Re: Python recompile

On 12/03/2025 16:48, David Brown wrote:

On 12/03/2025 16:32, Michael S wrote:

As already said by Bart, it is a bit of fun.
To me it sounds like sufficient reason.
>

 Fun - or learning - are always perfectly good reasons for doing almost anything in programming as far as I am concerned (baring things like writing malware for fun).  And I can entirely appreciate the fun in writing a bignum library, or supporting it in an interpreted language.
 But I have trouble seeing the fun in endless complaints, blaming GMP authors for not making Bart's life easier, insisting on limiting everything in his build environment for no good reason, and all the rest of it.  Maybe Bart /does/ find that fun.

GMP and LIBFFI were a couple of well-known examples that came to mind. I have grappled with *dozens* of such projects.
Basically ANYTHING that requires on a shell-script that only works on Unixes and will not work on native Windows is going to be troublesome to build on the latter.
It doesn't just affect me either.
I also feel it is wrong to paint Windows as a poor machine for development, simply because it is not a Unix clone.
You make a choice to rely heavily (30K+ lines is 'heavily'!) on scripting features of a particular OS, which IMO would be a bad choice anyway, and then criticise some other OS for not being able to run that exact same script, or for not having the exact same bunch of utilities.
It would be like me saying I find Linux rubbish for development because it doesn't run my BAT scripts.