Why Linux is Becoming Impossible for Audio Developers to Ignore

[ruralaudio]

For better or worse, Linux can, and will be ignored until capitalism dictates otherwise.

You’d be suprised how many is dedicated to developing Linux. I have been really positive suprised, and it’s getting better every week.

But for most vendors: You are absolutely right. Which just makes it even more meaningful supporting the smaller vendors and solo developers actually supporting Linux.

[audiojunkie]

You can compile your plugins and you can test your plugins on the same machine. Just don't expect software that was compiled on a newer library to work on a distro with an older library. It's a simple rule that isn't that hard.

No, I need to compile on an older machine as you say, but the DAW I need to test in doesn't work on that machine.

Then you need to use a VM with the older distro and compile from there. Run your work system from a modern distro. Problem solved. It's really easy to do.

All this for an extra 5% in sales? f**k it, I'm out.

Do you even sell anything? I'm only aware of you having free software. Setting up a VM is standard operating procedure for development. You aren't supposed to be using the same system you compile on for anything else, in order to keep your packages clean. I'm sorry if you feel that way, but that's they way most development is done. I find it hard to believe you wouldn't know this, as an "experienced" developer. Maybe you need to look to another field? You obviously aren't happy doing the steps a normal developer would take.

Jeez, I wonder why devs avoid Linux and its community... Do As We WiSh Or YoU'rE nOt A rEaL dEv

I'm sorry that people might feel this way. Many of Figbug's constant complaints about Linux come from not doing things the proper way with Linux. Things like backwards library compatibility and how it works in Linux, using VMs and such for creating clean packages, and running a modern distro for normal use are all pre-101 essentials. Lately, he's been making a ton of complaints about Linux, when it's starting to look to me like his biggest problems are coming from his development environment and his attitude. He's provided Linux native binaries for years, which I think is awesome, but he's clearly not happy with doing it. No one wants someone to be unhappy. But no one wants someone constantly "raining on parades" either.

[FigBug]

I'm sorry that people might feel this way. Many of Figbug's constant complaints about Linux come from not doing things the proper way with Linux. Things like backwards library compatibility and how it works in Linux, using VMs and such for creating clean packages, and running a modern distro for normal use are all pre-101 essentials. Lately, he's been making a ton of complaints about Linux, when it's starting to look to me like his biggest problems are coming from his development environment and his attitude. He's provided Linux native binaries for years, which I think is awesome, but he's clearly not happy with doing it. No one wants someone to be unhappy. But no one wants someone constantly "raining on parades" either.

I'm fine doing Linux builds for my open source stuff, and will continue to do so. They are just hobby projects anyway. I do builds on GitHub Actions ubuntu-latest. If that doesn't work for people, they can build themselves. If something else doesn't work, they can fix it and open a pull request. If that is above their skill level, then they are out of luck. I don't owe them anything, they didn't give me any money.

As for commercial projects, I will be steering my contracting clients away from releasing Linux builds in future. A lot of them see threads like this and think Linux is an easy win that will get them extra sales. I think a lot of the Linux evangelists aren't being honest with the amount of effort it takes to port a plugin to another platform (even if using a cross platform framework like JUCE) and the ongoing support costs of an additional platform.

I think there are selfish motives behind the advocacy. Linux users want more plugins, so they make promises to developers that the market can't back up.

[audiojunkie]

I'm sorry that people might feel this way. Many of Figbug's constant complaints about Linux come from not doing things the proper way with Linux. Things like backwards library compatibility and how it works in Linux, using VMs and such for creating clean packages, and running a modern distro for normal use are all pre-101 essentials. Lately, he's been making a ton of complaints about Linux, when it's starting to look to me like his biggest problems are coming from his development environment and his attitude. He's provided Linux native binaries for years, which I think is awesome, but he's clearly not happy with doing it. No one wants someone to be unhappy. But no one wants someone constantly "raining on parades" either.

I'm fine doing Linux builds for my open source stuff, and will continue to do so. They are just hobby projects anyway. I do builds on GitHub Actions ubuntu-latest. If that doesn't work for people, they can build themselves. If something else doesn't work, they can fix it and open a pull request. If that is above their skill level, then they are out of luck. I don't owe them anything, they didn't give me any money.

As for commercial projects, I will be steering my contracting clients away from releasing Linux builds in future. A lot of them see threads like this and think Linux is an easy win that will get them extra sales. I think a lot of the Linux evangelists aren't being honest with the amount of effort it takes to port a plugin to another platform (even if using a cross platform framework like JUCE) and the ongoing support costs of an additional platform.

I think there are selfish motives behind the advocacy. Linux users want more plugins, so they make promises to developers that the market can't back up.

Well of course there are selfish motives behind the advocacy! And we DO want more plugins! But aside from noting the constant Linux adoption and growth rates, and trying to educate users (and developers) about Linux and how it works, there's not much more that we can do but advocate--except to purchase. And I purchase everything I can afford, as much as I can afford. I simply can't speak for others and their choices. No promises are being made about anything.

[Tiles]

I'm sorry that people might feel this way. Many of Figbug's constant complaints about Linux come from not doing things the proper way with Linux. Things like backwards library compatibility and how it works in Linux, using VMs and such for creating clean packages, and running a modern distro for normal use are all pre-101 essentials. Lately, he's been making a ton of complaints about Linux, when it's starting to look to me like his biggest problems are coming from his development environment and his attitude. He's provided Linux native binaries for years, which I think is awesome, but he's clearly not happy with doing it. No one wants someone to be unhappy. But no one wants someone constantly "raining on parades" either.

I'm fine doing Linux builds for my open source stuff, and will continue to do so. They are just hobby projects anyway. I do builds on GitHub Actions ubuntu-latest. If that doesn't work for people, they can build themselves. If something else doesn't work, they can fix it and open a pull request. If that is above their skill level, then they are out of luck. I don't owe them anything, they didn't give me any money.

As for commercial projects, I will be steering my contracting clients away from releasing Linux builds in future. A lot of them see threads like this and think Linux is an easy win that will get them extra sales. I think a lot of the Linux evangelists aren't being honest with the amount of effort it takes to port a plugin to another platform (even if using a cross platform framework like JUCE) and the ongoing support costs of an additional platform.

I think there are selfish motives behind the advocacy. Linux users want more plugins, so they make promises to developers that the market can't back up.

I agree with pretty much all of this.

I think the biggest misunderstanding in these discussions is that people confuse "a Linux build exists" with "Linux support is solved".

A binary is only the first step. The real cost comes afterwards: testing, reproducing bugs on different distributions, dealing with different library versions, installation issues and support requests. That ongoing responsibility is what makes an additional platform expensive.

For open source projects the equation is different because users can build from source, fix issues themselves, contribute patches, or accept that their specific setup is not supported.

That said, even open source projects have limits. The limiting factor is often maintainer time. A contribution does not automatically reduce workload. Someone still has to review it, understand it, test it and maintain it in the future. A random CMake change or platform-specific workaround can easily create more work than it saves.

[audiojunkie]

For better or worse, Linux can, and will be ignored until capitalism dictates otherwise.

You’d be suprised how many is dedicated to developing Linux. I have been really positive suprised, and it’s getting better every week.

But for most vendors: You are absolutely right. Which just makes it even more meaningful supporting the smaller vendors and solo developers actually supporting Linux.

Agreed. The biggest problem right now with Linux in the audio world is the "chicken and egg" problem. Linux users want more plugins, but developers don't want to support a small group. But then, those who want to use Linux won't because there aren't enough plugins. It's a cycle that will go round and round until something changes. The good news, is that whereas Windows and MacOS adoption is dropping, Linux adoption is growing. But it's growing slowly. So year after year, there are more plugins. And year after year, there are more users. But it will take a long while before the situation reaches critical mass to the point where the bigger businesses will want to support Linux. So, in the meantime, we continue to help, and educate, and advocate.

[ruralaudio]

Jeez, I wonder why devs avoid Linux and its community... Do As We WiSh Or YoU'rE nOt A rEaL dEv

You can find 29 interesting interviews with devs who doesn’t avoid Linux here: https://linuxaudio.dev/

[Tiles]

Wow, 29 developers. I guess Linux audio is saved then and doesn't need any other developers anymore.

Jokes aside, those interviews are a great showcase of people who decided that Linux support makes sense for their projects. But they don't change the fact that every developer still has to make their own cost-benefit calculation.

[DRMR]

But they don't change the fact that every developer still has to make their own cost-benefit calculation.

It is the same for other platforms no?

MacOS particularly has been mostly cost with no benefit for me .. so far.

[ruralaudio]

Wow, 29 developers. I guess Linux audio is saved then and doesn't need any other developers anymore.

Jokes aside, those interviews are a great showcase of people who decided that Linux support makes sense for their projects. But they don't change the fact that every developer still has to make their own cost-benefit calculation.

Yes, agreed.

To me it’s very much build on values, like the developers who use Linux themselves, like it for music production and want to support the community, or the developers who want to give their users the free choice of where to use their plugins (Thanks!). Several also specifically mentions the low effort it takes them, because of their framework (mostly JUCE).

I see no one mention any kind of financial benefit as the primary reason, and if that’s the only thing driving you as a developer, then Linux is definitely not the way to look - for now.

We should also add, that you can’t expect to sell copies, just because you have a Linux version of your plugin. People need to find it, trust it and like it enough to pay the money. I have 283 native Linux plugins installed, many of these are paid plugins of very high quality (just got a couple of DDMF plugins, who would have thought they made my SuperPlate replacement!!!). I don’t need more plugins, so I won’t go out and buy anything that is released for Linux, just because I’m a Linux user. It takes time to enter a new market, so I would advise for all developers to find an incentive that is not money at first. I’m pretty sure it will turn into money down the line though, but only time will tell for the individual developer. No one is secured sales on Windows or MacOS either and companies come and go.

[Tiles]

But they don't change the fact that every developer still has to make their own cost-benefit calculation.

It is the same for other platforms no?

MacOS particularly has been mostly cost with no benefit for me .. so far.

Yup, of course. This applies to other platforms as well.

The difference is that Windows and macOS usually provide a much larger market and audience for audio software, often with less additional effort required to reach those users.

And if macOS does not provide enough value for your specific project, then it is absolutely reasonable to reconsider whether continuing to support it makes sense. That is exactly how platform decisions should be made.

[keys_au1]

just go and hire out several online "influencers" - that will certainly increase the profile and perhaps even uptake

[cpr]

And in this particular case, Waveform 8 specifically, released in 2017, was notorious for having problems with Missing or incompatible library versions (glibc, libstdc++, Qt-related dependencies,

<sidetrack>Sorry for being pedantic, but Waveform was always built on JUCE, not Qt.</sidetrack>

[audiojunkie]

And in this particular case, Waveform 8 specifically, released in 2017, was notorious for having problems with Missing or incompatible library versions (glibc, libstdc++, Qt-related dependencies,

<sidetrack>Sorry for being pedantic, but Waveform was always built on JUCE, not Qt.</sidetrack>

No, you are correct. I was totally going off of memory, and memory failed me.