Live data from Hacker News

Discipline Doesn’t Scale

sicpers.info

171–173 of 173 posts

Re: Discipline Doesn’t Scale

#171
post #117

Earlier quoted context omitted.

In practice, type inference for return types works fine for the work I do (iOS development). Initializers are blindingly obvious anyway, and other functions are often members of an object or a struct. There are barely any bare functions.

Can’t members return other objects?

Yeah but coding style dictates that the member is then named "makeWidget()". Any confusion about the returned type is minimal, IMHO.

Re: Discipline Doesn’t Scale

#172

Earlier quoted context omitted.

Our contemporary corporate and social culture doesn't encourage or reward discipline. We produce mostly throwaway stuff that has to be released as soon as possible on the market, or else the company goes out of business. The incentives for craftsmanship and diligence are just not there.

Sadly, you are correct, but that’s not new. This has been the case throughout history. I was just thinking about the difference between “good,” and “excellent.” If we look at the difference between a Henkels kitchen knife, and a Japanese Takayuki knife, we are talking exponentially divergent costs. The Takayuki is definitely a lot better than the Henkels, but is it twenty times as good, as the price difference might…

I just realized I left out the "c" in "Henckels." My bad...

Re: Discipline Doesn’t Scale

#173
post #52

One of the most important reasons why we write programs, is because computers are slow and have limited memory. Computers only appear fast due to a host of tricks. The most important trick is the memory pyramid: with at the top a small number of very fast accessible memory (think CPU registers) and at the bottom a very fast but many orders of slower accessible memory (think the whole of the internet). And a lot of bu…

What makes a double-linked list better?

The double-linked list was just an example. Could also have been any other list container for Unicode code points, where each code point uses a fixed size of memory (at least 21 bits). You might even use a data structures where a 'base' character with (optional) combining characters are treated as a single unit, because for many operations they should be treated as a unit.
Post reply on HN