For the first time in years I went to an in-person conference. I took that opportunity in my hometown, at the Devfest Toulouse. Last edition I went to was in 2019 and I wrote an article about it back then. A lot of things have changed, and (gen)AI has entered the game, leading to an agenda full of sessions around this topic. Fortunately, there were other talks far from focusing on AI, and I must say, it was a relief : I attended them.

Content

I must say I wasn't expecting so many talks about AI. Somehow it makes me sad. I think Alexandru Nedelcu expressed it quite well in his blog. I almost escaped the AI hype : I started the day with a conference about BERT family of models and I have to admit I learned a lot throughout the session. And then I was done with AI !

This year my favorite talks were "Peut-on RIIR de tout" by Igor Laborie and the closing keynote (more on it later). During the event, I also remembered why I did not rush back to conferences like this earlier : fundamentally it is hard to find talks going real deep in a rabbit hole and those conferences are not designed for that. Time slots are too short, audience is expecting a show,... Speakers need to find more than just the tech to be selected. Yet I found that Igor did well in his talk. I might be biased because he's a person I know and like, to be fair. Despite some technical issues (with displays : common it's 2025 and in the area of AI, we still cannot handle video streams on cable properly 😂), and without share much code, it really resonated for me. I would summarize his talk by being the quest to the answer of "why isn't Rust more used everywhere ?" And sometimes I feel the same about Scala. One big difference though is that Rust is powering some low level libraries, tools, and basic blocks to build upon... Maybe I also liked the contrast with AI, because Rust is more frugal. Anyway, the talk gave me a lot to think about. So far the (unstructured) gist of what I had in mind afterward is :

• Go exists and is popular because the largest part of the engineering crowd is not passionate : at the end of the day, programming is a job. For lots of us, learning something new is not desirable.
• Passion ain't something strictly required for the business : although a language like Rust will reduce costs, improve efficiency, reduce the amount of bugs,... I wondered if Igor truly joked when he said something like "maybe they [companies] like bugs, that's why they don't use Rust"
• That brought me back to (gen)AI and the why it sucks the joy of programming, I mentioned earlier.
• Are we not using it much more because of the initial investment ? Scala blew my mind quite fast, when I discovered it. It unlocked something and things I wasn't interested in, or judged as non-interesting were suddenly extremely attractive. I'd say, Rust has the power to trigger some switches. Most certainly, if that's desired, learning Rust will make a better developer.
• Can we find meaning in Rewriting something in Rust ? Or is it just for the sake of expanding the language ?

The closing keynote, about Home assistant and crazy use cases, was something. As a newcomer to Home Assistant (I set it up at home a few months ago), I was eager to see what creative folks were doing with it. Structured with several speakers, presenting their automations/devices/use cases, it was super fun. Although some jokes were borderline (I'd rather prefer a Partner Acceptance Factor instead of Wife Acceptance Factor), I was delighted by how far we can go with such tools : Raspeberry Pi, ESP32, cheap raw material... Somewhere it's the encounter of hardware and software. I must say it reminded me of the days where I was tinkering with the 3D printer with my brother, or when I was playing with Arduino and sensors, without the industrial thinking, just to prototype, and get something done.

Venue and all

The venue itself, near Toulouse wasn't that great. I mean, it's located in an area where there is only offices and from 8:30am to 9am, it's rush hour ! I almost made it on time but then I queued for a long time before entering. The opening keynote (which sounded very interesting, and completely unrelated to AI !) had started for 20 minutes yet, and the speaker encountered a bunch of technical issues when I sat down. Downside of that was that I rushed to the room, and overall this "I have to rush" feeling did not leave me that day. Maybe the agenda was packed, maybe it was just a feeling. I was glad to run into some former colleagues, friends, and having chats about work / technologies,... I guess it is one of the reason I also purchased a ticket. Despite the entrance issue, staff was super nice and friendly. Lunch time was frustrating a bit, because quickies was interleaved in the meantime... Why not have those short sessions somewhere else in the agenda ? Unlike in past editions, I avoided wandering around the stands of sponsors (although I saw familiar faces). Most certainly because I had no reason to engage, I mean, I'm not seeking for a job.

Stressing it again, staff did a great job. For sure there are things to improve : I did not understand why it took so long to enter, no central place with the (printed) schedule (except for the app/website). I remember some some conferences printing it on the badge. Maybe it could also be a good place to put the type of meal chosen ? Having checked vegetarian on the registration form, I was nicely surprised to not struggle to get food. It was smooth experience (and food was good).

Closing words

It was a nice day to get some food for the brain, reconnect with former teammates and people I had not seen in a long time. I'm not sure I learned a lot, but I went to sessions where speakers were well prepared and overall managed to deliver the message to an heterogeneous audience. I knew it wasn't going to be about the internals of Tokio or how the cats-effect loop works, or those kind of topics, to be fair. And talking to such a large audience and being understood is an art, so kudos to the speakers I went to see. And again to the staff !