Find a flaw.You can't.
interpreted
>>108488111It’s absolute garbage and if you don’t realize this you should get checked for a brain tumour.
>>108488111If it's not possible to make highly optimized games using it, so it's bad.
>>108488116>>108488125>>108488130Hardware issue
>>108488111The logo is two faggots joined together by their penes and sphincters. They should have asked /g/ for help with it.
cpython is dogshit slow because some genius thought it was a good idea to let modules depend on internal implementation
>>108488111It's slow and doesn't work. Use C, oh wait you're indian.
>>1084881111. horribly slow, slower than fucking js of all languages, a language that is worse in most other aspects2. dynamic typing, which allows the code to have type errors during runtime because rakesh forgot to cast a float to an int and now some random function is trying to index an array with a float3. flabby syntax that looks like pseudocode, which encourages the programmer to code sloppilyOther than these aspects, it's actually not that bad.
>>108488111
>>108488111Dependency issues. You should be able to have multiple versions of the same package if needed instead of needing to upgrade/downgrade and then break shit for something else
>>108488353C is for Indians like (You) who feel the need to larp as memory managing wizards.C++, Java, C# and Rust solved all the problems of C and introduced no new ones that C had the answer to already.
>>108488111Dumb bait thread.Python is great for scripting. Not great for "software"If you don't know the difference or think there isn't one, you shouldn't be programming.
>>108488111Performance
significant whitespace is just retarded
>>108488130I think all browsers are written in python aren’t they?
>>108488769Nope. Wrong answer. Of all the many reasons to hate Python, this is not one of them.
Here's a concise steelman of Python's key flaws:GIL: Prevents true multi-threaded CPU parallelism in CPython, forcing workarounds for compute-heavy tasks.Dynamic typing: Bugs (wrong types, typos) only appear at runtime; static analysis is weak without heavy type hints.Performance: Significantly slower than C/Rust/Go due to interpreter overhead, especially in loops and hot paths.Packaging: Fragmented dependency management, virtualenvs, and binary compatibility issues remain painful.Excessive dynamism: Runtime mutability and monkey-patching make large codebases harder to reason about and maintain.Python excels at rapid development but hits these limits as scale, performance, or team size grows.
>>108488111>pypi goes offline>becomes unusable
>>108488111Versioning
>>108488734I like you. You doesn't afraid of anything. >thinks scripting is programming>probably thinks markup is also programmingI think this as well, anon. We should start a general.
>>108489030Scripting is programming. Markup isn't.You post like someone who produces only shit no matter what the objective or tools available.
>>108489054Aw, you got offended, that's cute & sad at the same time.Markup is programming, though, despite your Dunning-Kruger protestations. Smell you later, snowflake.
>>108489066Markup isn't a turing-complete language.
>>108489100We are talking about programming, brainiac, not Turing completeness; though . . .
>>108489121Programming implies use of a turing-complete programming language and has for half a century. Markup is not such a language.
>>108488111Global. Interpreter. Lock./thread
>>108488111>stable diffusion python: dozens of gigabytes>stable diffusion c++: dozens of megabyteswhat a pile of garbage
>>108489134>impliesYes, and being able to solve the captcha in order to post here implies one isn't a moran, yet there you go posting replies. ;3>markup is not such a language>but it is, you just don't like itokie dokie, my poorly coded LLM newfriend.
>>108488111>want to do one little thing>HERE'S A FUCKHUGE LIBRARY THAT HAS THE ONE LITTLE THING YOU WANT LOL ALSO IT'S SLOW AS FUCK AND PARTIALLY BROKEN LMAO SAARI like the language itself, but the ecosystem is severely jeeted.
>>108489294Then code that little thing yourself.
>Find a flaw.>You can't.No idea what's this.>>108488111Such a waste of good numbers.
>>108488770ywe
>>108488111Slow as piss.>B-but you can just write a C extension -You could just write your entire program in C.
>>108488389>recursionpython is not designed around it but you can up that 1k limit to whatever you want using sys.something iirc, the actual flaw is speed and bloat the recursion limit is a workaround that not a flaw on itself>0/0, sqrt(-1)that's why try except is for, implicit type casting on a weakly typed language is stupid, both functions are meant to return floats and do exactly that
>>108488111Why>len(myList)and not>myList.len?>inb4 make your own subclassI can't force all libraries I am using to use my subclass
people say its only for scripting but I've been making QT QML gui with python as the backend.I could use C++ but development will be slower and I won't have the massive libraries as my disposal (fastAPI, SQLAlchemy ORM etc..)
>>108488111No types
>>108488111Requiring indentation as part of interpretation.Other wise for it's scripting utility, it's quite fine.
>>108488443>C is for Indians like (You) who feel the need to larp as memory managing wizards.>C++, Java, C# and Rust solved all the problems of C and introduced no new ones that C had the answer to already.Correct answer:Sfx: https://youtu.be/A_SdDSGBenU
It's an interpreted language and thus not suited for quite a few use cases. Python tries to get around this by making it relatively simple to use C or even Fortran in libraries, but this does demonstrate the point that it just shouldn't be used for some things. Python happens to be one of the most common interpreted languages so there's a critical mass of people using it for things they shouldn't be, which fucks with you down the line if you work in some technical field. If you're doing bioinformatics, for example, you get to deal with how much existing work uses Python.In terms of implementation I'm not a big fan of GILs.Most annoying of all, I think, is package management. Perhaps to Windows users it makes sense to have a package manager for a language, but on a system that already has a native package manager this is just a simple way to enter dependency hell. This is particularly the case if, as I said, you're in some technical field. Such dependency hell also makes reproducing research a pain in the ass.
>>108488111It's slow-ish
people complain about GIL when you can just spawn subprocesses
>>108488779why not?
>>108488111The people using it are fucking insufferable. Also not viable for scale. Also, cheched
>>108488158programming issue
>>108488770fucking lol @ pythontards
Python's static type system needs local type evidence (e.g. GADTs or at least type indexed enums) and ideally HKTsAlso Python really needs ways of writing fully optimised kernels without allocation/dynamic calls/attributes
>>108493123Oh and even just default enums should have some way of indicating what type the values are. But they don't do that because of the dumb enum[...] shit, which should be deprecated honestly with temporary typing.XEnum or whatever that don't have [...] class operatorsAlso would be good too if flags enums had a way of typing single flags
>>108488389luau dont have this problemtoo slow in the browser tho
>do my homeworklel no
>>108493225lua is insane and defies rational explanationits probably divinely inspired
>>108492505It's a strongly typed language.
