Today I migrated some Go code to Python and made it 590x FASTER 😎
Remote
Anže
@anze3db@fosstodon.org
mastodon 4.7.3Writing Python 
, Django 
and surfing waves 🏄♂️
Projects:
🧑💻 https://fedidevs.com find awesome devs across the fediverse
💻 https://github.com/anze3db/django-tui a TUI for all your Django commands
673 Followers
577 Following
30 Posts
Joined November 06, 2022
GitHub:
Website:
Blog:
Twitter:
Quite a way to start a Monday 😅
With recent Python supply chain attacks (Trivy/LiteLLM), it’s worth mentioning uv’s `exclude-newer = "x days"` config.
It forces uv to only installs packages published more than x days ago, reducing risks since problematic packages should be yanked by then.
I asked Claude to come up with writing guidelines based on my blog posts:
> A few things I deliberately did not encode as guidelines: the frequent typos (characteristic of your fast publishing cadence)
88 blog post in 14 years is not fast publishing cadence 😅
My dog kept me in Zone 2 for 3kms. Then I had to drop him off and immediately made bad decisions. 💛
With all the supply chain attacks going around, we have to be very careful about how we update our dependencies. I've written a full blog post about it, but here is the TLDR:
1. Pin to hashes, not just versions
2. Automate the updates
3. Use dependency cooldowns
https://blog.pecar.me/how-to-safely-update-your-dependencies/
Everyone’s favorite web framework is now also in the gelato business 🤤
We benchmarked 15 models for triaging vulnerabilities.
* Kimi K3 came out on top but marked a true positive as a false positive ☠️
* Opus 5 refused to give a verdict on 11 vulnerabilities
* Sonnet 5 and Luna outperformed Terra and Sol
Full post 👇
https://www.fencer.dev/blog/llm-triage-sast-false-positives
I wrote a blog post about some of the options for adding type checking to your Django project!
Historically, Django projects have been tough to type-check, but the future feels bright, especially if PEP 827 gets accepted!
I wrote a post on how we tracked down slow imports in our Django app, fixed them by making imports lazy, and added a lint check to prevent regressions.
Now I’m waiting for Python 3.15 to make all of this easier! 🚀
https://blog.pecar.me/speeding-up-django-startup-times-with-lazy-imports/
We reviewed 4,978 mypy runs in our Django app:
• 4,731 green
• 246 failed
Of the 246 failures:
• 173 false positives
• 73 real issues
• 39 caught only by mypy and no other check!
Some noise, but it prevented real bugs from reaching production. 🎉
After 9 meetups full of talks, lightning talks, and even lightning turtles, it's time for the Python Lisbon Meetup to take it easy.
Join us at Linha d'Água on Thursday to relax and enjoy the summer in Lisbon with Python friends 🐢
https://www.meetup.com/python-lisbon-meetup/events/315420890/
Can't wait for lazy imports in Python 3.15! I spent a bunch of time and tokens over the weekend inlining heavy imports so that they don't get loaded during initial start of a Django app.
The `https://manage.py check` command is now 3.8x faster (9.45s down to 2.48s).
A bug that survived 15 years of upgrades.
One of my servers lost DNS after every reboot. ping 8.8.8.8 worked; ping https://google.com didn't resolve. I'd been "fixing" it by hand-editing /etc/resolv.conf, which, of course, was reset on the next boot. 🙈
Turned out to be a config from 2010. I finally debugged it properly with Claude.
https://blog.pecar.me/the-15-year-old-iptables-rule-that-broke-my-dns
I've finished #adventofcode 2025 and this is the first year that I managed to get all solutions under 100ms ⚡
All but two are under 10ms. Not bad for the slowest popular language out there 🐍, although I did have to use numpy a few times.
Replying to
@erwinrossen@mas.to good point. I think the cooldown period works for regular updates but not for security updates. If you are impacted by a security vuln you should upgrade immediately. That's not something you can easily automate though...
What you can do is set a lower cooldown period for packages you trust more. E.g. we have a general 3 day cooldown for all packages but 1 day for Django so that we bring in security updates the same week. We would patch immediately if the CVE was impacting us.
Another Python Lisbon Meetup in the books! Looking forward to the next one on May 7 👀
I'll be giving a talk about lazy imports at the next Python Lisbon Meetup!
See you there? 😀
Make sure to always reuse your botoclients. Initializing a fresh one on each task has A LOT of overhead!
I wrote a blog post about this year's Advent of Code. I had a blast and I am already looking forward to 2026!
* All my Python solutions were under 1s using pure Python and 100ms with a little help from mypy/scipy
• The AoC community is the best ❤️
Replying to
@jonathanbiemond yes, lazy imports are used heavily in larger web applications. Although they are a double edged sword becasue they can offload slow imports into the request/response cycle. So in prod you usually have to then "warm up" the process before it starts serving requests.
In our case, most of those heavy imports are used by celery tasks so nothing to really no need to import them when starting the web server.
Replying to
@jack I agree. I would even go as far and say that imports should have been lazy by default. But having the ability to opt into lazy imports by not needing to move imports into functions is going to be a great addition!
Like pretty much everyone with a static blog site in the last 6 months, I asked Claude to migrate it to a different static site generator.
Here's a blog post about it, along with some of my general thoughts on agentic code.
Replying to
See you on Mar 5 @ 19:00 at IST, Pavilhão de Matemática, room 3.10!
Check out our website for more info on how to join our community!
https://python-lisbon-meetup.github.io/
New board!
The Claude usage page is very confusing. Since Sonnet has a separate bar I thought I'd be able to switch to it after I max out Opus, but nope.
Of course this all makes sense if you actually read the tooltip/labels but who does that today anyway? 😅
Oh, I didn’t know GitHub doesn’t have pagination in the Commits tab of a pull request. 😅
Replying to
@trey@indieweb.social @stuartm@fosstodon.org I haven't heard it until now but out of the box it seems to have the same limitations as pyright, ty, pyrefly that I describe in the blog post.
With zuban I'm seeing over 7k type errors in my codebase where mypy + django-stubs has 0.
I'll play around a bit more with it when I have more time, but I think the general rule of the blog post still stands: You either choose completeness with mypy+django-stubs or you choose speed but then have to deal with attr-defined errors.