Physics, coding and black metal.

Vyssiikkaa, koodausta ja bläck metallia.

Apparently also politics when it doesn’t devolve into screaming into aether.

  • 0 Posts
  • 17 Comments
Joined 1 year ago
cake
Cake day: June 13th, 2023

help-circle


  • I guess that list could be helpful for some, but for me (and IMO, music production in general), it’s woefully inadequate to the point of hilarity.

    Pro audio has been a complete mess in Linux for ages, and it’s not even close to where it should be in order to be generally usable. Every 7-8 years or so when my old music computer starts to die I try and check if it has made substantial improvement, but apart from Musescore actually being good, it is hard to find any tangible progress from 15 years ago. Pipewire gives me some hope, but it’s far from production-ready in Pro audio world. And I’m not really going to get rid of all the VST stuff I’ve bought in the last 20 years (all of which still works out of the box on a new computer!)

    In addition, making music is the one hobby I have to get me away from tinkering with computers. I am not interested if I could make my Linux setup equally good if I spent weeks tinkering on it, when it’s literally easier for me to work for a week and buy a Macbook Air (or whatever crappy windows PC), where I get all of my old work ready for action in under a day, and I can trust that everything I do will just work, and work well at that. And it does it while allowing me to work remotely with other musicians since we can all use the same stuff.

    I’m pretty sure I’ll be in my grave before FOSS Pro Audio ever gets there, unfortunately.

    Edit: Ironically, the one FOSS thing I would love to use in my audio stuff is Guitarix, which is then the thing that doesn’t interop well with anything else. And I would love to have easy way to do all that I do on (Win/Mac Os) on Linux, but 20 years of disappointment is pretty hard to overcome at this point.








  • I am placing careful (nevermind that, this seems very nice) interest in this.

    Few questions (since I’m on mobile, and it’ll take me a while to get back to my computer to find out for myself):

    • How does managing sieve work with this?
    • Does it play along with rspamd?
    • Is it tested on x64_64 only?
    • Does it support PGP, can email be encrypted-at-rest using this?
    • Is there a way to run this behind a reverse proxy that handles the certificates? I’m not too keen on dealing with two separate sets of those in separate places.
    • Does this require LDAP?

    If missing, are those on roadmap?




  • I’d recommend going with the vanilla Raspberry Pi OS then. Sure, it’s not as lightweight as one would usually hope from a SBC OS, and it has the usual problems that apt has, but it general, it works. It has the firmware stuff ready, so no hassle with that. It has device trees set up in a generally-usable way from the get go, etc.

    I didn’t go that route myself and spent couple of days trying to get hardware acceleration to work where I wanted with the VideoCore chip, after which I gave up. VideoCore just isn’t that well supported by the general software stacks, but this was a year or so ago, so it might’ve improved.

    Also note that this is all RPi4 specific. Older RPis work quite well.


  • RPi uses a lot of software hacks to get its low-cost hardware running. It is certainly doable on other distros, but using anything but the official ones on RPi is asking for trouble, and you better know how to deal with device trees, etc.

    If you want SBC that is more standard-compliant and has better mainline driver support you should look at e.g. Pine64’s SBCs, such as RockPro64.


  • I’m glad people want to conribute. But everybody has ideas.

    You have to realise that “contributing an idea” for developers without any of your own work sounds awfully lot like asking people to work for you for free. That is not going to make you popular in FOSS circles. Most FOSS projects are undermanned as-is and maintaining is a thankless task.

    Like others have said, the best way would be to just start coding it yourself. People see you put work into something, they can get more excited about it. Advertising is fine, but unless you have something to show, it’s unlikely to attract much attention.

    There is a reason “a platform where regular people can suggest FOSS ideas to developers” doesn’t really exist. We have our own ideas, which take more time than we have already. A platform such as that would likely be full of people throwing out ideas and close to zero developers willing to work on them.