From my take, order is no meaningful:
- Installations are a pain point as people try to install on any Linux distro outhere.
- Building system for a few distributions is a pain point as libs like gstreamer remove capabilities, etc
- Build system for 3 installations is a pain point, they build a zipped archive, an appimage and an install script.
- Users email support for far more things on Linux than other platforms mostly because of installing on unsupported platforms. First level support just can’t say “we don’t support that, good luck”. Yes they are that kind of company/people.
- They are a small company and their upgrade and new feature cycle is rapid so dev days are full.
- They have a new product in the wings for CNC machines called Mill Mage and refactoring LightBurn code for reuse takes up even more dev hours/days.
- Some of their licensed libraries have weak Linux support AND they’ve been pushing Qt framework outside its capabilities requiring more OS/platform specific code eating more dev hours/days.
- Linux userbase is 1% of the Mac+Windows base( ~3,000 vs 500,000 ).
- Linux user base does not renew licenses at same rate as Mac+Windows with only 50% renewals for Linux licenses.
- Financially the Linux user base and license renewals don’t fund the platform effort.
- Apple has shifted to ARM and devs need to get going on native ARM while x86 emulation is still supported.
- Linux user base has been shrinking instead of growing while Mac+Windows base has been growing
- Something I’m missing but this is all I got right now.
A LightBurn for Raspberry Pi version changes things like the supported platforms list, installation, building and email support items. It would bump ‘Linux’ installation numbers but even a doubling would not change the financial aspect and likely not even with 100% annual renewals. I don’t know what the threshold would be(3x, 4x, ?).
For those wanting an rPi image and have a LightBurn license, I can build that up on a stock rPiOS image, re-image it and post to a google drive what’s been done via the x86 emulator for rPi.
It comes down to too few dev hours for too many platforms AND product lines(LightBurn and Mill Mage). Not enough revenue from Linux to justify keeping Linux devs on Linux and add Mac+Windows devs and just adding more devs isn’t something management wants to do with the current finance picture. I’ve offered things like dropping pain-point features, doing a LightBurn for Linux distribution(iso) or picking just one distro and the other items in the list still exist. Seems there is no acceptable answer/solution but ending development and support at the next release.