Wow, they (apparently) finally made the REPL not suck! I always thought it was weird how shit it was given that it’s one of the big reasons Python has become as popular as it is.
Maybe in another 20 years they can make the package tooling not suck too.
poetry
has made the package tooling generally not suck for me, anduv
seems to be getting better. Just a few more PEPs to go untiluv
does what I want. Here’s hoping.Poetry is frigging great.
Never heard of uv.uv
is basically a super fast Python tool, written in Rust. Looks promising, but it seems to be missing some features I really like from Poetry, but I’m keeping my eye on it.
Tbf,
uv
is trying to solve packaging nowYeah it’s definitely a vast improvement on previous attempts (Poetry et al).
I dunno if it can be called solved until it’s officially sanctioned and installed by default though, and I don’t see that happening for a very long time.
How is it better than poetry?
It actually works. I tried poetry once and it failed to resolve some simple dependency specification.
Ok you can say it was a fluke bug but it didn’t inspire confidence.
uv
is also freakishly fast. My venv setup times went from 57 seconds to 7 seconds. I seriously doubt Poetry can do that because it’s written in Python.I abandoned poetry after it was unable to install a specific version of pytorch I was using.
In pip I would do something like
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
, but IIRC poetry didn’t support the--index-url
option.
The Python REPL was always sort of minimal when used from the command line, but is quite usable in an Emacs window. IDLE is also useful some of the time. I never felt the need for anything like Eclipse because of it.
Maybe because people who needed it knew there were better ways to do it, like ipython and Jupyter. I’ve never heard of anyone gushing about the stock REPL.
Now only have to wait for:
- pyenv release
- pycharm update (including terminal)
- 3rd party libraries
to catch up…
Once that happens it’ll be just couple of years until trickles down to corpo I work at :(
We got Python 3.10 in our Hadoop/Spark setup recently. I’m really enjoying those improved debug messages, man.
Protip:
pip install pyupgrade
And thenfind . -name '*.py' -not -path '*.tox*' -print0 | xargs -0 pyupgrade --py310-plus
in your repo to update what can be updated.BTW, pyupgrade’s creator,
asottile
(that’s his name) also has an informative channel: Anthony Writes Code where he explains Python features, or goes into interesting bugs he ran into, etc. The good stuff.You assume that I can access PIP on a big data cluster in a financial institution ;) Even updating packages there requires me to ask for a custom image. I’m a data analyst so I just transform and extract what I can in a way that reduces size of the output and do cool stuff on my machine that has Python 3.11 and access to validated PyPI mirror. ETL that happens entirely on the cluster needs to be so optimised that I don’t need anything fancy thankfully.
Wait how does one make enough money to afford the JetBrains suite? I just do everything in VSCode.
I know some people who have their work pay for it. I pay for the all products pack and it decreases in cost each year until a certain point. Not sure if I’m on some extra discount or whatnot but I only pay $18/mo and it’s easily worth it.
Is that on your personal machine, though?
I have the jetbrains toolbox on like 4 of my machines at this point and three are personal and one is work. It’s a great experience but I pay for it personally because I value it.
If I end up getting a promotion I’ll have to consider that. It’s just a lot of money for me right now, because my current employer doesn’t pay shit at my title/rank.
Oh if only I could get my work to pay for it. Unfortunately, I’m in a megacorp that would shove said request so far down into bureaucracy hell…
I use and enjoy VS codium as well, but PyCharm has a community edition that is free.
VSCodium doesn’t have the Python plugin, does it? It also misses the config sync when you’re logged in (IIRC). Not the worst to miss if you start out, but I’ll take it over having to track EVERYTHING in my code-workspace file.
Ah
Nice I guess it’s time to check if my daily used libraries have stable 3.12 releases already.
I guess the free-threaded mode and the JIT compiler will be the most important features from what I read, but their significance is out of my expertise.
My absolute favorite with this update is the new REPL! It features Multiline-editing and a paste mode for easier pasting code. It also added the spaces automatically in my example.
Sometimes I want to make some quick tests on some data in the terminal without installing IPython to my environment first, this is great news!
This new error message will also be very useful for beginners and relieve StackOverflow:
AttributeError: module 'numpy' has no attribute 'array' (consider renaming '/home/me/numpy.py' if it has the same name as a third-party module you intended to import)
That last one is going to be so good. Months ago I ran into that while porting the “Crafting Interpreters” java-based interpreter into python. It took me a few hours to figure out that one of my modules was colliding with “token” in the stdlib, a module I didn’t even know existed. Glad it’s being made clearer.
Woo!
We’re still on 3.11, but it’s been some time since I last checked compatibility with the later releases. Good job everyone, I’m going to be playing with this over the next couple weeks to see if we can upgrade to it.
deleted by creator
Docstrings now have their leading indentation stripped
I foolish thought that it meant that finally python introduced a hassle free simple way to have indented triple-quoted literal strings. But no. It baffles me that you cannot have simple literal strings that are indented. This is specially annoying if you are using them as templates to output multiline text.
So I remember the plan to improve Python’s speed from 3.9 to 3.13… Has there been an updated plan since? I presume the JIT will likely be faster in 3.14 (it’;s already at parity - pretty impressive for a first release), but is there anything else planned?