What's up Python: Jupyter more popular than Vim, VSCode in Excel and a button

Ok, not really VSCode. But close.

Summary

• Python developer survey 2022

• Microsoft introduces a Python Editor in Excel

• Raspberry Pi 5 revealed

Python developer survey 2022

September was a slow month and the biggest news is, of course, coming up tomorrow, with the Python 3.12 release.

Still, we had the results of the annual Jetbains' Python developer server.

While I personally love this survey (the django one too!), it's good to remember it's extremely biased, especially because of the population sample, and the way the questions are laid out makes things confusing.

Don't take it an ultimate truth, but it hints at interesting trends.

Here are a few you'll find:

• OS: 59% on Linux, 58% on Windows, 26% on Mac. Yes, some stats don't add up to 100%, because you could answer multiple times.

• IDE/Editor: VSCode 37%, PyCharm 29% and the 3rd one is... Jupyter? At 5%. Vim after it, at 3%.

• Data analysis, then Web dev and finally Machine Learning are the most popular topics when coding with Python.

• 7% still use Python 2, 3.10 being the most used (in 2022, year of the survey). 11% don’t update their version.

• 37% install Python via python.org, 26% using an OS provided tool (or homebrew, which was mistakenly in this categ) and 17% anaconda. You know what I think about that.

• 49% isolate their projects using virtual environments (venv (43%), virtualenv (37%), poetry (14%), but they used 'virtualenv' in the question so they messed up, and 31% docker.

• 35% don't use a venv in a container. It's sad to see so many people that will suffer for no good reason. Use a venv everytwhere, people. Even in minimal images. It's very hard to get it right.

• 76% use pip, 23% conda, 15% poetry. Good.

• PyPI is the master source, but 35% install packages from ... Github! More than Anaconda (16%). Wow.

• setuptools (40%) is still very much used, despite the slow removal. I’m in there.

• 38% publish on a private PyPI. Didn't think it was that much; I think it says more about the population sample than the community in general.

• Web techs: 39% Flask, 29% Django, 25% Fast API.

• request is still the most used lib.

Microsoft introduces a Python Editor in Excel

I was wondering how Python in Excel was going to feel like since a space significant language with statements/expressions divide is not great for putting in a cell.

Now I have the answer.

Basically, they added components from VSCode. Since you could already use pandas to load the cells, I'm more inclined to believe in a good experience now.

Raspberry Pi 5 revealed

You could argue this is not really a Python news, but I still very much see the raspy as a Python box.

I was trying to shop for one V4 to make a little DIY Python project for a party, but it's out of stock almost everywhere.

So the version 5, announced 3 days ago, is very welcome:

Same form factor and connectivity, 2 times faster and at last, a power button on the board. But honestly I think most people would not care and be happy with the exact same unit, that at least they could buy.

Available in late October, if the foundation is not late, which it usually is.