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.
Tell people that unsupported platforms are unsupported.
I manage a devops pipeline that compiles and packages to Debian/Ubuntu, RHEL and its variants, appimage, and flatpaks. Using QT6. It’s a solved problem.
Again, unsupported platforms are unsupported.
Don’t add features until you have the time to do it right with maintaining current features.
Pushing QT out of its boundaries is usually a sign of a programming anti-pattern.
Userbase is less than 1%. But, according to the author, they have a ton of Linux users.
I was pretty cool with everything until I read through more comments. And more and more I get the feeling that this decision was forced by a few people to the chagrin of the development team.
Some of the responses are bordering on, you don’t matter.
But, how many of the libraries in use by LightBurn are open source with permissive licensing? It feels really hypocritical to be told on one hand that my silly little hobbyist OS isn’t good enough for real work and that I should just nerd up and buy an operating system that treats me like an idiot at best, and a commodity to be milked at worst; and then have that same company gleefully use the contributions that I personally make to the Linux kernel, OpenSSL, and other core technologies.
I was happy to be quiet and figure out a new way of doing things. But after reading some of the downright hostile responses from (presumably) employees at the company on other forums, Hackernoon, etc., I’m starting to re-think my conclusion that this was a business decision and not the result of a devops person being in over their heads and telling everybody else that Linux is too hard, too fragmented, and has too small of a userbase.
Thanks for this @DougL , this is a really great summary.
To expand on this, and I’m not trying to put down anyone’s ideas, but I want to reiterate for those reading this great summary: There’s not a single one of the provided options that we haven’t actually considered (or tried!) during the nearly 2 years we’ve been considering this. I promise everyone here that we truly tried to avoid this. My only real regret from that process is that we didn’t just go ahead and make the decision final and public several months ago. Instead of continuing to try to find a way to avoid it and finally being faced with an impending deadline which we couldn’t move past without deciding that Linux ends with v1.7
Not sure where you are getting that. As stated, we have only a little over 3000 active Linux users. Certainly not a “ton” compared to the rest of our userbase.
Nope. Even our most die-hard Linux devs on the team have all agreed this was the only way to go.
This one is at least partly because Qt relies on other libs for some things. GStreamer is used on Mac and Linux, and Windows goes native to the hardware, but poorly, so we’re writing our own platform specific libs to address this. Qt 6 broke camera support entirely on MacOS for quite a while, and it’s only been in recent Qt 6 versions that it’s come back and is working well.
If we adopt Qt 6 there’s a bunch of code we need to refactor and it drops all Windows versions before 10, and a large swath of MacOS versions as well. We’re toying with building on Qt 6 and Qt 5 to alleviate this. It won’t be a small amount of effort.
The camera issue is only one thing. We rely on a bunch of libraries (licensing, Qt, BSL support, PoTrace, LibUSB, OpenCV, and more) and every one of those has its own dependencies. On Linux in particular they often mismatch, so we have to pick one. Qt wants a different version of LibSSL than the licensing system we use, so certain functions in LightBurn that use HTTPS fail. Building a partial version or using a different packaging method doesn’t solve this.
Dropping features or leaving bugs intact isn’t really an option because users complain or submit bug reports about it, and it burns time.
I’d be curious which forums / responses you’re referring to, but I doubt they’re us. Justin and I have both posted on the Hackaday thread with our own names, Joe and I have replied to a few Reddit comments, and in all cases we’ve been civil.
We know people are upset, and you have a right to be. We haven’t locked or censored this thread because I personally think you all have a right to complain, and we chose this, and should absolutely hear those complaints. There’s a lot of back chatter here at work, and it’s pretty empathetic - I know you can’t see that, but we’re not making fun of people. We’re commiserating with you and trying to help you understand why and how we got here.
Meerk40t is the closest I know of. I think it only xlates Ruida commands from LightBurn for Ruida over to M2Nano and GRBL controllers. I think.
It would likely be the place to start but then again, I just found this which resides in the ‘see if there’s an Inkscape plugin for that’ department:
It ran through my head someday LightBurn might work as a plugin for Inkscape and provide far less of the things which might be too OS specific.
The Linux community will eventually come up with a nice solution without LightBurn around to solve the problem. Yes I know, LightBurn versions upto 1.7 will not stop working, they’ll just be no more support/development. Linux users are interesting bunch. I’ll take a project with active development over others if it not only gets the job done but the devs are adding and supporting the code.
if the shoe doesn’t fit great and there is no way I’m talking about ALL users but the general population which I have dealt with supporting and helping the majority never got OS 101 classes and just don’t know what’s going on. Some corporations do a better job with educating OS users but the majority of the consumer public gets nothing but maybe a class on “The Word”, “The Powerpoint” and “The Excel” in high school.
Again, a generalization regarding those who actively seek out to learn Linux or another OS and those who take what arrived and try to click their way through to the internet, etc. Sorry to have hurt your feeling. I was not talking about you.
You didn’t “hurt my feelings”, you just blatantly and outright insulted the majority of Lightburn users. Reading the responses in this thread, on Reddit and on Hackaday really shows the sense of self-importance the Linux users of Lightburn have, despite accounting for a tiny amount of the actual user base. The disingenuous arguments that Linux users are a fast and quickly growing group is getting tiring.
To top it off, people on Hackaday are throwing around accusations of an imaginary move to a subscription based model, that they wanted to do this to fire their Linux developers, that the developers are lazy/stupid or even mismanaging the business and application.
We have been told that the decision has been made, that they have already considered all the alternatives, that they aren’t going to reverse their decision, and yet you continue to argue about it, hoping to achieve what, exactly?
All this talk of which distros will work with LB, which versions, WMI, Vine, etc and the issues each raises that must be negotiated in order to run specific programs are a good example of why folks like myself stay as far away from Linux as possible. I’ve never had a major problem of installing or running a program in Windows. If Linux users want make it more popular with the general public, they should use their influence to make it easier to use and install.
Well, hackaday and even /. have long ago been over run by non geek types so what goes on there should be considered like it were Twitter, Facebook, etc. And ya, hackaday posts went off the rails quickly and somehow did not ruffle a single of my feathers. And I except and understand where LightBurn is coming from and somehow you missed where I was trying to summarize to help others know the complexity of the task they have and have been dealing with.
Sorry to have offended all of those windows user and kinda wondering why they are even HERE. The many I’ve worked with in software development, UNIX, would be here to gloat.
I’m enjoying hearing all of the business users who have switched to Linux on the floor. I will work on getting LightBurn running on the rPi for those who want to quickly expand using version 1.7 for a long while until another option gets developed.
I am a recent convert, and had some Linux-related difficulties getting started. I got wonderful support from the community.
I am also a software developer and have worked extensively in both Windows and Linux for at least 30 years. I am a very loyal Linux user. I don’t touch Windows or Macs (especially Macs) unless I absolutely must.
But I understand the issues clearly. It seems to me that the company needs to do what it needs to continue producing product.
PC’s are really cheap. (See Beelink, etc.) A Windows license is (for me) distasteful but endurable. I already have one that I run in a ProxMox VM just to use when nothing else will do. I will eventually buy another small box and Windows license to run LightBurn 1.8 when the time comes.
I know that there are many users in situations different than mine. The cost of LightBurn, Windows, and the hardware to run them is very small compared to my investment in tools and materials, though. And, blessedly, I no longer have to fight middle management and the corporate machinations they enforce. (And so glad not to be one of them anymore).
I will do what it takes to continue using LightBurn’s products as long as they continue their good work. I much prefer that they continue to exist and succeed - even if not on my preferred platform.
This is horribly disappointing, especially when combined with the end of those computers that are still running strong with LINUX but are incompatible with Windows 11. What a scam that is. I am relatively new to Linux and had just gotten it all where it was working great: MPLAB X IDE for my microcontrollers, Inkscape for my drawing, Audacity for my audio work, GIMP and RawTherapee for my photography, and, of course, Lightburn for my laser work. Now I will have to scrap two computers just so I can get Windows 11 to work. Pretty disappointed.
Or you can continue to use LightBurn 1.7.x and below indefinitely on Linux - if there’s a feature or bug fix after 1.7.x that really shakes the earth, then at that point, you may want to consider one of the future-supported OS’es.
count me in on this effort. Maybe I can get a camera to finally work over ethernet. And maybe I won’t have to have a windows computer physically attached to my laser in order todo so. Right now my laser sits aprox 20ft from my desk where I design. USB doesn’t work over that distance, therefore no camera. And yes I’ve tried various USB extenders and USB to ethernet options. A rPI attached to the laser running a VM with LB on it, might just be the best solution of all with Remote Desktop access and shared directories. I happen to have an extra rPI 5 sitting here waiting.
Sad, but I get it. There aren’t a lot of Linux desktop users and we tend not to spend a lot on software. Just how it is. Could we trouble you to please try to make sure it will work on Wine or similar? I mean, we can run AAA games on it, Lightburn can’t be more demanding than those. Though that sort of thing is not always the issue.
Unfortunately it’s not that easy. And this is not just me trying to brush off the question with hypotheticals, I’ve spent quite a lot of time researching WINE as an option before we made the call about Linux.
There are two things that prevent us from making WINE work…
WINE doesn’t support WMI on 64bit applications. WMI is required for our licensing to work and without it that system is unable to link your license to that specific machine. The failure mode of WMI not being available is that the licensing system will simply not allow you to activate, which would prevent you from being able to run the application (even in trial mode).
WINE doesn’t have very good support for USB devices beyond serial and a few HID devices (mouse, keyboard, gamepad). This would let some gcode devices that only use serial work (though the user would have to manually map the Linux serial port to the “Windows” serial port. But many DSP and basically any galvo devices we support would not work.
This is what I do, Linux on the cheap NUC attached to the laser, and MacOS (or sometimes Linux) laptop for design work.
Hmm, I wonder if this is a possible middle ground for LightBurn… can we have a Linux ‘server’ application which doesn’t have the graphical features they are struggling to make cross-platform, and instead is just about driving the laser…