4 ms·

> The interpreter is not that fast, the language is not that expressive, what passes for package management is a bad joke, etc. I can work with it but I'm not n

by bormaj 2y ago

> The interpreter is not that fast, the language is not that expressive, what passes for package management is a bad joke, etc. I can work with it but I'm not necessarily loving it. These are rather disingenuous criticisms of python. The utility of any language is measured by its use as a means to some end. It's clear that Python's ecosystem has prospered not just because it's a fad, but because it's actually useful. Not perfect, but useful. Python can be improved, but I'm sure most other tools/languages can be equally criticized.

lolinder 2y ago
Python in data is similar to C in systems: it's not the best language for the job, but it was good enough at a critical point in history that it became the standard by happenstance—there weren't enough things wrong with it to repel the Brownian motion of developers from seeding the ecosystem—and now it's the best language for the job because it's the standard. > These are rather disingenuous criticisms of python. The utility of any language is measured by its use as a means to some end. It's clear that Python's ecosystem has prospered not just because it's a fad, but because it's actually useful. Not perfect, but useful. Your criticisms of OP's argument are more disingenuous than OP: no one said Python was a fad or that the ecosystem wasn't useful. See here: > Python is where all the action is when it comes to data science, llms, and all the rest. So it's the path of the least resistance. And there's a great argument to not challenge that and just do what everybody else does and put your head down and not criticize any of that. Which is why I use it on a few projects. The library ecosystem is great. They're just saying that Python-the-language isn't their favorite (largely subjective but with a lot of agreement from much of the community) and they mention offhand that Python's package management is a disaster (which is 100% accurate).
- bormaj 2y ago
  I have no qualms with the OPs comment other than the same criticisms I highlighted could be applied to just about any other language in a similarly vague context. So what justice do those criticisms do if we're not being specific about them? For example > Package management is a disaster > The interpreter is slow > The syntax is not expressive Is your particular experience with pip bad? Or is it poetry/conda or another upstart project? Are you working with a monorepo? In what use case is the interpreter slow? The syntax is not expressive (for what)? My point about tooling/utility is that how you use a tool matters a lot (and maybe I didn't express that in my original reply).
- bormaj 2y ago
  > People use python because everybody else uses python. That's it. It's not particularly good at anything it does. But it will get the job done and I can do it. But that just isn't good enough for me. It's the visual basic for data science. And that's not a compliment. Being idiot proof is it's main feature. But that doesn't make it the smart choice. I'll leave this here as well, I think it sufficiently implies the notion of Python being a fad and not particularly useful.

by bormaj 2y ago

lolinder 2y ago

Python in data is similar to C in systems: it's not the best language for the job, but it was good enough at a critical point in history that it became the standard by happenstance—there weren't enough things wrong with it to repel the Brownian motion of developers from seeding the ecosystem—and now it's the best language for the job because it's the standard. > These are rather disingenuous criticisms of python. The utility of any language is measured by its use as a means to some end. It's clear that Python's ecosystem has prospered not just because it's a fad, but because it's actually useful. Not perfect, but useful. Your criticisms of OP's argument are more disingenuous than OP: no one said Python was a fad or that the ecosystem wasn't useful. See here: > Python is where all the action is when it comes to data science, llms, and all the rest. So it's the path of the least resistance. And there's a great argument to not challenge that and just do what everybody else does and put your head down and not criticize any of that. Which is why I use it on a few projects. The library ecosystem is great. They're just saying that Python-the-language isn't their favorite (largely subjective but with a lot of agreement from much of the community) and they mention offhand that Python's package management is a disaster (which is 100% accurate).

bormaj 2y ago
I have no qualms with the OPs comment other than the same criticisms I highlighted could be applied to just about any other language in a similarly vague context. So what justice do those criticisms do if we're not being specific about them? For example > Package management is a disaster > The interpreter is slow > The syntax is not expressive Is your particular experience with pip bad? Or is it poetry/conda or another upstart project? Are you working with a monorepo? In what use case is the interpreter slow? The syntax is not expressive (for what)? My point about tooling/utility is that how you use a tool matters a lot (and maybe I didn't express that in my original reply).
bormaj 2y ago
> People use python because everybody else uses python. That's it. It's not particularly good at anything it does. But it will get the job done and I can do it. But that just isn't good enough for me. It's the visual basic for data science. And that's not a compliment. Being idiot proof is it's main feature. But that doesn't make it the smart choice. I'll leave this here as well, I think it sufficiently implies the notion of Python being a fad and not particularly useful.