge, referrals, glue, TTL), 0002-dns-wire-format (the bytes of a question and an answer) and… https://chalkbrain.com/lessons/networking/build-only-the-lost-page/ https://chalkbrain.com/lessons/networking/build-only-the-lost-page/ Before you start. You need 0626-build-the-page-that-gets-through: the same test road, the same five-byte packets, and the stop-and-wait and go-back-N senders… https://chalkbrain.com/lessons/networking/build-the-page-that-gets-through/ https://chalkbrain.com/lessons/networking/build-the-page-that-gets-through/ Before you start. 0417-the-page-that-never-arrived told the idea: number every page, answer with "got it", resend after a timeout, and keep several pages on… https://chalkbrain.com/lessons/from-scratch/dns-everything/ https://chalkbrain.com/lessons/from-scratch/dns-everything/ Before you start. This series has built a DNS server in six stages. By now it can do six things, one per earlier lesson: https://chalkbrain.com/lessons/from-scratch/dns-forwarding-resolver/ https://chalkbrain.com/lessons/from-scratch/dns-forwarding-resolver/ Before you start. 0623-dns-compressed-names taught the server to read names written as pointers, and gave it a small client that decodes any reply. Until now… https://chalkbrain.com/lessons/from-scratch/dns-compressed-names/ https://chalkbrain.com/lessons/from-scratch/dns-compressed-names/ Before you start. 0622-dns-question-and-answer wrote names as length-prefixed labels (03 www 07 example 04 test 00) and built answers whose name field was an… https://chalkbrain.com/lessons/from-scratch/dns-question-and-answer/ https://chalkbrain.com/lessons/from-scratch/dns-question-and-answer/ Before you start. 0621-dns-the-udp-socket-and-header built a socket and a parser for the 12-byte header: an ID, a flags word, and four counts, the first of… https://chalkbrain.com/lessons/from-scratch/dns-the-udp-socket-and-header/ https://chalkbrain.com/lessons/from-scratch/dns-the-udp-socket-and-header/ Before you start. 0620-dns-server-from-scratch set out the build: a DNS server answers one question packet with one answer packet, over UDP, and every packet… https://chalkbrain.com/lessons/from-scratch/dns-server-from-scratch/ https://chalkbrain.com/lessons/from-scratch/dns-server-from-scratch/ Before you start. This series builds a small DNS server, the program that turns a name like www.example.test into the number a computer needs to reach it. You… https://chalkbrain.com/lessons/from-scratch/git-everything/ https://chalkbrain.com/lessons/from-scratch/git-everything/ Before you start. This series has built a small git in five stages. Each one is a layer you can name: https://chalkbrain.com/lessons/from-scratch/git-log-and-checkout/ https://chalkbrain.com/lessons/from-scratch/git-log-and-checkout/ Before you start. 0617-git-commits-refs-and-head made the commit: a top tree, a parent hash, an author and a message, with a branch file holding the newest one… https://chalkbrain.com/lessons/from-scratch/git-commits-refs-and-head/ https://chalkbrain.com/lessons/from-scratch/git-commits-refs-and-head/ Before you start. 0616-git-the-tree turned a folder into one top hash: blobs for files, trees for folders, every tree named by its contents. A tree is a… https://chalkbrain.com/lessons/from-scratch/git-the-tree/ https://chalkbrain.com/lessons/from-scratch/git-the-tree/ Before you start. 0615-git-the-blob built the blob: a file's bytes, named by the hash of a header plus those bytes, stored zlib-compressed under .git/objects.… https://chalkbrain.com/lessons/from-scratch/git-the-blob/ https://chalkbrain.com/lessons/from-scratch/git-the-blob/ Before you start. 0614-git-from-scratch set out the build: git stores a version of your project as objects named by the hash of their contents. This stage… https://chalkbrain.com/lessons/from-scratch/git-from-scratch/ https://chalkbrain.com/lessons/from-scratch/git-from-scratch/ Before you start. This series builds a small copy of git, the tool that keeps every saved version of a folder of files. You need one idea, refreshed here: a… https://chalkbrain.com/lessons/from-scratch/lsm-everything/ https://chalkbrain.com/lessons/from-scratch/lsm-everything/ Before you start. This series built a small copy of a log-structured storage engine in seven stages. By now it can: keep recent writes in a sorted memtable… https://chalkbrain.com/lessons/from-scratch/lsm-compaction-and-levels/ https://chalkbrain.com/lessons/from-scratch/lsm-compaction-and-levels/ Before you start. This series builds a small copy of a storage engine that never edits a file in place. 0610-lsm-sstable-and-index made flushing safe and… https://chalkbrain.com/lessons/from-scratch/lsm-bloom-filter/ https://chalkbrain.com/lessons/from-scratch/lsm-bloom-filter/ Before you start. This series builds a small copy of a storage engine that never edits a file in place. 0610-lsm-sstable-and-index made each table file cheap… https://chalkbrain.com/lessons/from-scratch/lsm-sstable-and-index/ https://chalkbrain.com/lessons/from-scratch/lsm-sstable-and-index/ Before you start. This series builds a small copy of a storage engine that never edits a file in place. 0609-lsm-memtable-and-wal built the sorted memtable and… https://chalkbrain.com/lessons/from-scratch/lsm-memtable-and-wal/ https://chalkbrain.com/lessons/from-scratch/lsm-memtable-and-wal/ Before you start. This series builds a small copy of a storage engine that never edits a file in place. 0608-lsm-from-scratch has the map. You need one idea… https://chalkbrain.com/lessons/from-scratch/lsm-from-scratch/ https://chalkbrain.com/lessons/from-scratch/lsm-from-scratch/ Before you start. This series builds a small copy of a storage engine: the part of a database that stores keys and values on disk and finds them again. The… https://chalkbrain.com/lessons/from-scratch/raft-everything/ https://chalkbrain.com/lessons/from-scratch/raft-everything/ Before you start. This series built a small copy of Raft in four stages. By now your five nodes can: pick one leader by numbered terms and votes… https://chalkbrain.com/lessons/from-scratch/raft-a-kv-on-top/ https://chalkbrain.com/lessons/from-scratch/raft-a-kv-on-top/ Before you start. This series builds a small copy of Raft, the algorithm a group of machines uses to keep one shared list of commands. You have a leader… https://chalkbrain.com/lessons/from-scratch/raft-persistence/ https://chalkbrain.com/lessons/from-scratch/raft-persistence/ Before you start. This series builds a small copy of Raft, the algorithm a group of machines uses to keep one shared list of commands.… https://chalkbrain.com/lessons/from-scratch/raft-log-replication/ https://chalkbrain.com/lessons/from-scratch/raft-log-replication/ Before you start. This series builds a small copy of Raft, the algorithm a group of machines uses to keep one shared list of commands.… https://chalkbrain.com/lessons/from-scratch/raft-leader-election/ https://chalkbrain.com/lessons/from-scratch/raft-leader-election/ Before you start. This series builds a small copy of Raft, the algorithm a group of machines uses to keep one shared list of commands. This first stage builds… https://chalkbrain.com/lessons/from-scratch/raft-from-scratch/ https://chalkbrain.com/lessons/from-scratch/raft-from-scratch/ Before you start. This series builds a small copy of a consensus system: a group of five programs that keep one identical list of commands, even when some of… https://chalkbrain.com/lessons/how-computers-work/build-malloc/ https://chalkbrain.com/lessons/how-computers-work/build-malloc/ Before you start. A running program keeps two work areas in memory. The stack holds calls that have not finished, and empties itself as they return. The heap… https://chalkbrain.com/lessons/how-computers-work/the-program-that-starts-programs/ https://chalkbrain.com/lessons/how-computers-work/the-program-that-starts-programs/ Before you start. A process is one running program, walled off so nothing else can touch its memory. The kernel is the one program that can make new ones… https://chalkbrain.com/lessons/how-computers-work/build-a-stack-machine/ https://chalkbrain.com/lessons/how-computers-work/build-a-stack-machine/ Before you start. A chip runs one loop forever: fetch the next instruction, decode which operation it is, execute it, with one number (the program counter)… https://chalkbrain.com/lessons/how-computers-work/waiting-without-spinning/ https://chalkbrain.com/lessons/how-computers-work/waiting-without-spinning/ Before you start. When two threads touch the same memory at once, an update can be lost, and a lock fixes that by letting only one thread at a time near the… https://chalkbrain.com/lessons/how-computers-work/the-names-a-program-needs-to-find/ https://chalkbrain.com/lessons/how-computers-work/the-names-a-program-needs-to-find/ Before you start. A compiler turns the code you write into the chip's own instructions, one file at a time, and a running program keeps its short-lived work on… https://chalkbrain.com/lessons/how-computers-work/whose-turn-and-who-decides/ https://chalkbrain.com/lessons/how-computers-work/whose-turn-and-who-decides/ Before you start. The CPU is the chip that runs a program's instructions; a core is one of several copies of that machinery inside it, each running one program… https://chalkbrain.com/lessons/how-computers-work/the-address-book-every-process-carries/ https://chalkbrain.com/lessons/how-computers-work/the-address-book-every-process-carries/ Before you start. RAM is fast and small, the disk is huge and far slower, and a program runs well only when what it needs sits in the fast part… https://chalkbrain.com/lessons/database-internals/the-engine-in-production/ https://chalkbrain.com/lessons/database-internals/the-engine-in-production/ Before you start. 0256-wal-and-recovery showed the write-ahead log: every change goes into an append-only log before it touches a data page, so a crash can be… https://chalkbrain.com/lessons/database-internals/the-plan-running/ https://chalkbrain.com/lessons/database-internals/the-plan-running/ Before you start. 0259-the-query-planner showed the planner: the part of Postgres that reads a summary of your data and bets on one way to answer a query.… https://chalkbrain.com/lessons/database-internals/which-shortcut-to-build/ https://chalkbrain.com/lessons/database-internals/which-shortcut-to-build/ Before you start. 0257-b-tree-on-disk showed what an index is: a second, sorted copy of one column that points back at the table, so a lookup reads a few pages… https://chalkbrain.com/lessons/networking/the-key-at-the-door/ https://chalkbrain.com/lessons/networking/the-key-at-the-door/ Before you start. From 0007-the-whisper-code you have the split lock: a key in two halves, a public half anyone may see and a private half its owner never… https://chalkbrain.com/lessons/networking/the-hold-line/ https://chalkbrain.com/lessons/networking/the-hold-line/ Before you start. From 0003-tcp-udp-handshake you have the phone call: TCP opens with three messages ("can you hear me?", "yes, and can you hear me?", "yes"),… https://chalkbrain.com/lessons/networking/where-the-time-goes/ https://chalkbrain.com/lessons/networking/where-the-time-goes/ Before you start. From 0004-how-packets-find-their-way you have the picture this lesson stands on: a router is a depot, a packet is a parcel, and every parcel… https://chalkbrain.com/lessons/networking/inside-one-postal-company/ https://chalkbrain.com/lessons/networking/inside-one-postal-company/ Before you start. From 0004-how-packets-find-their-way you have the depot picture: a router is a depot that only knows its own short list of rows, each row… https://chalkbrain.com/lessons/networking/the-corridor-inside/ https://chalkbrain.com/lessons/networking/the-corridor-inside/ Before you start. From 0005-inside-your-building you have the building: the router is a doorman who hands every device a flat number (its IP address on the… https://chalkbrain.com/lessons/security/the-risk-register/ https://chalkbrain.com/lessons/security/the-risk-register/ Before you start. 0066-the-security-mindset gave you the threat model, a written list of ways a system could be abused. 0072-application-hardening gave you the… https://chalkbrain.com/lessons/security/the-door-you-forgot-was-a-door/ https://chalkbrain.com/lessons/security/the-door-you-forgot-was-a-door/ Before you start. 0069-authorization-and-identity taught the ownership check: logging in proves who you are, not that order 124 is yours.… https://chalkbrain.com/lessons/security/the-key-that-has-to-change/ https://chalkbrain.com/lessons/security/the-key-that-has-to-change/ Before you start. 0067-cryptography-foundations gave you symmetric encryption: one secret key locks and unlocks, so whoever holds the key holds the data.… https://chalkbrain.com/lessons/security/the-badge-you-cannot-take-back/ https://chalkbrain.com/lessons/security/the-badge-you-cannot-take-back/ Before you start. 0068-authentication gave you two ways to remember that someone already logged in. A session is a logbook entry kept at the front desk, and… https://chalkbrain.com/lessons/security/the-trace-you-would-actually-see/ https://chalkbrain.com/lessons/security/the-trace-you-would-actually-see/ Before you start. 0073-anatomy-of-a-breach showed that a breach is a chain, that the defender only has to catch the attacker at one link, and that most real… https://chalkbrain.com/lessons/how-things-work/your-home-estimated/ https://chalkbrain.com/lessons/how-things-work/your-home-estimated/ Before you start. This lesson uses the others on this shelf, one idea each. From 0550-the-socket-is-a-waterfall: watts are joules per second, and the bill… https://chalkbrain.com/lessons/how-things-work/why-you-can-see-through-sand/ https://chalkbrain.com/lessons/how-things-work/why-you-can-see-through-sand/ Before you start. You need two ideas. From 0141-remote-invisible-light: each colour of light comes in packets called photons, and a photon's energy, in… https://chalkbrain.com/lessons/how-things-work/the-lid-that-cooks-faster/ https://chalkbrain.com/lessons/how-things-work/the-lid-that-cooks-faster/ Before you start. You need one idea from 0140-cold-is-a-lie: a liquid's boiling point is not fixed. Squeeze it and it boils hotter; let the pressure crash and… https://chalkbrain.com/lessons/how-things-work/every-plug-ends-at-a-spinning-magnet/ https://chalkbrain.com/lessons/how-things-work/every-plug-ends-at-a-spinning-magnet/ Before you start. You need three earlier lessons, each in one line. 0550-the-socket-is-a-waterfall: power is volts times amps, and the bill counts… https://chalkbrain.com/lessons/how-things-work/three-lights-pretending-to-be-every-colour/ https://chalkbrain.com/lessons/how-things-work/three-lights-pretending-to-be-every-colour/ Before you start. You need 0141-remote-invisible-light: light is one kind of wave on a long ruler, its wavelength decides its colour, and your eyes see only a… https://chalkbrain.com/lessons/how-things-work/twice-the-speed-four-times-the-stop/ https://chalkbrain.com/lessons/how-things-work/twice-the-speed-four-times-the-stop/ Before you start. You need nothing but the experience of being in a car or on a bike when it stopped hard. One idea from 0136-entropy-arrow-of-time helps:… https://chalkbrain.com/lessons/how-things-work/spin-from-a-wire/ https://chalkbrain.com/lessons/how-things-work/spin-from-a-wire/ Before you start. You need 0550-the-socket-is-a-waterfall: voltage is the push, current is the flow, power is the two multiplied, and the grid's current flips… https://chalkbrain.com/lessons/how-things-work/the-socket-is-a-waterfall/ https://chalkbrain.com/lessons/how-things-work/the-socket-is-a-waterfall/ Before you start. You need one idea from 0140-cold-is-a-lie: a heater has to make every joule of heat it gives you, while a heat pump only moves heat that… https://chalkbrain.com/lessons/finance/the-ledger-that-cannot-lie/ https://chalkbrain.com/lessons/finance/the-ledger-that-cannot-lie/ Before you start. From 0527-every-litre-leaves-one-tank-and-enters-another: a ledger is accounts and entries, every transaction sums to zero, and the trial… https://chalkbrain.com/lessons/finance/the-stranger-with-your-card-number/ https://chalkbrain.com/lessons/finance/the-stranger-with-your-card-number/ Before you start. From 0532-push-pull-and-the-instant-transfer: a card payment is a pull, so a stranger holding your card details can start one, and the… https://chalkbrain.com/lessons/finance/one-point-nine-billion-that-was-never-there/ https://chalkbrain.com/lessons/finance/one-point-nine-billion-that-was-never-there/ Before you start. From 0527-every-litre-leaves-one-tank-and-enters-another: every movement is a pair of entries, so a book can add up and still be wrong about… https://chalkbrain.com/lessons/finance/the-bank-that-closed-at-half-past-three/ https://chalkbrain.com/lessons/finance/the-bank-that-closed-at-half-past-three/ Before you start. From 0526-your-balance-is-the-banks-debt: a payment between customers of two banks is the two banks owing each other, and at some point they… https://chalkbrain.com/lessons/finance/push-pull-and-the-instant-transfer/ https://chalkbrain.com/lessons/finance/push-pull-and-the-instant-transfer/ Before you start. From 0526-your-balance-is-the-banks-debt: your balance is a promise your bank owes you, so paying someone at another bank means your bank… https://chalkbrain.com/lessons/finance/the-hold-that-is-not-money-yet/ https://chalkbrain.com/lessons/finance/the-hold-that-is-not-money-yet/ Before you start. From 0530-four-parties-to-every-tap: a card payment runs from the merchant to the acquirer, across a network, to the issuer, which approves… https://chalkbrain.com/lessons/finance/four-parties-to-every-tap/ https://chalkbrain.com/lessons/finance/four-parties-to-every-tap/ Before you start. From 0526-your-balance-is-the-banks-debt: your balance is a promise your bank owes you, so paying someone means getting your bank to move… https://chalkbrain.com/lessons/finance/half-a-point-a-day-gone/ https://chalkbrain.com/lessons/finance/half-a-point-a-day-gone/ Before you start. From 0528-a-book-you-never-erase: a ledger is entries you only add to, and every balance is the sum of its entries, so any error in a single… https://chalkbrain.com/lessons/finance/a-book-you-never-erase/ https://chalkbrain.com/lessons/finance/a-book-you-never-erase/ Before you start. From 0527-every-litre-leaves-one-tank-and-enters-another: a ledger is accounts and entries, every transaction sums to zero, and the trial… https://chalkbrain.com/lessons/finance/every-litre-leaves-one-tank-and-enters-another/ https://chalkbrain.com/lessons/finance/every-litre-leaves-one-tank-and-enters-another/ Before you start. 0161-stripe-case-study gave you the habit: never write one number for a movement of money, write a pair, one line out of somewhere and one… https://chalkbrain.com/lessons/finance/your-balance-is-the-banks-debt/ https://chalkbrain.com/lessons/finance/your-balance-is-the-banks-debt/ Before you start. One idea from 0505-owning-a-slice-of-the-ship: there are two ways to put money into something. You can own a piece of it, or you can lend to… https://chalkbrain.com/lessons/finance/the-honest-backtest/ https://chalkbrain.com/lessons/finance/the-honest-backtest/ Before you start. From 0514-a-day-that-should-never-have-happened: volatility is how widely returns swing. From 0515-one-cloud-over-every-river: mixing assets… https://chalkbrain.com/lessons/finance/the-website-priced-at-the-moon/ https://chalkbrain.com/lessons/finance/the-website-priced-at-the-moon/ Before you start. From 0513-a-drunk-walk-through-paris: in an efficient market, news is priced in so fast that there is no easy free money left to pick up.… https://chalkbrain.com/lessons/finance/when-every-model-agreed/ https://chalkbrain.com/lessons/finance/when-every-model-agreed/ Before you start. From 0515-one-cloud-over-every-river: mixing many bets cuts the swing, down to a floor set by how much they move together (their… https://chalkbrain.com/lessons/finance/borrowed-water/ https://chalkbrain.com/lessons/finance/borrowed-water/ Before you start. From 0514-a-day-that-should-never-have-happened: volatility is how widely returns swing. From 0203-compound-interest: a wild run of returns… https://chalkbrain.com/lessons/finance/the-dial-nobody-can-see/ https://chalkbrain.com/lessons/finance/the-dial-nobody-can-see/ Before you start. From 0520-a-price-with-no-opinion-in-it: an option's price is forced by copying it with the underlying thing and a loan, and a binomial tree… https://chalkbrain.com/lessons/finance/a-price-with-no-opinion-in-it/ https://chalkbrain.com/lessons/finance/a-price-with-no-opinion-in-it/ Before you start. You need one idea from 0519-the-right-not-the-duty: a call option is a paid right to buy something later at a price fixed today (the strike),… https://chalkbrain.com/lessons/finance/the-right-not-the-duty/ https://chalkbrain.com/lessons/finance/the-right-not-the-duty/ Before you start. From 0518-a-price-agreed-for-next-seasons-rice you need the future: an agreement to trade later at a price fixed today, binding on both… https://chalkbrain.com/lessons/finance/a-price-agreed-for-next-seasons-rice/ https://chalkbrain.com/lessons/finance/a-price-agreed-for-next-seasons-rice/ Before you start. From 0512-under-a-buttonwood-tree you need the exchange: one place where buyers and sellers of the same thing meet, through an order book.… https://chalkbrain.com/lessons/finance/the-bank-that-bought-the-safest-thing/ https://chalkbrain.com/lessons/finance/the-bank-that-bought-the-safest-thing/ Before you start. From 0501-why-a-coin-today-beats-a-coin-next-year you need discounting: money promised later is worth less today, and the higher the going… https://chalkbrain.com/lessons/finance/what-the-valley-pays-you-to-carry-the-drought/ https://chalkbrain.com/lessons/finance/what-the-valley-pays-you-to-carry-the-drought/ Before you start. From 0515-one-cloud-over-every-river you need the two kinds of risk. A river's local weather (idiosyncratic risk) cancels when you own many… https://chalkbrain.com/lessons/finance/one-cloud-over-every-river/ https://chalkbrain.com/lessons/finance/one-cloud-over-every-river/ Before you start. From 0514-a-day-that-should-never-have-happened you need volatility: the typical size of a return's swing, measured as a standard deviation.… https://chalkbrain.com/lessons/finance/a-day-that-should-never-have-happened/ https://chalkbrain.com/lessons/finance/a-day-that-should-never-have-happened/ Before you start. From 0513-a-drunk-walk-through-paris you need the random walk: a price that moves each day by chance, with no memory, whose spread grows with… https://chalkbrain.com/lessons/finance/a-drunk-walk-through-paris/ https://chalkbrain.com/lessons/finance/a-drunk-walk-through-paris/ Before you start. From 0512-under-a-buttonwood-tree you need the stall: buyers queue with bids, sellers with asks, and the price sits in the gap between them,… https://chalkbrain.com/lessons/finance/under-a-buttonwood-tree/ https://chalkbrain.com/lessons/finance/under-a-buttonwood-tree/ Before you start. You need two ideas. From 0505-owning-a-slice-of-the-ship: a share is a slice of a company that you can sell to someone else, which only helps… https://chalkbrain.com/lessons/finance/one-page-for-the-whole-tank/ https://chalkbrain.com/lessons/finance/one-page-for-the-whole-tank/ Before you start. All of chapter one, in order: 0500-nineteen-and-six-where-the-water-goes (cash flow), 0501-why-a-coin-today-beats-a-coin-next-year (a coin… https://chalkbrain.com/lessons/finance/returns-too-smooth-to-be-true/ https://chalkbrain.com/lessons/finance/returns-too-smooth-to-be-true/ Before you start. You need two ideas from earlier in the track. 0505-owning-a-slice-of-the-ship showed that owning shares means owning something whose value… https://chalkbrain.com/lessons/finance/the-sale-made-at-the-bottom/ https://chalkbrain.com/lessons/finance/the-sale-made-at-the-bottom/ Before you start. Three earlier lessons do the groundwork. 0087-cognitive-biases showed that a bias is a mental shortcut that leans the same way every time, so… https://chalkbrain.com/lessons/finance/the-same-seed-in-a-sheltered-pot/ https://chalkbrain.com/lessons/finance/the-same-seed-in-a-sheltered-pot/ Before you start. You need compounding from 0203-compound-interest: growth earns growth, so anything taken out early takes its whole future with it, which is… https://chalkbrain.com/lessons/finance/thirty-four-years-to-break-even/ https://chalkbrain.com/lessons/finance/thirty-four-years-to-break-even/ Before you start. You need a share and a bond from 0505-owning-a-slice-of-the-ship: a share is a slice of a company that rises and falls with it, a bond is a… https://chalkbrain.com/lessons/finance/the-bet-on-the-whole-haystack/ https://chalkbrain.com/lessons/finance/the-bet-on-the-whole-haystack/ Before you start. You need a share from 0505-owning-a-slice-of-the-ship: a slice of ownership in one company, worth more when it does well and less when it… https://chalkbrain.com/lessons/finance/owning-a-slice-of-the-ship/ https://chalkbrain.com/lessons/finance/owning-a-slice-of-the-ship/ Before you start. You need the tank from 0500-nineteen-and-six-where-the-water-goes: a surplus is water left over after the taps. You also need two ideas from… https://chalkbrain.com/lessons/finance/paying-a-little-so-a-lot-cannot-sink-you/ https://chalkbrain.com/lessons/finance/paying-a-little-so-a-lot-cannot-sink-you/ Before you start. You need the tank from 0500-nineteen-and-six-where-the-water-goes (money as water, cash flow as in minus out) and the rain barrel from… https://chalkbrain.com/lessons/finance/the-pipe-that-drains-faster-than-it-fills/ https://chalkbrain.com/lessons/finance/the-pipe-that-drains-faster-than-it-fills/ Before you start. From 0500-nineteen-and-six-where-the-water-goes: the tank, and cash flow as what comes in minus what goes out. From… https://chalkbrain.com/lessons/finance/the-rain-barrel-for-the-dry-season/ https://chalkbrain.com/lessons/finance/the-rain-barrel-for-the-dry-season/ Before you start. From 0500-nineteen-and-six-where-the-water-goes: cash flow is what comes in minus what goes out, and a deficit has to be covered from savings… https://chalkbrain.com/lessons/finance/why-a-coin-today-beats-a-coin-next-year/ https://chalkbrain.com/lessons/finance/why-a-coin-today-beats-a-coin-next-year/ Before you start. One idea from 0203-compound-interest: money left to grow earns interest on its interest, so 1,000 at 5% becomes 1,050 after a year and… https://chalkbrain.com/lessons/finance/nineteen-and-six-where-the-water-goes/ https://chalkbrain.com/lessons/finance/nineteen-and-six-where-the-water-goes/ Before you start. Nothing. You need to be able to subtract one number from another, and that is the whole of the maths here. This is the first lesson of the… https://chalkbrain.com/lessons/portfolio-projects/s6-build-a-replicated-key-value-store/ https://chalkbrain.com/lessons/portfolio-projects/s6-build-a-replicated-key-value-store/ Before you start. You've built S2, a door in front of identical servers, and S3, a key-value server. Both lessons end by suggesting you put the first in front… https://chalkbrain.com/lessons/portfolio-projects/s5-build-a-message-log/ https://chalkbrain.com/lessons/portfolio-projects/s5-build-a-message-log/ Before you start. You've built S1, so an append-only log, a length in front of each record, and replay on startup are already yours. This project points that… https://chalkbrain.com/lessons/portfolio-projects/s4-build-an-http-server/ https://chalkbrain.com/lessons/portfolio-projects/s4-build-an-http-server/ Before you start. You've built S2, which let Go's net/http library handle every byte of every request. This project takes that library away. You need ports and… https://chalkbrain.com/lessons/data-systems/data-about-people/ https://chalkbrain.com/lessons/data-systems/data-about-people/ Before you start. 0044-stream-processing and 0045-where-data-systems-are-heading built a design around one add-only list of everything that happened, the log,… https://chalkbrain.com/lessons/data-systems/the-choices-before-the-requirements/ https://chalkbrain.com/lessons/data-systems/the-choices-before-the-requirements/ Before you start. Nothing from earlier lessons; this one opens the track. You only need to have used an app that remembers things, an order history, a feed, a… https://chalkbrain.com/lessons/personal-development/how-to-read-a-self-help-claim/ https://chalkbrain.com/lessons/personal-development/how-to-read-a-self-help-claim/ Before you start. This track has already debunked three famous claims: the 10,000-hour rule in 0121-deliberate-practice, the 21-day habit in… https://chalkbrain.com/lessons/personal-development/wishing-is-not-planning/ https://chalkbrain.com/lessons/personal-development/wishing-is-not-planning/ Before you start. 0123-habits-and-behaviour-change showed that a habit runs on a cue: something in the world that sets off the routine without a decision.… https://chalkbrain.com/lessons/personal-development/the-alarm-that-wont-switch-off/ https://chalkbrain.com/lessons/personal-development/the-alarm-that-wont-switch-off/ Before you start. 0121-deliberate-practice asked you to practise where you fail some of the time, because that is where feedback is richest.… https://chalkbrain.com/lessons/personal-development/what-the-raise-buys/ https://chalkbrain.com/lessons/personal-development/what-the-raise-buys/ Before you start. 0126-energy-goals-and-recovery treated your energy as something you spend and refill, and sleep as the floor under everything else. That… https://chalkbrain.com/lessons/personal-development/the-task-you-keep-not-starting/ https://chalkbrain.com/lessons/personal-development/the-task-you-keep-not-starting/ Before you start. 0123-habits-and-behaviour-change showed that a habit is a loop: a cue sets off a craving, the craving drives a routine, and the routine pays… https://chalkbrain.com/lessons/philosophy/what-a-word-points-at/ https://chalkbrain.com/lessons/philosophy/what-a-word-points-at/ Before you start. You need nothing but the ordinary sense that a word like "cat" is about cats: it is a sound or a mark that somehow reaches out to a thing in… https://chalkbrain.com/lessons/philosophy/when-the-evidence-cant-decide/ https://chalkbrain.com/lessons/philosophy/when-the-evidence-cant-decide/ Before you start. 0103-logic-truth-and-the-limits-of-reason showed that a rule built from past cases is never proved, only made likely, because the next case… https://chalkbrain.com/lessons/philosophy/is-death-bad/ https://chalkbrain.com/lessons/philosophy/is-death-bad/ Before you start. 0105-free-will-identity-and-the-self asked what makes you the same person over time, and settled on a chain of memories and character that… https://chalkbrain.com/lessons/philosophy/the-oldest-argument/ https://chalkbrain.com/lessons/philosophy/the-oldest-argument/ Before you start. 0103-logic-truth-and-the-limits-of-reason split a good argument into two checks. An argument's shape is valid when the conclusion has to… https://chalkbrain.com/lessons/philosophy/why-obey-anyone/ https://chalkbrain.com/lessons/philosophy/why-obey-anyone/ Before you start. 0104-ethics-and-the-good-life ended on John Rawls's veil: pick a society's rules before you know who you will be in it, and you will protect… https://chalkbrain.com/lessons/philosophy/is-anything-actually-wrong/ https://chalkbrain.com/lessons/philosophy/is-anything-actually-wrong/ Before you start. 0104-ethics-and-the-good-life gave you three ways to judge an act (by its outcome, by the duty it keeps, by the character of the person… https://chalkbrain.com/lessons/philosophy/argue-it-yourself/ https://chalkbrain.com/lessons/philosophy/argue-it-yourself/ Before you start. 0103-logic-truth-and-the-limits-of-reason gave you two checks on any argument. Its shape is valid when the conclusion has to follow if the… https://chalkbrain.com/lessons/psychology/the-words-you-think-in/ https://chalkbrain.com/lessons/psychology/the-words-you-think-in/ Before you start. 0085-how-memory-works pictured memory as an office: a desk (working memory, what you are holding in mind right now) with about four trays,… https://chalkbrain.com/lessons/psychology/when-the-mind-hurts/ https://chalkbrain.com/lessons/psychology/when-the-mind-hurts/ Before you start. 0470-how-psychology-knows-what-it-knows gave you the habit this lesson needs most: a famous study is a claim until someone checks it, and an… https://chalkbrain.com/lessons/psychology/measuring-a-person/ https://chalkbrain.com/lessons/psychology/measuring-a-person/ Before you start. 0470-how-psychology-knows-what-it-knows showed that a correlation is two things moving together, not one causing the other, and that a claim… https://chalkbrain.com/lessons/psychology/the-mind-being-built/ https://chalkbrain.com/lessons/psychology/the-mind-being-built/ Before you start. 0091-learning-and-expertise showed knowledge piling up as chunks, so an expert sees more in the same glance. 0471-the-wet-hardware showed a… https://chalkbrain.com/lessons/psychology/what-a-feeling-is-for/ https://chalkbrain.com/lessons/psychology/what-a-feeling-is-for/ Before you start. 0087-cognitive-biases showed a fast, automatic mind that answers before the slow one checks, and 0090-motivation-and-habit showed a craving… https://chalkbrain.com/lessons/psychology/the-room-that-agreed-with-the-wrong-line/ https://chalkbrain.com/lessons/psychology/the-room-that-agreed-with-the-wrong-line/ Before you start. 0087-cognitive-biases showed that a fast, automatic guess can pull the same wrong way every time, without you noticing.… https://chalkbrain.com/lessons/psychology/the-wet-hardware/ https://chalkbrain.com/lessons/psychology/the-wet-hardware/ Before you start. You need the idea from 0470-how-psychology-knows-what-it-knows that a striking result is worth checking before you trust it; it comes back… https://chalkbrain.com/lessons/psychology/how-psychology-knows-what-it-knows/ https://chalkbrain.com/lessons/psychology/how-psychology-knows-what-it-knows/ Before you start. You need two things you already have: a rough feel for percentages, and the experience of hearing "studies show..." and wondering whether to… https://chalkbrain.com/lessons/mind-blown/blue-with-no-blue-in-it/ https://chalkbrain.com/lessons/mind-blown/blue-with-no-blue-in-it/ Before you start. One idea helps: light is a wave, and each colour is a different length of wave, red the longest and violet the shortest.… https://chalkbrain.com/lessons/mind-blown/the-kitchen-table-lab/ https://chalkbrain.com/lessons/mind-blown/the-kitchen-table-lab/ Before you start. You need none of the other Mind Blown lessons first: each check below names the lesson that explains it, and you can do the check before or… https://chalkbrain.com/lessons/mind-blown/heavier-does-not-fall-faster/ https://chalkbrain.com/lessons/mind-blown/heavier-does-not-fall-faster/ Before you start. You need a heavy book, a sheet of paper and a coin. No physics is assumed. If you have read 0218-how-planes-actually-fly, you have met air… https://chalkbrain.com/lessons/mind-blown/what-fire-actually-is/ https://chalkbrain.com/lessons/mind-blown/what-fire-actually-is/ Before you start. You need a candle, a match and a glass jar, used with an adult's care on a heatproof surface. One idea from elsewhere helps: everything is… https://chalkbrain.com/lessons/mind-blown/a-blanket-that-lets-light-in/ https://chalkbrain.com/lessons/mind-blown/a-blanket-that-lets-light-in/ Before you start. One idea from elsewhere helps: everything warm gives off light of a kind our eyes cannot see, called infrared. 0141-remote-invisible-light… https://chalkbrain.com/lessons/mind-blown/the-ground-moves-as-fast-as-your-fingernails/ https://chalkbrain.com/lessons/mind-blown/the-ground-moves-as-fast-as-your-fingernails/ Before you start. You need a world map and your own hands. 0458-earth-history-as-one-year gives the time scale this lesson leans on: Earth is 4.54 billion… https://chalkbrain.com/lessons/mind-blown/why-the-night-is-dark/ https://chalkbrain.com/lessons/mind-blown/why-the-night-is-dark/ Before you start. One idea carries this lesson: light takes time to travel, so when you look far away you see the past. 0197-the-night-sky-is-a-time-machine… https://chalkbrain.com/lessons/mind-blown/earth-history-as-one-year/ https://chalkbrain.com/lessons/mind-blown/earth-history-as-one-year/ Before you start. You need a calendar and the ability to divide. If you have read 0457-nobody-designed-the-eye, you know natural selection needs a great deal… https://chalkbrain.com/lessons/mind-blown/nobody-designed-the-eye/ https://chalkbrain.com/lessons/mind-blown/nobody-designed-the-eye/ Before you start. You need no biology. One idea helps: living things carry a written recipe that is copied every time they have offspring, and copies are not… https://chalkbrain.com/lessons/mind-blown/the-moon-is-not-in-a-shadow/ https://chalkbrain.com/lessons/mind-blown/the-moon-is-not-in-a-shadow/ Before you start. You need a lamp, a ball (an orange works) and a dark room. If you have read 0455-summer-is-not-closer-to-the-sun, you have already met the… https://chalkbrain.com/lessons/mind-blown/summer-is-not-closer-to-the-sun/ https://chalkbrain.com/lessons/mind-blown/summer-is-not-closer-to-the-sun/ Before you start. You need a torch, a sheet of paper, and nothing else. If you have read 0197-the-night-sky-is-a-time-machine you already know sunlight takes… https://chalkbrain.com/lessons/game-theory/solve-these-cold/ https://chalkbrain.com/lessons/game-theory/solve-these-cold/ Before you start. This is the last lesson of the track, and it uses all of it: dominance and Nash equilibrium from… https://chalkbrain.com/lessons/game-theory/who-paid-for-the-taxi/ https://chalkbrain.com/lessons/game-theory/who-paid-for-the-taxi/ Before you start. From 0092-what-is-a-game you have a game: what you get depends on what others do too. From 0447-who-gets-the-bigger-slice you have bargaining… https://chalkbrain.com/lessons/game-theory/three-friends-one-restaurant/ https://chalkbrain.com/lessons/game-theory/three-friends-one-restaurant/ Before you start. From 0098-auctions-and-mechanism-design you have mechanism design: choose a game's rules so that people acting in their own interest produce… https://chalkbrain.com/lessons/game-theory/what-they-know-that-you-dont/ https://chalkbrain.com/lessons/game-theory/what-they-know-that-you-dont/ Before you start. From 0088-decision-making-under-uncertainty you have expected value: weigh each outcome by how likely it is and add them up. From… https://chalkbrain.com/lessons/game-theory/hawks-and-doves/ https://chalkbrain.com/lessons/game-theory/hawks-and-doves/ Before you start. From 0096-mixed-strategies-and-zero-sum you have the indifference trick: a mix is stable when every move in it earns the same, so nobody… https://chalkbrain.com/lessons/game-theory/who-gets-the-bigger-slice/ https://chalkbrain.com/lessons/game-theory/who-gets-the-bigger-slice/ Before you start. From 0446-draw-the-tree-start-from-the-end you have backward induction: draw who moves when, then settle the last choice first and step back.… https://chalkbrain.com/lessons/game-theory/draw-the-tree-start-from-the-end/ https://chalkbrain.com/lessons/game-theory/draw-the-tree-start-from-the-end/ Before you start. Every game so far had players choosing at the same moment, so a grid held it. From 0097-coordination-schelling-points-and-credible-threats… https://chalkbrain.com/lessons/game-theory/nash-on-a-number-line/ https://chalkbrain.com/lessons/game-theory/nash-on-a-number-line/ Before you start. From 0093-dominant-strategies-and-nash-equilibrium you have the Nash equilibrium: a set of choices where nobody gains by changing only their… https://chalkbrain.com/lessons/how-computers-work/the-drawer-that-says-it-saved/ https://chalkbrain.com/lessons/how-computers-work/the-drawer-that-says-it-saved/ Before you start. A file is only bytes plus an agreement about how to read them (0046-everything-is-a-number), and a program never touches the disk itself: it… https://chalkbrain.com/lessons/database-internals/the-database-that-wont-answer/ https://chalkbrain.com/lessons/database-internals/the-database-that-wont-answer/ Before you start. The last seven lessons opened one database a layer at a time: a table is a pile of 8 KB pages (0254-pages-and-heap-files), an index is a… https://chalkbrain.com/lessons/database-internals/the-sentence-that-says-what/ https://chalkbrain.com/lessons/database-internals/the-sentence-that-says-what/ Before you start. You have used an app that saves things: a sign-up form, a bank balance, a list of past orders. That is all this lesson needs; it comes first… https://chalkbrain.com/lessons/from-scratch/build-your-own-dynamo/ https://chalkbrain.com/lessons/from-scratch/build-your-own-dynamo/ Before you start. This series builds a small copy of Dynamo, the store Amazon wrote to keep each saved value on several machines and keep answering while some… https://chalkbrain.com/lessons/from-scratch/enough-go-to-read-these-builds/ https://chalkbrain.com/lessons/from-scratch/enough-go-to-read-these-builds/ Before you start. This track builds seven real systems from an empty file, all in the Go language: a memory store like Redis, a message log like Kafka, a web… https://chalkbrain.com/lessons/security/the-shop-you-secure/ https://chalkbrain.com/lessons/security/the-shop-you-secure/ Before you start. 0066-the-security-mindset gave you the threat model: a written list of the ways a system could be made to misbehave, walked door by door with… https://chalkbrain.com/lessons/security/spot-the-bug/ https://chalkbrain.com/lessons/security/spot-the-bug/ Before you start. 0066-the-security-mindset gave you the trust boundary: the line a request crosses when it comes from someone you do not control. Lessons… https://chalkbrain.com/lessons/security/whose-page-is-this/ https://chalkbrain.com/lessons/security/whose-page-is-this/ Before you start. 0068-authentication gave you the session: after you log in, the site hands your browser a small pass (a cookie), and the browser shows it on… https://chalkbrain.com/lessons/networking/it-wont-load-the-ladder/ https://chalkbrain.com/lessons/networking/it-wont-load-the-ladder/ Before you start. 0418-one-click-every-layer walked one page load through five stops: the concierge turns a name into a number (0001-dns-resolution), depots… https://chalkbrain.com/lessons/networking/one-click-every-layer/ https://chalkbrain.com/lessons/networking/one-click-every-layer/ Before you start. This lesson holds the whole track at once. The concierge who turns a name into an address is 0001-dns-resolution. The depots that pass a… https://chalkbrain.com/lessons/networking/the-page-that-never-arrived/ https://chalkbrain.com/lessons/networking/the-page-that-never-arrived/ Before you start. You know from 0003-tcp-udp-handshake that the network only carries postcards: chunks of data, called packets, that arrive or silently don't.… https://chalkbrain.com/lessons/networking/what-a-request-actually-says/ https://chalkbrain.com/lessons/networking/what-a-request-actually-says/ Before you start. You know that before a website sends anything, your laptop and the site open a phone call that delivers every byte in order, from… https://chalkbrain.com/lessons/ai/the-returns-desk-for-real/ https://chalkbrain.com/lessons/ai/the-returns-desk-for-real/ Before you start. The returns desk is a support assistant that can look an order up with lookuporder and open a return with startreturn, on top of a model that… https://chalkbrain.com/lessons/ai/rag-over-your-notes/ https://chalkbrain.com/lessons/ai/rag-over-your-notes/ Before you start. A language model, the AI that writes the answer, only knows what is in the text you hand it. So you cut your notes into chunks of a few… https://chalkbrain.com/lessons/ai/sampling-and-what-scale-buys/ https://chalkbrain.com/lessons/ai/sampling-and-what-scale-buys/ Before you start. In 0412-the-gpt-block you trained a small GPT, the kind of model built in the last two lessons, which writes one character at a time. Trained… https://chalkbrain.com/lessons/ai/the-gpt-block/ https://chalkbrain.com/lessons/ai/the-gpt-block/ Before you start. A weight is one adjustable number inside a model, and training nudges every weight, again and again, to shrink the loss, one number for how… https://chalkbrain.com/lessons/ai/attention-by-hand/ https://chalkbrain.com/lessons/ai/attention-by-hand/ Before you start. A weight is one adjustable number inside a model, and training nudges every weight to make the loss smaller (0408-autograd-by-hand). The loss… https://chalkbrain.com/lessons/ai/the-mlp-language-model/ https://chalkbrain.com/lessons/ai/the-mlp-language-model/ Before you start. A weight is one adjustable number inside a network; training nudges every weight, over and over, to shrink the loss, one number for how wrong… https://chalkbrain.com/lessons/ai/the-bigram-model/ https://chalkbrain.com/lessons/ai/the-bigram-model/ Before you start. A weight is one adjustable number inside a model. Its slope says how much the loss, the model's wrongness, moves if you nudge that weight.… https://chalkbrain.com/lessons/ai/autograd-by-hand/ https://chalkbrain.com/lessons/ai/autograd-by-hand/ Before you start. A weight is one adjustable number inside a network. The loss is one number for how wrong the network's answer is. A weight's slope says how… https://chalkbrain.com/lessons/ai/red-teaming/ https://chalkbrain.com/lessons/ai/red-teaming/ Before you start. The returns desk is this run's support assistant: a language model that looks up orders with lookuporder and opens returns with startreturn.… https://chalkbrain.com/lessons/ai/looking-inside-the-model/ https://chalkbrain.com/lessons/ai/looking-inside-the-model/ Before you start. 0075-what-a-neural-network-is showed that a network is layers of neurons, each multiplying inputs by weights and adding them up. Training… https://chalkbrain.com/lessons/ai/four-ways-to-split-a-model/ https://chalkbrain.com/lessons/ai/four-ways-to-split-a-model/ Before you start. From 0404-the-gpu: a GPU is thousands of simple calculators with memory that is very fast but small, 80 GB on an A100. From… https://chalkbrain.com/lessons/ai/the-gpu/ https://chalkbrain.com/lessons/ai/the-gpu/ Before you start. A matrix multiply is many dot products at once, multiply the pairs and add, and one neural-network layer is one, from… https://chalkbrain.com/lessons/ai/a-lora-run/ https://chalkbrain.com/lessons/ai/a-lora-run/ Before you start. A weight is one adjustable number inside a model, and the loss is one number for how wrong it is; training nudges the weights to shrink the… https://chalkbrain.com/lessons/ai/rnns-and-lstms/ https://chalkbrain.com/lessons/ai/rnns-and-lstms/ Before you start. From 0075-what-a-neural-network-is you need one unit: inputs multiplied by weights, added up, then passed through an activation. From… https://chalkbrain.com/lessons/ai/convolutional-networks/ https://chalkbrain.com/lessons/ai/convolutional-networks/ Before you start. From 0075-what-a-neural-network-is: a neuron multiplies each input by a weight, adds the results and a bias, and passes zero if the total is… https://chalkbrain.com/lessons/ai/why-the-loss-wont-go-down/ https://chalkbrain.com/lessons/ai/why-the-loss-wont-go-down/ Before you start. From 0075-what-a-neural-network-is: a neuron is a unit that adds up its inputs, each multiplied by a weight. A weight is one adjustable… https://chalkbrain.com/lessons/dsa/build-an-autocomplete/ https://chalkbrain.com/lessons/dsa/build-an-autocomplete/ Before you start. 0345-clinic-23-implement-trie built a trie: every stored word is a path from one starting node, one letter per step, words that begin alike… https://chalkbrain.com/lessons/dsa/build-a-diff-the-greedy-walk/ https://chalkbrain.com/lessons/dsa/build-a-diff-the-greedy-walk/ Before you start. 0382-build-a-diff-the-table built a diff from a table with one cell for every pair of lines: correct, but 16,000-line files already needed… https://chalkbrain.com/lessons/dsa/build-a-diff-the-table/ https://chalkbrain.com/lessons/dsa/build-a-diff-the-table/ Before you start. 0112-dynamic-programming-worked filled a grid for two protein chains: each cell answered "how much do these two beginnings share, in order?",… https://chalkbrain.com/lessons/dsa/dp-five-more-shapes/ https://chalkbrain.com/lessons/dsa/dp-five-more-shapes/ Before you start. 0063-recursion-and-dynamic-programming gave you the trick: answer each small copy of a question once, write it in a notebook, and build… https://chalkbrain.com/lessons/dsa/the-clock-that-wraps-around/ https://chalkbrain.com/lessons/dsa/the-clock-that-wraps-around/ Before you start. 0135-bit-manipulation showed that a computer stores a whole number as a row of 1s and 0s, and that the row has a fixed length, often 32… https://chalkbrain.com/lessons/dsa/why-the-loop-is-right/ https://chalkbrain.com/lessons/dsa/why-the-loop-is-right/ Before you start. 0062-sorting-and-searching gave you binary search: on a sorted list, look at the middle, throw away the half that cannot hold what you want,… https://chalkbrain.com/lessons/dsa/trees-asked-five-ways/ https://chalkbrain.com/lessons/dsa/trees-asked-five-ways/ Before you start. A binary search tree is a chart of boxes (nodes) where each box has at most two boxes below it, every smaller number sits somewhere down its… https://chalkbrain.com/lessons/devops/ship-it-the-alarm/ https://chalkbrain.com/lessons/devops/ship-it-the-alarm/ Before you start. 0376-ship-it-the-sync put the dock app on a laptop cluster, run by Argo CD from a git repository, and ended on a gap: a version that starts… https://chalkbrain.com/lessons/devops/ship-it-the-sync/ https://chalkbrain.com/lessons/devops/ship-it-the-sync/ Before you start. 0375-ship-it-the-line built the app, dock, into a small image, a sealed copy of the program that any machine can run.… https://chalkbrain.com/lessons/devops/ship-it-the-line/ https://chalkbrain.com/lessons/devops/ship-it-the-line/ Before you start. 0026-the-assembly-line-cicd gave you the line itself: every change is built once into an image, a sealed copy of the app and everything it… https://chalkbrain.com/lessons/devops/the-bill-nobody-reads/ https://chalkbrain.com/lessons/devops/the-bill-nobody-reads/ Before you start. 0249-ec2-and-instances gave you the rented machine: a virtual server is a slice of a real computer in someone else's building, and you pay… https://chalkbrain.com/lessons/devops/the-part-someone-else-made/ https://chalkbrain.com/lessons/devops/the-part-someone-else-made/ Before you start. 0026-the-assembly-line-cicd gave you the image, a sealed copy of the app and everything it needs, and two ways to name it: a tag, which is a… https://chalkbrain.com/lessons/devops/where-a-ship-may-berth/ https://chalkbrain.com/lessons/devops/where-a-ship-may-berth/ Before you start. A pod is one running copy of a program, a ship, and a node is one machine in the cluster (the group of machines Kubernetes runs on), a quay… https://chalkbrain.com/lessons/devops/a-harbour-on-your-laptop/ https://chalkbrain.com/lessons/devops/a-harbour-on-your-laptop/ Before you start. 0370-the-rooms-aws-runs-for-you gave you two rooms the harbour runs: a queue, a board where one program leaves a message until another takes… https://chalkbrain.com/lessons/devops/the-rooms-aws-runs-for-you/ https://chalkbrain.com/lessons/devops/the-rooms-aws-runs-for-you/ Before you start. 0249-ec2-and-instances gave you the instance: a slice of a big shared computer that you rent by the hour, where the operating system, the… https://chalkbrain.com/lessons/devops/one-cargo-plan-three-ports/ https://chalkbrain.com/lessons/devops/one-cargo-plan-three-ports/ Before you start. 0021-the-harbor-master gave you the harbour's ledger: you write a short text file saying what you want running (a YAML file, which is just… https://chalkbrain.com/lessons/from-scratch/kv-everything/ https://chalkbrain.com/lessons/from-scratch/kv-everything/ Before you start. 0367-kv-hinted-handoff kept the store taking writes while machines were down: a healthy machine stood in for a dead one and kept the value as… https://chalkbrain.com/lessons/from-scratch/kv-hinted-handoff/ https://chalkbrain.com/lessons/from-scratch/kv-hinted-handoff/ Before you start. 0366-kv-quorum stored every key on three machines, its replicas, and made a write wait for W of them to confirm and a read ask R of them,… https://chalkbrain.com/lessons/from-scratch/kv-quorum/ https://chalkbrain.com/lessons/from-scratch/kv-quorum/ Before you start. 0365-kv-the-ring built the ring: every key (the name a value is stored under) hashes to a spot on a clock face, and walking clockwise from it… https://chalkbrain.com/lessons/from-scratch/kv-the-ring/ https://chalkbrain.com/lessons/from-scratch/kv-the-ring/ Before you start. 0029-splitting-the-ledger cut one book into several because one book, on one machine, could not keep up: every name had to live in exactly… https://chalkbrain.com/lessons/system-design/the-stall-built-rush-hour/ https://chalkbrain.com/lessons/system-design/the-stall-built-rush-hour/ Before you start. 0363-the-stall-built-copies-and-shards finished the stall's back room: a copy of the book that runs a few milliseconds behind, and the book… https://chalkbrain.com/lessons/system-design/the-stall-built-copies-and-shards/ https://chalkbrain.com/lessons/system-design/the-stall-built-copies-and-shards/ Before you start. 0362-the-stall-built-the-spine built the stall as a Go program: a gate that sends each customer to the least busy of three stalls, a slate… https://chalkbrain.com/lessons/system-design/the-stall-built-the-spine/ https://chalkbrain.com/lessons/system-design/the-stall-built-the-spine/ Before you start. You need the chai stall from 0011-the-chai-stall: one famous stall became three, and a man at the gate sends each customer to one of them,… https://chalkbrain.com/lessons/system-design/a-lock-on-every-door/ https://chalkbrain.com/lessons/system-design/a-lock-on-every-door/ Before you start. From 0130-the-contract-at-the-counter you need the API, the written menu of what a caller may ask for, and its two refusals: 401 means "I… https://chalkbrain.com/lessons/system-design/the-customer-who-wont-hang-up/ https://chalkbrain.com/lessons/system-design/the-customer-who-wont-hang-up/ Before you start. From 0130-the-contract-at-the-counter: two programs meet at a counter, and every exchange so far was one question, one answer. From… https://chalkbrain.com/lessons/system-design/the-branch-near-your-house/ https://chalkbrain.com/lessons/system-design/the-branch-near-your-house/ Before you start. From 0011-the-chai-stall you need the famous stall that became many stalls, and from 0012-the-slate-next-to-the-stove the slate: a small… https://chalkbrain.com/lessons/system-design/the-last-vada/ https://chalkbrain.com/lessons/system-design/the-last-vada/ Before you start. From 0128-the-right-book-for-the-job you need the book itself: the database, where the stall writes its facts down and reads them back, and… https://chalkbrain.com/lessons/dsa/the-practice-ladders/ https://chalkbrain.com/lessons/dsa/the-practice-ladders/ Before you start. 0117-interview-patterns gave you the idea this page leans on: a pattern is a problem shape that keeps coming back with a solution you can… https://chalkbrain.com/lessons/dsa/clinic-34-linked-list-cycle/ https://chalkbrain.com/lessons/dsa/clinic-34-linked-list-cycle/ Before you start. 0059-linked-lists-stacks-queues built the chain: each item sits in a box called a node, and each node carries a note, the pointer, saying… https://chalkbrain.com/lessons/dsa/clinic-33-koko-eating-bananas/ https://chalkbrain.com/lessons/dsa/clinic-33-koko-eating-bananas/ Before you start. 0062-sorting-and-searching showed binary search on a sorted list: look at the middle item, throw away the half that cannot hold what you… https://chalkbrain.com/lessons/dsa/clinic-32-single-number/ https://chalkbrain.com/lessons/dsa/clinic-32-single-number/ Before you start. A computer stores every whole number as a row of ones and zeros, a row of switches read in twos instead of tens, so 101 is one four and one… https://chalkbrain.com/lessons/dsa/clinic-31-pow-x-n/ https://chalkbrain.com/lessons/dsa/clinic-31-pow-x-n/ Before you start. 0135-bit-manipulation showed that every whole number is stored as a row of 1s and 0s, and that sliding the row one place to the right halves… https://chalkbrain.com/lessons/dsa/clinic-30-rotate-image/ https://chalkbrain.com/lessons/dsa/clinic-30-rotate-image/ Before you start. 0057-arrays-and-strings gave you a row of numbered boxes, where you can reach any box by its number without walking past the others, and a… https://chalkbrain.com/lessons/dsa/clinic-29-non-overlapping-intervals/ https://chalkbrain.com/lessons/dsa/clinic-29-non-overlapping-intervals/ Before you start. 0350-clinic-28-merge-intervals took a pile of road closures, sorted them by where they start, and walked the road once, widening one sign at… https://chalkbrain.com/lessons/dsa/clinic-28-merge-intervals/ https://chalkbrain.com/lessons/dsa/clinic-28-merge-intervals/ Before you start. Sorting a list takes a little more than one pass over it: ten times the items costs about thirteen times the work, which engineers write as… https://chalkbrain.com/lessons/dsa/clinic-27-jump-game/ https://chalkbrain.com/lessons/dsa/clinic-27-jump-game/ Before you start. 0064-algorithm-design-patterns showed that "grab the best-looking choice and never look back" can be flatly wrong: with coins of 1, 3 and 4,… https://chalkbrain.com/lessons/dsa/clinic-26-redundant-connection/ https://chalkbrain.com/lessons/dsa/clinic-26-redundant-connection/ Before you start. A graph is a set of dots joined by lines; 0061-graphs builds it from scratch, and here the dots are houses and the lines are cables.… https://chalkbrain.com/lessons/dsa/clinic-25-network-delay-time/ https://chalkbrain.com/lessons/dsa/clinic-25-network-delay-time/ Before you start. You need a list, and the habit of keeping a running smallest number while you read it. Dots joined by lines (a graph) are introduced in… https://chalkbrain.com/lessons/dsa/clinic-24-word-search-ii/ https://chalkbrain.com/lessons/dsa/clinic-24-word-search-ii/ Before you start. You need the tree of letters from 0060-trees: every word is one path down from the top, one letter per step, and words that start the same… https://chalkbrain.com/lessons/dsa/clinic-23-implement-trie/ https://chalkbrain.com/lessons/dsa/clinic-23-implement-trie/ Before you start. A tree, from 0060-trees, is a shape with one starting point at the top, branches going down, and nothing ever looping back up. A hash set,… https://chalkbrain.com/lessons/ai/logistic-regression-precision-recall/ https://chalkbrain.com/lessons/ai/logistic-regression-precision-recall/ Before you start. You need 0075-what-a-neural-network-is and 0342-vectors-and-matrix-multiply. The first showed that one neuron is a weighted sum: each input… https://chalkbrain.com/lessons/ai/probability-softmax-temperature/ https://chalkbrain.com/lessons/ai/probability-softmax-temperature/ Before you start. 0342-vectors-and-matrix-multiply showed that a layer is a row of dot products, so the model's last layer hands back one plain number, a… https://chalkbrain.com/lessons/ai/vectors-and-matrix-multiply/ https://chalkbrain.com/lessons/ai/vectors-and-matrix-multiply/ Before you start. You need school arithmetic and nothing more: multiply two numbers, add a short column, and remember that a minus times a minus is a plus.… https://chalkbrain.com/lessons/devops/git-the-ledger/ https://chalkbrain.com/lessons/devops/git-the-ledger/ Before you start. Two lessons lean on this one without explaining it. 0025-iac-and-gitops keeps the whole system's blueprint "in git" and lets a patrol, a… https://chalkbrain.com/lessons/devops/linux-and-the-shell/ https://chalkbrain.com/lessons/devops/linux-and-the-shell/ Before you start. Read 0052-processes-threads-and-the-kernel and 0017-running-a-container first, or take these two sentences from them. From the first: the… https://chalkbrain.com/lessons/devops/the-page-at-3am/ https://chalkbrain.com/lessons/devops/the-page-at-3am/ Before you start. From 0027-the-watchtower-observability you need one idea: a pager is an alarm that wakes a person when a number that users feel, like the… https://chalkbrain.com/lessons/devops/the-restore-nobody-tested/ https://chalkbrain.com/lessons/devops/the-restore-nobody-tested/ Before you start. 0023-config-secrets-storage gave you the warehouse: a pod can die, but its data sits on a real disk that stays, and the next pod picks up the… https://chalkbrain.com/lessons/system-design/counting-the-crowd-without-a-register/ https://chalkbrain.com/lessons/system-design/counting-the-crowd-without-a-register/ Before you start. You need the stall from 0011-the-chai-stall, which got famous and became ten stalls, and the slate from 0012-the-slate-next-to-the-stove: a… https://chalkbrain.com/lessons/system-design/the-stall-that-turns-people-away/ https://chalkbrain.com/lessons/system-design/the-stall-that-turns-people-away/ Before you start. You need 0013-the-order-slip-rail: orders go on a wire as paper slips, one stove cooks them in turn, and the wait takes off as the stove… https://chalkbrain.com/lessons/system-design/thirty-stalls-on-one-order/ https://chalkbrain.com/lessons/system-design/thirty-stalls-on-one-order/ Before you start. You need the stall from 0011-the-chai-stall: one famous stall that became a chain, with the crowd shared between stalls so no single one… https://chalkbrain.com/lessons/security/a-shop-built-to-be-broken/ https://chalkbrain.com/lessons/security/a-shop-built-to-be-broken/ Before you start. You need two ideas from earlier in this track. From 0066-the-security-mindset: every place where something from outside reaches a system is a… https://chalkbrain.com/lessons/devops/the-cluster-that-wont-sail/ https://chalkbrain.com/lessons/devops/the-cluster-that-wont-sail/ Before you start. Kubernetes is the harbour master: you hand it a written order ("two copies of this program, always"), and it keeps making that true. Each… https://chalkbrain.com/lessons/system-design/one-wala-many-kettles/ https://chalkbrain.com/lessons/system-design/one-wala-many-kettles/ Before you start. You need the stall from 0011-the-chai-stall: one wala at one stove, a queue that grew when the stall got famous, and more stalls added so the… https://chalkbrain.com/lessons/ai/one-knob-one-slope/ https://chalkbrain.com/lessons/ai/one-knob-one-slope/ Before you start. You need 0076-how-it-learns, which gave you the loop: boil wrongness down to one number called the loss, find each dial's slope, and step… https://chalkbrain.com/lessons/ai/guardrails-cost-and-latency/ https://chalkbrain.com/lessons/ai/guardrails-cost-and-latency/ The returns desk is a support assistant built on a large language model (LLM) - software trained to read and write text - that helps an online shop's customers… https://chalkbrain.com/lessons/ai/evals-for-apps/ https://chalkbrain.com/lessons/ai/evals-for-apps/ Before you start. 0324-graphs-and-state-machines gave the desk a fixed set of states to move through on a return, and had it name which one it landed in at the… https://chalkbrain.com/lessons/ai/multi-agent/ https://chalkbrain.com/lessons/ai/multi-agent/ A large language model, an LLM, answers by predicting the next word from the text it can see at that moment. Picture one running a shop's returns desk, able to… https://chalkbrain.com/lessons/ai/graphs-and-state-machines/ https://chalkbrain.com/lessons/ai/graphs-and-state-machines/ Before you start. The example running through this lesson is a support assistant for an online shop's returns desk: a program a customer types at, which can… https://chalkbrain.com/lessons/ai/memory/ https://chalkbrain.com/lessons/ai/memory/ Before you start. The assistant in this lesson works an online shop's returns desk, where it can look up an order and start a return. Around the model sits… https://chalkbrain.com/lessons/ai/the-harness/ https://chalkbrain.com/lessons/ai/the-harness/ Before you start. 0321-the-loop built the twenty lines that assemble a window, call the model, run the tool it asks for, and check the answer - call that the… https://chalkbrain.com/lessons/ai/the-loop/ https://chalkbrain.com/lessons/ai/the-loop/ Before you start. The running example is a support assistant for an online shop's returns desk: software that answers customer messages, and that can look up… https://chalkbrain.com/lessons/ai/structured-output-and-tool-schemas/ https://chalkbrain.com/lessons/ai/structured-output-and-tool-schemas/ Before you start. A model writes one token at a time, a token being a word or a piece of one, the unit it writes in. Everything it has read so far -… https://chalkbrain.com/lessons/ai/context-engineering/ https://chalkbrain.com/lessons/ai/context-engineering/ Before you start. A token is a word or a piece of one, and the model can see only a fixed number of them at a time: 8,000 for the returns desk. That whole… https://chalkbrain.com/lessons/ai/prompt-engineering/ https://chalkbrain.com/lessons/ai/prompt-engineering/ Before you start. 0317-what-the-model-sees built the string the desk sends every turn: the whole conversation flattened into one long piece of text, then cut… https://chalkbrain.com/lessons/ai/what-the-model-sees/ https://chalkbrain.com/lessons/ai/what-the-model-sees/ Before you start. A model reads and writes in tokens, chunks of text about four characters long, and it is paid per token it reads and per token it writes.… https://chalkbrain.com/lessons/ai/routing-and-cascades/ https://chalkbrain.com/lessons/ai/routing-and-cascades/ Before you start. A model charges you for every word it reads and every word it writes. A bigger model charges more for each of those words, and answers harder… https://chalkbrain.com/lessons/ai/constrained-decoding/ https://chalkbrain.com/lessons/ai/constrained-decoding/ Before you start. This track is about serving large language models, and one of them writes its reply a token at a time, where a token is a word or a piece of… https://chalkbrain.com/lessons/ai/prefill-decode-split/ https://chalkbrain.com/lessons/ai/prefill-decode-split/ Before you start. 0313-chunked-prefill left you in a room where 32 people share one card. The card is a GPU: the chip built for doing the same arithmetic on… https://chalkbrain.com/lessons/ai/chunked-prefill/ https://chalkbrain.com/lessons/ai/chunked-prefill/ Before you start. 0311-the-two-clocks gave you the room: 32 people sharing one GPU, each getting a new token every 39 ms, a promise called TPOT that a busy GPU… https://chalkbrain.com/lessons/ai/prefix-caching/ https://chalkbrain.com/lessons/ai/prefix-caching/ Before you start. A model reads and writes in tokens: word-pieces, a few characters each, so "cancelling" might reach it as two of them. While it reads, it… https://chalkbrain.com/lessons/ai/the-two-clocks/ https://chalkbrain.com/lessons/ai/the-two-clocks/ Before you start. 0310-inference-engineering walked the whole trip a request takes through a serving system and pointed out that a reply hides two different… https://chalkbrain.com/lessons/ai/inference-engineering/ https://chalkbrain.com/lessons/ai/inference-engineering/ Before you start. 0109-inference-at-scale showed you how to make one GPU answer one request fast: handle several requests together in one go, shrink the… https://chalkbrain.com/lessons/from-scratch/k8s-everything/ https://chalkbrain.com/lessons/from-scratch/k8s-everything/ Before you start. This series builds a small copy of Kubernetes: a program that keeps a fleet of machines running the programs you asked for, and restarts them… https://chalkbrain.com/lessons/from-scratch/k8s-services-and-networking/ https://chalkbrain.com/lessons/from-scratch/k8s-services-and-networking/ Before you start. This series builds a small copy of Kubernetes, a program that watches over a fleet of machines and keeps every program you asked for running… https://chalkbrain.com/lessons/from-scratch/k8s-the-kubelet/ https://chalkbrain.com/lessons/from-scratch/k8s-the-kubelet/ Before you start. This series builds a small copy of Kubernetes: a program that keeps a fleet of machines running the programs you asked for, and restarts them… https://chalkbrain.com/lessons/from-scratch/k8s-the-scheduler/ https://chalkbrain.com/lessons/from-scratch/k8s-the-scheduler/ Before you start. This series builds a small copy of Kubernetes, a program that keeps a fleet of machines running the programs you asked for, and restarts them… https://chalkbrain.com/lessons/from-scratch/k8s-objects-and-controllers/ https://chalkbrain.com/lessons/from-scratch/k8s-objects-and-controllers/ Before you start. This series builds a small copy of Kubernetes: a program that keeps a fleet of machines running the programs you asked for, and restarts them… https://chalkbrain.com/lessons/from-scratch/k8s-store-and-watch/ https://chalkbrain.com/lessons/from-scratch/k8s-store-and-watch/ Before you start. This series builds a small copy of Kubernetes: a program that keeps a fleet of machines (a cluster) running the programs you asked for, and… https://chalkbrain.com/lessons/from-scratch/k8s-the-reconcile-loop/ https://chalkbrain.com/lessons/from-scratch/k8s-the-reconcile-loop/ Before you start. This series builds a small copy of Kubernetes: a program that keeps a fleet of machines running the programs you asked for, and restarts them… https://chalkbrain.com/lessons/from-scratch/k8s-imperative-to-declarative/ https://chalkbrain.com/lessons/from-scratch/k8s-imperative-to-declarative/ Before you start. This series builds a small copy of Kubernetes, a program that keeps a fleet of machines running the programs you asked for and restarts them… https://chalkbrain.com/lessons/from-scratch/sched-everything/ https://chalkbrain.com/lessons/from-scratch/sched-everything/ Before you start. This lesson assumes you built the small scheduler across the seven lessons before it, so here is what it can already do, one line each: - It… https://chalkbrain.com/lessons/from-scratch/sched-the-hard-part-time/ https://chalkbrain.com/lessons/from-scratch/sched-the-hard-part-time/ Before you start. This lesson is part of a build that makes a small distributed job scheduler - a program that fires jobs at the right moment, on the right… https://chalkbrain.com/lessons/from-scratch/sched-idempotency-and-fencing/ https://chalkbrain.com/lessons/from-scratch/sched-idempotency-and-fencing/ Before you start. This series builds a small scheduler: a program that runs jobs at set times across many machines instead of one, the way a single computer's… https://chalkbrain.com/lessons/from-scratch/sched-dispatch-queue/ https://chalkbrain.com/lessons/from-scratch/sched-dispatch-queue/ Before you start. This lesson is part of a series that builds a small copy of a job scheduler: a program that runs other programs at the exact right time, the… https://chalkbrain.com/lessons/from-scratch/sched-leader-election/ https://chalkbrain.com/lessons/from-scratch/sched-leader-election/ Before you start. This series builds a small copy of a job scheduler: a program whose whole purpose is to fire other programs at the right moment, the way a… https://chalkbrain.com/lessons/from-scratch/sched-heap-and-timing-wheel/ https://chalkbrain.com/lessons/from-scratch/sched-heap-and-timing-wheel/ Before you start. This series builds a small job scheduler: a program you hand a piece of work and a time, and it runs that work when the time comes, again and… https://chalkbrain.com/lessons/from-scratch/sched-durable-job-store/ https://chalkbrain.com/lessons/from-scratch/sched-durable-job-store/ Before you start. This series builds a small distributed job scheduler, a program whose only task is to fire off other jobs - a nightly backup, an hourly… https://chalkbrain.com/lessons/from-scratch/sched-the-tick-loop/ https://chalkbrain.com/lessons/from-scratch/sched-the-tick-loop/ Before you start. This series builds a small scheduler, a program whose only job is running other programs at the right moment instead of a person watching a… https://chalkbrain.com/lessons/from-scratch/db-everything/ https://chalkbrain.com/lessons/from-scratch/db-everything/ Before you start. This series has been building a small database of our own: a program that stores rows on disk, finds one fast, and survives a crash without… https://chalkbrain.com/lessons/from-scratch/db-the-query-layer/ https://chalkbrain.com/lessons/from-scratch/db-the-query-layer/ Before you start. This series builds a small copy of a database: a program that stores rows of data in files on disk and answers questions about them. Six… https://chalkbrain.com/lessons/from-scratch/db-transactions-mvcc/ https://chalkbrain.com/lessons/from-scratch/db-transactions-mvcc/ Before you start. This series builds a small database, a program that stores rows on disk and answers queries the way Postgres or SQLite do. So far it writes… https://chalkbrain.com/lessons/from-scratch/db-write-ahead-log/ https://chalkbrain.com/lessons/from-scratch/db-write-ahead-log/ Before you start. This series builds a small copy of a database, a program that stores your data and can find any piece of it again fast, the way a real… https://chalkbrain.com/lessons/from-scratch/db-the-b-tree/ https://chalkbrain.com/lessons/from-scratch/db-the-b-tree/ Before you start. This series builds a small database: a program that stores rows on disk and finds them again by key. So far it can split a file into… https://chalkbrain.com/lessons/from-scratch/db-pages/ https://chalkbrain.com/lessons/from-scratch/db-pages/ Before you start. This series builds a small database of our own: a program that stores rows of data on disk and can find any one of them again on demand. So… https://chalkbrain.com/lessons/from-scratch/db-the-index/ https://chalkbrain.com/lessons/from-scratch/db-the-index/ Before you start. This series builds a small database of our own: a program that saves pairs of a key (a name you look something up by) and a value (what's… https://chalkbrain.com/lessons/from-scratch/db-the-append-only-file/ https://chalkbrain.com/lessons/from-scratch/db-the-append-only-file/ Before you start. This series builds a small copy of a database: a program whose only job is to remember things you hand it, and hand them back correctly… https://chalkbrain.com/lessons/from-scratch/http-everything/ https://chalkbrain.com/lessons/from-scratch/http-everything/ Before you start. This series has been building a small HTTP server, a program that answers a browser's request for a page the same way every real website… https://chalkbrain.com/lessons/from-scratch/http-chunked-encoding/ https://chalkbrain.com/lessons/from-scratch/http-chunked-encoding/ Before you start. This series builds a small copy of a web server: a program that listens on the network for a request - a short message asking for something -… https://chalkbrain.com/lessons/from-scratch/http-concurrency/ https://chalkbrain.com/lessons/from-scratch/http-concurrency/ Before you start. This series builds a small copy of a web server, the program that answers when a browser asks for a page. Think of a connection as a phone… https://chalkbrain.com/lessons/from-scratch/http-keep-alive/ https://chalkbrain.com/lessons/from-scratch/http-keep-alive/ Before you start. This series builds a small copy of a web server - a program that sits waiting on the network, reads one request (one question, like "give me… https://chalkbrain.com/lessons/from-scratch/http-routing-and-response/ https://chalkbrain.com/lessons/from-scratch/http-routing-and-response/ Before you start. This series builds a small copy of a web server: a program that reads a request coming in over a network and sends back the right answer. So… https://chalkbrain.com/lessons/from-scratch/http-frame-the-body/ https://chalkbrain.com/lessons/from-scratch/http-frame-the-body/ Before you start. This series builds a small copy of a web server: a program that listens on the network, reads a request, and sends back an answer. The toy… https://chalkbrain.com/lessons/from-scratch/http-parse-the-request/ https://chalkbrain.com/lessons/from-scratch/http-parse-the-request/ Before you start. This series builds a small copy of an HTTP server, the program that answers when a browser or a tool like curl asks a website for a page. The… https://chalkbrain.com/lessons/from-scratch/http-the-raw-socket/ https://chalkbrain.com/lessons/from-scratch/http-the-raw-socket/ Before you start. This series builds a small HTTP server, a program that answers a browser's request for a page the same way every real website already does.… https://chalkbrain.com/lessons/from-scratch/kafka-everything/ https://chalkbrain.com/lessons/from-scratch/kafka-everything/ Before you start. This series builds a small copy of Kafka, a program that lets one part of a system hand a stream of events to many other parts without losing… https://chalkbrain.com/lessons/from-scratch/kafka-exactly-once/ https://chalkbrain.com/lessons/from-scratch/kafka-exactly-once/ Before you start. This series builds a small copy of Kafka, a program that keeps one shared, ordered notebook of everything that happened and lets any number… https://chalkbrain.com/lessons/from-scratch/kafka-replication-and-durability/ https://chalkbrain.com/lessons/from-scratch/kafka-replication-and-durability/ Before you start. This series builds a small copy of Kafka, a program that other programs send records to, in order, and read back later in that same order. So… https://chalkbrain.com/lessons/from-scratch/kafka-consumer-groups/ https://chalkbrain.com/lessons/from-scratch/kafka-consumer-groups/ Before you start. 0273-kafka-partitions split one topic, a named stream of messages such as every order a shop takes, into several partitions - separate… https://chalkbrain.com/lessons/from-scratch/kafka-partitions/ https://chalkbrain.com/lessons/from-scratch/kafka-partitions/ Before you start. You need to know that Kafka keeps every message it's given in one long, append-only file - a log - and that each message gets a number called… https://chalkbrain.com/lessons/from-scratch/kafka-segments-and-offset-index/ https://chalkbrain.com/lessons/from-scratch/kafka-segments-and-offset-index/ Before you start. This series builds a small copy of Kafka, the program big companies use to pass events (every click, every order) from the programs that make… https://chalkbrain.com/lessons/from-scratch/kafka-the-log/ https://chalkbrain.com/lessons/from-scratch/kafka-the-log/ Before you start. This series builds a small copy of Kafka, a program that keeps a stream of events and lets many other programs read that stream at their own… https://chalkbrain.com/lessons/from-scratch/kafka-the-naive-queue/ https://chalkbrain.com/lessons/from-scratch/kafka-the-naive-queue/ Before you start. This series builds a small copy of Kafka, a program that lets one part of a system hand a stream of events to many other parts, without… https://chalkbrain.com/lessons/from-scratch/redis-everything/ https://chalkbrain.com/lessons/from-scratch/redis-everything/ Before you start. This series builds a small copy of Redis, a program that keeps everything it stores in memory instead of on disk, so that asking it something… https://chalkbrain.com/lessons/from-scratch/redis-surviving-a-crash/ https://chalkbrain.com/lessons/from-scratch/redis-surviving-a-crash/ Before you start. The six stages before this one built a key-value store that keeps every key in RAM, fast because RAM never has to wait on a disk, and… https://chalkbrain.com/lessons/from-scratch/redis-eviction-memory-is-full/ https://chalkbrain.com/lessons/from-scratch/redis-eviction-memory-is-full/ Before you start. You know that Redis keeps everything in RAM, and that a key can carry its own expiry date so it deletes itself later - that was… https://chalkbrain.com/lessons/from-scratch/redis-keys-that-die/ https://chalkbrain.com/lessons/from-scratch/redis-keys-that-die/ Before you start. 0265-redis-one-thread-many-clients set the rule this whole lesson bends around: Redis answers every request through one worker, and that… https://chalkbrain.com/lessons/from-scratch/redis-one-thread-many-clients/ https://chalkbrain.com/lessons/from-scratch/redis-one-thread-many-clients/ Before you start. This series builds a small copy of Redis, a program that stores keys and values in memory and answers other programs over the network. So far… https://chalkbrain.com/lessons/from-scratch/redis-the-map-set-get-del/ https://chalkbrain.com/lessons/from-scratch/redis-the-map-set-get-del/ Before you start. You already have a command in hand - SET name garv - because the wire-protocol stage parsed it out of the raw bytes for you. This lesson… https://chalkbrain.com/lessons/from-scratch/redis-resp-the-wire-protocol/ https://chalkbrain.com/lessons/from-scratch/redis-resp-the-wire-protocol/ Before you start. You need to be able to count the characters in a word and add two small numbers - that's the whole of the arithmetic here. The toy so far is… https://chalkbrain.com/lessons/from-scratch/redis-the-socket-that-answers/ https://chalkbrain.com/lessons/from-scratch/redis-the-socket-that-answers/ Before you start. 0182-redis-from-scratch is the project lesson: a toy key-value server, grown one stage at a time, with a real fix at each stage for something… https://chalkbrain.com/lessons/database-internals/database-internals-everything/ https://chalkbrain.com/lessons/database-internals/database-internals-everything/ Before you start. Picture the database as a library: a room of shelves holding the actual books, a front desk that keeps the busiest ones close by instead of… https://chalkbrain.com/lessons/database-internals/locks-latches-and-transactions/ https://chalkbrain.com/lessons/database-internals/locks-latches-and-transactions/ Before you start. The last lesson watched the database plan how to answer a query, the question your app sends it, before reading a single row (one record,… https://chalkbrain.com/lessons/database-internals/the-query-planner/ https://chalkbrain.com/lessons/database-internals/the-query-planner/ Before you start. The last lesson showed that Postgres, the open-source database this series opens up, never deletes a row on update - it writes a fresh copy… https://chalkbrain.com/lessons/database-internals/mvcc-and-vacuum/ https://chalkbrain.com/lessons/database-internals/mvcc-and-vacuum/ Before you start. The last lesson showed you the index: a wide, shallow tree of pages that finds any row in three or four page reads. This one is about what… https://chalkbrain.com/lessons/database-internals/b-tree-on-disk/ https://chalkbrain.com/lessons/database-internals/b-tree-on-disk/ Before you start. 0256-wal-and-recovery showed you that every change is written to a log and flushed to disk before its commit is confirmed, so a crash can… https://chalkbrain.com/lessons/database-internals/wal-and-recovery/ https://chalkbrain.com/lessons/database-internals/wal-and-recovery/ Before you start. The last lesson showed that the database keeps its own copies of pages in RAM and changes them there, not on disk, because touching RAM is… https://chalkbrain.com/lessons/database-internals/the-buffer-pool/ https://chalkbrain.com/lessons/database-internals/the-buffer-pool/ Before you start. 0254-pages-and-heap-files showed you that a table on disk is a pile of fixed 8 KB pages, and every page holds its rows behind a slot list at… https://chalkbrain.com/lessons/database-internals/pages-and-heap-files/ https://chalkbrain.com/lessons/database-internals/pages-and-heap-files/ Before you start. You've used an app that saves things - a login, a cart, a comment - and somewhere behind it sat a database. This lesson opens that box for… https://chalkbrain.com/lessons/devops/the-sandbox-architecture/ https://chalkbrain.com/lessons/devops/the-sandbox-architecture/ Before you start. Kubernetes is a system that runs programs across a group of machines, called a cluster, and keeps them running. You need one picture from… https://chalkbrain.com/lessons/devops/vm-vs-k8s-and-the-cluster/ https://chalkbrain.com/lessons/devops/vm-vs-k8s-and-the-cluster/ Before you start. Two ideas from earlier lessons, in one line each. A team soon needs more than one machine, and a group of machines run together is a fleet… https://chalkbrain.com/lessons/devops/object-storage-and-cdn/ https://chalkbrain.com/lessons/devops/object-storage-and-cdn/ Note on names. Every hostname, distribution id and bucket name below is a placeholder (app.example.com, d111111abcdef8.cloudfront.net, assets-prod). Nothing… https://chalkbrain.com/lessons/devops/nginx-reverse-proxy/ https://chalkbrain.com/lessons/devops/nginx-reverse-proxy/ Before you start. You need to know that a running program listens on a numbered port (0010-one-address-a-thousand-doors), and that a web address is really just… https://chalkbrain.com/lessons/devops/ec2-and-instances/ https://chalkbrain.com/lessons/devops/ec2-and-instances/ Before you start. A virtual machine is a slice of a real computer that answers every command as if it owned the whole thing. This lesson is about how thin that… https://chalkbrain.com/lessons/devops/connectivity/ https://chalkbrain.com/lessons/devops/connectivity/ Before you start. A VPC is an island: a private network with its own range of addresses that nothing outside it can see or reach on its own, the thing… https://chalkbrain.com/lessons/devops/services-and-ingress/ https://chalkbrain.com/lessons/devops/services-and-ingress/ Before you start. Kubernetes runs your program as many copies spread over many machines. One running copy is called a pod. Pods die and get replaced all the… https://chalkbrain.com/lessons/devops/pods-replicasets-deployments/ https://chalkbrain.com/lessons/devops/pods-replicasets-deployments/ Before you start. Kubernetes is a system that runs your programs across a fleet of machines: you write down what you want running, and it keeps making that… https://chalkbrain.com/lessons/devops/load-balancing-and-dns/ https://chalkbrain.com/lessons/devops/load-balancing-and-dns/ Before you start. You already know how a name resolves - recursive resolver, root, TLD, authoritative server (0001-dns-resolution), and what's on the wire… https://chalkbrain.com/lessons/devops/routing-and-gateways/ https://chalkbrain.com/lessons/devops/routing-and-gateways/ Before you start. You need to know what a VPC and a subnet are: a network you control, fenced off from everyone else's, cut into smaller pieces you can treat… https://chalkbrain.com/lessons/devops/iam-and-identity/ https://chalkbrain.com/lessons/devops/iam-and-identity/ Before you start. AWS Identity and Access Management (IAM) is the part of Amazon Web Services that decides who may do what, and it keeps two things apart: who… https://chalkbrain.com/lessons/devops/security-groups-vs-nacls/ https://chalkbrain.com/lessons/devops/security-groups-vs-nacls/ Before you start. You need to know that a TCP connection opens with a SYN and gets an answering SYN-ACK back - 0003-tcp-udp-handshake covered that handshake.… https://chalkbrain.com/lessons/devops/vpc-and-subnets/ https://chalkbrain.com/lessons/devops/vpc-and-subnets/ Before you start. You need to know that a computer's address is a number written as four groups, like 10.42.0.4, and that adding one bit to a binary count… https://chalkbrain.com/lessons/devops/regions-azs-and-the-backbone/ https://chalkbrain.com/lessons/devops/regions-azs-and-the-backbone/ Before you start. You don't need any cloud experience for this one - every idea here maps onto a building, a city, and the roads between them. If you've read… https://chalkbrain.com/lessons/mind-blown/the-moon-illusion/ https://chalkbrain.com/lessons/mind-blown/the-moon-illusion/ Before you start. You need one everyday fact: things look smaller the farther away they are, which is why a plane shrinks to a speck at altitude and fills your… https://chalkbrain.com/lessons/mind-blown/why-onions-make-you-cry/ https://chalkbrain.com/lessons/mind-blown/why-onions-make-you-cry/ Before you start. You don't need any chemistry for this. All you need is one fact: every living cell is a tiny sealed room, and a knife is the only thing that… https://chalkbrain.com/lessons/mind-blown/the-off-switch-for-you/ https://chalkbrain.com/lessons/mind-blown/the-off-switch-for-you/ Before you start. You need no biology beyond what a surgery already is: cutting someone open while they don't move and don't hurt. The one habit this lesson… https://chalkbrain.com/lessons/mind-blown/a-wanted-poster-for-your-immune-system/ https://chalkbrain.com/lessons/mind-blown/a-wanted-poster-for-your-immune-system/ Before you start. You don't need any science background here, just curiosity. One fact helps: DNA is a code, written in four letters, that your cells read to… https://chalkbrain.com/lessons/mind-blown/why-you-cannot-remember-being-a-baby/ https://chalkbrain.com/lessons/mind-blown/why-you-cannot-remember-being-a-baby/ Before you start. Nothing technical needed, only your own memory, or the total lack of it, from before you turned about three. If you've ever tried to recall… https://chalkbrain.com/lessons/mind-blown/why-you-spend-a-third-of-life-unconscious/ https://chalkbrain.com/lessons/mind-blown/why-you-spend-a-third-of-life-unconscious/ Before you start. You don't need any science background here. All the maths is a handful of percentages, and it starts with something you can picture: a… https://chalkbrain.com/lessons/mind-blown/why-you-cannot-tickle-yourself/ https://chalkbrain.com/lessons/mind-blown/why-you-cannot-tickle-yourself/ Before you start. You already know the fact this lesson is built from: you can't tickle yourself, but someone else can, with the exact same finger on the exact… https://chalkbrain.com/lessons/system-design/why-data-is-the-hard-part/ https://chalkbrain.com/lessons/system-design/why-data-is-the-hard-part/ Before you start. You need only arithmetic - multiplying, dividing, and reading a percentage as "so many out of a hundred." Nothing here leans on another… https://chalkbrain.com/lessons/mind-blown/why-time-speeds-up-as-you-age/ https://chalkbrain.com/lessons/mind-blown/why-time-speeds-up-as-you-age/ Before you start. You need nothing more than your own memory of a childhood summer that felt endless, and a recent year that felt like it vanished in a few… https://chalkbrain.com/lessons/mind-blown/you-have-a-hole-in-your-vision/ https://chalkbrain.com/lessons/mind-blown/you-have-a-hole-in-your-vision/ Before you start. You need a blank page, something to draw with, and one eye you can close - keep them ready, since the first section puts them to use. One… https://chalkbrain.com/lessons/mind-blown/why-the-ocean-is-blue/ https://chalkbrain.com/lessons/mind-blown/why-the-ocean-is-blue/ Before you start. You need nothing but a swimming pool and a torch; there is no maths in the story half of this lesson. 0209-why-the-sky-is-blue already… https://chalkbrain.com/lessons/mind-blown/the-map-is-not-the-territory/ https://chalkbrain.com/lessons/mind-blown/the-map-is-not-the-territory/ Before you start. You don't need geometry or economics for the first half. If you've ever used a subway map, or watched a number everyone was told to hit stop… https://chalkbrain.com/lessons/mind-blown/no-one-is-in-charge/ https://chalkbrain.com/lessons/mind-blown/no-one-is-in-charge/ Before you start. Nothing technical is needed for the story below - if you've watched a flock of birds swirl at dusk, or ants crossing a path, you already have… https://chalkbrain.com/lessons/mind-blown/a-battery-stores-a-chemical-cliff/ https://chalkbrain.com/lessons/mind-blown/a-battery-stores-a-chemical-cliff/ Before you start. You need to know that electric current is a flow of something through a wire, the way water flows through a pipe. That is the only piece of… https://chalkbrain.com/lessons/mind-blown/the-less-you-know-the-more-sure-you-feel/ https://chalkbrain.com/lessons/mind-blown/the-less-you-know-the-more-sure-you-feel/ Before you start. You need to know what a percentile is: if you score in the 80th percentile on a test, you beat 80 of every 100 people who took it. That's the… https://chalkbrain.com/lessons/mind-blown/how-soap-destroys-germs/ https://chalkbrain.com/lessons/mind-blown/how-soap-destroys-germs/ Before you start. You already know that oil and water refuse to mix, and that germs like the flu virus or a cold virus are too small to ever see. That is the… https://chalkbrain.com/lessons/mind-blown/hot-water-can-freeze-faster/ https://chalkbrain.com/lessons/mind-blown/hot-water-can-freeze-faster/ Before you start. All you need is a thermometer and the patience to compare two numbers. The one idea this story leans on hardest is that a single reading on a… https://chalkbrain.com/lessons/mind-blown/watch-a-particle-and-it-changes/ https://chalkbrain.com/lessons/mind-blown/watch-a-particle-and-it-changes/ Before you start. You need nothing more than the picture of ripples spreading out from two gaps in a harbour wall, the kind you get by throwing two stones into… https://chalkbrain.com/lessons/mind-blown/cut-one-ball-into-two/ https://chalkbrain.com/lessons/mind-blown/cut-one-ball-into-two/ Before you start. All you need is the everyday idea that two solid balls can be "the same size," and enough trust in a proof to follow it even when it fights… https://chalkbrain.com/lessons/mind-blown/why-a-mirror-flips-left-right/ https://chalkbrain.com/lessons/mind-blown/why-a-mirror-flips-left-right/ Before you start. All you need is a mirror, or a clear memory of standing in front of one, and five minutes to test what you think you know instead of trusting… https://chalkbrain.com/lessons/mind-blown/the-sugar-pill-that-works/ https://chalkbrain.com/lessons/mind-blown/the-sugar-pill-that-works/ Before you start. You don't need any medical training or statistics for this one. You just need to be willing to watch one number get taken apart, piece by… https://chalkbrain.com/lessons/mind-blown/how-planes-actually-fly/ https://chalkbrain.com/lessons/mind-blown/how-planes-actually-fly/ Before you start. You don't need any physics for this lesson, just a memory: a hand held out of a moving car window, tilted up into the wind. If you've felt… https://chalkbrain.com/lessons/mind-blown/why-the-microwave-heats-the-food/ https://chalkbrain.com/lessons/mind-blown/why-the-microwave-heats-the-food/ Before you start. You already know that rubbing a balloon on your hair makes it stick to a wall - that's electric charge pulling on electric charge, nothing to… https://chalkbrain.com/lessons/mind-blown/where-is-everybody/ https://chalkbrain.com/lessons/mind-blown/where-is-everybody/ Before you start. Two ideas carry this story. First, even light takes time to cross space: a light-year is the distance light covers in a year, so a star 10… https://chalkbrain.com/lessons/mind-blown/how-long-is-a-coastline/ https://chalkbrain.com/lessons/mind-blown/how-long-is-a-coastline/ Before you start. You need to be able to multiply and divide, and to picture a number that keeps growing every time you check it more closely, without ever… https://chalkbrain.com/lessons/mind-blown/real-numbers-start-with-one-too-often/ https://chalkbrain.com/lessons/mind-blown/real-numbers-start-with-one-too-often/ Before you start. You need to be comfortable turning a percentage into a count out of 100, and to know that multiplying a number by 10 just adds a zero. That… https://chalkbrain.com/lessons/mind-blown/some-infinities-are-bigger/ https://chalkbrain.com/lessons/mind-blown/some-infinities-are-bigger/ Before you start. You need to be able to read a row of decimals and count on your fingers, nothing past that. The only other thing you need is patience for one… https://chalkbrain.com/lessons/mind-blown/the-statistic-that-reverses-itself/ https://chalkbrain.com/lessons/mind-blown/the-statistic-that-reverses-itself/ Before you start. You need to be able to read a percentage - 82 admitted out of 100 applicants is 82%, nothing more. Everything else in the first half is… https://chalkbrain.com/lessons/mind-blown/the-place-where-light-cannot-escape/ https://chalkbrain.com/lessons/mind-blown/the-place-where-light-cannot-escape/ Before you start. You need two things already in your head: gravity is mass pulling on mass, harder the closer and heavier things get, and nothing outruns… https://chalkbrain.com/lessons/mind-blown/why-you-cannot-fold-paper-past-seven-times/ https://chalkbrain.com/lessons/mind-blown/why-you-cannot-fold-paper-past-seven-times/ Before you start. All you need is that folding something in half doubles its thickness - check it on a folded napkin if that isn't already obvious. Everything… https://chalkbrain.com/lessons/mind-blown/why-the-sky-is-blue/ https://chalkbrain.com/lessons/mind-blown/why-the-sky-is-blue/ Before you start. You need to know that sunlight looks white but is secretly every color mixed together, the way a prism splits it into a rainbow. Nothing else… https://chalkbrain.com/lessons/mind-blown/spooky-action-at-a-distance/ https://chalkbrain.com/lessons/mind-blown/spooky-action-at-a-distance/ Before you start. You need nothing beyond counting to a hundred and knowing what a coin flip looks like. Nobody in this story does any maths you can't check in… https://chalkbrain.com/lessons/mind-blown/your-phone-runs-on-einstein/ https://chalkbrain.com/lessons/mind-blown/your-phone-runs-on-einstein/ Before you start. You don't need physics from anywhere else to follow this one; every number here is worked out from nothing. If you've already read about how… https://chalkbrain.com/lessons/mind-blown/silence-built-from-more-noise/ https://chalkbrain.com/lessons/mind-blown/silence-built-from-more-noise/ Before you start. You already know the one physics fact this needs: in a tug of war, if both sides pull exactly as hard, in exactly opposite directions, at… https://chalkbrain.com/lessons/mind-blown/you-are-written-in-four-letters/ https://chalkbrain.com/lessons/mind-blown/you-are-written-in-four-letters/ Before you start. You need to know that your body is built from trillions of tiny cells, and that each one carries its own full copy of the instructions for… https://chalkbrain.com/lessons/mind-blown/why-the-siren-drops/ https://chalkbrain.com/lessons/mind-blown/why-the-siren-drops/ Before you start. A sound's pitch, the note you hear, is just how close together its wave crests reach your ear: crests packed close sound higher, crests… https://chalkbrain.com/lessons/mind-blown/compound-interest/ https://chalkbrain.com/lessons/mind-blown/compound-interest/ Before you start. You need to be able to multiply and read a percentage as "this fraction of." 0192-fold-a-paper-to-the-moon already showed you the fun part,… https://chalkbrain.com/lessons/mind-blown/a-transistor-is-just-a-switch/ https://chalkbrain.com/lessons/mind-blown/a-transistor-is-just-a-switch/ Before you start. You need to know that electricity is just a crowd of electrons moving through a wire, and that a computer stores everything as one of two… https://chalkbrain.com/lessons/mind-blown/why-ice-floats/ https://chalkbrain.com/lessons/mind-blown/why-ice-floats/ Before you start. All you need is this: whether something floats depends on how heavy it is for its size, not its shape - a coin sinks, a hollow plastic duck… https://chalkbrain.com/lessons/mind-blown/one-row-in-ten-million/ https://chalkbrain.com/lessons/mind-blown/one-row-in-ten-million/ Before you start. You need to be comfortable counting, and with the idea that multiplying a number by itself a few times gets big fast - the same idea behind… https://chalkbrain.com/lessons/mind-blown/truths-math-cannot-prove/ https://chalkbrain.com/lessons/mind-blown/truths-math-cannot-prove/ Before you start. You already know what a proof is: a chain of steps from statements you start with to a conclusion, with no step allowed to sneak in something… https://chalkbrain.com/lessons/mind-blown/the-road-that-slows-everyone/ https://chalkbrain.com/lessons/mind-blown/the-road-that-slows-everyone/ Before you start. You need to be able to add, divide, and compare two numbers - nothing past what you'd use splitting a dinner bill four ways. Every word this… https://chalkbrain.com/lessons/mind-blown/the-night-sky-is-a-time-machine/ https://chalkbrain.com/lessons/mind-blown/the-night-sky-is-a-time-machine/ Before you start. You need nothing you do not already have: dividing a distance by a speed gives you a time, the way it does when you work out how long a drive… https://chalkbrain.com/lessons/mind-blown/you-are-empty-space/ https://chalkbrain.com/lessons/mind-blown/you-are-empty-space/ Before you start. You need only what's left over from school: that everything is made of atoms, and that an atom has a heavy speck at its middle with something… https://chalkbrain.com/lessons/mind-blown/monty-hall-the-illegal-switch/ https://chalkbrain.com/lessons/mind-blown/monty-hall-the-illegal-switch/ Before you start. You need nothing beyond fractions like "2 out of 3". If you've read 0189-the-birthday-paradox you already know the habit this note leans on… https://chalkbrain.com/lessons/mind-blown/the-question-no-computer-can-answer/ https://chalkbrain.com/lessons/mind-blown/the-question-no-computer-can-answer/ Before you start. You don't need to know how to program, or anything about computers beyond this: a program is a fixed list of instructions, and running it on… https://chalkbrain.com/lessons/mind-blown/a-secret-shouted-across-a-room/ https://chalkbrain.com/lessons/mind-blown/a-secret-shouted-across-a-room/ Before you start. You need nothing more than a calculator's four buttons - add, multiply, and the idea that some sums "wrap around" once they pass a limit, the… https://chalkbrain.com/lessons/mind-blown/fold-a-paper-to-the-moon/ https://chalkbrain.com/lessons/mind-blown/fold-a-paper-to-the-moon/ Before you start. You need to be comfortable doubling a number in your head, again and again - 2, 4, 8, 16 - and nothing else. That's the whole of the maths in… https://chalkbrain.com/lessons/mind-blown/every-sound-is-sine-waves/ https://chalkbrain.com/lessons/mind-blown/every-sound-is-sine-waves/ Before you start. Nothing. There is no maths here beyond doubling and tripling a number, and no physics you need to bring: the lesson builds every sound up… https://chalkbrain.com/lessons/mind-blown/the-99-percent-test/ https://chalkbrain.com/lessons/mind-blown/the-99-percent-test/ Before you start. You need to be able to work out a percent of a round number, the kind of arithmetic you'd use to check a bill. That is the whole of the maths… https://chalkbrain.com/lessons/mind-blown/the-birthday-paradox/ https://chalkbrain.com/lessons/mind-blown/the-birthday-paradox/ Before you start. All the arithmetic here is counting: how many people, how many pairs they make between them, nothing harder than that. Every number below is… https://chalkbrain.com/lessons/from-scratch/kubernetes-from-scratch/ https://chalkbrain.com/lessons/from-scratch/kubernetes-from-scratch/ Before you start. This series builds a small copy of Kubernetes: a program you tell what you want running, and which then keeps that true on its own, without… https://chalkbrain.com/lessons/from-scratch/distributed-scheduler-from-scratch/ https://chalkbrain.com/lessons/from-scratch/distributed-scheduler-from-scratch/ Code: teaching/code/from-scratch/scheduler/ - one Go module, go run ./NN-stage. Go is the programming language every stage is written in. The demos run on a… https://chalkbrain.com/lessons/from-scratch/database-from-scratch/ https://chalkbrain.com/lessons/from-scratch/database-from-scratch/ Code: teaching/code/from-scratch/database/ - one Go module, go run ./NN-stage. Real files, a real B-tree (verified: 1000 keys, height 6, in-order sorted), real… https://chalkbrain.com/lessons/from-scratch/http-server-from-scratch/ https://chalkbrain.com/lessons/from-scratch/http-server-from-scratch/ Before you start. This series builds a small copy of an HTTP server: a program that answers when a browser (or curl, or another program) asks it for something,… https://chalkbrain.com/lessons/from-scratch/kafka-from-scratch/ https://chalkbrain.com/lessons/from-scratch/kafka-from-scratch/ Before you start. This series builds a small copy of Kafka: a program that keeps a list of everything that happened, in order, and lets any number of programs… https://chalkbrain.com/lessons/from-scratch/redis-from-scratch/ https://chalkbrain.com/lessons/from-scratch/redis-from-scratch/ Before you start. You need to know that a running program holds data in its own memory, and that memory empties the moment the program stops. That gap is what… https://chalkbrain.com/lessons/dsa/foundations-10-dynamic-programming/ https://chalkbrain.com/lessons/dsa/foundations-10-dynamic-programming/ Before you start. 0173-foundations-02-recursion-memo-table ne ek engine diya tha: recursion se sochna (chhote sawaal mein tod ke), jawab ek notebook mein likh… https://chalkbrain.com/lessons/dsa/foundations-09-graphs/ https://chalkbrain.com/lessons/dsa/foundations-09-graphs/ Before you start. Pichhle lesson mein dekha tree bhi ek graph hai, bas usme cycle nahi hoti - ghoom ke wapas usi jagah aane wala raasta… https://chalkbrain.com/lessons/dsa/foundations-08-trees-recursion/ https://chalkbrain.com/lessons/dsa/foundations-08-trees-recursion/ Before you start. Foundations #02 mein recursion aur call stack dekh chuke ho - ek function khud ko chhote input pe bulata hai, jab tak sabse chhota case (base… https://chalkbrain.com/lessons/dsa/foundations-07-linked-lists-pointers/ https://chalkbrain.com/lessons/dsa/foundations-07-linked-lists-pointers/ Before you start. Pichhle lesson mein dekha tha ki array mein har cheez ek line mein, agal-bagal rakhi hoti hai, aur beech mein kuch daalna sabko khiska deta… https://chalkbrain.com/lessons/dsa/foundations-06-stacks-queues/ https://chalkbrain.com/lessons/dsa/foundations-06-stacks-queues/ Before you start. Array yaad karo: dabbon ki ek seedhi row, har dabbe ka apna number, jispe seedhe pahunchna sasta hota hai -… https://chalkbrain.com/lessons/dsa/foundations-05-sorting-searching/ https://chalkbrain.com/lessons/dsa/foundations-05-sorting-searching/ Before you start. Foundations ki pichli lesson mein O(n) dekha tha - list do guni, kaam bhi do guna, ek seedhi lakeer jaisa kaam. Is lesson mein do naye rules… https://chalkbrain.com/lessons/dsa/foundations-04-hashing/ https://chalkbrain.com/lessons/dsa/foundations-04-hashing/ Pehle ye jaan lo. 0174-foundations-03-arrays-pointers-window ka array scan (ek-ek karke check karna) yehi lesson tez banata hai. Bas itna pata ho ki % (modulo)… https://chalkbrain.com/lessons/dsa/foundations-03-arrays-pointers-window/ https://chalkbrain.com/lessons/dsa/foundations-03-arrays-pointers-window/ Before you start. Foundations #01 mein O(n) sikha tha - "ek baar chhua, ek baar gina" (0172-foundations-01-big-o-touched-once). Yahan wahi rule chaar naye… https://chalkbrain.com/lessons/dsa/foundations-02-recursion-memo-table/ https://chalkbrain.com/lessons/dsa/foundations-02-recursion-memo-table/ Before you start. Foundations #01 se itna kaafi hai: jab size double hoti hai aur kaam bhi double ho jaata hai, use exponential kehte hain, aur wo bahut tezi… https://chalkbrain.com/lessons/dsa/foundations-01-big-o-touched-once/ https://chalkbrain.com/lessons/dsa/foundations-01-big-o-touched-once/ Before you start. Bas itna chahiye: tum gin sakte ho, aur tumne kabhi na kabhi ek for-loop dekha hai - ek list ke har element ko ek baar chhoo kar aage badhna.… https://chalkbrain.com/lessons/dsa/clinic-22-longest-common-subsequence/ https://chalkbrain.com/lessons/dsa/clinic-22-longest-common-subsequence/ Before you start. Ek 2D grid mein padhna-likhna aata ho, aur "yahan tak pahunchne ka aakhri kadam kya tha" wala sawaal poochhna aata ho - wahi tarkeeb… https://chalkbrain.com/lessons/dsa/clinic-21-unique-paths/ https://chalkbrain.com/lessons/dsa/clinic-21-unique-paths/ Before you start. Pichhli clinic 0151-clinic-02-climbing-stairs mein har kadam pe do choice tha, aur har jagah ko ek number se pehchaante the (kaunsi seedhi) -… https://chalkbrain.com/lessons/dsa/clinic-20-top-k-frequent/ https://chalkbrain.com/lessons/dsa/clinic-20-top-k-frequent/ Before you start. 0154-clinic-05-two-sum mein tumne dekha tha ki ek hash map kisi cheez ko turant (ek hi kadam mein) yaad rakh sakta hai - yahan wahi table use… https://chalkbrain.com/lessons/dsa/clinic-19-kth-largest/ https://chalkbrain.com/lessons/dsa/clinic-19-kth-largest/ Before you start. Bas itna pata ho ki array numbers ki list hai, har number ki apni jagah (index) hoti hai, aur sort karna matlab chhote se bade order mein… https://chalkbrain.com/lessons/devops/lambda-cross-account-s3/ https://chalkbrain.com/lessons/devops/lambda-cross-account-s3/ Before you start. 0163-irsa-cross-account left you one idea: a program that needs to get into somebody else's space doesn't have to carry a permanent key to… https://chalkbrain.com/lessons/dsa/clinic-18-3sum/ https://chalkbrain.com/lessons/dsa/clinic-18-3sum/ Before you start. Tumhe list ko chhote se bade tak laga ke crawl karna aana chahiye - bas itna. 0166-clinic-17-container-with-most-water ne "list ke dono siron… https://chalkbrain.com/lessons/devops/folder-level-s3/ https://chalkbrain.com/lessons/devops/folder-level-s3/ Before you start. You need to know that an IAM policy is a JSON document attached to an identity, allowing or denying specific actions on specific things -… https://chalkbrain.com/lessons/dsa/clinic-17-container-with-most-water/ https://chalkbrain.com/lessons/dsa/clinic-17-container-with-most-water/ Before you start. Array ek line mein rakhe numbers hote hain, aur index unki position ka number hai - pehla number index 0 pe baithta hai, doosra index 1 pe.… https://chalkbrain.com/lessons/devops/why-so-many-boxes/ https://chalkbrain.com/lessons/devops/why-so-many-boxes/ Before you start. 0162-the-two-planes gave you two guards at AWS's front door: one checking whether a packet can reach a machine, the other checking whether an… https://chalkbrain.com/lessons/dsa/clinic-16-course-schedule/ https://chalkbrain.com/lessons/dsa/clinic-16-course-schedule/ Before you start. 0164-clinic-15-number-of-islands mein tumne graph traversal dekha tha - nodes, edges, aur unpe BFS/DFS chalana - jahan teer dono taraf jaate… https://chalkbrain.com/lessons/devops/oidc-passport/ https://chalkbrain.com/lessons/devops/oidc-passport/ Before you start. In 0163-irsa-cross-account, a program in one account showed a signed slip to another account and got a short-lived pass back, and nobody… https://chalkbrain.com/lessons/dsa/clinic-15-number-of-islands/ https://chalkbrain.com/lessons/dsa/clinic-15-number-of-islands/ Naksha mein har khaana zameen ya paani hai, aur board dikhata hai ki paas-paas ki zameen mil kar kaise ek hi tapu ban jaati hai - poora sabak isi ek… https://chalkbrain.com/lessons/system-design/google-search-case-study/ https://chalkbrain.com/lessons/system-design/google-search-case-study/ Before you start. You already know the feeling of typing a couple of words into a search box and getting a good answer back before you've finished sitting… https://chalkbrain.com/lessons/devops/irsa-cross-account/ https://chalkbrain.com/lessons/devops/irsa-cross-account/ Before you start. An AWS account is one company's or team's walled section of AWS, with its own bill and identities. A container is one program sealed with… https://chalkbrain.com/lessons/dsa/clinic-14-combination-sum/ https://chalkbrain.com/lessons/dsa/clinic-14-combination-sum/ Yeh drawing wahi pedh dikhati hai jo pichli clinic mein tha - chuno, aage jao, wapas aao - bas ab har shaakha par ek shart hai, aur toot-te hi turant kaat di… https://chalkbrain.com/lessons/system-design/google-maps-case-study/ https://chalkbrain.com/lessons/system-design/google-maps-case-study/ Before you start. A website that makes you wait for an answer is showing you latency - the time between asking and getting a reply. That's the only idea this… https://chalkbrain.com/lessons/devops/the-two-planes/ https://chalkbrain.com/lessons/devops/the-two-planes/ Before you start. You need to know what a pod is: one running copy of your program, given its own address on the network. And what an API call is: one program… https://chalkbrain.com/lessons/dsa/clinic-13-subsets/ https://chalkbrain.com/lessons/dsa/clinic-13-subsets/ Before you start. Array mein end pe daalna push aur end se nikaalna pop kehlata hai - bas itna pehle se pata hona chahiye. Function apne aap ko dobara bula… https://chalkbrain.com/lessons/system-design/zomato-case-study/ https://chalkbrain.com/lessons/system-design/zomato-case-study/ Before you start. You should already have a picture of a two-sided marketplace app in your head: a customer requesting something, one nearby provider, a system… https://chalkbrain.com/lessons/dsa/clinic-12-search-rotated/ https://chalkbrain.com/lessons/dsa/clinic-12-search-rotated/ Ye drawing ek sorted list dikhata hai jise beech se kaat ke uska pehla tukda peeche jod diya gaya hai - jahan sabse bada number achanak sabse chhote se takrata… https://chalkbrain.com/lessons/system-design/stripe-case-study/ https://chalkbrain.com/lessons/system-design/stripe-case-study/ Before you start. You need to know what a retry is: a computer that got no answer, so it asks again. That is the only background this note assumes; everything… https://chalkbrain.com/lessons/dsa/clinic-11-binary-search/ https://chalkbrain.com/lessons/dsa/clinic-11-binary-search/ Ye board dikhata hai ek sorted list ko baar-baar beech se kaatte jaana - har compare pe aadhi list hamesha ke liye phenk di jaati hai, jab tak ek hi ghar na… https://chalkbrain.com/lessons/system-design/twitter-case-study/ https://chalkbrain.com/lessons/system-design/twitter-case-study/ Before you start. All you need is the shape of a social feed: you post something, and the people who follow you expect to see it soon after. This lesson is… https://chalkbrain.com/lessons/dsa/clinic-10-diameter-of-binary-tree/ https://chalkbrain.com/lessons/dsa/clinic-10-diameter-of-binary-tree/ Before you start. 0158-clinic-09-maximum-depth mein ek kaam seekha tha: har node (tree ka ek dabba, jiske neeche zyada se zyada do bachche latak sakte hain) se… https://chalkbrain.com/lessons/system-design/whatsapp-case-study/ https://chalkbrain.com/lessons/system-design/whatsapp-case-study/ Before you start. You need to know that a phone app talking to the internet is really talking to a computer elsewhere called a server, and that anything you… https://chalkbrain.com/lessons/dsa/clinic-09-maximum-depth/ https://chalkbrain.com/lessons/dsa/clinic-09-maximum-depth/ Before you start. Itna kaafi hai: upar ek box hota hai, uske neeche zyada se zyada do box latak sakte hain, aur yehi shape neeche tak dohrati hai.… https://chalkbrain.com/lessons/system-design/uber-case-study/ https://chalkbrain.com/lessons/system-design/uber-case-study/ Before you start. You need to already be comfortable with the idea that a system can be forced to choose between always answering and always being right the… https://chalkbrain.com/lessons/dsa/clinic-08-daily-temperatures/ https://chalkbrain.com/lessons/dsa/clinic-08-daily-temperatures/ Board mein har din ek bar hai, uski height us din ka temperature. Jab tak koi lamba (garam) bar nahi aata, chhote bars ek ke upar ek dher mein khade rehte… https://chalkbrain.com/lessons/system-design/netflix-case-study/ https://chalkbrain.com/lessons/system-design/netflix-case-study/ Before you start. You need to know that a video file is enormous and a web page is tiny, and that somebody pays for every mile a file travels. That is all; the… https://chalkbrain.com/lessons/dsa/clinic-07-valid-parentheses/ https://chalkbrain.com/lessons/dsa/clinic-07-valid-parentheses/ Ye board ek hi sawaal dikhata hai: jab brackets ek doosre ke andar nested hon, abhi is pal sabse haal mein kaunsa khula bracket apne jode (partner) ka intezaar… https://chalkbrain.com/lessons/system-design/claude-case-study/ https://chalkbrain.com/lessons/system-design/claude-case-study/ Before you start. You need to know that a program can be handed some text and hand back some text. You also need to know that a chip must fetch a number out of… https://chalkbrain.com/lessons/dsa/clinic-06-longest-substring-no-repeat/ https://chalkbrain.com/lessons/dsa/clinic-06-longest-substring-no-repeat/ Yeh board ek string ke upar do pointer dikhata hai jo aage khiskte hain, jaise ek chhoti khidki jo repeat milte hi thoda simat jaati hai - par kabhi peeche… https://chalkbrain.com/lessons/system-design/openrouter-case-study/ https://chalkbrain.com/lessons/system-design/openrouter-case-study/ Before you start. When an app answers you with AI, it is sending your text over the internet to a machine somebody else owns and paying that owner for the… https://chalkbrain.com/lessons/dsa/clinic-05-two-sum/ https://chalkbrain.com/lessons/dsa/clinic-05-two-sum/ Before you start. List mein har number ki apni jagah hoti hai - use uska index kehte hain, 0 se ginkar. 0175-foundations-04-hashing ya 0058-hashing ne hash map… https://chalkbrain.com/lessons/system-design/amazon-case-study/ https://chalkbrain.com/lessons/system-design/amazon-case-study/ Before you start. Amazon is two businesses sharing one set of machines: the shop you buy from, and AWS, which rents storage and computers to everybody else.… https://chalkbrain.com/lessons/dsa/clinic-04-meeting-rooms-ii/ https://chalkbrain.com/lessons/dsa/clinic-04-meeting-rooms-ii/ LC 253 LeetCode par premium hai. Wahi sawaal free mein: LC 2406 Divide Intervals Into Minimum Number of Groups (bas wahan end time bhi shamil hai, to jo… https://chalkbrain.com/lessons/system-design/dropbox-case-study/ https://chalkbrain.com/lessons/system-design/dropbox-case-study/ Before you start. You need one fact: Dropbox keeps a copy of your files on your own machine and another copy on its machines, and those two copies drift apart… https://chalkbrain.com/lessons/dsa/clinic-03-coin-change/ https://chalkbrain.com/lessons/dsa/clinic-03-coin-change/ Pehle padho. 0151-clinic-02-climbing-stairs mein tumne "aakhri kadam kya tha?" poochna seekha, aur mile hue jawabon ko sirf jodna (SUM) tha. Yahan sawaal wahi… https://chalkbrain.com/lessons/system-design/cloudflare-case-study/ https://chalkbrain.com/lessons/system-design/cloudflare-case-study/ Before you start. You need to know that when you open a web address, some machine somewhere has to answer it, and that the nearer that machine is, the faster… https://chalkbrain.com/lessons/dsa/clinic-02-climbing-stairs/ https://chalkbrain.com/lessons/dsa/clinic-02-climbing-stairs/ Before you start. Tumhe bas itna chahiye: ek function ko apna hi chhota version phir se bulaana aana chahiye, aur number jodna-ghatana. Agar ye bilkul naya… https://chalkbrain.com/lessons/system-design/figma-case-study/ https://chalkbrain.com/lessons/system-design/figma-case-study/ Before you start. You need to have used an app where two people type into the same document at once and neither has to wait for the other. This lesson is the… https://chalkbrain.com/lessons/dsa/clinic-01-reverse-a-linked-list/ https://chalkbrain.com/lessons/dsa/clinic-01-reverse-a-linked-list/ Before you start. Itna kaafi hai: ek variable sirf ek jagah ka naam hota hai, chitthi par likhe pate jaisa - khud jagah nahi.… https://chalkbrain.com/lessons/system-design/slack-case-study/ https://chalkbrain.com/lessons/system-design/slack-case-study/ Before you start. Slack is the chat app companies use instead of email, and it was built on one idea: a company signs up, gets its own private space, and… https://chalkbrain.com/lessons/dsa/git-diff-case-study/ https://chalkbrain.com/lessons/dsa/git-diff-case-study/ Every time you run git diff, something underneath is solving a puzzle: what is the fewest line changes that turn the old file into the new one. It looks like… https://chalkbrain.com/lessons/portfolio-projects/s3-build-a-redis-server/ https://chalkbrain.com/lessons/portfolio-projects/s3-build-a-redis-server/ Before you start. You need S1's log: a file you only add to at the end, replayed from the top on startup, which lets a store of names and values survive being… https://chalkbrain.com/lessons/portfolio-projects/s2-build-a-load-balancer/ https://chalkbrain.com/lessons/portfolio-projects/s2-build-a-load-balancer/ Before you start. You've met ports and connections in the networking track, and you can run a terminal command and write a little code, but you've never built… https://chalkbrain.com/lessons/portfolio-projects/s1-build-your-own-database/ https://chalkbrain.com/lessons/portfolio-projects/s1-build-your-own-database/ Before you start. You can run a terminal command and write a little code, but you've never built a database - so far you've only read how real ones work, in… https://chalkbrain.com/lessons/system-design/youtube-case-study/ https://chalkbrain.com/lessons/system-design/youtube-case-study/ Before you start. You need to know that a website is a program running on rented computers, and that it keeps its facts in a database, a filing cabinet a… https://chalkbrain.com/lessons/devops/company-platforms-deep-set/ https://chalkbrain.com/lessons/devops/company-platforms-deep-set/ Before you start. 0025-iac-and-gitops showed you a pipeline that rebuilds a whole server from a text file, and 0026-the-assembly-line-cicd showed you the… https://chalkbrain.com/lessons/system-design/instagram-case-study/ https://chalkbrain.com/lessons/system-design/instagram-case-study/ Before you start. You need to know that a website is a program running on rented computers, and that it keeps its information in a database, a filing cabinet a… https://chalkbrain.com/lessons/how-things-work/fan-cools-you-not-the-room/ https://chalkbrain.com/lessons/how-things-work/fan-cools-you-not-the-room/ Before you start. You need one idea from earlier in this site: cooling never needs a source of cold, only heat leaving fast enough, which is why a fridge's… https://chalkbrain.com/lessons/how-things-work/remote-invisible-light/ https://chalkbrain.com/lessons/how-things-work/remote-invisible-light/ Before you start. You only need to know what a TV remote does and that your eyes can miss things that are still there; nothing about waves, light or radiation… https://chalkbrain.com/lessons/how-things-work/cold-is-a-lie/ https://chalkbrain.com/lessons/how-things-work/cold-is-a-lie/ Before you start. All you need is a fridge, an AC, or a memory of touching a cold metal railing. 0136-entropy-arrow-of-time showed that the universe drifts… https://chalkbrain.com/lessons/devops/ai-ml-ops-deep-set/ https://chalkbrain.com/lessons/devops/ai-ml-ops-deep-set/ Before you start. 0109-inference-at-scale and 0111-evaluating-models already showed you how a trained model turns a prompt into an answer; this note is about… https://chalkbrain.com/lessons/system-design/discord-case-study/ https://chalkbrain.com/lessons/system-design/discord-case-study/ Before you start. You need one idea: a chat app must remember everything anyone ever typed, and must also hand what someone typed one second ago to everyone… https://chalkbrain.com/lessons/database-internals/database-internals-deep-set/ https://chalkbrain.com/lessons/database-internals/database-internals-deep-set/ Before you start. 0128-the-right-book-for-the-job picked Postgres as the database under the app. 0035-storage-engines showed the two ways an engine stores… https://chalkbrain.com/lessons/how-things-work/entropy-arrow-of-time/ https://chalkbrain.com/lessons/how-things-work/entropy-arrow-of-time/ Before you start. You need to be comfortable counting: if you have three coins, how many ways can they land? That is the only maths in the story half, and… https://chalkbrain.com/lessons/dsa/bit-manipulation/ https://chalkbrain.com/lessons/dsa/bit-manipulation/ Before you start. One fact from 0047-numbers-for-real: a computer stores every whole number as a row of ones and zeros, read like ordinary counting but in twos… https://chalkbrain.com/lessons/dsa/complexity-theory/ https://chalkbrain.com/lessons/dsa/complexity-theory/ Before you start. You already know how to read a cost off a line of code - count the loops, and see whether doubling the input doubles the work… https://chalkbrain.com/lessons/dsa/trees-worked/ https://chalkbrain.com/lessons/dsa/trees-worked/ Before you start. You need a binary search tree: a box holding one key, pointing at up to two others, smaller keys down the left and bigger down the right, so… https://chalkbrain.com/lessons/dsa/string-algorithms/ https://chalkbrain.com/lessons/dsa/string-algorithms/ Before you start. You need to be comfortable with an index - a number naming one slot in a list, counting up from the start - and a loop that walks one array… https://chalkbrain.com/lessons/system-design/four-systems-in-one-night/ https://chalkbrain.com/lessons/system-design/four-systems-in-one-night/ Before you start. You need the six steps every design question runs through: clarify what is being asked, estimate the scale, sketch the API, choose the data… https://chalkbrain.com/lessons/system-design/the-contract-at-the-counter/ https://chalkbrain.com/lessons/system-design/the-contract-at-the-counter/ Before you start. You need two facts: that programs talk to each other over a network, and that a network sometimes loses the message, so the sender is left… https://chalkbrain.com/lessons/system-design/the-slate-properly/ https://chalkbrain.com/lessons/system-design/the-slate-properly/ Before you start. You need one picture from 0012-the-slate-next-to-the-stove: the chai stall chalks the day's prices on a slate by the stove, so nobody walks… https://chalkbrain.com/lessons/system-design/the-right-book-for-the-job/ https://chalkbrain.com/lessons/system-design/the-right-book-for-the-job/ Before you start. You need one idea from earlier in this track: a system keeps its facts somewhere and reads them back later, which is the ledger… https://chalkbrain.com/lessons/personal-development/your-personal-operating-system/ https://chalkbrain.com/lessons/personal-development/your-personal-operating-system/ Before you start. The last lesson's whole point was that energy is not willpower - it is a resource you either protect or spend down, and a bad night's sleep… https://chalkbrain.com/lessons/personal-development/energy-goals-and-recovery/ https://chalkbrain.com/lessons/personal-development/energy-goals-and-recovery/ Before you start. The last lesson showed that even a good decision can get hijacked by a shortcut your brain reaches for without asking you… https://chalkbrain.com/lessons/personal-development/mental-models-and-decisions/ https://chalkbrain.com/lessons/personal-development/mental-models-and-decisions/ Before you start. The last lesson showed that a page you've re-read until it feels familiar is not one you can actually reproduce cold - that feeling of "I… https://chalkbrain.com/lessons/personal-development/how-to-learn/ https://chalkbrain.com/lessons/personal-development/how-to-learn/ Before you start. Deliberate practice said a skill grows at the edge of what you can already do, with fast feedback on exactly what went wrong, never from the… https://chalkbrain.com/lessons/personal-development/habits-and-behaviour-change/ https://chalkbrain.com/lessons/personal-development/habits-and-behaviour-change/ Before you start. The last lesson showed that protecting your attention works by removing the thing that distracts you, not by resisting it in the moment -… https://chalkbrain.com/lessons/personal-development/deep-work-and-attention/ https://chalkbrain.com/lessons/personal-development/deep-work-and-attention/ Before you start. 0121-deliberate-practice told you that skill grows only when you are working right at the edge of what you can do, pushed by real feedback.… https://chalkbrain.com/lessons/personal-development/deliberate-practice/ https://chalkbrain.com/lessons/personal-development/deliberate-practice/ Before you start. The first read needs no maths beyond a percentage; the second adds an exponent and a logarithm, each explained where it appears. You need… https://chalkbrain.com/lessons/system-design/design-the-whole-thing/ https://chalkbrain.com/lessons/system-design/design-the-whole-thing/ Before you start. You need the five things the earlier lessons built, and one clause each is enough to bring them back: a load balancer (one public address in… https://chalkbrain.com/lessons/system-design/the-ledger-everyone-reads/ https://chalkbrain.com/lessons/system-design/the-ledger-everyone-reads/ Before you start. You need the order-slip rail from 0013-the-order-slip-rail: the order goes on a spike, the kitchen takes the next slip whenever it is free,… https://chalkbrain.com/lessons/system-design/charged-twice/ https://chalkbrain.com/lessons/system-design/charged-twice/ Before you start. 0030-the-cut-phone-line gave you the one fact everything here rests on: a network drops messages, and it drops them silently, so you have to… https://chalkbrain.com/lessons/dsa/interview-patterns/ https://chalkbrain.com/lessons/dsa/interview-patterns/ Before you start. 0064-algorithm-design-patterns gave you the four families a solution can come from - divide and conquer, greedy, dynamic programming,… https://chalkbrain.com/lessons/dsa/shortest-paths-worked/ https://chalkbrain.com/lessons/dsa/shortest-paths-worked/ Before you start. 0061-graphs showed you BFS, which finds the shortest path by counting turns, one road at a time - it assumes every road takes the same effort… https://chalkbrain.com/lessons/ai/mixture-of-experts/ https://chalkbrain.com/lessons/ai/mixture-of-experts/ Before you start. A language model is a pile of numbers, called its weights or its dials; the formal word is parameters. There are billions of them, and the… https://chalkbrain.com/lessons/ai/multimodality/ https://chalkbrain.com/lessons/ai/multimodality/ Before you start. From 0077-tokens-and-embeddings: a word never enters a language model as letters, it is swapped for a long list of numbers first, and every… https://chalkbrain.com/lessons/ai/diffusion-and-image-generation/ https://chalkbrain.com/lessons/ai/diffusion-and-image-generation/ Before you start. One thing from 0075-what-a-neural-network-is: a network is a pile of dials, and turning them needs examples where somebody already knows the… https://chalkbrain.com/lessons/dsa/dynamic-programming-worked/ https://chalkbrain.com/lessons/dsa/dynamic-programming-worked/ Before you start. You need 0063-recursion-and-dynamic-programming's idea already: solve a small version of a question once, write the answer down, and build… https://chalkbrain.com/lessons/ai/evaluating-models/ https://chalkbrain.com/lessons/ai/evaluating-models/ Before you start. You need to know a language model writes open-ended text rather than picking one label from a list, which is what 0080-inference showed it… https://chalkbrain.com/lessons/ai/reasoning-and-test-time-compute/ https://chalkbrain.com/lessons/ai/reasoning-and-test-time-compute/ Before you start. Three words from earlier lessons. Layers are the rows a model is built from, and it has a fixed number of them. Dials are the adjustable… https://chalkbrain.com/lessons/ai/inference-at-scale/ https://chalkbrain.com/lessons/ai/inference-at-scale/ Before you start. From 0080-inference: a model writes a reply one token at a time, a token being a chunk of text about three quarters of a word, running the… https://chalkbrain.com/lessons/devops/under-the-hood/ https://chalkbrain.com/lessons/devops/under-the-hood/ Before you start. A pod is one running container, or a few that share a machine, that the cluster promises to keep running; 0021-the-harbor-master showed that,… https://chalkbrain.com/lessons/system-design/the-two-ledgers-must-agree/ https://chalkbrain.com/lessons/system-design/the-two-ledgers-must-agree/ Before you start. 0029-splitting-the-ledger cut the chai stall's single ledger into two books, so two boys could write at once, each book holding half the… https://chalkbrain.com/lessons/security/llm-and-agent-security/ https://chalkbrain.com/lessons/security/llm-and-agent-security/ Before you start. 0070-web-vulnerabilities showed a website taking a stranger's typed text and running it as if it were a command - a login box or a comment… https://chalkbrain.com/lessons/philosophy/free-will-identity-and-the-self/ https://chalkbrain.com/lessons/philosophy/free-will-identity-and-the-self/ Before you start. The last lesson, on ethics, gave you three tests for whether an act is right: does it produce good outcomes, does it follow the right rule,… https://chalkbrain.com/lessons/philosophy/ethics-and-the-good-life/ https://chalkbrain.com/lessons/philosophy/ethics-and-the-good-life/ Before you start. The last lesson looked at how arguments work, and where reasoning hits walls it can prove exist but never get past. This lesson is about a… https://chalkbrain.com/lessons/philosophy/logic-truth-and-the-limits-of-reason/ https://chalkbrain.com/lessons/philosophy/logic-truth-and-the-limits-of-reason/ Before you start. 0099-what-is-knowledge asked what you can know, and settled on a belief you can actually justify as true. 0102-can-machines-think asked… https://chalkbrain.com/lessons/philosophy/can-machines-think/ https://chalkbrain.com/lessons/philosophy/can-machines-think/ Before you start. The last lesson left the hard problem open: nothing you can observe from outside proves whether a system has an inner life, only what it… https://chalkbrain.com/lessons/philosophy/consciousness-and-the-hard-problem/ https://chalkbrain.com/lessons/philosophy/consciousness-and-the-hard-problem/ Before you start. 0100-the-mind-body-problem left you with functionalism: a mind is defined by what it does, so a brain and a machine that behaved identically… https://chalkbrain.com/lessons/philosophy/the-mind-body-problem/ https://chalkbrain.com/lessons/philosophy/the-mind-body-problem/ Before you start. 0099-what-is-knowledge asked how sure you can ever be that you know something; this lesson asks a different question about the same "you" -… https://chalkbrain.com/lessons/philosophy/what-is-knowledge/ https://chalkbrain.com/lessons/philosophy/what-is-knowledge/ Before you start. You don't need anything from an earlier lesson - this is the first stop in Philosophy. All you need is a habit you already have: trusting a… https://chalkbrain.com/lessons/game-theory/auctions-and-mechanism-design/ https://chalkbrain.com/lessons/game-theory/auctions-and-mechanism-design/ Before you start. The last lesson showed how two strangers who must meet in New York without phoning end up at the same spot: noon at Grand Central, the… https://chalkbrain.com/lessons/game-theory/coordination-schelling-points-and-credible-threats/ https://chalkbrain.com/lessons/game-theory/coordination-schelling-points-and-credible-threats/ Before you start. You already have a Nash equilibrium: an outcome where nobody wants to switch, given what everyone else does. Last lesson, being predictable… https://chalkbrain.com/lessons/game-theory/mixed-strategies-and-zero-sum/ https://chalkbrain.com/lessons/game-theory/mixed-strategies-and-zero-sum/ Before you start. The last lesson (0095-repeated-games-and-cooperation) showed that when you expect to face the same player again and again, being predictable… https://chalkbrain.com/lessons/game-theory/repeated-games-and-cooperation/ https://chalkbrain.com/lessons/game-theory/repeated-games-and-cooperation/ Before you start. 0094-the-prisoners-dilemma found the trap: questioned once, in separate rooms, ratting out your partner beats staying silent no matter what… https://chalkbrain.com/lessons/game-theory/the-prisoners-dilemma/ https://chalkbrain.com/lessons/game-theory/the-prisoners-dilemma/ Before you start. 0093-dominant-strategies-and-nash-equilibrium showed you how to read a table of outcomes and pick the one move that beats every other move no… https://chalkbrain.com/lessons/game-theory/dominant-strategies-and-nash-equilibrium/ https://chalkbrain.com/lessons/game-theory/dominant-strategies-and-nash-equilibrium/ Before you start. You need one thing from 0092-what-is-a-game: a game is any situation where what you get depends on a choice someone else makes too, and every… https://chalkbrain.com/lessons/game-theory/what-is-a-game/ https://chalkbrain.com/lessons/game-theory/what-is-a-game/ Before you start. 0088-decision-making-under-uncertainty showed that an uncertain choice has a best answer once you write down what could happen and what each… https://chalkbrain.com/lessons/psychology/learning-and-expertise/ https://chalkbrain.com/lessons/psychology/learning-and-expertise/ Before you start. The last lesson showed you why you do things at all: a brain signal that tracks surprise more than pleasure, and a four-part loop - a cue, a… https://chalkbrain.com/lessons/psychology/motivation-and-habit/ https://chalkbrain.com/lessons/psychology/motivation-and-habit/ Before you start. The last lesson, 0089-attention-and-cognitive-load, showed that your working memory is a small tank that fills up fast under focused thought.… https://chalkbrain.com/lessons/psychology/attention-and-cognitive-load/ https://chalkbrain.com/lessons/psychology/attention-and-cognitive-load/ Before you start. 0085-how-memory-works showed you that what you can hold in mind right now is a small working space, not the same thing as everything you… https://chalkbrain.com/lessons/psychology/decision-making-under-uncertainty/ https://chalkbrain.com/lessons/psychology/decision-making-under-uncertainty/ Before you start. 0087-cognitive-biases named loss aversion and framing as two of the mind's shortcuts. This lesson turns those into the full arithmetic of a… https://chalkbrain.com/lessons/psychology/cognitive-biases/ https://chalkbrain.com/lessons/psychology/cognitive-biases/ Before you start. 0085-how-memory-works and 0086-spaced-repetition were about getting facts to stick; this lesson is about a different failure - a guess that… https://chalkbrain.com/lessons/psychology/spaced-repetition/ https://chalkbrain.com/lessons/psychology/spaced-repetition/ Before you start. 0085-how-memory-works showed you that a memory isn't a recording you replay, it's a guess you rebuild each time, and that guess fades on a… https://chalkbrain.com/lessons/psychology/how-memory-works/ https://chalkbrain.com/lessons/psychology/how-memory-works/ Before you start. You need nothing but your own memory: the kind of phone number you can repeat back this second and lose in a minute, and an old memory you'd… https://chalkbrain.com/lessons/ai/capstone-anatomy-of-an-llm-app/ https://chalkbrain.com/lessons/ai/capstone-anatomy-of-an-llm-app/ Before you start. You need the returns desk from 0083-agents-and-tool-use, where the clerk who can only talk is the large language model itself, the printed… https://chalkbrain.com/lessons/ai/agents-and-tool-use/ https://chalkbrain.com/lessons/ai/agents-and-tool-use/ Before you start. You need one fact from 0080-inference: a language model is a text predictor. You hand it a string, it hands back the next stretch of text,… https://chalkbrain.com/lessons/ai/rag/ https://chalkbrain.com/lessons/ai/rag/ Before you start. You need 0077-tokens-and-embeddings, which turned any piece of text into a point in space, positioned so that two texts about the same thing… https://chalkbrain.com/lessons/ai/fine-tuning-and-alignment/ https://chalkbrain.com/lessons/ai/fine-tuning-and-alignment/ Before you start. You need one fact from 0079-from-transformer-to-llm: training a language model means hiding the next word in a mountain of text and turning… https://chalkbrain.com/lessons/ai/inference/ https://chalkbrain.com/lessons/ai/inference/ Before you start. You need to know that the model is a huge pile of fixed numbers, its parameters (a 7-billion-parameter model is a pile of 7 billion of them),… https://chalkbrain.com/lessons/ai/from-transformer-to-llm/ https://chalkbrain.com/lessons/ai/from-transformer-to-llm/ Before you start. The last lesson built a machine that takes a row of word-pieces and lets each absorb meaning from the ones before it, and 0076-how-it-learns… https://chalkbrain.com/lessons/ai/attention-and-the-transformer/ https://chalkbrain.com/lessons/ai/attention-and-the-transformer/ Before you start. 0077-tokens-and-embeddings turned every word into a list of numbers, each looked up on its own, so the "bank" in river bank and the "bank" in… https://chalkbrain.com/lessons/ai/tokens-and-embeddings/ https://chalkbrain.com/lessons/ai/tokens-and-embeddings/ Before you start. From 0075-what-a-neural-network-is you need one fact: a neural network is votes times dials, added up, row after row, and it eats numbers and… https://chalkbrain.com/lessons/ai/how-it-learns/ https://chalkbrain.com/lessons/ai/how-it-learns/ Before you start. You need 0075-what-a-neural-network-is, which gave you this: a network is rows of small units, each adding up weighted votes and passing the… https://chalkbrain.com/lessons/ai/what-a-neural-network-is/ https://chalkbrain.com/lessons/ai/what-a-neural-network-is/ Before you start. You need to be able to multiply two numbers and add the results. That is the whole of the maths in the story half of this lesson. Everything… https://chalkbrain.com/lessons/security/anatomy-of-a-breach/ https://chalkbrain.com/lessons/security/anatomy-of-a-breach/ Before you start. 0072-application-hardening gave you the individual locks: patch known bugs, watch your own logs, and let no login reach more than it needs… https://chalkbrain.com/lessons/security/application-hardening/ https://chalkbrain.com/lessons/security/application-hardening/ Before you start. The last lesson forced every connection to the app onto an encrypted, verified line, so nobody between you and it can read or tamper with a… https://chalkbrain.com/lessons/security/transport-and-network-security/ https://chalkbrain.com/lessons/security/transport-and-network-security/ Before you start. 0067-cryptography-foundations gave you the three tools this lesson wires together: a shared key that locks and unlocks the same way on both… https://chalkbrain.com/lessons/security/web-vulnerabilities/ https://chalkbrain.com/lessons/security/web-vulnerabilities/ Before you start. The last lesson showed a trusted helper doing something for the wrong person because it never checked who really asked -… https://chalkbrain.com/lessons/security/authorization-and-identity/ https://chalkbrain.com/lessons/security/authorization-and-identity/ Before you start. 0068-authentication proved who is asking, a token that says "this really is you". This lesson answers the harder question that token can't:… https://chalkbrain.com/lessons/security/authentication/ https://chalkbrain.com/lessons/security/authentication/ Before you start. 0067-cryptography-foundations gave you two tools this lesson spends. A hash scrambles anything into a fixed-length string with no way back to… https://chalkbrain.com/lessons/security/cryptography-foundations/ https://chalkbrain.com/lessons/security/cryptography-foundations/ Before you start. 0066-the-security-mindset gave you three things worth protecting: keep a message secret, keep it unchanged, know who sent it. This lesson… https://chalkbrain.com/lessons/security/the-security-mindset/ https://chalkbrain.com/lessons/security/the-security-mindset/ Before you start. You need two things from earlier in this site: that a program is just instructions a machine runs one at a time… https://chalkbrain.com/lessons/dsa/capstone-attack-an-unseen-problem/ https://chalkbrain.com/lessons/dsa/capstone-attack-an-unseen-problem/ Before you start. You've already met a judge for speed (Big-O), a shelf of structures to hold data, and a handful of strategies for attacking it, one lesson at… https://chalkbrain.com/lessons/dsa/algorithm-design-patterns/ https://chalkbrain.com/lessons/dsa/algorithm-design-patterns/ Before you start. You've already picked up five tools in this track - arrays, hash maps, trees, graphs, and the trick of a problem asking a smaller copy of… https://chalkbrain.com/lessons/dsa/recursion-and-dynamic-programming/ https://chalkbrain.com/lessons/dsa/recursion-and-dynamic-programming/ Before you start. You need to know what happens when one function calls another: it pauses, hands control over, and picks back up exactly where it left off… https://chalkbrain.com/lessons/dsa/sorting-and-searching/ https://chalkbrain.com/lessons/dsa/sorting-and-searching/ Before you start. You need an index - a position in a list, counted from zero - and what 0056-complexity-and-big-o already gave you: that doubling the input… https://chalkbrain.com/lessons/dsa/graphs/ https://chalkbrain.com/lessons/dsa/graphs/ Before you start. You need only arrays: a numbered list where you can jump straight to slot i. Nothing about graphs is assumed here - every new word below is… https://chalkbrain.com/lessons/dsa/trees/ https://chalkbrain.com/lessons/dsa/trees/ Before you start. You already have the hash table from 0058-hashing: instant lookup, no memory of order. This lesson is the opposite trade - a shape that… https://chalkbrain.com/lessons/dsa/linked-lists-stacks-queues/ https://chalkbrain.com/lessons/dsa/linked-lists-stacks-queues/ Before you start. You need one fact from 0057-arrays-and-strings: an array keeps its values side by side in memory (why that is the fast kind of memory:… https://chalkbrain.com/lessons/dsa/hashing/ https://chalkbrain.com/lessons/dsa/hashing/ Before you start. You need arrays from 0057-arrays-and-strings: a numbered row of slots you can jump straight to. You also need O(1) and O(n) from… https://chalkbrain.com/lessons/dsa/arrays-and-strings/ https://chalkbrain.com/lessons/dsa/arrays-and-strings/ Before you start. You need 0056-complexity-and-big-o's two words: O(1) means the cost never changes as the pile grows, O(n) means it climbs in step with it.… https://chalkbrain.com/lessons/dsa/complexity-and-big-o/ https://chalkbrain.com/lessons/dsa/complexity-and-big-o/ Before you start. You need to be able to multiply two numbers and read a simple table - that's the whole of the maths in the story half. This is the first… https://chalkbrain.com/lessons/how-computers-work/capstone-keypress-to-pixel/ https://chalkbrain.com/lessons/how-computers-work/capstone-keypress-to-pixel/ Before you start. The lesson on I/O, interrupts and the bus gave you the last piece: a device raises a flag instead of waiting to be asked, and a DMA… https://chalkbrain.com/lessons/how-computers-work/io-interrupts-and-the-bus/ https://chalkbrain.com/lessons/how-computers-work/io-interrupts-and-the-bus/ Before you start. You already know a CPU can be structured to juggle many tasks at once without finishing any of them first - that's concurrency, and it's what… https://chalkbrain.com/lessons/how-computers-work/concurrency-and-the-hardware/ https://chalkbrain.com/lessons/how-computers-work/concurrency-and-the-hardware/ Before you start. 0052-processes-threads-and-the-kernel gave you a thread, one flow of instructions moving through a program, and a process, the box that can… https://chalkbrain.com/lessons/how-computers-work/processes-threads-and-the-kernel/ https://chalkbrain.com/lessons/how-computers-work/processes-threads-and-the-kernel/ Before you start. The last lesson, 0051-how-code-becomes-execution, left one program's instructions running as a single stream on the CPU, fetched and carried… https://chalkbrain.com/lessons/how-computers-work/how-code-becomes-execution/ https://chalkbrain.com/lessons/how-computers-work/how-code-becomes-execution/ Before you start. 0049-the-cpu left you with a chip that does exactly one thing, forever: fetch an instruction, decode it, run it - it has no idea what a word,… https://chalkbrain.com/lessons/how-computers-work/the-memory-hierarchy/ https://chalkbrain.com/lessons/how-computers-work/the-memory-hierarchy/ Before you start. 0049-the-cpu gave you the loop: fetch, decode, execute, using registers built into the chip. Almost every step eventually needs data from… https://chalkbrain.com/lessons/how-computers-work/the-cpu/ https://chalkbrain.com/lessons/how-computers-work/the-cpu/ Before you start. 0048-logic-from-switches built an ALU, a circuit of gates that adds and compares numbers, and flip-flops, single bits of memory that hold a… https://chalkbrain.com/lessons/how-computers-work/logic-from-switches/ https://chalkbrain.com/lessons/how-computers-work/logic-from-switches/ Before you start. 0046-everything-is-a-number showed a computer stores everything as switches wired into bytes, each one only on or off. 0047-numbers-for-real… https://chalkbrain.com/lessons/how-computers-work/numbers-for-real/ https://chalkbrain.com/lessons/how-computers-work/numbers-for-real/ Before you start. You need what 0046-everything-is-a-number gave you: a computer stores everything as a fixed row of on-or-off switches, bits, and eight of… https://chalkbrain.com/lessons/how-computers-work/everything-is-a-number/ https://chalkbrain.com/lessons/how-computers-work/everything-is-a-number/ Before you start. Nothing from earlier in this track - this is where it begins. All the arithmetic in the story half is counting to eight and doubling;… https://chalkbrain.com/lessons/data-systems/where-data-systems-are-heading/ https://chalkbrain.com/lessons/data-systems/where-data-systems-are-heading/ Before you start. The previous lesson, 0044-stream-processing, showed you the log: an add-only list of what happened, read from the start whenever you want to… https://chalkbrain.com/lessons/data-systems/stream-processing/ https://chalkbrain.com/lessons/data-systems/stream-processing/ Before you start. The previous lesson, 0043-batch-processing, waited until a whole day's file sat still on disk, then worked through it once, start to finish -… https://chalkbrain.com/lessons/data-systems/batch-processing/ https://chalkbrain.com/lessons/data-systems/batch-processing/ Before you start. You need to know what a hash function does. Feed it anything and it gives back a number: feed it the word "the" and it might give back 10.… https://chalkbrain.com/lessons/data-systems/consistency-and-consensus/ https://chalkbrain.com/lessons/data-systems/consistency-and-consensus/ Before you start. The previous lesson, 0041-when-machines-fail, showed why no machine in a group can fully trust the network: a message can be late, lost, or… https://chalkbrain.com/lessons/data-systems/when-machines-fail/ https://chalkbrain.com/lessons/data-systems/when-machines-fail/ Before you start. The previous lesson, 0038-partitioning, cut one dataset into pieces and gave each machine only its piece. Earlier, 0040-transactions gave you… https://chalkbrain.com/lessons/data-systems/transactions/ https://chalkbrain.com/lessons/data-systems/transactions/ This lesson is about what happens when two things touch the same row, or the same rule, at once, and why the fix that catches all of it is expensive on purpose. https://chalkbrain.com/lessons/data-systems/storage-extras/ https://chalkbrain.com/lessons/data-systems/storage-extras/ Before you start. The last lesson found two ways a database keeps data findable without reading every row - a sorted tree of pages, or sorted files merged in… https://chalkbrain.com/lessons/data-systems/partitioning/ https://chalkbrain.com/lessons/data-systems/partitioning/ Before you start. The previous lesson, replication 0037-replication, put a copy of the whole dataset on every machine, so losing one machine cost you nothing.… https://chalkbrain.com/lessons/data-systems/replication/ https://chalkbrain.com/lessons/data-systems/replication/ Before you start. The previous lesson, 0040-transactions, was one machine's promise: a group of writes lands whole or not at all, even with two people touching… https://chalkbrain.com/lessons/data-systems/encoding-and-schema-change/ https://chalkbrain.com/lessons/data-systems/encoding-and-schema-change/ Before you start. The previous lessons, 0035-storage-engines and 0039-storage-extras, were about how a database lays bytes out on disk so it can find them… https://chalkbrain.com/lessons/data-systems/storage-engines/ https://chalkbrain.com/lessons/data-systems/storage-engines/ Before you start. The previous lesson, 0034-data-models-and-query-languages, was about the shapes you're allowed to store your data in - tables, documents,… https://chalkbrain.com/lessons/data-systems/data-models-and-query-languages/ https://chalkbrain.com/lessons/data-systems/data-models-and-query-languages/ Before you start. The previous lesson, 0033-what-a-data-system-promises, gave you the three things a system owes you: correct, fast, still standing at 3am.… https://chalkbrain.com/lessons/data-systems/what-a-data-system-promises/ https://chalkbrain.com/lessons/data-systems/what-a-data-system-promises/ Before you start. You already know a system can only handle so much before something gives; 0011-the-chai-stall showed a food stall's queue outgrowing what one… https://chalkbrain.com/lessons/system-design/who-holds-the-pen/ https://chalkbrain.com/lessons/system-design/who-holds-the-pen/ Before you start. The stall in this lesson is the chai stall the System Design spine has been running: five clerks behind the counter, each keeping a copy of… https://chalkbrain.com/lessons/system-design/the-bouncer-at-the-door/ https://chalkbrain.com/lessons/system-design/the-bouncer-at-the-door/ Before you start. You need one fact about a server: it finishes a fixed amount of work per second, and a request is one job handed to it. And one about… https://chalkbrain.com/lessons/system-design/the-cut-phone-line/ https://chalkbrain.com/lessons/system-design/the-cut-phone-line/ Before you start. The copy clerks gave you a second book that copies from the first and can fall behind by seconds; splitting the ledger spread the stall's… https://chalkbrain.com/lessons/system-design/splitting-the-ledger/ https://chalkbrain.com/lessons/system-design/splitting-the-ledger/ Before you start. You need the copy clerks: one master book, several clerks holding copies, so a hundred people read at once. That lesson left one thing… https://chalkbrain.com/lessons/devops/least-privilege-12-factor/ https://chalkbrain.com/lessons/devops/least-privilege-12-factor/ Before you start. You need to know what a container is: a program shipped sealed in its own box, so the same box runs the same way everywhere it lands.… https://chalkbrain.com/lessons/devops/the-watchtower-observability/ https://chalkbrain.com/lessons/devops/the-watchtower-observability/ Before you start. You need to know what a pod is: Kubernetes's word for one running copy of your program, which it can kill and restart elsewhere without… https://chalkbrain.com/lessons/devops/the-assembly-line-cicd/ https://chalkbrain.com/lessons/devops/the-assembly-line-cicd/ Before you start. A commit is one saved change to the code. An image is a frozen, runnable copy of the app and everything it needs, sealed once and never… https://chalkbrain.com/lessons/devops/iac-and-gitops/ https://chalkbrain.com/lessons/devops/iac-and-gitops/ Before you start. You need to know what a shared, versioned file is - a document with a history everyone can see, which is what people mean by keeping… https://chalkbrain.com/lessons/devops/getting-traffic-in/ https://chalkbrain.com/lessons/devops/getting-traffic-in/ Before you start. 0021-the-harbor-master gave you the Service: a stable name inside the cluster that always points at whichever pods are alive right now, no… https://chalkbrain.com/lessons/devops/config-secrets-storage/ https://chalkbrain.com/lessons/devops/config-secrets-storage/ Before you start. 0021-the-harbor-master showed you that Kubernetes can kill any pod and start an identical one in its place, unasked. This lesson covers what… https://chalkbrain.com/lessons/devops/self-healing-and-autoscaling/ https://chalkbrain.com/lessons/devops/self-healing-and-autoscaling/ Before you start. 0020-one-crate-to-a-fleet gave you the loop that keeps checking what's running against what you asked for and fixes the gap;… https://chalkbrain.com/lessons/devops/the-harbor-master/ https://chalkbrain.com/lessons/devops/the-harbor-master/ Before you start. A container is one program sealed in a box with everything it needs, running on a machine; an image is the frozen copy of that box you start… https://chalkbrain.com/lessons/devops/one-crate-to-a-fleet/ https://chalkbrain.com/lessons/devops/one-crate-to-a-fleet/ Before you start. Compose's whole world was one docker-compose.yml, one command, one machine: docker compose up starts every container you listed and then it's… https://chalkbrain.com/lessons/devops/many-at-once-compose/ https://chalkbrain.com/lessons/devops/many-at-once-compose/ Before you start. An image is the frozen box the last lesson built, docker run starts one, and its flags are the extra words on that command that punch holes… https://chalkbrain.com/lessons/devops/the-recipe-dockerfile/ https://chalkbrain.com/lessons/devops/the-recipe-dockerfile/ Before you start. You've already built an image and watched a container start from it - that image is a stack of layers, frozen once each one is written. This… https://chalkbrain.com/lessons/devops/running-a-container/ https://chalkbrain.com/lessons/devops/running-a-container/ Before you start. Lesson 0016 sealed your app into an image - a frozen box on disk that does nothing alone - and introduced the daemon, the engine that pulls… https://chalkbrain.com/lessons/devops/docker-shipping-container/ https://chalkbrain.com/lessons/devops/docker-shipping-container/ Before you start. This is the first lesson of the DevOps track, so nothing earlier is assumed. You only need the ordinary experience of a program that runs… https://chalkbrain.com/lessons/system-design/the-copy-clerks/ https://chalkbrain.com/lessons/system-design/the-copy-clerks/ Before you start. You need the chai stall the first episode built, which grew from one counter into several, and the slip rail from 0013-the-order-slip-rail,… https://chalkbrain.com/lessons/system-design/the-order-slip-rail/ https://chalkbrain.com/lessons/system-design/the-order-slip-rail/ Before you start. You need one idea from 0011-the-chai-stall: when a single stall could not keep up, you put several stalls behind one counter and shared the… https://chalkbrain.com/lessons/system-design/the-slate-next-to-the-stove/ https://chalkbrain.com/lessons/system-design/the-slate-next-to-the-stove/ Before you start. You need the one move from 0011-the-chai-stall: every wala at the counter was made forgetful on purpose, so any wala can serve any customer… https://chalkbrain.com/lessons/system-design/the-chai-stall/ https://chalkbrain.com/lessons/system-design/the-chai-stall/ Before you start. You need one idea from 0010-one-address-a-thousand-doors: a single address can be answered by machines in many cities, because the network… https://chalkbrain.com/lessons/networking/one-address-a-thousand-doors/ https://chalkbrain.com/lessons/networking/one-address-a-thousand-doors/ Before you start. 0004-how-packets-find-their-way gave you the routing table: a list of rows, each one a range of addresses and a next hop, and a router that… https://chalkbrain.com/lessons/networking/write-the-telegram/ https://chalkbrain.com/lessons/networking/write-the-telegram/ Before you start. 0001-dns-resolution showed you a DNS lookup happening. 0002-dns-wire-format took its 29 bytes apart on the wire. Here you write those same 29… https://chalkbrain.com/lessons/networking/outgrowing-the-phone-call/ https://chalkbrain.com/lessons/networking/outgrowing-the-phone-call/ Before you start. You need that TCP promises to deliver bytes in the exact order they were sent - the promise this whole lesson is about, set up in… https://chalkbrain.com/lessons/networking/the-whisper-code/ https://chalkbrain.com/lessons/networking/the-whisper-code/ Before you start. You already know that no data moves until one full round trip happens first - the TCP handshake, from 0003-tcp-udp-handshake - and that a… https://chalkbrain.com/lessons/networking/the-long-distance-reading/ https://chalkbrain.com/lessons/networking/the-long-distance-reading/ Before you start. You need the handshake from 0003-tcp-udp-handshake: before any of your data moves, both ends already spend one round trip agreeing the call… https://chalkbrain.com/lessons/networking/inside-your-building/ https://chalkbrain.com/lessons/networking/inside-your-building/ Before you start. You already met your router as the concierge in 0001-dns-resolution, looking up a name and handing back an address, and as the first hop in… https://chalkbrain.com/lessons/networking/how-packets-find-their-way/ https://chalkbrain.com/lessons/networking/how-packets-find-their-way/ Before you start. Nothing here leans on another lesson. One picture runs all the way through: a router is a depot that only knows its own street, a packet is a… https://chalkbrain.com/lessons/networking/tcp-udp-handshake/ https://chalkbrain.com/lessons/networking/tcp-udp-handshake/ Before you start. You need to know that every device on a network has an address, and that a message crossing the internet travels broken into chunks rather… https://chalkbrain.com/lessons/networking/dns-wire-format/ https://chalkbrain.com/lessons/networking/dns-wire-format/ Before you start. 0001-dns-resolution showed your resolver doing the asking for you, one hop at a time. Here you watch what that one question, and its answer,… https://chalkbrain.com/lessons/networking/dns-resolution/ https://chalkbrain.com/lessons/networking/dns-resolution/ Before you start. This is the first lesson in the series, so nothing earlier is assumed. You've typed a web address into a browser and watched a page appear;… https://chalkbrain.com/videos/ai-astra-means/ https://chalkbrain.com/videos/ai-astra-means/ Episode three ended on a checker. Ten thousand agents wrote a proof, and what made anyone trust it was seventeen hours in which a program with no opinion checked every step. Hold that thought. This ep https://chalkbrain.com/videos/ai-astra-math/ https://chalkbrain.com/videos/ai-astra-math/ Last time we ended on a problem with every number in the launch: each benchmark was a target while the model was trained. So here is the one kind of test nobody can train for. An open problem. A quest https://chalkbrain.com/videos/ai-astra-claims/ https://chalkbrain.com/videos/ai-astra-claims/ Last time we set out five definitions of AGI, and found that every one of them is a test somebody chose. A press release cannot pass a test. Evidence can. So this episode is the evidence, laid in a le https://chalkbrain.com/videos/ai-astra-agi/ https://chalkbrain.com/videos/ai-astra-agi/ September third, twenty twenty six. OpenAI ships GPT-6 Astra, and its president opens with four words: welcome to the AGI era. Nvidia's chief executive posts that AGI has arrived. OpenAI's own chief e https://chalkbrain.com/videos/am-everything/ https://chalkbrain.com/videos/am-everything/ Nine episodes, one canvas. Amazon is two companies: a shop whose whole instinct is availability at any cost, and a cloud built out of the shop's own leftovers. Everything on this board falls out of on https://chalkbrain.com/videos/am-amplifiers/ https://chalkbrain.com/videos/am-amplifiers/ February two thousand and seventeen. An operator, debugging a billing system, ran a command from an established playbook to remove a few servers. One input was entered incorrectly, and a much larger s https://chalkbrain.com/videos/am-chaotic-storage/ https://chalkbrain.com/videos/am-chaotic-storage/ Two ways to store a million different things. The obvious way is what every shop and every library does: a shelf for books, a shelf for toys, a shelf for cables. You always know where a book goes. Wit https://chalkbrain.com/videos/am-eleven-nines/ https://chalkbrain.com/videos/am-eleven-nines/ Eleven nines. That is a durability target, per object, per year: store ten million objects and you should expect to lose one, every ten thousand years. But the number is not a promise anybody makes. I https://chalkbrain.com/videos/am-constant-work/ https://chalkbrain.com/videos/am-constant-work/ A drip machine and an urn. The drip machine makes a cup when you ask for it, so a queue of thirsty people makes it slower and slower. The urn brewed everything up front. Serving the first cup and the https://chalkbrain.com/videos/am-blast-radius/ https://chalkbrain.com/videos/am-blast-radius/ Somewhere out there is a request that kills whatever handles it. Corrupted, or it trips a specific bug — AWS calls it a poison pill. Eight workers, all shared between all customers. The pill lands on https://chalkbrain.com/videos/am-cart/ https://chalkbrain.com/videos/am-cart/ The shopping cart had one requirement. Customers must be able to add an item even if disks are failing, network routes are flapping, or data centres are being destroyed by tornadoes. That is not senti https://chalkbrain.com/videos/am-mandate/ https://chalkbrain.com/videos/am-mandate/ Before the memo there was one binary. The original storefront was a single program with all the logic in it and every team reaching straight into the same database. By two thousand and one, one change https://chalkbrain.com/videos/am-map/ https://chalkbrain.com/videos/am-map/ Amazon is two companies wearing one name. A shop that must never refuse an order, and a computing business built out of the shop's own leftovers. Three million checkouts on a single day, and five hund https://chalkbrain.com/videos/gs-everything/ https://chalkbrain.com/videos/gs-everything/ The whole thing, on one canvas. Along the top, a query: understand it, scatter it to the shards, merge, rank, assemble the page. Along the bottom, a pipeline that never stops: crawl, index, score, dis https://chalkbrain.com/videos/gs-beyond-links/ https://chalkbrain.com/videos/gs-beyond-links/ Everything so far produced a ranked list of pages. Go and look at an actual results page though, and the ranked list is a strip down the middle of something much busier. There is an answer lifted out https://chalkbrain.com/videos/gs-infra/ https://chalkbrain.com/videos/gs-infra/ Everything in this series assumes somewhere to put the web and something to process it with. In two thousand and three there was nothing to buy that could do either. So the bet was made the other way https://chalkbrain.com/videos/gs-freshness/ https://chalkbrain.com/videos/gs-freshness/ Something on the web changed while I said that sentence. Now here is the problem: for years, the only way a change reached search was the next full rebuild of the index, and a full rebuild of the whol https://chalkbrain.com/videos/gs-ranking/ https://chalkbrain.com/videos/gs-ranking/ The index has just handed back everything that matches. For a normal query that is millions of pages, and every one of them is a legitimate answer. Almost nobody will look past the first three. So the https://chalkbrain.com/videos/gs-query-serving/ https://chalkbrain.com/videos/gs-query-serving/ The index is sharded by document, which means no single machine can answer anything. Every query has to visit all of them. And it has about two hundred milliseconds, total, including the ranking that https://chalkbrain.com/videos/gs-inverted-index/ https://chalkbrain.com/videos/gs-inverted-index/ Three documents, and a search for the word pizza. The obvious way to answer is to open each document and look. That works here. Now make it the whole web, and every query reads every page ever written https://chalkbrain.com/videos/gs-crawler/ https://chalkbrain.com/videos/gs-crawler/ A crawler is a loop around a queue that never empties. Every page you fetch hands you its outlinks, and those go straight back into the queue. So the web crawls itself, forever, and the frontier is ef https://chalkbrain.com/videos/gs-map/ https://chalkbrain.com/videos/gs-map/ Two systems wear one trench coat here. One is gigantic and offline: it crawls the whole web, indexes it, scores every page, and it never stops. The other is online, and it has about two hundred millis https://chalkbrain.com/videos/yt-everything/ https://chalkbrain.com/videos/yt-everything/ Everything, on one canvas. Five hundred hours arrive every minute. Each upload is cut into segments, transcoded in parallel into a ladder of resolutions and codecs, and packaged. Copies of the popular https://chalkbrain.com/videos/yt-the-cheat/ https://chalkbrain.com/videos/yt-the-cheat/ This is YouTube's own description of how they engineer, from a talk in twenty twelve. Choose the simplest solution possible, with the loosest guarantees that are practical. And the corollary they put https://chalkbrain.com/videos/yt-content-id/ https://chalkbrain.com/videos/yt-content-id/ Two thousand seven. YouTube is a year into being owned by Google and is being sued for a billion dollars. Somebody proposes reviewing uploads for copyright. So do the arithmetic. Five hundred hours ar https://chalkbrain.com/videos/yt-recommendations/ https://chalkbrain.com/videos/yt-recommendations/ For years the thing being optimised was the click. Did they arrive. And a metric like that has a champion content type: a shocking thumbnail, a three second payoff, and instant regret. It wins on the https://chalkbrain.com/videos/yt-billion-hours/ https://chalkbrain.com/videos/yt-billion-hours/ A billion hours of video are watched every day. The obvious mental picture is enormous datacentres pushing that out across the internet, and it is wrong. Most of those bytes never reach a YouTube buil https://chalkbrain.com/videos/yt-transcode-factory/ https://chalkbrain.com/videos/yt-transcode-factory/ You upload one file. What has to exist afterwards is dozens of them. Every resolution from a hundred and forty four pixels up to eight K, times three codecs, times a set of bitrates, plus the manifest https://chalkbrain.com/videos/yt-vitess/ https://chalkbrain.com/videos/yt-vitess/ Sharding by user bought them four years. Then three things happened at once. Write traffic outgrew a single primary per shard. Adding read replicas stopped helping, because reads had outgrown them too https://chalkbrain.com/videos/yt-nine-engineers/ https://chalkbrain.com/videos/yt-nine-engineers/ March two thousand six: thirty million views a day. July: a hundred million. More than three times, in four months. And here is everybody who worked on it. Two sysadmins, two scalability architects, t https://chalkbrain.com/videos/yt-map/ https://chalkbrain.com/videos/yt-map/ Five hundred hours of video are uploaded every minute. A billion hours are watched every day. Everything else on this map falls out of the gap between them. A video is written once and watched up to a https://chalkbrain.com/videos/fg-everything/ https://chalkbrain.com/videos/fg-everything/ One canvas tag, a GPU context, a websocket and a byte array. That is everything the browser gives them, and inside it runs a C plus plus engine with its own scene graph, compositor, text layout and re https://chalkbrain.com/videos/fg-plugin-sandbox/ https://chalkbrain.com/videos/fg-plugin-sandbox/ Running a plugin comes down to one line: eval, on somebody else's code. Their words — that is the quintessential definition of insecure. And this plugin is inside your document. It can read any data i https://chalkbrain.com/videos/fg-undo-multiplayer/ https://chalkbrain.com/videos/fg-undo-multiplayer/ Layer order is an array, and arrays have indices zero, one, two. There is no integer between one and two, so drop a layer between two others and you renumber everything after it. Figma's answer: do no https://chalkbrain.com/videos/fg-shard-one-table/ https://chalkbrain.com/videos/fg-shard-one-table/ Twenty twenty. All of Figma's metadata — permissions, file info, comments — on one Postgres instance, the biggest box their cloud sold. Peak CPU over sixty five percent. Their own sentence: if the dat https://chalkbrain.com/videos/fg-drawing-a-letter/ https://chalkbrain.com/videos/fg-drawing-a-letter/ A GPU draws triangles. A letter O is not triangles, and worse, it has a hole in the middle. One warning before we start: Figma's entire public record on their renderer is a single sentence from twenty https://chalkbrain.com/videos/fg-one-process-per-doc/ https://chalkbrain.com/videos/fg-one-process-per-doc/ Two thousand sixteen. One TypeScript worker, single threaded, owning a whole pile of documents at once. Then somebody opens a five hundred megabyte file. Encoding it takes an arbitrarily long time, an https://chalkbrain.com/videos/fg-neither-ot-nor-crdt/ https://chalkbrain.com/videos/fg-neither-ot-nor-crdt/ There are two famous answers to how do many people edit one thing at once. Operational transform, which is what a document editor uses. And conflict free replicated data types, which is what everyone https://chalkbrain.com/videos/fg-browser-inside/ https://chalkbrain.com/videos/fg-browser-inside/ In twenty fifteen, a design tool in a browser was a joke. Figma's answer was not to use the web platform better. It was to refuse almost all of it. HTML and SVG re-tessellate every time you zoom. Canv https://chalkbrain.com/videos/fg-map/ https://chalkbrain.com/videos/fg-map/ Open a Figma file and almost nothing on that page is the web. The browser hands over four things: one canvas tag, a GPU context, a websocket, and a byte array. Everything you see inside that rectangle https://chalkbrain.com/videos/dc-everything/ https://chalkbrain.com/videos/dc-everything/ Everything, on one canvas. Your app holds a socket to the gateway. A message goes by REST to the Python monolith, which mints its id, through the Rust shield that collapses the crowd, and into the sto https://chalkbrain.com/videos/dc-api/ https://chalkbrain.com/videos/dc-api/ The unfashionable middle. One Python codebase, seventeen hundred endpoints and seven hundred kinds of background job, and it never got rewritten. But it is not deployed once. The same image goes out a https://chalkbrain.com/videos/dc-search/ https://chalkbrain.com/videos/dc-search/ The message store answers one question beautifully: the latest fifty in this channel. Ask it which messages contain the word pizza and it has nothing. Text search needs the index inside out — a word, https://chalkbrain.com/videos/dc-fanout/ https://chalkbrain.com/videos/dc-fanout/ Here is the trap. A message costs one send times the number of people watching. But an event everybody emits — typing, going online, going away — costs that number times itself. On a server with ten m https://chalkbrain.com/videos/dc-voice/ https://chalkbrain.com/videos/dc-voice/ Ten people in a voice call, and there are three ways to wire it. Everyone sends to everyone: fine at five, melts a laptop at twenty five, and hands your address to every stranger in the room. Or a ser https://chalkbrain.com/videos/dc-data-services/ https://chalkbrain.com/videos/dc-data-services/ A viral moment does not send a million different questions. It sends the same question a million times. Every one of those clients wants the last fifty messages of the same channel, and if each of the https://chalkbrain.com/videos/dc-message-store/ https://chalkbrain.com/videos/dc-message-store/ Start with the workload, because it picks the database. Writes are tiny, constant, and never edited afterwards. Reads want the last fifty messages, almost always. And every so often a single channel b https://chalkbrain.com/videos/dc-gateway/ https://chalkbrain.com/videos/dc-gateway/ Open Discord and your app makes exactly one connection. A websocket to the gateway. Every single thing you watch happen arrives down that one pipe: messages, typing dots, someone going idle, the membe https://chalkbrain.com/videos/ub-everything/ https://chalkbrain.com/videos/ub-everything/ One tap. The gateway routes it, dispatch picks a driver out of a cell, the trip service runs the ride as a state machine, payments charges once it ends, and the whole time, underneath, five million ph https://chalkbrain.com/videos/ub-cells/ https://chalkbrain.com/videos/ub-cells/ Everything so far has described one copy of this system. Now put every city on earth through it. One shared stack means one bad deploy, one poisoned cache, one overloaded database takes down riders an https://chalkbrain.com/videos/ub-eta/ https://chalkbrain.com/videos/ub-eta/ Under every spatial question in this system there is one graph. Intersections are nodes, road segments are edges, and here is the part people skip: the weight on an edge is not how long the road is. I https://chalkbrain.com/videos/ub-payments/ https://chalkbrain.com/videos/ub-payments/ Everywhere else in this system, being slightly wrong is fine. A stale location, an optimistic arrival time, a price from four seconds ago. Here, being slightly wrong means a stranger was charged twice https://chalkbrain.com/videos/ub-trip-state/ https://chalkbrain.com/videos/ub-trip-state/ Everything else in this system answers in milliseconds. A trip does not. It is requested, matched, driven and paid over the better part of an hour, and two different people are sending it events the w https://chalkbrain.com/videos/ub-surge/ https://chalkbrain.com/videos/ub-surge/ A fare is mostly fixed. A base amount, a rate per minute, a rate per kilometre, some fees. None of those can move while you are standing on the street. Exactly one term in that formula is allowed to c https://chalkbrain.com/videos/ub-dispatch/ https://chalkbrain.com/videos/ub-dispatch/ The index just handed dispatch forty drivers near this rider. Now the hard part. Exactly one of them gets this trip. That choice has to be good, it has to happen in seconds, and it must never, ever be https://chalkbrain.com/videos/ub-geo-index/ https://chalkbrain.com/videos/ub-geo-index/ This one subsystem does two incompatible things at once. It absorbs more than a million location writes every second, and it answers, in milliseconds, which drivers are within two kilometres of a poin https://chalkbrain.com/videos/ub-map/ https://chalkbrain.com/videos/ub-map/ Five million drivers are pushing their location every four seconds. Somewhere a rider taps a button, and this system has under a second to answer two questions: who is near me, and what does this cost https://chalkbrain.com/videos/iv-web-foundations/ https://chalkbrain.com/videos/iv-web-foundations/ The first twenty minutes are the easy questions. What is a div. What is the box model. What does four oh three mean. Easy to answer badly. A one word answer is correct, and it reads as someone who has https://chalkbrain.com/videos/iv-react-build/ https://chalkbrain.com/videos/iv-react-build/ A React coding round is almost always three things. Build a form. Fetch something and show it. And render a tree. Everything else is a follow up question about those three. This video writes all three https://chalkbrain.com/videos/iv-react-hooks/ https://chalkbrain.com/videos/iv-react-hooks/ React hooks is where a front end interview catches people. Not because the ideas are hard. Because the questions are about behaviour, not syntax. Why did this counter stop at one. Why does this effect https://chalkbrain.com/videos/iv-react-core/ https://chalkbrain.com/videos/iv-react-core/ Every React interview opens the same way. What is React, and why does it exist. The answer they want is not a library for building components. It is a different way of changing a screen from the one t https://chalkbrain.com/videos/iv-js-browser/ https://chalkbrain.com/videos/iv-js-browser/ The browser half of the interview is where most people lose points. Not because it is hard, but because the answers are one sentence deep. This is that half, in order. How a page gets drawn. How an ev https://chalkbrain.com/videos/iv-js-async/ https://chalkbrain.com/videos/iv-js-async/ Two days before a JavaScript interview, this is the area you get asked about. Not closures. The event loop, promises, async await. It is the part you cannot fake. They ask what this prints, and you ei https://chalkbrain.com/videos/iv-output-puzzles/ https://chalkbrain.com/videos/iv-output-puzzles/ Forty questions. Each one is a few lines of JavaScript and one question. What does this log. The code appears, you get a moment to answer, then the real output lands. Green means you predicted it. Red https://chalkbrain.com/videos/iv-js-types/ https://chalkbrain.com/videos/iv-js-types/ Two days from an interview, and this question opens most of them. How many types does JavaScript have. Eight. Seven primitives and one object type. Say the number first, then the list. What is graded https://chalkbrain.com/videos/iv-js-engine/ https://chalkbrain.com/videos/iv-js-engine/ Here is the first thing an interviewer will hand you. A function that logs two variables before either is declared. One log prints undefined. The other throws. Same function, two lines apart. That gap https://chalkbrain.com/videos/mb-spooky-action/ https://chalkbrain.com/videos/mb-spooky-action/ One event makes two particles. Send them as far apart as you like — different buildings, different continents, in principle different star systems. Measure one, and you get a completely random answer. https://chalkbrain.com/videos/mb-moon-illusion/ https://chalkbrain.com/videos/mb-moon-illusion/ You have seen it. A full moon sitting on the horizon, enormous, far too big to be the same moon that is a small pale disc overhead. People describe it as twice the size. Photograph both, on the same n https://chalkbrain.com/videos/mb-emergence/ https://chalkbrain.com/videos/mb-emergence/ Watch this flock turn. Every bird changes direction at almost the same instant, and the shape holds. So which one decided? Which bird is at the front, calling it? There is no such bird. Nothing in thi https://chalkbrain.com/videos/mb-sky-blue/ https://chalkbrain.com/videos/mb-sky-blue/ The sky has no colour of its own. It is not the ocean, and it is not the air being blue. It is sunlight taking detours. White light arrives carrying every wavelength at once, and it meets air molecule https://chalkbrain.com/videos/mb-braess/ https://chalkbrain.com/videos/mb-braess/ A city builds one new road. No extra cars. No roadworks. And every single driver's journey gets longer. Sixty five minutes before it opened. Eighty minutes after. Nobody made a mistake, nobody drove b https://chalkbrain.com/videos/mb-mirror/ https://chalkbrain.com/videos/mb-mirror/ Raise your right hand in front of a mirror. Watch which hand your reflection raises. Here you are from above, and here is your double, and the marked hand is on the same side of the room in both. Noth https://chalkbrain.com/videos/mb-night-sky/ https://chalkbrain.com/videos/mb-night-sky/ Look up tonight and you are not seeing the sky. You are seeing a stack of old photographs, taken at different moments, printed on top of each other. Light is quick — three hundred thousand kilometres https://chalkbrain.com/videos/mb-gps-relativity/ https://chalkbrain.com/videos/mb-gps-relativity/ Your phone does not know where it is. It works it out by listening. Every satellite overhead is broadcasting one thing, over and over: the time on my clock, right now. Your phone compares when those a https://chalkbrain.com/videos/mb-seven-folds/ https://chalkbrain.com/videos/mb-seven-folds/ You have heard that paper cannot be folded more than seven times. Try it. Around the sixth or seventh fold the sheet simply stops. Everyone says the paper got too strong. That is not what happened. Th https://chalkbrain.com/videos/mb-dunning-kruger/ https://chalkbrain.com/videos/mb-dunning-kruger/ You have seen this curve. A spike of confidence called Mount Stupid, a crash into the valley of despair, a long climb back up. It is on a thousand slides. It is not in the paper. Nobody ever measured https://chalkbrain.com/videos/mb-compound-interest/ https://chalkbrain.com/videos/mb-compound-interest/ Two people, one illustrative rate, round numbers all the way through. The first adds a thousand a year for ten years, then stops forever. The second starts ten years later and never stops — thirty pay https://chalkbrain.com/videos/mb-fermi/ https://chalkbrain.com/videos/mb-fermi/ The galaxy has a hundred billion stars, and it has been here for thirteen and a half billion years. Nearly every one of those stars has planets. And in all that time, in all that room, nobody has call https://chalkbrain.com/videos/mb-empty-space/ https://chalkbrain.com/videos/mb-empty-space/ Hold up your hand. It looks solid. It feels solid. It is almost entirely nothing. Take one atom of hydrogen: a nucleus at the middle, one electron out at the edge. The gap between them is about sixty https://chalkbrain.com/videos/mb-coastline/ https://chalkbrain.com/videos/mb-coastline/ How long is the coast of Britain? That sounds like a fact — a number you could look up in an atlas. It is not. There is no such number. And the reason is the strangest thing I know about measuring any https://chalkbrain.com/videos/mb-monty-hall/ https://chalkbrain.com/videos/mb-monty-hall/ Three doors. One car, two goats. You pick door one. Now the host, who knows exactly what is behind all three, opens door three and shows you a goat. He offers you the switch. Take it. Switching wins t https://chalkbrain.com/videos/mb-benford/ https://chalkbrain.com/videos/mb-benford/ Take a pile of real numbers — river lengths, city populations, the totals on a year of invoices — and look only at the first digit. You would expect each of the nine to lead about eleven percent of th https://chalkbrain.com/videos/mb-simpsons-paradox/ https://chalkbrain.com/videos/mb-simpsons-paradox/ Treatment A beat treatment B on the small stones. Treatment A beat treatment B on the large stones. There are only two groups, and every patient is in one of them. Add the two groups together and trea https://chalkbrain.com/videos/mb-time-speeds-up/ https://chalkbrain.com/videos/mb-time-speeds-up/ A summer when you were eight lasted forever. A year at thirty five is gone before you finish noticing it started. Same three hundred and sixty five days. Same planet, same orbit. So what actually chan https://chalkbrain.com/videos/mb-bigger-infinities/ https://chalkbrain.com/videos/mb-bigger-infinities/ Count the whole numbers: one, two, three, forever. Now count only the even ones. It feels like you just threw half of them away. But pair them off. One with two, two with four, three with six, and not https://chalkbrain.com/videos/sd-ledger-everyone-reads/ https://chalkbrain.com/videos/sd-ledger-everyone-reads/ Last time you made a consumer safe to receive the same message twice. The chai stall's order slip rail was still one producer and one kitchen, and a slip torn off that rail is gone the moment somebody https://chalkbrain.com/videos/sd-two-ledgers/ https://chalkbrain.com/videos/sd-two-ledgers/ Last time a majority decided who holds the pen. One leader, one book, every write landing in a single place. Now the money has to leave that book and land in another one, on a different machine. The c https://chalkbrain.com/videos/dc-map/ https://chalkbrain.com/videos/dc-map/ One Discord server can hold ten million members, and a football match can spike its traffic nine times in a single night. The same company also keeps every message anyone has ever sent, forever. Those https://chalkbrain.com/videos/sd-who-holds-pen/ https://chalkbrain.com/videos/sd-who-holds-pen/ The bouncer at the door could decide alone — one machine, one counter, one instant no. But the chai stall's ledger has only one pen, and four episodes back we watched the clerk holding it collapse. Wh https://chalkbrain.com/videos/sd-right-book/ https://chalkbrain.com/videos/sd-right-book/ Back at the stall there was one book, and every number in the shop went into it. A real system asks fourteen different questions before lunch, and no single book answers all of them well. So the usefu https://chalkbrain.com/videos/sd-slate-properly/ https://chalkbrain.com/videos/sd-slate-properly/ Back at the stall we chalked a slate and called it caching. That version was a toy. Hit rate, a lifetime on each line, and erase whatever nobody asks for. It skipped three questions, and those are the https://chalkbrain.com/videos/se-auth/ https://chalkbrain.com/videos/se-auth/ Every login you have ever done answers two completely different questions. Who are you? And what are you allowed to do? Authentication, and authorization. Almost every real data leak is a system that https://chalkbrain.com/videos/sd-contract-counter/ https://chalkbrain.com/videos/sd-contract-counter/ At the stall a deal was a sentence across the counter and nobody wrote it down. A published interface is the opposite of that. It is a promise in writing to callers you will never meet, including code https://chalkbrain.com/videos/dsa-segment-tree/ https://chalkbrain.com/videos/dsa-segment-tree/ Here is a row of numbers. Eight of them. Somebody asks for the total between position two and position five. Then somebody changes position four. Then they ask again. Two operations, forever, interlea https://chalkbrain.com/videos/dsa-graphs/ https://chalkbrain.com/videos/dsa-graphs/ Pin some dots on a wall and draw string between them. That is a graph. The dots are vertices. The strings are edges. That is the entire definition, and it is why graphs turn up in problems that look n https://chalkbrain.com/videos/dsa-dp/ https://chalkbrain.com/videos/dsa-dp/ Here is a page of scratch work with every answer you have already computed written down the side. That is dynamic programming. It is not a clever trick and it is not a data structure. It is recursion, https://chalkbrain.com/videos/dsa-greedy/ https://chalkbrain.com/videos/dsa-greedy/ A greedy algorithm makes one promise. At every step, take the option that looks best right now, and never reconsider it. Picture a walker who only ever steps uphill. They always reach a hilltop. They https://chalkbrain.com/videos/dsa-trie/ https://chalkbrain.com/videos/dsa-trie/ Down the edge of a paper dictionary there are thumb tabs, and every word behind a tab starts with the same letter. Now cut that idea finer. One tab per letter, then a tab inside that tab, all the way https://chalkbrain.com/videos/dsa-heap/ https://chalkbrain.com/videos/dsa-heap/ A triage desk at a hospital. People arrive in no particular order, and every time a room opens you take the most urgent one waiting. Nobody sorts the whole room. You only ever need one fact: which is https://chalkbrain.com/videos/dsa-trees/ https://chalkbrain.com/videos/dsa-trees/ Look at a tournament bracket. The final sits at the top, and underneath it are two smaller brackets that look exactly like the whole thing. That is the entire definition of a binary tree. A tree is ei https://chalkbrain.com/videos/dsa-stack-queue-deque/ https://chalkbrain.com/videos/dsa-stack-queue-deque/ Picture a stack of plates by the sink. You add to the top, you take from the top, and the plate you washed last is the one you dry first. That is a stack. Last in, first out, and nothing else is allow https://chalkbrain.com/videos/dsa-linked-list/ https://chalkbrain.com/videos/dsa-linked-list/ Picture a train. Each carriage carries cargo, and each carriage is coupled to exactly one carriage behind it. That is a linked list. A node holds a value and the address of the next node, and one poin https://chalkbrain.com/videos/dsa-two-pointers/ https://chalkbrain.com/videos/dsa-two-pointers/ Here is a sorted row of numbers, and a question. Which two of them add to sixteen? The obvious answer is two nested loops, checking every pair. That is order n squared. There is a way to answer it in https://chalkbrain.com/videos/dsa-recursion-backtracking/ https://chalkbrain.com/videos/dsa-recursion-backtracking/ Stand in a queue and ask how many people are ahead of you. You do not count them. You ask the person in front, and add one. The person at the very front answers zero. That is recursion, and it is alre https://chalkbrain.com/videos/dsa-number-theory/ https://chalkbrain.com/videos/dsa-number-theory/ Take a strip of length twelve and lay tiles of equal length along it. Tiles of length three fit exactly, four of them. Tiles of length five do not, they leave two cells over. That is all divisibility https://chalkbrain.com/videos/dsa-bit-manipulation/ https://chalkbrain.com/videos/dsa-bit-manipulation/ Picture a wall of thirty two light switches. That is an integer: thirty two independent on or off flags, and every arithmetic operator you know is just a rule for flipping them. Bit manipulation is wh https://chalkbrain.com/videos/dsa-prefix-window/ https://chalkbrain.com/videos/dsa-prefix-window/ An index at the back of a book costs an afternoon to build and nothing at all to use. That is the whole of this video. Every technique here pays a fixed price once, up front, so that a question you wi https://chalkbrain.com/videos/dsa-sets-maps/ https://chalkbrain.com/videos/dsa-sets-maps/ A row of numbered boxes. Box zero, box one, box two. Give an array a position and it hands you the value back instantly, because the address is arithmetic. Now the question this whole video answers. W https://chalkbrain.com/videos/dsa-complexity/ https://chalkbrain.com/videos/dsa-complexity/ Here is a stopwatch and two laptops. The same sorting program takes four seconds on one and one second on the other. So which number goes in the textbook? Neither. Seconds measure the machine. We want https://chalkbrain.com/videos/dsa-pointers/ https://chalkbrain.com/videos/dsa-pointers/ Memory is one long row of numbered boxes. Every box has a number, and that number is its address. A variable is only a name your compiler hands to one of those boxes. By the time the program is runnin https://chalkbrain.com/videos/dsa-sorting-searching/ https://chalkbrain.com/videos/dsa-sorting-searching/ Here is a drawer with ten thousand receipts in it, in no order at all. Ask it three questions. Does this one exist? Are there duplicates? Group these by month. Every answer means touching every receip https://chalkbrain.com/videos/dsa-strings/ https://chalkbrain.com/videos/dsa-strings/ A string is an array. That is the whole starting point. In memory, hello is five numbers sitting next to each other, and the letter at index two is one addition away. Everything easy about strings com https://chalkbrain.com/videos/dsa-arrays/ https://chalkbrain.com/videos/dsa-arrays/ Picture a street of identical houses, numbered from zero, with no gaps between them. That is an array. One block of memory, every element exactly the same size, laid end to end. Almost everything else https://chalkbrain.com/videos/dsa-loops-patterns/ https://chalkbrain.com/videos/dsa-loops-patterns/ A loop is the first thing you write and the last thing you stop getting wrong. Strip it down and it is exactly three decisions. Where it starts. When it stops. And what changes on every single step. G https://chalkbrain.com/videos/dsa-cpp-intro/ https://chalkbrain.com/videos/dsa-cpp-intro/ Every competitive problem comes with a time limit printed at the top of the page. Often two seconds. Your solution is correct or it is not, and then it is fast enough or it is not. C plus plus is chos https://chalkbrain.com/videos/do-sandboxing/ https://chalkbrain.com/videos/do-sandboxing/ Someone hands you code and asks you to run it. A pull request from a stranger. A cell in a shared notebook. A function a customer uploaded. Something a model just generated. You have to execute it on https://chalkbrain.com/videos/do-cdn-object-storage/ https://chalkbrain.com/videos/do-cdn-object-storage/ Start with what it is not. Object storage is not a disc, and it is not a filesystem. There are no directories, no file handles, no seeking to byte four thousand. There is a bucket, a key, and a whole https://chalkbrain.com/videos/do-iac/ https://chalkbrain.com/videos/do-iac/ Somebody opened the cloud console at two in the morning, ticked one box, and the outage stopped. Eight months later nobody can say which box, or why, and staging does not have it ticked. That is click https://chalkbrain.com/videos/do-observability/ https://chalkbrain.com/videos/do-observability/ Every dashboard on the wall is green. The graphs are flat, the checks pass, and the support queue is filling with people who cannot check out. Nothing is broken, according to anything you thought to m https://chalkbrain.com/videos/do-cicd/ https://chalkbrain.com/videos/do-cicd/ There is a lamp on the wall in some engineering offices. Green means the main branch builds and passes its tests. Red means it does not, and nobody starts anything new until it is green again. That la https://chalkbrain.com/videos/do-kubernetes-2/ https://chalkbrain.com/videos/do-kubernetes-2/ A node has a fixed amount of room, and the scheduler packs pods onto it by reading a label on the box, never by weighing the box. That label is two numbers per container. Requests, and limits. Almost https://chalkbrain.com/videos/do-kubernetes-1/ https://chalkbrain.com/videos/do-kubernetes-1/ A thermostat has one job. You set a number, and it keeps working until the room matches it. You never tell it to switch the heater on. Kubernetes is that idea applied to running software, and everythi https://chalkbrain.com/videos/do-docker/ https://chalkbrain.com/videos/do-docker/ Think of a ticket window. You slide a request through the slot, and the workshop behind it does the work. That is Docker. The command you type is a client. It posts to a daemon over a local socket, an https://chalkbrain.com/videos/do-containers/ https://chalkbrain.com/videos/do-containers/ A virtual machine boots its own kernel. A container does not. It is one Linux process, running on the host's kernel, that has been lied to about what it can see and how much it can use. That is the wh https://chalkbrain.com/videos/do-autoscaling/ https://chalkbrain.com/videos/do-autoscaling/ Traffic is not flat. It has a morning ramp, an evening peak, and a quiet middle of the night. Fixed capacity gives you two bad choices. Buy for the peak and pay for idle machines all day, or buy for t https://chalkbrain.com/videos/do-certificates/ https://chalkbrain.com/videos/do-certificates/ A padlock in the address bar means the traffic is encrypted. It does not mean you are talking to who you think you are. Encryption without identity is a private conversation with a stranger. Nobody ca https://chalkbrain.com/videos/do-processes-proxies/ https://chalkbrain.com/videos/do-processes-proxies/ Two things stand between your code and a working service. Something that keeps the process running when it dies. And something that stands in front of it, so the outside world only ever sees one door. https://chalkbrain.com/videos/do-vms-baremetal/ https://chalkbrain.com/videos/do-vms-baremetal/ Take the lid off a server and there is no cloud in there. There are sockets holding processors, sticks of memory, a network card, and drives. Everything you will ever run is that hardware, shared out https://chalkbrain.com/videos/do-bash-terminal/ https://chalkbrain.com/videos/do-bash-terminal/ That blinking prompt is a program. Not the operating system, and not the window it is drawn in. It is a program called a shell, whose whole job is to read the line you type, work out what it means, an https://chalkbrain.com/videos/wd-queues-pubsub/ https://chalkbrain.com/videos/wd-queues-pubsub/ You click sign up. The server creates your account, then sends a welcome email, then generates an avatar thumbnail, then pushes a record into the sales system. Only after all four does the page come b https://chalkbrain.com/videos/wd-websockets-webrtc/ https://chalkbrain.com/videos/wd-websockets-webrtc/ The web was built on one shape. You ask, the server answers, and it is over. Nothing the server learns afterwards can reach you. A price changed, a message arrived, a job finished, and your page has n https://chalkbrain.com/videos/wd-nextjs/ https://chalkbrain.com/videos/wd-nextjs/ A web page has to become H T M L somewhere, and there are only three places that can happen. On a build machine, before anyone asks. On a server, the moment someone asks. Or in the browser, after an e https://chalkbrain.com/videos/wd-tailwind/ https://chalkbrain.com/videos/wd-tailwind/ Before you write a single style, C S S asks you for a name. What do you call this? Card header title. Nav item active. You are inventing a private vocabulary that nobody else knows, for a thing that a https://chalkbrain.com/videos/wd-react/ https://chalkbrain.com/videos/wd-react/ There are two ways to change what is on a screen. Find the part that is wrong and edit it in place. Or describe the whole screen for the new moment and let something else work out the difference. Reac https://chalkbrain.com/videos/wd-bun/ https://chalkbrain.com/videos/wd-bun/ Start a Node project and count the programs you now depend on. Node itself. N P M. The TypeScript compiler. A test runner. A bundler. Five separate tools that all have to agree with each other. Bun's https://chalkbrain.com/videos/wd-turborepo/ https://chalkbrain.com/videos/wd-turborepo/ Two repositories. One shared type definition, copied into both. Someone edits it in one place, and the other keeps compiling happily against the old shape until production disagrees. A monorepo is the https://chalkbrain.com/videos/wd-typescript/ https://chalkbrain.com/videos/wd-typescript/ A proofreader marks up a manuscript in red. The printed book carries none of those marks. TypeScript is that proofreader. It reads your JavaScript, argues with you about it, and then deletes every sin https://chalkbrain.com/videos/wd-postgres-orm/ https://chalkbrain.com/videos/wd-postgres-orm/ Every new project asks the same question. Which database? For almost everything with users, orders, and things that belong to other things, the answer is still a relational one. Postgres. And the reas https://chalkbrain.com/videos/wd-databases-mongo/ https://chalkbrain.com/videos/wd-databases-mongo/ Every data model answers the same two questions. What do you read together in one go? And how often does that thing change? Get those two answers right and the storage engine almost picks itself. Get https://chalkbrain.com/videos/wd-http-express/ https://chalkbrain.com/videos/wd-http-express/ Open a connection to a web server, type a few lines of text, and press enter twice. That is a request. H T T P is not a mystery format. It is plain text with a shape, and the shape is the same in both https://chalkbrain.com/videos/wd-node-vs-browser/ https://chalkbrain.com/videos/wd-node-vs-browser/ An engine on a workshop bench does nothing. No wheels, no fuel line, no steering. V eight is that engine. It runs the language and nothing more. Everything you actually use, a button, a file, a socket https://chalkbrain.com/videos/wd-async-js/ https://chalkbrain.com/videos/wd-async-js/ Open any web page and everything you write runs on a single thread. One thing at a time, in order, with no help. And yet a page can wait on the network, animate, and answer a click at once. The event https://chalkbrain.com/videos/wd-js-architecture/ https://chalkbrain.com/videos/wd-js-architecture/ Here is the one image for this chapter. A light switch on a wall. You see two positions, up and down. You do not see the wiring, the breaker, or the gauge of the cable. A module is that switch, and it https://chalkbrain.com/videos/wd-js-basics/ https://chalkbrain.com/videos/wd-js-basics/ Think of a variable as a label, not a box. Copy a number into a second label and change it, and the first is untouched. Copy an object and change it, and the first changed too. Same syntax, opposite r https://chalkbrain.com/videos/wd-html-css/ https://chalkbrain.com/videos/wd-html-css/ Open any web page and what the browser actually holds is not a page. It is a tree. One root element, branches nested inside branches, and every piece of text hanging off one of them. Everything else y https://chalkbrain.com/videos/ai-multimodal/ https://chalkbrain.com/videos/ai-multimodal/ A photograph, a voice note, and a paragraph of text. Three completely different files. Three completely different shapes on disk. And one model answers questions about all three of them. That is not t https://chalkbrain.com/videos/ai-frontier/ https://chalkbrain.com/videos/ai-frontier/ Plot a language model's loss against the compute spent training it, on log axes, and you do not get a cloud. You get a straight line. That one finding organised the last few years, because a straight https://chalkbrain.com/videos/ai-evals/ https://chalkbrain.com/videos/ai-evals/ One case is failing on your screen. You reword the prompt, run it again, and it passes. That feels like progress. It is one data point, chosen by you, graded by you, on the example you happened to be https://chalkbrain.com/videos/ai-rl-finetuning/ https://chalkbrain.com/videos/ai-rl-finetuning/ Picture an apprentice copying a master's brushstrokes, line for line. That is supervised fine tuning. You show the model a prompt and one written answer, and it learns to reproduce it. Copying works, https://chalkbrain.com/videos/ai-finetune-build/ https://chalkbrain.com/videos/ai-finetune-build/ A lift engineer finishes a job and types a note on her phone. Somewhere downstream that note has to become four boxes on a parts order: part code, quantity, urgency, site. Four boxes, filled the same https://chalkbrain.com/videos/ai-finetuning/ https://chalkbrain.com/videos/ai-finetuning/ Three dials sit on the desk in front of you. Prompting. Retrieval. Fine tuning. People reach for the third one first, and almost always for a job the first two would have done. So start by naming exac https://chalkbrain.com/videos/ai-computer-use/ https://chalkbrain.com/videos/ai-computer-use/ Here is the task, and it does not change for the next ten minutes. There is a receipt sitting in the downloads folder. Open it, type its total into the expenses app, attach the file, and submit it for https://chalkbrain.com/videos/ai-mcp/ https://chalkbrain.com/videos/ai-mcp/ Open the back of any A I application and you find a drawer of adapters. One for the file system. One for the ticket tracker. One for the database. Every one hand built, every one different, and none o https://chalkbrain.com/videos/ai-memory/ https://chalkbrain.com/videos/ai-memory/ At the end of a session, a whiteboard gets wiped. Everything the agent knew is gone, and the next conversation starts from nothing. That is the default. Memory is the decision to keep something, on pu https://chalkbrain.com/videos/ai-agent-frameworks/ https://chalkbrain.com/videos/ai-agent-frameworks/ Every agent framework you have ever installed is the same loop you already wrote, with things bolted on. Call the model. Read what it asked for. Run the tool. Put the result back. Go round again. Say https://chalkbrain.com/videos/ai-agents/ https://chalkbrain.com/videos/ai-agents/ A pocket calculator answers once and then stops. A person holding one keeps going: press, read, press again, until the number is right. That second thing is an agent. Not a smarter model. A model put https://chalkbrain.com/videos/ai-context/ https://chalkbrain.com/videos/ai-context/ Picture one spotlight on a dark stage. It has a fixed amount of light. Every extra thing you drag into the beam gets a thinner share of it. A context window works the same way. It is a budget you spen https://chalkbrain.com/videos/ai-rag/ https://chalkbrain.com/videos/ai-rag/ A model ships the way a printed encyclopaedia ships. Every fact it has was set on the day it went to press, and the press has already run. The weights are frozen numbers. Nothing that happened after t https://chalkbrain.com/videos/ai-observability/ https://chalkbrain.com/videos/ai-observability/ Your dashboard is entirely green. Two hundred OK. Ninety milliseconds. Zero errors. And the answer the model just gave a customer was confidently, fluently wrong. Every signal you have says healthy, b https://chalkbrain.com/videos/ai-huggingface/ https://chalkbrain.com/videos/ai-huggingface/ Start with a filing cabinet where nothing is ever thrown away. That is the Hugging Face Hub underneath. Every model, every dataset, every demo is a git repository. Ordinary version control, holding fi https://chalkbrain.com/videos/ai-attention-industry/ https://chalkbrain.com/videos/ai-attention-industry/ Here is the thing that decides every design in this video. Training a language model and serving one are not the same problem. Training is bound by arithmetic. Serving is bound by memory. Same weights https://chalkbrain.com/videos/ai-attention/ https://chalkbrain.com/videos/ai-attention/ Read this sentence. The bank was steep and muddy. Now read this one. The bank refused the loan. Same four letters, two unrelated meanings, and nothing inside the word itself told you which one. The ot https://chalkbrain.com/videos/ai-convnets/ https://chalkbrain.com/videos/ai-convnets/ Here is a photograph, two hundred pixels square, in colour. That is one hundred and twenty thousand numbers. Feed it to an ordinary fully connected layer of a thousand units and you have just asked fo https://chalkbrain.com/videos/ai-sequence-models/ https://chalkbrain.com/videos/ai-sequence-models/ Three words. Dog bites man. Now the same three words in the other order. Man bites dog. Nothing in the vocabulary changed. Everything in the meaning did. Order is information, and a model that sees a https://chalkbrain.com/videos/ai-neural-nets/ https://chalkbrain.com/videos/ai-neural-nets/ Picture a mixing desk. Every input arrives on its own channel, and every channel has a fader that decides how much of it reaches the output. That is a neuron. The faders are called weights, and learni https://chalkbrain.com/videos/ai-deep-learning/ https://chalkbrain.com/videos/ai-deep-learning/ Here is a program that converts miles to kilometres. One line, one constant, and you knew the constant before you started. Now write one that decides whether a photograph contains a cat. Same shape, s https://chalkbrain.com/videos/ai-history/ https://chalkbrain.com/videos/ai-history/ Artificial intelligence has meant three completely different things. First, rules that a human wrote down. Then weights that a machine learned from examples. And now scale, applied to one architecture https://chalkbrain.com/videos/st-everything/ https://chalkbrain.com/videos/st-everything/ One more time, all of it. A card number arrives, the merchant's server posts one payment intent with one key, the front door dedupes it, the machine takes its first legal hop, fraud scoring waves it t https://chalkbrain.com/videos/st-beyond/ https://chalkbrain.com/videos/st-beyond/ Stripe ships a whole new financial product in the time a competitor spends rebuilding the plumbing for one. That is not a hiring story. Once you own the ledger, the rails, and the primitives underneat https://chalkbrain.com/videos/st-data-infra/ https://chalkbrain.com/videos/st-data-infra/ You get a handful of innovation tokens, and you spend them on the one thing that makes your company different. Everything else, you pick boring. For a payments company that is not a preference, it is https://chalkbrain.com/videos/st-exactly-once/ https://chalkbrain.com/videos/st-exactly-once/ Last episode, one endpoint became safe to retry. But a payment is not one endpoint. It touches fraud, the ledger, balances and notifications, and every wire between them can lose a message or deliver https://chalkbrain.com/videos/st-rails/ https://chalkbrain.com/videos/st-rails/ You tap your card, and a second later the screen says approved. Nothing has been paid. What just happened is a message crossing four separate institutions, in a format designed in the 1980s, and comin https://chalkbrain.com/videos/mb-base-rate/ https://chalkbrain.com/videos/mb-base-rate/ A test that is ninety nine percent accurate says you are positive. You are probably fine. Not maybe fine. Around ninety one percent likely to be perfectly well. Same test, same result, and nothing abo https://chalkbrain.com/videos/st-ledger/ https://chalkbrain.com/videos/st-ledger/ Here is the rule that everything else in this episode obeys. Money is never created, and money is never destroyed. It only ever moves from one account to another. Which means every entry in the book h https://chalkbrain.com/videos/st-state-machine/ https://chalkbrain.com/videos/st-state-machine/ Your server charges a card. The card succeeds. And then, before it can write down that it succeeded, the process dies. Now nobody in the world can answer the only question that matters: did that custo https://chalkbrain.com/videos/st-idempotency/ https://chalkbrain.com/videos/st-idempotency/ You press pay, the spinner turns, and then nothing comes back. Did the charge go through, or did it die on the way? From where you are sitting those two look identical. Last episode I promised you the https://chalkbrain.com/videos/sl-everything/ https://chalkbrain.com/videos/sl-everything/ One more time, all of it. You press enter, the nearest edge takes it, the monolith writes it down first, Vitess stores it under the channel, a channel server works out who is subscribed, and their gat https://chalkbrain.com/videos/sl-search/ https://chalkbrain.com/videos/sl-search/ Google search has to reject a trillion documents cheaply, and that pruning is the entire game. Slack searches one workspace. Small, bounded, and yours. It can afford to spend real money thinking about https://chalkbrain.com/videos/st-map/ https://chalkbrain.com/videos/st-map/ One API call moves money between two strangers' banks, and it must never lose a cent. Here is the thing people get wrong about Stripe: it is not really a payments company. It is a ledger company, with https://chalkbrain.com/videos/sl-two-owners/ https://chalkbrain.com/videos/sl-two-owners/ Slack's original design had one column doing three completely different jobs. Where do these rows live. Are you allowed to see them. And who even are you. All three answered by the same word, and it w https://chalkbrain.com/videos/mb-birthday/ https://chalkbrain.com/videos/mb-birthday/ There are three hundred and sixty five days in a year. So how many people do you need in a room before two of them share a birthday, with a fifty fifty chance? Have a number in your head. It is twenty https://chalkbrain.com/videos/sl-shard-key/ https://chalkbrain.com/videos/sl-shard-key/ In twenty thirteen Slack picked a shard key. Everything a company owned lived together, found by asking one question: which team? It was the right answer for four years. Then nobody changed it, nothin https://chalkbrain.com/videos/sl-monolith/ https://chalkbrain.com/videos/sl-monolith/ Five million lines, written in twenty fourteen, and it is still the front door of Slack today. Every architecture talk of the last decade said break the monolith up. Slack read the same advice, ran th https://chalkbrain.com/videos/mb-sine-waves/ https://chalkbrain.com/videos/mb-sine-waves/ Every sound you have ever heard — a chord, a voice, a drum hit — is pure tones stacked on top of each other. Nothing else. And there is a piece of maths that can take any sound apart and hand you the https://chalkbrain.com/videos/sl-thundering-herd/ https://chalkbrain.com/videos/sl-thundering-herd/ Opening Slack used to cost one enormous HTTP call. Before you could read a single message, the client asked for the entire workspace: every user, every channel, every membership. Seven megabytes at fo https://chalkbrain.com/videos/sl-message-path/ https://chalkbrain.com/videos/sl-message-path/ Open Slack and the first thing that happens has nothing to do with messages. Your laptop opens one long-lived socket to a gateway server and leaves it open all day. That server now holds your connecti https://chalkbrain.com/videos/sl-map/ https://chalkbrain.com/videos/sl-map/ You type a message in a channel with two thousand people in it, and press enter. Half a second later it is on every one of their screens, on four continents, whether their laptop is awake or their pho https://chalkbrain.com/videos/sd-bouncer-door/ https://chalkbrain.com/videos/sd-bouncer-door/ Last time the wire was cut and you had to choose: answer anyway, or refuse to be wrong. This time nothing is broken. One regular at the chai stall is knocking five hundred times a second, and everyone https://chalkbrain.com/videos/mb-secret-shouted/ https://chalkbrain.com/videos/mb-secret-shouted/ Two strangers stand at opposite ends of a crowded room. They have never met, and they cannot whisper. Every single word they say, they have to shout, and the whole room hears it. By the end of this, t https://chalkbrain.com/videos/sd-cut-phone-line/ https://chalkbrain.com/videos/sd-cut-phone-line/ Splitting the ledger bought you machines. It also bought you the wire between them, and that wire is not a detail. Two branches of the famous chai stall, one book each, kept in step over a phone line. https://chalkbrain.com/videos/wa-everything/ https://chalkbrain.com/videos/wa-everything/ All of it, one more time. Your phone holds a socket open, the message is encrypted before it leaves, a registry finds the recipient's server, the mailbox writes it down, and their phone acknowledges i https://chalkbrain.com/videos/wa-erlang/ https://chalkbrain.com/videos/wa-erlang/ Some numbers worth sitting with. Four hundred and fifty million users. Thirty two engineers. And around two million live connections held open on a single server. Those numbers are not a boast, they a https://chalkbrain.com/videos/wa-encryption/ https://chalkbrain.com/videos/wa-encryption/ The claim is strong enough to be worth checking. The company running the servers cannot read the messages passing through them. Not will not. Cannot. The private keys never leave the two phones having https://chalkbrain.com/videos/wa-groups/ https://chalkbrain.com/videos/wa-groups/ A message to one person is one delivery. A message to a group of two hundred is two hundred deliveries, two hundred mailbox writes, and two hundred phones to wake up. One tap on your screen became a s https://chalkbrain.com/videos/wa-databases/ https://chalkbrain.com/videos/wa-databases/ Which database should this use is the wrong question, and it is the one everybody asks first. A chat app does not have one kind of data. It has six, and they want completely different things from a st https://chalkbrain.com/videos/wa-retention/ https://chalkbrain.com/videos/wa-retention/ Messages do not arrive in the order they were sent. One retries, another takes a different path, and they land shuffled. Which is fine on the wire, and completely unacceptable in a chat window. The fi https://chalkbrain.com/videos/wa-journey/ https://chalkbrain.com/videos/wa-journey/ One message, a phone in Mumbai to a phone in Virginia. It has to arrive in under a second, and it has to survive a server dying halfway. Those two requirements pull in opposite directions. The first p https://chalkbrain.com/videos/wa-connection/ https://chalkbrain.com/videos/wa-connection/ A server cannot start a conversation with your phone. There is no number to call. It sits behind a carrier's network with no public address, and whatever address it does have will be different in an h https://chalkbrain.com/videos/wa-map/ https://chalkbrain.com/videos/wa-map/ Netflix was a system that keeps everything and serves it from as close to you as possible. WhatsApp is the opposite. It holds your message only until it is delivered, and then it throws it away. Here https://chalkbrain.com/videos/nf-everything/ https://chalkbrain.com/videos/nf-everything/ Here is all of it at once. Press play, and a gateway routes you, a service authorizes you and builds a list of streams for your exact device, steering picks a box, and that box streams chunks until yo https://chalkbrain.com/videos/nf-data-platform/ https://chalkbrain.com/videos/nf-data-platform/ Every play, every pause, every scroll past something you did not click is an event. Billions of them a day, emitted by clients and services alike, all headed for the same place. From there it splits b https://chalkbrain.com/videos/nf-resilience/ https://chalkbrain.com/videos/nf-resilience/ One home screen is not one request. It fans out to dozens of separate services, each with its own database, each deployed and scaled and broken independently of the others. Which creates a problem tha https://chalkbrain.com/videos/nf-recommendations/ https://chalkbrain.com/videos/nf-recommendations/ About eighty percent of what people watch on Netflix started on the home screen, not in the search box. Which makes that screen the product. Every row on it, and every tile inside every row, was chose https://chalkbrain.com/videos/nf-control-plane/ https://chalkbrain.com/videos/nf-control-plane/ Between your thumb hitting play and the first frame appearing, there is a gap. Nothing in that gap is video. It is a handful of cheap questions being asked and answered, and the whole design goal is t https://chalkbrain.com/videos/nf-encoding/ https://chalkbrain.com/videos/nf-encoding/ A studio hands over one enormous master file. What has to come out the other end is thousands of files: every bitrate, every codec, every resolution, every language, packaged and encrypted and ready t https://chalkbrain.com/videos/nf-abr/ https://chalkbrain.com/videos/nf-abr/ Every title exists at several qualities at once, and each one is cut into chunks a few seconds long. The chunks line up in time, which is the whole trick, because it means you can change your mind bet https://chalkbrain.com/videos/nf-cdn/ https://chalkbrain.com/videos/nf-cdn/ Netflix is about fifteen percent of everything downloaded on the internet. Serve that from a data centre and every stream crosses the public backbone, which nobody can afford and no origin can sustain https://chalkbrain.com/videos/nf-map/ https://chalkbrain.com/videos/nf-map/ Two hundred million people press play, often on the same evening, and each expects a film to start in about a second. Almost everything Netflix is follows from taking that sentence seriously. The path https://chalkbrain.com/videos/sd-split-ledger/ https://chalkbrain.com/videos/sd-split-ledger/ The copy clerks scaled the reads and left the writes exactly where they were. Every change still lands in one book. Replication copies the same data many times. To scale writes you have to split diffe https://chalkbrain.com/videos/sd-copy-clerks/ https://chalkbrain.com/videos/sd-copy-clerks/ Every wala walks to one register. That register is a single database, and it is two problems wearing one coat. It is the bottleneck every read must pass through, and it is the one place all of your da https://chalkbrain.com/videos/sd-order-slip-rail/ https://chalkbrain.com/videos/sd-order-slip-rail/ At the chai stall this was a rail of paper slips. In a system it is a queue, and the problem it solves is this. When taking work and doing work are the same step, one slow job freezes everyone standin https://chalkbrain.com/videos/sd-cache-slate/ https://chalkbrain.com/videos/sd-cache-slate/ Last time the walas were made to remember nothing, and that is exactly what made them interchangeable. But it left a bill. Every single question now walks to the register in the back room. How much is https://chalkbrain.com/videos/mb-fold-to-moon/ https://chalkbrain.com/videos/mb-fold-to-moon/ One sheet of paper, a tenth of a millimetre thick. Fold it in half and you have two layers. Fold it again, four. Again, eight. Then sixteen. Doubling feels boring, and that is the trap. It is not bori https://chalkbrain.com/videos/sd-chai-stall/ https://chalkbrain.com/videos/sd-chai-stall/ One stall. One stove. One person making chai. It is perfect, and it is cheap, right up until the cricket match ends and five hundred people want chai in the same minute. There are only two ways out of https://chalkbrain.com/videos/sd-queue-deadlock/ https://chalkbrain.com/videos/sd-queue-deadlock/ Slack runs about nine billion background jobs a day. Push notifications, link previews, search indexing, and every single message you post. In twenty sixteen, all of it stopped, and it could not resta https://chalkbrain.com/videos/mb-halting-problem/ https://chalkbrain.com/videos/mb-halting-problem/ Imagine one perfect program. Feed it any code at all, and it tells you whether that code will finish, or spin forever. Call it H, the judge. It would be the most useful tool ever written. And it seems https://chalkbrain.com/videos/ai-neural-net/ https://chalkbrain.com/videos/ai-neural-net/ Stack a hundred layers in a neural network. Forget one tiny detail, and all a hundred of them collapse into a single one. You would have learned nothing at all. Start with the atom. One neuron takes i https://chalkbrain.com/videos/net-packet-journey/ https://chalkbrain.com/videos/net-packet-journey/ You type an address and hit enter. Your data has to cross the planet to a machine you have never heard of. And here is the strange part: nobody involved knows the route. There is no map of the interne https://chalkbrain.com/videos/dsa-big-o/ https://chalkbrain.com/videos/dsa-big-o/ One million items. Two ways to search them. One finishes before you blink. With the other one, you will die of old age first. Watch them race. Linear scan checks every item, one at a time. Binary sear