lec01.txt

---------------
Lecture 1

Is programming a discipline?
- clearly there is a discipline called "computer science"
- is it the same as programming?
  - there are several great computer scientists who never wrote code
  - and there are good programmers who don't know computer science
    - does Ken Thompson know type theory?
    - does he follow research on complexity theory?
  - common view: programming is not a discipline, just grunt work

Claim: programming is a discipline
- a wonderful, glorious discipline
- a view not commonly held or stated
- for example, Ken Thompson stated, "it's all solved. study newer emerging
  fields instead"
- not just great programming; day-to-day programming is a discipline

Have other people held this view?
- Once upon a time, there people who held this view
- 1960-70
- Dijkstra: "The Humble Programmer"
- Nicklaus Wirth: "Alex, write the code yourself. Let your students write
  papers."

Great books
- Wirth, Systematic Programming
- Wirth, Algorithms + Data Structures = Programs
- Dijkstra, Discipline of Programming
- Dijkstra, EWD
- Dahl, Dijkstra, Hoare, On Structured Programming
- Knuth, The Art of Computer Programming

Why classic books?
- Good taste
- Should aspire to them

Knuth
- Bible
  - Starts with math for programmers - Discrete Mathematics
  - Memory Management
  - 2: Random number generation
  - 2: Arithmetic algorithms
  - 3: Sorting & Searching
- then... TeX, Metafont
- and... Literate Programming (web)
- Stanford Graph Base
- No programming language - MIX, MMIX
- Need to know computer architecture
  - Hennessy & Patterson, Computer Architecture
- After Knuth: belief that the computer must be hidden

Computer Science
- Created out of Math and Electrical Engineering
- Europe: "Informatics"
- If a discipline has science after its name...
- 1965-68: departments founded by programmers
- No split between CS and programmers
- Knuth, at symposium on theoretical cs: All of you should write code during
  your sabbatical. You should also consider writing code between your
  sabbaticals...

Programmer?
- No, I'm an architect
- Programmer has become a disreputable title
- Earlier, "analyst"
- No! - We should be proud to be programmers

Profession
- Canon
- Ethics
- Mastery

ACM
- Used to be a professional organization for programmers
- Was hijacked by CS professors
- Used to publish code
  - ACM Collection of Algorithms
    - was taken over by numerical analysts
- We need a professional organization
  - with journals
- Programmers often describe themselves as Artists - by which they mean they
  don't have to follow any rules. That is bad for the field, and bad for them.
  We need t be professional and "boring" like lawyers

What is Programming?
- What unites us?
  - Doctors: "Health"
  - Lawyers: "Justice"
  - Us?
- Write lots of code?
- for digital computers?
- to solve problems
- modeling reality?
- map problem into bits

BITS! - 0s and 1s
  - great discovery
  - Leibniz
  - Turing
- real-world problem -> mathematics -> bits
- manipulate the bits with code
- bits are hidden these days
  - we use "objects"
  - can't even get to bits on modern platforms
- portability?
  - no, bits are the same everywhere
  - c abstract machine
    - sequence of bytes

We must study real machines
- individual brilliant people can make mistakes
- but eventually bad architectures die out
- market dynamics
- Intel reacts to customer demands
  - we we must take it seriously
- Late 90s - x86 is dead, long live itanium

Programmer needs to know the whole stack
- problem
- math
- bits
- algorithms
- compilers
- processors
- it's wonderful!

---------------
Lecture 2