Also, look at codetriage.com.

1. Start with smaller projects with simpler codebases. Do not try to understand what LibreOffice or Jellyfin are doing.
2. Maybe try to understand how one specific part of that software works. Chatbots are really, really good at explaining existing code, so ask your bot where the feature is located and let it explain it to you

Take note of the version of the affected program you are using (likely depends on your distro). It could be that the bug has been fixed in a more recent release, and you don't want to create noise for an issue that has already been resolved and hasn't made it to you, or is specific to the build for your distro.

Once you've figured out where the issue must lie, make a report with detailed steps on how to reproduce the issue. Don't ghost the thread, if a dev asks followup questions, try to help give feedback.

If you're lucky, and they post a commit to fix the issue, take a look at it and try to understand what it is doing. Over time, you'll start to get a feel for how open development works.

Translating documentation

And translations for the application itself in some cases, so that it can be ran and used in different languages. Human contributions to the language files of the application is what my project needs more than anything else, and if you do speak a non-English language fluently, I'd be very grateful! Here's how I've worded it for my project:

If you would like your language to be supported, please create a new pull request, or a new issue with your language file, using an existing language file as a template.

If you’re primarily interested in contributing by writing code, then you need to learn to code. Pick a language that a project you care about is written in and find the beginners tutorials.

You going to want to learn the basic tools the project uses even if you’re not writing code just so you can interact with the project’s community, so probably git, their bug tracker, and whatever they use for documentation and translation.

And don’t be surprised if it takes a while to get that first contribution in. FOSS developers tend to become grumpy with time and generally have high standards, and due to the massive amount of LLM slop being thrown their way these last few years have often become even more so. But your work will be appreciated if you stick with it.

If you want to work with projects that use C++, then by all means.

I do appreciate Mistral being European and maintaining better values than American competition. That's why I feel bad rating it so low. Although as you said, many tasks that LLMs are useful for don't require the cutting edge at all.

In fact, Mistral Vibe (I preferred the name Le Chat, but oh well) is probably more than enough for most people. It's just that availability of much better models makes it look worse than it is.

There's also the question of system prompts. ChatGPT has been known to work considerably better with a good prompt setup, to the point where it's not one of the worst chatbots anymore. I was only testing the defaults. It's entirely possible that setting the right system prompt for Mistral improves it noticeably.

At the end of the day, however, if I can run a small, open-weight model locally with comparable results - using a huge, cloud-based and closed model makes no sense to me. I'm not connecting to any cloud unless I really need to.

Alibaba is releasing Qwen 3.8 27B in a little over 3 hours from now, so I'll have to try that too. Though I'll probably wait for fine-tuned community releases to make a more detailed comparison.

Did you create an account with Mistral? Without it they only expose a low-end model to the web, which isn't properly declared last time I checked. Can be the free plan, I guess they just want to avoid their infrastructure being hammered.

Personally all those Silicon Valley pricks can offer whatever, I don't even have a Google account anymore and couldn't be happier with that. Not saying Mistral is optimal, but at least they're not part of the current US swamp. And I at least somewhat believe they actually read the damn GDPR. I'm totally fine with it not being the "cutting-edge", especially given the cutting-edge cuts straight through culture and people's lifes on purpose and with maximum brutality, and I seriously liked to see their image generator being way worse as well as Mistral refusing to open sites that banned it in the past (if they changed that behaviour please let me know). Not to mention their site isn't even remotely as infected with trackers and other junk as the others, at least that's what I've read (forgot to bookmark that hackers' analysis I've read a year ago unfortunately).

The data structures, macro basics, and core concepts are OK, but for example ADCs have literally 0 documentation at all, documentation about DMAs is barebones at absolute best, non-led PWMs is not good, etc...

Pretty much every single peripheral needs 5x-100x the documentation it currently has lol.

From my experience, they work fairly well when asked to explain code or error logs. For reference, I mainly tested Gemini and Deepseek. Apparently Claude is the best for coding, but I haven't really tried it yet.

Personally I'd avoid both Mistral and Copilot. I used to work on Copilot, so unfortunately had a lot of exposure to it. Tried Mistral for a while too.

These two are the worst LLMs I've used so far. Honestly probably worse than local Qwen with 27-35B parameters. I feel a little bad hating on a European company, but unfortunately it's way behind competition. At least from my testing.

Not saying they can't be useful for OP. Just that they're like a handheld circular saw*, a powertool better only used by people who already know sufficient basics about wood.

*where the salesman got rid of the safety handle and tells you to "just let it go on its own".

That said, you can absolutely teach yourself to code! Hell, if you figured out self hosting I'm willing to bet you've probably already got a handful of transferable skills.

In the interest of disclosure, I learned programming as part of my degree (took a Java class as a resume booster my first semester and ended realizing that I preferred coding to calculus and switched my major), so I'm not exactly a wealth of knowledge in the self study department. However I would recommend a course that has labs/ challenges/ projects to practice what your learning. Everyone learns differently, but I've personally found that I picked up programming concepts better when I got hands-on and applied them in my assignments.

The programmer's open secret is that generally once you know one programming language, it's pretty easy to pick up another. For example, Java and Python might write them in different ways, but at the end of the day a loop is a loop and as long as you know you need to repeat some code until some condition is met, it's perfectly acceptable to Google "loop in Python" or "[Java concept] in Python". In other words, it doesn't super matter which language you start with; just pick something and stick with it (bonus points if you can turn around and tinker with an OSS app with what you're learning).

Next, you'd probably want to keep learning the language (e.g., by solving LeetCode challenges), read more about programming and software development in general, as well as check out another course or a tutorial about the specific frameworks/libraries or even just paradigms that the project works with.

A good starting point is FreeCodeCamp on YouTube. Courses there are excellent for the proce of 0€. If you have difficulties with anything - you can always ask a chatbot. If you opt for courses on something like Udemy - you can always just ask the instructor and they'll usually respond within 2 hours, timezone permitting.

Can you look at the source code in the repos and make any sense of it?

Are there open issues connected to the repo, that you think you could make some headway with?

If you have 0 experience writing code, I think contributing to an existing project isn't a good way to start. Better to text a textbook and work through some exercises, and do a few small projects of your own, before trying to join an existing team where as a total newbie, you'll be a burden. Note that "small project of your own" can involve forking something that exists and making your own changes or additions. That doesn't require much attention from other people.

Then, contribute once you feel you are ready to contribute.