they all seemingly make really really strange decisions when doing so.
Spoiler alert, it happens too in proprietary software, physical engineering, and as soon as there is a corporate structure and a quality department it’s even worse because you need to explain why you want to spend more money, and document the impact which means do a shit load of paperwork for every change
TranquilTurbulence@lemmy.zip 16 hours ago
The idea behind Python is to get the community to contribute. More people know Python than Assembly or Fortran. At some point, running a FOSS project like Piefed becomes a numbers game. Having more developers is useful in the beginning.
If Piefed grows significantly, it might make sense to rewrite the whole thing in a different language, but right now, contributions matter more than efficiency.
tyler@programming.dev 5 hours ago
So you set up a nice strawman with assembly and fortran there (which would never be used for a web server) instead of suggesting a realistic option like C# or the JVM, both of which have much larger communities of people that actually know what they’re doing.
You’d get just as many contributions in Java or Kotlin and the quality would be higher as well.
The decisions at the start of the project have the most influence on the project, more so than anything ever will later down the line.
Munkisquisher@lemmy.nz 16 hours ago
Or you follow the python ethos and when it matters, you profile the code, and rewrite only the modules that need it in a lower level language.
TranquilTurbulence@lemmy.zip 11 hours ago
That would make more sense. Best of the both worlds.