Course Capstones
The narration between entries paraphrases; it never reads an entry aloud before the entry plays. Every entry in this lecture plays exactly as it does on its own page — same file, same speed, nothing laid over it.
What you can do after this
- Diagnose a slow application by naming which layer each candidate cause lives on
- State a trade in the course's grammar — name both sides, price both sides, choose on purpose
- Say what abstraction does, and recognise the same move across every unit
- Say what shape the course's canonical errors share
- Answer what computer science is, with the curriculum as the evidence
What it covers
6 entries in 6 topics. Each number links to that entry on its own.
-
1. The diagnostic
-
2. The ledger
-
3. The one idea
-
4. The museum
-
5. The app, complete
-
6. The answer
Check yourself
How do you diagnose a slow application?
By layer. The complaint has a candidate cause on every level the course covered — the algorithm's order of growth, the data structure chosen, the memory hierarchy, the database's normalisation, the network round trip, the deployment. Diagnosing top to bottom uses the whole curriculum as one instrument.
What grammar do the course's trades share?
Name both sides, price both sides, choose on purpose. Time against space, performance against productivity, normalising against query speed, transaction guarantees against processing cost — all of them take that form, and the discipline is largely the management of them.
What is the single idea underneath every unit?
Abstraction — hide the machinery, expose the contract. The operating system does it to hardware, a language does it to machine code, a declared type does it to bits, virtual memory does it to physical addresses. One idea in eleven costumes.
What shape do the course's canonical errors share?
Each one is a true thing stretched one step too far. Bits do carry data, but not meaning. A cast does convert, but not without cost. NP is about polynomial time, but not the negation of it. The error is always the extra step.
What is computer science?
The study of computation across every layer where computation lives — from the theory of what can be computed, through algorithms and languages, down to the hardware, and out to the systems and networks running on it. The course opened by declining to answer this and closes with eleven units as the evidence.
Lecture transcript
The narration as recorded, in order. Where an entry plays, the transcript links to it rather than reprinting it — the entry plays from its own file, unaltered, and says the same thing on its own page.
This is Unit Twelve of Computer Science, and it is unlike every unit before it. There is nothing new here. No definitions, no mechanisms, no procedures. Every entry in this unit is a capstone, and every capstone reaches back across the whole course rather than closing a block of its own. The first takes the most ordinary complaint anybody has about software — this is slow — and shows that the question has a real answer on every single level the course covered, from the algorithm to the memory hierarchy to the network. Two of them are about how the course thinks rather than what it said. One gathers every trade the course made and shows they share one grammar. Another argues that eleven units leaned on a single idea in eleven costumes. One is a gallery of every error the course marked, and it finds a shape they all share — each is a true thing stretched one step too far. And two of them close threads: the transit app that has run since Unit One, and the question the course opened by refusing to answer. By the end you will be able to do five things. One: diagnose a slow application by layer. Two: state a trade in the course's grammar. Three: recognise abstraction as the same move everywhere. Four: say what shape the course's errors share. Five: answer what computer science is.
One capstone, and it starts from a complaint rather than a concept. Somebody says an application is slow. That sentence has a candidate answer in the algorithm, in the data structure, in the memory hierarchy, in the database, in the network, and in how the thing was deployed — and knowing which layer to look at first is what the course was for.
The most ordinary complaint there is.
CS 0671 One Slow App, Every Layer
The question has a floor on every level of the course, and diagnosing it top to bottom is the whole curriculum working as one instrument rather than as eleven separate subjects.
One capstone, and it makes an argument about the discipline itself. Nearly every decision this course described was a trade — time against space, performance against productivity, normalising against querying speed, guarantees against processing cost. Gathered in one place, they share a single grammar.
Every trade, in one place.
CS 0675 The Tradeoff Ledger
Computer science run end to end is trade management, and the course's recurring decisions share one grammar — name both sides, price both sides, and choose on purpose rather than by default.
One capstone, and it is the deepest thing in the course. Every unit made the same move — hide the machinery, expose the contract. The operating system did it to hardware. The language did it to machine code. The type did it to bits. Eleven costumes, one idea underneath.
The move every unit made.
CS 0680 Abstraction, the Course's One Idea
Every unit leaned on the same move, and seeing abstraction as one idea wearing eleven costumes is the deepest version of the whole curriculum — the thing that makes the units a subject rather than a list.
One capstone, and it does something the individual error entries could not. Each of those corrected one belief in isolation. Walked together, the whole gallery reveals that they fail the same way — every one of them is something true, carried one step past where it stops being true.
Every error, in one gallery.
CS 0685 The Misconception Museum
The course kept a museum of canonical errors, each numbered beside its correct twin — and walking the whole gallery at once shows they share a shape: a true thing, stretched one step too far.
One capstone, and it closes the longest thread in the course. The transit app was experienced in Unit One, analysed, built in Unit Nine, deployed in Unit Ten, and defended in Unit Eleven. Its final pass runs a rider's half-second through all of it at once.
The app, one last time.
CS 0690 The App, Complete
One transit app has carried the course from its first unit to its last, and its final pass runs the rider's half-second through every layer simultaneously rather than one at a time.
One capstone, and it answers the question the course opened by setting aside. Unit One separated computer science from programming and declined to say what the subject actually is. This entry answers it — and the answer is only available now, because the evidence is the eleven units behind it.
The question, finally answered.
CS 0695 What Computer Science Is, Answered
The course opened by separating computer science from programming, and closes by answering the larger question with the whole curriculum as evidence — the study of computation across every layer where computation lives.
Five things you can now do, and Computer Science is complete. You can diagnose a slow application by layer — algorithm, structure, memory, database, network, deployment — and know which floor to check first. You can state a trade in the course's grammar: name both sides, price both sides, and choose on purpose rather than by default. You can recognise abstraction as the single move underneath every unit — hide the machinery, expose the contract — and name where each unit made it. You can say what shape the course's canonical errors share: each one is something true, carried one step past where it stops holding. And you can answer what computer science is, with eleven units as the evidence rather than with a definition offered on trust.
This is free, and stays free. If it helped, here is what keeps it running.