Developing Windows only plugins

[Urs]

Zebra had several complete UI overhauls, it went from PowerPC to Intel (and now to Arm), from 32 bit to 64 bit, from gcc to MS Compiler to Clang, from C++99 to C++2017, from hand coded abstraction of filesystem to boost to std, from AU to VST2 to AAX to VST3 to AUv3 and so on and so on...

I'm curious as how you managed to use std::filesystem on macOS.
XCode complains that using it requires a 10.15 deployment target, which is not something most could afford now. It's the same with variant and other handy C++17/20 features unfortunately.

Not sure we use std::filesystem. I did see std paths and files and stuff... hardly have anything to do with it.

[S0lo]

If you want to make a living having fun, I personally think it is possible with windows-only music software but could be wrong. Making money is maybe a different skill set than coding.

But if you do lots of non-fun stuff just to make a modest living, that would suck. If non-fun is programming non-Apple, then stick with Apple only. If non-fun is programming Apple, then avoid Apple. Life is too short. Enter a market where the work is "interesting and fun" fully aware it won't make the same paycheck as some deadly-boring but higher-paid non-fun specialty, then torture yerself half of the work year doing a platform that ain't fun. How does that make sense?

Some folks are platform-agnostic and the "fun" is working on the plugin or the music app, and the platform doesn't matter. Different strokes.

It seams what we value changes over a life time. For some people, the fun becomes the money. For others it remains programming. For some it's a combination. For some, its the peoples attention and followers. Some value intellect, others professionalism. So the "fun" part becomes related to those things. Some may value all with different priorities.

Allot of these things are subconscious and sometimes very hard to put your hands one. Our inner motives is a fascinating mystery. Yet we're usually unaware of how they work.

[AndyMusician]

But if you're super partisan about Mac OS, then please, do not make a plug in for Macs, then spend all your time here posting about how much you hate Mac OS. It just gets really old to sift through all that nonsense.

I also don't want to develop for Linux and Android. Did you notice no one is bothered by that?

Most linux (or even android) users do not ask because we know that the answer will be "there is no market, so no" and honestly we have many good plugins and DAWs to work with like U-He stuff, Surge, Bitwig, Waveform etc. Also most windows plugins work fine in linux through wine and hence its easier to just use that.
Meaning we are just used to developers saying there is no market and there wont be linux support soon, and it kinda makes sense why that is the case, but like the other person pointed out, 5% market share isn't that bad. I'm also sure that most people producing on linux have atleast 1 plugin from U-he, so there is a niche. Also smaller markets are easier to get into as competition is pretty less.

The problem with Linux is making sure that the plugin is compatible with the different flavors of Linux. I've read stories of apps running well on Ubuntu but not on Debian, or running well on Linux Mint but not Xubuntu, etc. Looks like a big puzzle to me.

[Paree]

But if you're super partisan about Mac OS, then please, do not make a plug in for Macs, then spend all your time here posting about how much you hate Mac OS. It just gets really old to sift through all that nonsense.

I also don't want to develop for Linux and Android. Did you notice no one is bothered by that?

Most linux (or even android) users do not ask because we know that the answer will be "there is no market, so no" and honestly we have many good plugins and DAWs to work with like U-He stuff, Surge, Bitwig, Waveform etc. Also most windows plugins work fine in linux through wine and hence its easier to just use that.
Meaning we are just used to developers saying there is no market and there wont be linux support soon, and it kinda makes sense why that is the case, but like the other person pointed out, 5% market share isn't that bad. I'm also sure that most people producing on linux have atleast 1 plugin from U-he, so there is a niche. Also smaller markets are easier to get into as competition is pretty less.

The problem with Linux is making sure that the plugin is compatible with the different flavors of Linux. I've read stories of apps running well on Ubuntu but not on Debian, or running well on Linux Mint but not Xubuntu, etc. Looks like a big puzzle to me.

I haven't experienced any such issues personally, but you can try to have a beta build for Linux and if it works, great. Else most plugins designed for windows can be bridged using wine+LinVST and it works.

[mystran]

The problem with Linux is making sure that the plugin is compatible with the different flavors of Linux. I've read stories of apps running well on Ubuntu but not on Debian, or running well on Linux Mint but not Xubuntu, etc. Looks like a big puzzle to me.

More often than not these types of thing comes down to what libraries (and/or what versions of libraries) are installed on the system. In theory, if you statically link everything (and don't make assumptions about filesystem organization, which can also vary) then things should work more or less the same everywhere. In practice, the OSS ecosystem is not really designed for statically linking everything, because as far as the OSS ecosystem goes you are supposed to distribute your stuff as source code instead and let the distributions build the binaries for you, so they can then let their package managers figure out what other stuff needs to be installed.

The Linux-kernel system-call API is very stable and the X11 protocol is also very stable, but more or less anything else is up to the distribution maintainers and/or the user (since some distributions give the user a lot of freedom in choosing what they want and don't want), so if you want to release Linux binaries that "just work" everywhere, you pretty much have to throw the usual workflows of developing (open-source) Linux-software right out of the window and instead build all you actually need directly into the binary itself.

Now, there is a rather large additional body of software that is usually installed on most systems, but that's the issue: there is no true standard, so you're essentially rolling dice.

[wrl]

The problem with Linux is making sure that the plugin is compatible with the different flavors of Linux. I've read stories of apps running well on Ubuntu but not on Debian, or running well on Linux Mint but not Xubuntu, etc. Looks like a big puzzle to me.

Haven't heard this one in a while – the waters have been pretty calm lately. I do my Linux builds from a Debian 10 VM and that's sorted out all of my compatibility issues.

[Benutzername]

I'm using several Arch and Debian based distros and I had very little problems with closed source applications in the last years and I'm using a lot of apps that deeply integrate into the desktop like Zoom, Teamviewer, Teams, Seafile and so on. Also development systems, CAD and audio applications like REAPER, Bitwig, Waveform and many plugins run perfectly fine natively. Since Steam added proton my games library is five times as big as on MacOs Mojave and twenty times as big as on Catalina (no 32 bit) and I have yet to find a game that doesn't work as good as with Windows on the same machine if it is officially supported.

The main problem with Linux is that failing developers don't know what they are doing. I don't mean that in a disrespectful way but there seems to be a schema in many failed releases: Most of the time the main developers don't do the port themselves and often they don't even use Linux on their computers at all. Usually these companies get a new employee, trainee or student apprentice and they do the port as a project. This project gets done up to a state where it is barely running on the computer of the student and then it's released into the wild with the same price tag as the Windows and Mac versions. This is not a problem for a first alpha release but then the project is considered done and never touched again. This is where it starting to fall apart as every new platform needs time to mature and this is no different than porting a Windows app to MacOs or vice versa.

Linux apps often don't get this time because no one's spending money for it but who's going to spend money for a buggy piece of crap, especially if it's really expensive? Linux users have been burned many many times by 'commercial' developers who release paid software and then never update or fix it again so they will watch out carefully where they put the money. However, if they decide to spend the money then Linux users usually spend most (eg on a Humble Bundle). This means that developers will need some endurance to get up and running on Linux. Linux users pay for software if it's worth it but they won't buy everything blindly that you throw at them, no matter how good your reputation is on the Windows or Mac side.

[SampleScience]

Due to compatibility problems with the exported plugins made with the software that I use to create plugins, I've been forced to sell Windows only plugins since early December 2020.

After a month of selling Windows only plugins, I can say that you can make a comfortable living out of it for sure. I'm making around 75%-80% of the income I used to make selling macOS/Windows plugins and it was already pretty high. So I'm good...

I now get almost 0 support requests and I can focus on creating new plugins instead of updating my plugin catalog to maintain compatibility with macOS new versions. It's a relief and it saves me a lot of time I can invest in more fruitful work.

Because of that, my income is slowly growing to what it used to be and I'm confident I'll make as much as I used to make in the upcoming months by converting some of my plugins to Kontakt.

[licasto2]

Please, please, don't be THAT one stereotypical butthurt Mac guy. It's a fact that Mac users pay more and ARE willing to pay more.

It's because we CAN pay more;-) Please, please, don't be THAT one stereotypical butthurt PC guy that's broke and has to "roll his own". See how dumb that sounds? It's true , but it's still dumb.

I'm not broke and I can have a better, more performant machine for the same (or less) amount of cash Apple is asking for their "Pro" range.

Hey, if Logic was still available for Windows I’d probably be in the same boat as you.