On: How the append-only btree works (2010)

[Hnrobert42]

As Phil Karlton, “There are only two hard things in computer science: cache validation and naming things.” Seems like you found a way to work on both!

https://skeptics.stackexchange.com/questions/19836/has-phil-...

[timeimp → Hnrobert42]

I always have to add the "and off-by-one" at the end.

Nothing like reading memory you don't own!

[airstrike → timeimp]

There are actually only two hard problems in computer science:

0) Cache invalidation

1) Naming things

5) Asynchronous callbacks

2) Off-by-one errors

3) Scope creep

6) Bounds checking

[Scarblac → airstrike]

There is only one hard problem in software engineering: people.

[alternative_a → airstrike]

Let’s be pedantic here and get all religious over the words “Science” and “hard”.

Computer “science” has a difficult conceptual problem with caching. The optimal cache, this science tells us, is indistinguishable from a fortune teller who is never wrong (oracle). Fortune telling is a “hard” problem for a science based on reasoning. The best we can do is hedge bets (which is what the science of caching focuses on).

This same science also has a difficulty with naming things. Now numbering things is easy and science loves maths and maths love sciences, but science and letters have a more difficult history. Science would approve of “factoryfactoryfactoryImpl” btw ... it’s “a rational scheme of naming”. .

Here we see a “science” that is facing actual difficulties.

The rest of your list are difficult but not “hard”. The science of these matters is clear and the rest is up to the “scientists” struggling with “scope creep” and “bounds checking” ..

[pmarreck → Hnrobert42]

invalidation?

[Hnrobert42 → pmarreck]

Yes. Stupid speech to text and I didn’t double check. Thanks.