Table of contents
  1. Week 1 intro talk
  2. Week 5 video 4
  3. Week 7 video 3
  4. Week 7 video 4
  5. Week 8 video 3
  6. Week 9 video 4

Any errata will be linked from the video itself on the front page and fixed in the slides, so there’s no need to keep checking back here.

Week 1 intro talk

Slide 7 referred to the “Tuesday live sessions” - Tuesday should have read Monday, these are the normal workshops you have timetabled from 1pm-3pm.

Week 5 video 4

Part of the video shows a version of slide 3 with 15 bits stored under “Contents” rather than 16 bits.

Slide 5 of the video says that “for example, @SCREEN followed by M=1 would colour the 16th pixel black” - “16th” should read “first”.

Week 7 video 3

Slide 4 makes the claim that in modern CPUs, usually CISC is better than RISC for high-speed high-power applications. In fact, the pendulum has been swinging back the other way in recent years and RISC is competitive with CISC in these domains as well.

Week 7 video 4

Slide 3 says we can set the clock speed to the propagation delay of the fastest stage; this should read the slowest stage.

Week 8 video 3

In EBNF, the A - B syntax for anything that matches A but doesn’t match B is only valid when B expands to a finite list of possible token sequences. So for example: ⟨number⟩ - ⟨digit⟩ would be valid because ⟨digit⟩ is one of ten possibilities, namely ‘0’ through ‘9’; but ⟨number⟩ - ⟨posNumber⟩ would not be valid because ⟨posNumber⟩ could expand to infinitely many different strings.

Week 9 video 4

Slide 8 says that “temp is allocated a fixed segment of length 8 and base address 5 = TEMP” - this is all true except that TEMP is not a valid keyword in Hack assembly.