Ask rationally difficult questions

Ask rationally difficult questions
Photo by Brecht Corbeel / Unsplash

I attended PyCon AU a bit over a week ago. The title of this post is from Linda McIver's (from the Australian Data Science Education Institute) fantastic keynote about the important of understanding data.

An interesting counterpoint was Kendra Vant's Keynote on the Sunday about the importance of benchmarking - particularly when it comes to AI projects. It really rang true for me. We need firm metrics and benchmarking for all things, particularly AI, where there's way too much reliance on a 'vibe'.

Luke Wiwatowski's talk about running a simple application on AWS for 'chump change' struck a chord with me (slides here). I've been trying to run things like this blog for chump change for ages so I appreciated the talk (and pretty loose suggestions to use S3 for SQLite databases which I've seriously considered in the past).

I've been more or less out of the Python landscape for a few years now, but uv seems like a good idea and appears to increasingly be the package manager of choice for Python devs. Rust all the way down as they say.

The other thing that kept coming up was htmx which I've been aware of but haven't seen in action before. It looks like a great way to inject a bit of interactivity without having to go down the route of a fully fledged SPA or something of that sort. It would be worth a look. This article I read recently reminds us that there are alternatives to React and that React is starting to look quite dated. It's just so ubiquitous and hard to avoid.

Evy looks like a pretty cool tool to teach kids how to code.

DuckDB seems like a good idea and certainly had a lot of people talking in the break.

There were lots of other great talks of course. What is so striking about the conference though is how militantly inclusive it is, in stark contrast to so many other tech conference. I expect it's no accident that it's grown out of what was once the fringe world of open source software unlike more 'enterprise' ecosystems. While open source is now very mainstream, it wasn't 20 years ago. But that community built a beautiful conference around a great accessible programming language that happened to be open source, when making the choice to use open source was more of a political statement. There's a reason I've still never written a line of Java.

Well played team. I can't wait for it next year.