Modernizing LMMS with AI (like Claude)?

[kitzuki]

Hi everyone! First of all, thank you so much for all the hard work from the community and the developers who keep this project alive.
I have a question coming from pure curiosity and enthusiasm: seeing how AI tools like Claude are gaining a lot of momentum for writing and understanding code, how difficult would it be to use these technologies to modernize LMMS a bit and add new features faster?
I don't know how to code (if I did, I would love to dive into the codebase and contribute), so I'm curious about how complex it really is for those who actually know their way around programming. I recently saw someone build a DAW from scratch with features very similar to FL Studio or Ableton, and it got me thinking about everything that could be achieved here.
I know the community already contributes a massive amount, and maybe this sounds a bit silly, but I would love to see certain things integrated directly into the base program without needing external resources. Honestly, there aren't many of us using this DAW compared to others, and I don't just want to sit on my hands—I truly have a lot of love for LMMS and believe it has huge potential.
I'd love to open up a discussion: if you could add something new today, what would you put into LMMS? Personally, I would love easier built-in options to customize the UI colors to my liking.
Sorry for the long post, and thanks for reading!

[NorrisFByrd]

It is a fair question, and you are not silly for asking it. AI tools are genuinely good at reading a big old codebase and explaining what a piece of it does, which is half the battle with a project this old. The slower part is still review and testing, because a wrong change in audio code shows up as clicks or crashes rather than a typo. I like your own pick, by the way. Built-in UI color options would be a small, visible win and a good first project for someone new. If I could add one thing today, it would be a better piano roll, the note editing is the part I fight with most. What would you change after the colors?

[musikbear]

how difficult would it be to use these (AI) technologies to modernize LMMS

Very difficult!
Not only in respect to code, where security would be huge, but also ethically.
Where does a project go from man-made to machine-made?
We will have to face this conundrum when the next BOL is announced, and it is going to be difficult!
Then there is development..
In the Devs group the thoughts when it comes to creating new features goes from embracing AI to a total ban. Several will leave the project if AI-code is used.
And then the money.. Bringing AI into a project is nót free, and it is 'funny' you mentioning Claude, because Claude is the absolutely most expensive model of all.
The bottomline is that it is not going to happen.

[musikbear]

If I could add one thing today, it would be a better piano roll, the note editing is the part I fight with most.

I am sorry to hear that! What more precise is giving you problems?
Can you explain some usecases where editing is cumbersome and not intuitive?
We would love to hear those, because only through user-input can we improve LMMS!
Please tell us.