Re: Python list micro-benchmarks

[Bruno Haible]

Hi Collin,

> >   - It explains ChatGPT's failure: Probably there are more explanations
> >     regarding += on strings, on the web, than regarding += on lists.
> >     So ChatGPT used the "common" explanation, for strings, and then
> >     substituted s/string/list/.
> 
> Interesting. I know very little about ChatGPT so that explanation
> works for me. :)

My experience is that ChatGPT (3.5) provides good answers for things that
have a lot of mentions on the web. Whereas for things that are rarely
mentioned, it starts to hallucinate and often provides wrong answers.
Therefore, for routine questions around Python or C++, it is perfectly
suited. Whereas for expert questions or things that require logical
reasoning, you better do a fact-checking on the answer.

> Sounds good. What do you think about two patches to normalize the
> existing code with the conventions we've agreed upon:
> 
>      1. Convert '+= [item]' to '.append(item)'

OK.

>      2. Use single quotes for string literals.

OK in places where the strings don't contain single-quotes. In class
GLEmiter, in particular, there are many strings that contain commands
in shell syntax, and these often contain single-quotes. Here, choose
the style that best avoids backslashing.

Bruno