Angry Birds, flash mobs, Russell Brand, fidget spinners. All of these were virtually unavoidable in the previous decade, and yet, like so many popular trends, have now largely faded into obscurity. But in a recent announcement, the developers of LightBurn have brought back a relic of the past that we thought was all but buried along with Harambe — popular software not supporting Linux.
But this isn’t a case of the developers not wanting to bring their software to Linux. LightBurn, the defacto tool for controlling hobbyist laser cutters and engravers, was already multi-platform. Looking forward, however, the developers claim that too much of their time is spent supporting and packaging the software for Linux relative to the size of the user base. In an announcement email sent out to users, they reached even deeper into the mid-2000s bag of excuses, and cited the number of Linux distributions as a further challenge:
The segmentation of Linux distributions complicates these burdens further — we’ve had to provide three separate packages for the versions of Linux we officially support, and still encounter frequent compatibility issues on those distributions (or closely related distributions), to say nothing of the many distributions we have been asked to support.
We’re not sure how much of their time could possibly be taken up by responding to requests for supporting additional distributions (especially when the answer is no), but apparently, it was enough that they finally had to put their foot down — the upcoming 1.7.00 release of LightBurn will be the last to run on Linux.
To really add insult to injury, LightBurn is paid software, with users having to purchase a yearly license after the time-limited demo period. Accordingly, any Linux users who recently purchased a year’s license for the software can ask for a refund. Oh, and if you’re holding out hope that the community can swoop in and take over maintaining the Linux builds, don’t — LightBurn is closed source.
While there are open source projects like LaserWeb that can be used to control these types of machines regardless of what operating system you’re running, losing LightBurn on Linux definitely hurts. While we try not to put our stamp on closed source proprietary software because of situations exactly like this one, we have to admit that LightBurn was a nice tool, especially when compared to the joke software that many of these lasers ship with.
The developers end their notice to Linux users with what seems like a particularly cruel kick while they’re already down:
Rest assured that we will be using the time gained by sunsetting Linux support to redouble our efforts at making better software for laser cutters, and beyond. We hope you will continue to utilize LightBurn on a supported operating system going forward, and we thank you for being a part of the LightBurn community.
So take comfort, Linux users — LightBurn will emerge from this decision better than ever. Unfortunately, you just won’t be able to use it.
I just want to know: why is the move sunsetting linux support rather than doing a flatpak?
Because it’s not just a packaging issue. We’ve also had lots of suggestions regarding how packaging works, everything from deb to flatpack to AppImage. We currently ship 3 different binaries as well, more than for any other platform.
So stop and ship one?
Ubuntu is the biggest distro, just stick with that.
Having watch other projects struggle with this recently, appimage is a trap. It subtlety ties you into supporting every distro while telling you you have to support none and not giving you the tools to manage it.
Deb and apt are tried and true but I get how depending on your application and what you plan to support this could be a burden.
So really, I see why flatpak is the only option for a lot of applications.
I think us FLOSS devs get really frustrated by these moves because we invest our time in the code that is used to build so many apps. It hurts when the things we want to use only run on other platforms.
Funny. So many open source applications support Linux and all its distros, plus Windows and Mac, while making practically no revenue, yet LightBurn can’t figure out how to continue supporting Linux at all. Don’t mind me, I’m just a senior dev with a PhD…. I certainly am incapable of understanding this decision with the excuses given.
Sounds like they haven’t got their automated build pipelines down. Packaging should be automated.
Sounds like bad product management.
How would you do good product management in this case?
Supporting multiple Linux distro, and I mean supporting rather than just throwing the packages out there and hoping for the best, requires you to actually test each configuration and variation and troubleshoot bugs that exist on Distro X version Y using library version Z with a desktop environment A…
It quickly becomes a combinatorics problem where the amount of working hours you spend on supporting a very small portion of your users becomes completely disproportional to the money they’re paying you for it.
So you narrow down your supported platform configuration to just one, e.g. “flatpak on stock Ubuntu”, and if people with other configurations want to make it work, they can figure it out. That’s how Linux works.
Which means flipping the bird to everyone else and forcing them to run the app in a VM/sandbox and use up more memory/cpu/disk by duplicating the system ABI. You might as well be running it under WINE at that point and just save your efforts on developing the Windows version.
Release just for one distribution that can run as a VM, problem solved.
If you choose Alpine (musl differences, yadda yadda yadda, I know), chances are that it would even run faster than the one running natively, with minimal overhead in resources needed.
It is just a packaging issue, or you just don’t know what you are doing.
Well my opinion is you should create a live disc / bootable image / frugal live install with persistence of your own linux OS distro specific to your product. Then users could install it, tun it or use it as a VM or live Distro etc…
Fantastic idea! Or just dockerise it.
Windows is not an option wine and playonlinux are? Windows is not really supported by many SBCs, it is not supported on any old computers. Windows 11 is just a big web of hacks and spyware. Do nothing but test run on playonlinux once in a while. make no promises just try not to prevent it. This is what LaserGBRL does.
Why don’t keep a support for only one platform?
maybe they lost their linux devs and don’t want to pay to replace them
That’s not the case.
So you’re firing your Linux devs?
Really? Why are you making stuff up? No one got fired over this or let go. We have one dev team that works on a ton of different stuff. Several of us, myself included, have spent a large amount of time trying to resolve Linux issues.
Can’t reply to your comment response but it sounded like you had Linux devs but your dropping support so I was asking if they were loosing their jobs.
@James Nope. In fact the person who ultimately made this call is the first dev to get hired on past it being a single developer project. And the first things he worked on was trying to fix some Linux issues.
I get that people are upset at the decision, but it’s really disheartening to have a ton of accusations thrown at us. We’re not a big company.
My inner cynic says its more likely they wanted to ‘let go’ of them for the short term profits bumps of not paying their wages, or they were paid off to drop support by one of the other platforms wanting to maintain market share in that segment… It wouldn’t be the first time for either move.
With 4% market share of the desktop operating systems, one has to ask why bother in the first place?
If the point is a user-friendly software targeted towards non-experts and people who prefer the convenience instead of hacking stuff, you’re unlikely to find many Linux users in your intended target audience. Of course everyone likes nice software, but those who use Linux have really made their own bed in this matter.
I would guess it’s much more than 4% amongst hackers/makers, which are the target audience for this.
Someone should tell them about flatpack, docker, etc, but I suspect their reasons are just excuses to save devloper budget anyway.
Jack, the thing is that the software is a paid product for people who are buying quicker, better and easier results. In other words, people who value their time.
We are talking about a software for makers. The average maker does not fill well into “non-experts and people who prefer the convenience instead of hacking stuff”.
I have to agree with the others in the world of folks wanting a laser cutter Linux is likely to at least be comfortable and preferred if not the only OS used by a much larger portion than your 4%.
Also at this point Linux (at least the most common desktop distros) are also targeted at convenience and just works type desktop users as well – the days you had to know anything about the terminal, networking, your hardware or what is going on underneath the GUI are long long gone for the Mint, Ubuntu, etc distro’s. Unless your hardware is so darn new its not made it to the stable kernel in use Linux is likely to just work and often work better than Windows with the older hardware.
I’d suggest the common distro are actually easier to use now with a better user experience than Windows and its ever growing pile of bloat, even for users that have historically only used Windows. So while many Linux users like to tinker, or at least have the option to change lots of the x configuration even if they will never get around to it…
I don’t think makers are synonymous with “software hackers” – not everyone is doing embedded programming or web developing – most are really doing arts and crafts and general DIY and experimenting that happens to involve 3D printing and laser cutting etc., and it’s this crowd that is willing to pay for professionally maintained software, since they’re not experienced software developers and system administrators themselves.
Just because you’re a maker doesn’t mean you’re willing to put up with Linux as an operating system.
> Linux (at least the most common desktop distros) are also targeted at convenience and just works type desktop users
They just happen to fail at it. The concept of “just works” in Linux is that some stuff just works out of the box, but you still don’t have proper tools or automation to make other stuff work easily or conveniently if you need anything else.
4% is inflated by all the IOT devices running Linux.
No credible source gets that number.
More like 1%
Came here to make the same point, although I’d be happy with an AppImage as well, or even a Snap if push came to shove. Complaining about package management in this day and age just highlights the strong smell of BS emanating from their rationalization.
There’s a good chance that it won’t be long before somebody comes up with an FOSS solution. Here’s hoping it’s feature-competitive enough with Lightburn to cost them a chunk of their business, if not their whole business.
To be fair, Linus Torvalds was complaining about this recently too. The gist of it was that Linux will never become mainstream as long as it requires so much more effort to support compared to Windows and other operating systems.
I remember years ago getting a demo of something on a CD, and it listed how to get it running on various *nix systems. A dozen variations on “Mount as a drive, etc etc.”
Windows: ” Put CD in drive”.
(Ok, autorun was a bad idea, but anyway.)
Defaulting to autorun instead of asking for confirmation first was the bad idea.
Literally, you pressed a button, put a physical thing into the drive and pressed a button again (or pushed the draw shut). What extra would you get from clicking to make something happen after that??
NKT: watch the file list, if the CD contains what it is supposed to contain. Or doing something different than the default (read the install script instead of executing it)
Windows install is built for the average person apprehensive about plugging their mouse in, in case they break something, or are the just plug it in and go types, having never even seen inside a computer case. Linux, as much as some distros try, is very much NOT for those people (there is nothing wrong with them, they’re just of a different kind of mindset is all. Like the person into cars and the person into fashion, neither is inherently better or smarter than the other, they’re just into different things). Linux is geared towards the kind of people that know short of slamming the thermal limit of your hardware the worst you can do is bork some system files and have to reinstall, the people who can tell which type of expansion slot a card is for, how to find the size of a hard drive from the serial number on the drives sticker, know why this old ddr2 ram stick won’t work with that thread ripper. Two very different use cases (and honestly the amount of handholding that has creeped into linux is a little bit maddening….used to be it did exactly as was instructed, it was on you the user to take responsibility for the commands you execute. Now theres stuff that refuses to even acknowledge you called it because you either are or are not root. Not warn you, not fire a permissions error, just “no, fuck you” and fail silently….when root cant even get audio running or play a file in vlc shits gone fucky….).
My first Linux install was in ’93 and I think I can exp[lain in one word… Fragmentation.
It used to be simply package for a the variant; RPM, DEB or tarball.
The bloody distro and upstream devs are racing so far and fast that it’s nearly impossible to be sure a given package run anymore. I’m involved in HAM radio on Linux and the amount of tom foolery showing us there is AMAZING.
Yeah, the harcore linux crowd are Linux’s worst enemy – fighting over some detail they don’t like about any given distro and then going off and making their own new fork with blackjack and hookers or whatever, it’s ridiculous and hurts the overall effort.
The reality is that most people don’t care and just want stuff that works – I’ve been dailying Mint for over a decade and it’s great precisely because I never have to fiddle with it and their entire ethos is don’t fix what isn’t broken so there’s no surprises or radical changes.
The point about breaking stuff and keeping things changing is exactly to deter the commercialization of Linux as a platform, so it could remain their personal playground without the hoi polloi forcing any rules on them. Standardization is death.
TempleOS it is then!
@HaHa
Gnu Hurd would be a better choice, been around longer and no fragmentation / distro wars.
well, changes are rather caused by hardware landscape moving. When you’ve seen last time CD/DVD drive in the box? When you printed last time anything? Drift from 1 screen to display+laptop to 2 screens, to 3 screens – noticed how environment changed? USB-ization of everything much? Kernel went through major changes from fs to memory mgmt, to miriads of new devices, yada yada yada… We have to change. I must we have also clowns like Lennart you-know-who, who had pushed systemd and whatnot else – but those are changes caused by server side for sake of fat cats needing launchers, monitors and runners. Anything not moving gets obsolete sooner than you blink your eye.
Yes, do a flatpak or one for current Debian build and that should cover most Linux users.
…eventually people will just stop buying into proprietary bovine fecal matter such as this and just turn to open source equivalents. This brings up a bit of irony that we all run into. People in general agree that we shouldn’t be filling landfills, yet the entire infrastructure they have chosen is nothing but. I run into this daily.
I honestly think supporting arch and debian would be the best two if i were choosing. I believe arch is becoming the de-facto due to it’s use on mobile devices such as steam deck.
Flatpaks weren’t around a decade ago?
Not quite
Yes, Ikea has been selling flatpaks for much longer than that.
Why don’t these machines just use a variant of G-Code and allow us to bring our own software?
They do, which is why you can use LightBurn (or LaserWeb) instead of the manufacturer software in the first place.
It depends on the laser. Not all of them run the g-code protocol. There’s lots of things that g-code isn’t perfect at. For example, doing a scan (think like an image) pass back and forth with g-code is far less efficient. Many times it can cause starvation where the controller moves faster than g-code can even be sent since each “laser on” and “laser off” command and move is its own line.
Where that ends up is producers of laser cutters want their own protocol. LightBurn even started primarily with trying to control Ruida based machines.
Read the support thread — it looks like they support 3 major Linux installs, and are struggling with saying “no” to hundreds of subvariant requests. So they chose the nuclear option instead of sticking to their “guns” on supporting just the latest LTS version (singular) of Ubuntu.
No personal skin in this game for me, but I feel for the company and employees feeling like their hand is being forced by economics to go this route. My day job is largely one of saying “no” to weird 1-offs, and I admit it hurts not being able to enable good work.
Cue in the arguments about “it should be open-source so the community can do the distribution for their own preferred versions”.
Yeah, true, but then you won’t be making any money out of it – so why pay any effort to develop the software? That’s the catch-22 of Linux.
Yeah, no… Dead wrong. Welcome to 2024.
Open source software makes money, no problem. But you do have to have the right model, develop and maintain a community.
Like how? Get sponsorship from Google and then beg for the rest?
Was Grafana funded by Google?
Grafana is essentially Business-to-Business software. It gets funding by corporations that want to use it for their own product or analytics etc. It runs by the “paid support” model.
See the discussion below about the failure of this model from an individual home user’s point of view.
Cue that XKCD about all of modern infrastructure teetering on a FOSS project being selflessly maintained by some guy in Nebraska since 2003. Not only do people do it, but literally everything relies on a few devs doing this and basically always has
Just like most commercial software… only thing differentiating closed source with FOSS is that it’s plain to see that it’s teetering with FOSS. Closed source software companies often fake it like there’s a huge team, when there’s really just a precious few.
You can make money while being opensource – you are paid for the support or services by the folks that actually rely on your product, by the otherwise monopoly so they can avoid scrutiny and of course by the usually one off donations of the smaller user.
The real argument here is nothing to do with open/closed source, it is you should not be allowed to change the terms of the sale at all after the fact? So if you sell software as supporting x it should have to stay supporting x at least until you retire that product. So if you want to change the terms you should have to be selling a new major version of the product and have retired the old. Which no company on a subscription model will want to do as they will probably haemorrhage customers who had kept paying for a service they hadn’t been using enough… And for company’s with a pay to own model for their user nothing will really change except whichever users have been abandoned will be stuck on the old version.
The latter does seem to be more the case for Lightburn from a skim read – you buy the software and it will just work forever and you just keep paying every year for a year of updates. Which means IMO at the very least they should have to support Linux for a full year of updates with clear markings Linux support is being dropped, and no new sales of the year of updates to the ‘L’ version – which they clearly are not doing on their website yet anyway, prominent ‘Linux support’ everywhere…
Which creates the perverse incentive that your software should suck, so people would have to buy support from you. After all, if it was easy and straightforward, you’d be out of business.
That is also my though when someone says you can make money off open source software through support.
It does seem that Lightburn doesn’t want to be just a support company, they want to make laser software…..
Don’t get me wrong, I love and use a LOT of open source software but that doesn’t mean everything should or could be open source.
It really doesn’t – the company’s paying you to do their computing just want 100% uptime so their production of whatever it is they do is not disturbed, and for many of them no matter how straightforward your software is they would rather pay your company to have a large team on call. So if/when something goes wonky its fixed really quickly rather than pay for one more staff member who being only one person can’t troubleshoot or fix anything as quickly, and who likely costs more than paying for the support as well.
If you give your customer a great experience they will stick with you, and keep doing what they are good at rather than try and add into their workforce a computer tech or 10… Heck some of the support companies my Dad ended up working for gave really shit service, pushing broken untested or key feature lacking updates out to clients every other week even though its been reported a million times. Which meant somebody like him on the troubleshooting side has to go back in an roll everything back to the working state, and still the company paying for their support apparently loved them as they were better than the previous mob somehow…
But we’re not talking about B2B software here.
Linux works well for the business case, where the relationship between the buyer and the seller is more like a contracted maintenance company in charge of an office building: the software is made to order and need, being made of a generic common solution and then individualized parts according to the order and support contract.
That model doesn’t work with regular consumer stuff where giving such support to individual users wouldn’t pay for the effort; the individual consumer couldn’t pay that much money, so the “paid support” model fails here. What ends up happening is that individual home users get the crumbs that fall off of the business users’ tables, which is the bare generic solution without any of the support.
That was supposed to be “FOSS” instead of “Linux” above, but you get the point.
The actual open part of the software is just a framework that is available to anyone in a form that is mostly incomplete, non-assembled or “empty”, and the value you add as a company is to help your customers use it by training or by offering tailored solutions built with the framework. That means the individual user who cannot or doesn’t want to pay your professional rates has to climb up that tree ass-first and spend years just learning the ropes.
Dude there are hundreds if not thousands of small business that rely on their CNC tools just working – its just as much B2B software as the Linux desktop can be – useful for the individual and if the software is actually good it can become the core part of a businesses workflow the way Adobe, Solidworks etc stuff is.
And in the case of Lightburn that is already the sort of model they are running it seems – you don’t have to pay ongoing, the ongoing payments are for updates and support. Yet the individual has been buying it..
Yes, but we’re talking about FOSS, and LightBurn isn’t. Their model is to provide software that already works out of the box, because that’s the value for their customers. That’s what they’re paying for, and the support and upgrades are an optional extra.
The open source “free to install, pay for support” model leverages the case where a business would need to spend so many labor-hours to actually implement the solution and train people to use it that it’s cheaper to spend $10k on support to get started. For many small businesses, this is a trap, because it would be genuinely cheaper and faster to just buy the closed source solution, even though the up-front price seems crazy.
That’s an impossible requirement if X keeps changing and breaking compatibility with your product on their side.
Breaking updates are so rare and basically don’t happen when your support the LTS and only the LTS – should not come up much if at all for the stuff they officially support. Also just put it in a flatpak or appimage or whatever similar format and then it almost does not matter at all, and with the added benefit that you only have to package it once for every distro. So even though you only officially support x their entire extended family should be able to just use it.
Every time I’ve tried to use Linux, the problem I’ve had is that the LTS distributions don’t get the latest software or updates, so you end up jumping through hoops to get the OTHER stuff you need working properly.
@Dude if you are using the LTS you want proven stable reliable functional software with the serious security issues that come up getting fixed – you are rather doing it wrong if you insist you must drag the cutting edge in from elsewhere! If you stick to the stuff in the package manager, compile a compatible version, or use the appimage/flatpak type concept…
In effect either run a rolling release or run an LTS – trying to turn a LTS into a rolling release you might as well just build your own distro from scratch anyway. It would probably be quicker and you’d get to strip out lots of stuff your off the shelf distro will bring that you don’t need…
Let me decide what I want.
That is the problem.
Can’t you see the contradiction we’re at? I need software that isn’t in the LTS, so I need to use some other distro that distributes it, but you want this company to support “LTS and only the LTS”.
Why can’t I just install whatever I want? Why do I have to go through these gatekeepers to use my computer and software?
@Dude you said yourself ‘the problem I’ve had is that the LTS don’t get the latest’ – if you want the latest you don’t want the LTS – that is actively a contradiction to the goal of LTS! The whole point of the LTS is the stable platform in which everything will just work with only the really important stuff being brought into it, and generally only after its been in the bleeding edge long enough to find and fix the breakages – the whole point is to not have the very latest of anything unless absolutely required as that is where the worst bugs are always going to be, as they haven’t yet been found let alone fixed…
If you insist you want the ‘OTHER stuff’ that isn’t just in the LTS repo, or provided for the current LTS of choice (which usually seems to be Ubuntu) you are far better off not starting with a LTS distro, as that would be like trying to build an F1 car out of a family hatchback, technically you could get somewhere close but nothing much of the original is actually compatible with your goals. Where if you start from the rolling release cutting edge stuff as newer library in 99.99% of cases will still do the exact same job as a version a few iterations old that ‘OTHER stuff’ probably just works, even if it was meant for and only officially supports the LTS…
You can install whatever you want however you want, and just like in the world of Windoze there is no certainty it will actually work right if you are trying to install it out of spec, as seen with how well older hardware tends to work on newer windon’t… And if you really really must run x even on a LTS that doesn’t have a new enough version of something in the world of containerised applications etc you can do it, it just needs to be packaged up in your method of choice.
Isn’t the point of LTS to keep the operating system stable? The user software shouldn’t have anything to do with that.
That’s the entire problem here: the software you can “officially” get is married to the version of the operating system you use when you’re running Linux. If program X is for the LTS distro only because that’s the only thing they’ll support, while programs Y, Z, A, B, C… require you to use something else, then for all practical purposes it’s a no-go.
It’s an extremely dumb situation, where you have to run multiple virtual operating systems and sandboxes etc. just to have the software you want. It’s complicated, inefficient, and the competing operating systems simply don’t have this issue.
Mind, whenever you’re dealing with Linux distros, it’s common that the official repository simply doesn’t HAVE a piece of software because nobody’s bothered to backport it. Not even an old version of something – you just can’t have it.
So what are you gonna do?
Someone said that the biggest problem of Linux is that the distro “ecosystem” tries to own and control all the software that the user has access to. There’s this gatekeeper community that decides what the ordinary dumb user is allowed to have. This is the greatest barrier for commercial paid software for Linux: you have to go through the gatekeepers who are not interested in you, who often actively shun you and refuse to deal with you.
Dude the only reason the Operating system matters at all to the programs you can run is because a native linux app isn’t shipping its personal version of every single darn thing it is built with…
If you want to have no shared resources you can use the Appimage/Flatpak or pretty much the Windows software model – nobody is stopping you.
Most native Linux software sold for Linux historically targets the LTS of one or more of the biggest names and ships in their packaging format – no gatekeeper at all – the packaging standards are open, make your package and ship it.. Though these days most things are appimage type it seems to me – why package things multiple ways if you don’t have to and the days of every byte of drive storage really mattering are over.
But because they target that LTS “old and stable” shared library that software will nearly always work entirely properly on any distro nearer the bleeding edge, even across a different packaging format (as translation of those is usually easy and the next few versions of a dependency will in nearly all cases function perfectly in place of the older – can’t use v 1.1 for something built on 1.2 (probably, though it will often still work if you actually try it) but almost always can use v1.5 for a program built expecting 1.1…
It is a pain on those occasions you do run into dependency issues, and I’m sure many of us here have bashed out heads into that at some point… But from the point of view of supporting your product support the LTS and its no different to supporting a particular version of Windoze – not much to do until the next major version comes around.
The Linux users can get a refund, which makes your whole disquisition moot.
A refund is not the same as a service you were expecting to have being removed. Now if its a full refund or at least a generous one for the time remaining that isn’t so bad, still very annoying if you were depending on it. (Though with how Lightburn seems to work at least your old version will stay working, and hopefully you can continue to activate it forever).
However nowhere on their website can I see any details on that, or so far even that they are dropping Linux support. And the email linked above just says ‘recently’ and contact us for a refund…
To be clear I’m not saying Lightburn is being particularly bad here, it seems they are at least trying to do things right for the effected user. But what they are doing is still kinda bad for the consumer, especially as their website doesn’t seem to be making clear Linux support is changing at all yet – so I might just buy Lightburn as a Linux user today entirely in ignorance! And it does open that question of should it be legal and should we just accept that the contract we buy into can be changed – which is much wider question than just Lightburn.
Written in ancient-ass BASIC because I don’t want to do a flowchart in ASCII art and a more C-like syntax is my day job.
10 REM SHOULD MY PROJECT BE OPEN SOURCE?
20 IF ARRAY_SIZE(EXISTING_SIMILAR_AND_GOOD_OSS_PROJECTS) < 1 THEN GOTO 50
30 PRINT "STOP! JUST CONTRIBUTE TO AN EXISTING PROJECT"
40 END
50 IF YOU_WANT_TO_OPEN_SOURCE_IT THEN GOTO 80
60 PRINT "RELEASE IT CLOSED SOURCE, CONSIDER OPENING IF IT COMMERCIALLY FAILS"
70 END
80 PRINT "GOOD LUCK AS THE LEADER OF YOUR NEW OPEN SOURCE PROJECT!"
90 END
Red herring. This is like saying your website will start blocking all visitors not using Microsoft Edge, because you are getting flooded with “requests” for better Opera support.
If the requests are intentionally annoying, it’s possible not all such requests are even genuine but are working to this outcome.
The simple thing to do is support ONE distro (Ubuntu, unless the landscape has changed).
Give the mundane task of “saying no” to automated scripts and low-level low-paid employees. If the company is using skilled engineers for such thing, management is the real problem (and the mis-used engineer should be questioning their life choices)
“The simple thing to do is support ONE distro (Ubuntu, unless the landscape has changed).”
It has. Linux Mint is now being recommended, especially for those moving from Windows.
By Lightburn?
I love Lightburn, and the developers do care about their product and customers. The update makes me sad, I hope between Linux users and the devs they can find a way forward that’s sustainable.
I’m also a paying customer and love the product. This announcement does feel like cutting off one of my arms, however.
probably different devs but maybe this will get MillMage moving forward…
For the record, the yearly LightBurn license is only needed for upgrades.
If you let your license lapse, you can still use the installed version, you just won’t get updates.
Unless that’s changed, but I don’t think it has.
This is an underated point!
I still use a 2015 licensed version of SolidWorks, and get along ok for the most part – but it’s also the only thing stopping me saying goodbye to windows forever.
Freecad seems to be improving rapidly. As soon as I’m comfortable with it I’ll make the switch to Linux.
Hopefully the current Linux builds of lightburn remain stable long enough for Laserweb to catch-up.
Same position here but without hope for freecad.
The way the dev went and still going give me less and less hope. But, no complain here, just a fact. their work, Their code Their way of thinking, take it or leave it.
Being stuck with windows for one software, just one, is frustrating and enraging but less than trying to use freecad in real world.
eh this is a pretty sound decision in my opinion, supporting UI on linux desktop is kind of a pain across all the flavors, my guess is that they are gonna try to go forward on a new version on .net which still has had lackluster support, however BLAZOR may prove a way forward that would offer everything we need, its pretty easy to use and renders reliably in modern browsers
Nope, we’re still using Qt with C++. Never used .Net / C#.
No, it’s not that hard to support UI on Linux. But because it’s different from Windows, it’s lost on some devs stuck in the past.
The premier UI in Linux is a web browser. The business logic is running as a local service.
Qt is pretty much the only cross platform OO framework to survive the onslaught of Microsoft shenanigans back in the 90s. Back when UNIX was a thing, OS/2 was a thing and Mac was a thing, Solaris was a thing, etc. Companies like Borland, Watcom, IBM and others provided cross platform compilers and application frameworks. Java even sprung from the ashes of the death of all those cross platform compilers and frameworks. Qt just kept going and still goes today.
I should probably move to laserweb. I never use the CAD functions in LightBurn, since it’s closed-source and proprietary (duh). I do all my CAD in, er, CAD software and just use LightBurn to drive the laser.
I did just this! Do all my setup with CAD, then massage it with Inkscape, then into LaserWeb it goes to finish the job.
Of course, LaserWeb has some issues and is not nearly as polished as Lightburn. But I know for my part I’m a little more motivated to pitch in with LaserWeb now that Lightburn is completely out of the picture.
The irony here is that LightBurn should already have a secret client model, since most burners should be running in another room. So instead of shifting to modern architecture, they’re going to go all the way backwards.
Glad I decided against Lightburn… Now I’ll give my full support to LaswerWeb.
Good. Maybe we’ll finally get some usable open-source laser software for linux now, instead of everyone using lightburn because “it just werks”
This is the way. The “We can’t afford to support ALL Linux distributions” is a like when you ask a slippery politician one question, and they choose to answer as if a different question were asked.
The OS is dedicated to the task, just pick the most popular one and problem solved.
Can we now all request weird versions of Windows support to get all Windows support dropped?
Windows 1.0 is number 1!
I really like that hackaday is a site that can broadcast such information to so many makers at one time. This is providing a valuable service, notifying users to basically prepare for doom or to avoid depending on closed source.
It doesn’t matter. But, that’s also not how spelling works either.
Are they going to abandon all non-Windows operating systems?
I suspect abandoning MacOs will be a bad move if so.
If they don’t abandon MacOs, people could always go down the Hackintosh route and set up a dual boit machine that runs both their favourite Linux distro and MacOs. Which opens up some interesting possibilities given that MacOs is POSIX compliant.
Hackintoshing isn’t all that viable anymore. With Apple not having made Intel machines in a while, you’ll be limited in what hardware you can use. Basically at least 5 years old. And forget about Nvidia GPUs or AMD CPUs!
Although it’s fun to dream!
I’d be more inclined to just try the newest versions of the software with WINE, the work put into the WINE/Proton packages lately is very impressive (and WINE has been very capable if occasionally requiring a few Winetricks for a long long time) so I’d not be that shocked if it just worked perfectly.
It’s not going to be ported to Raspberry Pi OS then, I suppose. I am disappoint.
Since once of the big issues is difficulty with all the variance, and we’d be happy even if it was at a loss but was more easily supported, it’s more likely than you think. We’ve talked about the possibility of shipping it in the style of an appliance. Think like OctoPi, I guess. But that would reduce a lot of the variance and hopefully the support burden.
I only 3D print but his feels like the right approach for nearly any CNC process by moving OS support to the browser and could enable you to sell an officially supported interface device for the best user experience
Its too early to jump ship from Linux. They should wait until 2026/2027 as millions of PC’s will not run Windows 11 and many users will move to Linux. Linux should see a huge increase in usage. Personally, I am moving to Linux as Win 10 nears it end. I am not interested in a Microsoft account to logon to my own computer. (I know there are currently ways around it, but the writings on the wall)
I can’t believe we’re still doing “the year of Linux on the desktop” in 2024.
It’s just so difficult to understand why so many Windows users bend over and say “Thank you Sir, May I have another.” over and over and over and over for years and years. Defies logic and therefore we hear, surely millions will start using Linux because of what Microsoft is doing with Windows ${current version here}.
It’s not that hard to understand – change is difficult, not everyone is a computer person even if they have to use a computer all day and even the really easy to use distros like Ubuntu and Mint have less comprehensive / user friendly support than Windows because MS have the money & the people to pour into that.
For 99.9% of people the idea of having to open a command window and type something is basically a hard no and if your OS even suggests that you might possibly need to do that they’re going to run a mile.
In my case working with a command window is not an issue, I do it in both Windows and Linux.
What really keeps me from moving to Linux is the fact that the commands needed to perform any task depend on the flavor of Linux you are using, and even the version. Another thing is also that you have to have a permanent internet connection so that you can have a stack exchange page open to find the commands you need to perform even the most simple of things, mostly because there is no logical naming or even consistency.
I dread installing anything in Linux, because 99% of times, it will flag some error, and it will require digging for the fix, just to move to the next error, and the next, before finally crashing into the wall of permissions, all this when the instructions of the Github page just state, just type this to install.
To put it simply, a lot of things are unnecessarily complicated.
I have done all of the tinkering back in the ’90, now I just want things to work, and put my time into doing things I actually like doing.
This. 100% this. Lightburn is simply focusing their desired consumers. The: “it just works” users. I am one of these. I make money from software that does not work (I work in IT) but at home I want it to just work. Linux as a daily driver kept up to date (all parts) in a similar fashion to main-line windows 11 Home (not pro, not LTSR. Those are the WRONG users) does not just work. Too much variables everywhere. When I work on a maker-project at home, last thing I want is to spend my limited time figuring out my Linux system’s updates and how they caused graphics or chipset issues.
The squeeze as they say is not worth the juice.
I get that all software can be said to have issues no matter the platform, but if the platform is built in a bazillion different ways, googling the solution is not going to always provide THE fix for whatever specific issue you are having. Too much effort.
I just value my time more than I value open source, just like Lightburn values the majority of their users with consistently working systems than what they perceive as a smaller set their users with a wider variety systems that do all the things ever so slightly different.
In the end they are a company, they will follow their own direction and see to it they make money the way they want to. Forcing a company/individuals to adopt what you perceive as the correct way to do things is a fools errand and will not result in desired effects.
Lastly I’ll end with this: you want Linux support for a laser software? Make your own. Or do what normal makers do get software that just works and get on with your project. Seriously the end result here is some cool laser cut whatever you are making. Getting hung up on the OS the software is distributed for sounds like project procrastination.
I’m gonna go make something cool now. Bye!
Maybe it’s because they want to use lightburn.
I’ve been on Windows 10 for three years now, I also haven’t paid for an OS since Win 95 on it’s release day.
I enjoy using Windows, not sure how I’m bending over and taking it. A $100 purchase I made 10 years ago is still running great for me, and it’s much easier to use and get support for than Linux. It’s a time versus money thing for me, I have a few Linux machines for various tasks, so I’m no Windows sycophant, but the Linux machines take way more effort for the most mundane tasks. I have money, I don’t have time.
I try Linux, and always run into a wall with something.
Thw surface is very polished, but you need to fully understand the underlying system when you go off the path.
gstreamer lib changes have been a problem and some licensed code has been an issue from what I heard. Packaging is just one of many pain points. But the 2nd thing that hurts from their decision is they will not have Linux support for the Mill Mage CNC software they are developing. Vectric’s supporters will be happy to hear.
Meh, personally it’s been no big loss on the rest of them or LightBurn but I’ve got much more time for Russell Brand now than at any time previously; for him, at least, “obscurity” is clearly an improvement.
Half our dev team runs Linux. One used to work for Red Hat (he’s my engineering lead, and no one has been let go – in fact, this was his call). If it was *just* about packaging and supporting a single distro this would be a non issue, but there’s much more to it than that.
Linux users account for slightly less than 1% of our user base, so it actively costs more to maintain than we’ve ever made from it, by a multiple, and this has always been true.
As we move forward, we’re running into the limits of the cross platform framework we’re built on, and are having to write more OS specific code. Spending 1/3 the dev time for less than 1/100th the customers is not sane.
There are a dozen more reasons, many documented in the thread on our forum, but it’s not a decision we came to quickly or easily. We’ve been batting this around for about a year now, and a slowly decreasing share of sales and increasing dev complexity finally pushed it over the edge.
We’re a small team with finite resources, and it sucks, but this has to be a business decision, not an emotional one.
Disappointing, but I for one understand why you made this decision. No offense, but this is why open source is the way to go. If a piece of software takes a turn users don’t like, they can just fork it.
Not a Lightburn user (no convenient place to put a laser cutter, sadly), but a Linux user, so please hear me out.
It is totally understandable to drop platforms that cost money instead of making money. However, maybe you should re-think your long-term software architecture and thus gain platform-independence in a later version. People put laser cutters where it’s convenient to put laser cutters, not necessarily where it is convenient to have a computer. Thus, the obvious choice is to have a small computer (such as a Raspberry Pi) as a “print server” directly attached to the laser cutter, and then just have a client program running on a computer elsewhere. This could be a Windows client, but it could also run in a browser, thus supporting multiple platforms without packaging hassle etc. You could even offer additional cloud-based business models where people can pay extra to use cloud features. Remote access comes to mind. Of course, this is nothing revolutionary and should be the kind of architecture that one employs for a modern piece of software. This is all somewhat like what the Octoprint ecosystem is doing.
Which browser though? Sounds to me like you’d be opening another can of compatibility and support issue worms there…
…What? No. Just use a Chromium based browser as your standard, it’s all fine. Just not ActiveX, m’kay?
What if I prefer Firefox? Why should I play into the Google ecosystem?
I’m a Linux user but I’m also a realist – for all the sh*t you’re getting for this it’s a pretty obvious and easy decision, you’re a business not a charity and I don’t understand folks getting butthurt about it.
If the Linux crowd want more software they need to stop fragmenting distros, stop arguing over technicalities, admit that there’s some stuff that Windows and Mac get right that keeps many folks away from Linux, and make Linux a more attractive choice so the userbase grows to the point where it’s a no-brainer for companies to support it.
I use LightBurn as part of a production process leveraging a custom workflow of python generated lightburn xml files and the primitive network interface to fully automate laser marking and serialization of my products. I had been happy and recommending a similar approach to my colleagues before now. Just because people weren’t using Linux doesn’t mean they weren’t referred by Linux users, and I also don’t know how you can track active users on non-Internet connected systems unless you’re just considering activations.
It’s 2024 not 1995, and while I get cross platform takes work, USB, Networking, and even UI are pretty solved problems these days as is packaging via something like an appimage or just targeting one distro and letting the others figure it out. I don’t have (and will not have) a single windows machine in my production pipeline because of its propensity for uncontrolled system configuration changes, so I guess I’ll be finding a new option and steering people away from a product that will drop platform support with a few months notice. Lightburn was one of the only closed source programs I considered valuable enough to buy, so yeah I’m pretty disappointed.
Business decisions are business decisions. Just understand I’m not going to invest time in installing windows to just run your software. I’ll invest that time and effort into improving the open source competition.
So…. first: It’s great that people from LightBurn post replies here – even if there’s most likely nothing to win for them. Chapeau, thanks!
If these are the actual numbers (1/3 dev time? That certainly hurts) the business point of “something has to change” is clear. Since I’m not familiar with the main pain points (is it hardware support? Packaging (most likely not)? Cross-platform GUI?), it’s hard to argue here… on the other hand, if I glance over to Steam: They manage to run games (and quite a lot of them) to run on Linux; whatever the problem is, I’d assume that games are a lot worse. Always ;-) So I wonder if you evaluated “going the Wine way” as an option.
And of course (from an emotional point of view), as a long time Linux user (25 years w/o Windows in my household, and as a professional developer I’d consider myself a power user) I’m heavily disappointed. I hope my current version will continue to work for a long time to come… and I hope that someone at Lightbend occasionally tries running your software with Wine. I have several highly frustrating cases where an application would run on Wine just fine, if they didn’t do some strange stuff with copy protection and similar (looking at you, Steuersparerklärung! And all that iZotope sh*t which works great in demo mode, but crashes when you bought it!).
So… good luck with LightBurn – though somewhere in my heart I hope that your decision may prove wrong (not as a way to bankruptcy, but with a turn back to Linux support) in the future :-)
Some of our issues are to do with libraries we depend on having their own dependencies that conflict with each other. Qt depends on certain versions of LibSSL, GStreamer, and others. The licensing solution we use depends on a different version of LibSSL, so whichever version we include breaks something. This is one of several dependency conflicts.
We’re starting to write custom, OS specific code to handle things QT doesn’t do well (like camera support), and that new code would need to written for Linux in addition to Mac & Windows.
It doesn’t make sense to expend this much effort to maintain a version of the app that goes to such a tiny fraction of our user base.
Keeping the Linux version going has always cost us money, but it was never a large enough expense to be worrisome, but that has changed.
Existing versions will continue to work, and we’re not turning off licenses or servers, just no longer updating the Linux version after the 1.7.xx releases are done.
Can you write the software core as headless, and have the UI as eg. a web interface?
Voila, instant support of everything; most of the dependencies are in the UI crap.
The way Octoprint works.
The RaspberryPi stuff sounds cool but what about just testing against wine/proton on one specific distro? That seems to work pretty well these days. Then it’s just some tweaks to the Windows build for any bugs.
Used to help catch bugs on the windows side of things too.
Having only seen the ancient coherent (brand) laser engraver that is in my shed (xy table) and an SLS at a place I worked almost 20 years ago (movable mirrors) I assume the current models use similar methods (maybe a movable fiber optic “hot end?” ) but isn’t the software basically a reworked cnc router equivalent? Or are they much more refined? Are there any open source CAM software projects out there that are ripe for modification? Just curious.
All this whining over Linux being dropped by a software developer is a lots of unfounded noise. I rather think that the underlying tone is that paying for closed-source software is always an iffy proposition for the license holder, which is a position on which I agree.
I love Linux and was certified in UNIX as well as Windows server and I believe from an engineering perspective, an OS is just something to make PC hardware usable. The Application GUI interface will be familiar if not exactly the same for a well written cross-platform code.
The current price of Lightburn is $60 the first year with a $30 renewal for 14 months.
For my 3D needs, I use Lightburn on a dedicated Windows-11 PC which is off-the-net to avoid any issues with driver/OS updates AND also, being off the Internet allows for better performance by not needing to run AV-security software and no firewall or network card drivers loaded in RAM. (Connect through your own firewall when you must handshake for license renewal.)
A dedicated HP PC (refurb) with Win-10 Professional license is under $140 from Amazon for a nice 6-core I5@3GHz with 16G RAM and 256G SSD.
As a long retired EE with a 30 year career in Enterprise IT Operations and 10 years as a LLC designing custom software solutions, I strongly believe in "staying away" from computer platforms that run too many software application – the 'environment' becomes too complex to support mainly due to application prerequisites such as common libraries and foundations. Today, Enterprises are using virtualization to minimize software conflicts but this is not IMO something one should do @home. Cheap h/w doing one or two basic requirements are amazingly simple to maintain and the newest, fastest machines are not required.
Hey, that sounds really familiar; I remember some folks saying some things about doing one thing really well.. maybe it’s a philosophy valuable on the macro scale too.
Although it sounds like you are far from ”it just works;” what you said about renewing licenses sounds pretty challenging and fraught with risk, unless you happen to already have technical competency to do fancy networking and security.
Containers? I’r is it end hardware calls the problem?
Given the accelerating frequency with which I see people bowing never to use Windows 11 and swearing to go Linux instead, The light burn people seem to be picking just exactly ironically the wrong time to move in a just exactly that ironically the wrong direction.
If anything, I’d say they should abandon Windows and concentrate on Linux, cuz that’s where the users will be TWO YEARS FROM NOW. Talk about cutting off your nose to spite your face!
Well, I guess this is what we get for allowing software to become a commercial endeavor instead of a labor of love, 35 years ago in the first place.
Depends on your circle. I bet if I asked 20 of my coworkers what Linux is maybe 4 would know.
Ah but is your job at all related to science/engineering/making? – if not its probably not the sort of job where folks that might just want a laser cutter will gravitate towards and so as a population sample not very applicable to the folks that may want to buy this software.
And like the guy from Lightburn said less than one percent of their users are on Linux….
In reality, these types of laser cutters have been around for some time now and still no real OSS software alternative, LightBurns authors saw this gap in the market, so good on them, they have created a fantastic piece of affordable software for makers and professionals alike and I have no problem paying for it, sad Linux support will be ending but is economical just like recent layoffs where I work – people will adapt one way or the other, if an OSS alternative comes out of it great.
The best opening sentences i’ve read in a long while, thanks Tom
All this reminds me of the Youtube video, of Linus Tovalds.
Explaining why Desktop Linux sucks: (Application Packaging).
https://youtu.be/Pzl1B7nB9Kc?si=1U9vUDT1ABkUxgxF
If you go on any of the laser FB pages, the first things people will tell you is that “YOU HAVE TO GET LIGHTBURN” – it’s so not true. 99% of people can do just fine with the stock software. It’s just become one of those things that people say to one another and gets repeated. It’s just a thing. Good for the delevoper. He’s getting tons of sales that aren’t even needed. Bad for lazy consumers. But oh well.
If the problem is with packaging, I hope they consider just shipping either an AppImage or a Flatpack. If you have one of these and you do not like them, you can usually repackage them in your flavor of choice. For example I am an Archlinux user, and in the AUR repository there are many recipes that will repackage AppImages into native arch packages. Even other non universal packages can be repackaged if you at least have something to tinker with. For example in AUR I have been mantaining for years Lattice Diamond for Archlinux, that Lattice provides only in RPM format. My PKGBUILD script downloads the RPM and service packs, merges them, fixes whatever problems that could appear (library stuff mainly) and repackages them in pacman native format.
Why would anyone want to stay exclusively running Linux, is my question. You have a laser, you need and pay for a software package that costs money to run that laser, then you adapt and install Windows. I bought a goddamn iPad so I could use Procreate, and I absolutely hate iOS.
Do you really want to pay a small fortune for Windows license if you don’t want to actually use windows? Or pay the usual Apple tax for just one thing you rarely do? If you are so made of money you don’t care about the cost, or extra space your rarely used alternative hardware takes up you are pretty lucky…
Folks that want to ‘exclusively’ run Linux are simply folks that don’t want to have to buy the ‘same’ thing twice, learn how to use the multiple systems or deal with M$ and Apple slapping them around anymore… Life is much simpler and cheaper if your one computer can do everything you want, all the time (so no dual booting), with you in control of what changes and when, and all while you don’t have to pay a small fortune to company that keep on insisting you must sign your life away to them…
To be clear though if you love the Windows or Apple experience and don’t want to move on I don’t care, you do whatever suits you best – I’m not going to suggest you should buy new hardware or software unless you actually want to play with it. Which is the point you shouldn’t HAVE to suddenly adapt and pay more money to in the process just to use something you previously did.
Small fortune? Let me introduce you to the “arrr” channel. Most common proprietary distribution means.
It took me longer than it should have to “get” your comment!
B^)
A couple hundred bucks isn’t a “small fortune”. That’s an interesting loss of perspective when talking about equipment and materials with a hobby that ends up costing thousands anyways.
Just buy a license from Thailand.
Last I bought was $25 for 10 pro. No problems.
Nothing changes.
Last Netmare license I bought was below the cost of the local CNE mill.
I suggested he start buying the licenses he sold from there.
But big red would have had kittens.
Would have interfered with the revenue he was making selling certs.
I ain’t going back to Windows.
If you have to, just provide a binary for a single distro then.
People with other distros will find a way to make it work.
I keep an Ubuntu LTS chroot inside a different distro which allows me to run things like the PicoScope software on unsupported distributions without problems.
And that, folks, is why you should never, ever use any closed-source software if you can avoid it — you’re always at the mercy of its developer’s ‘business sense’, and their rationale for making more money regardless. LightBurn is just the latest example.
Good riddance, I say — let’s all us makers just move en masse to Laserweb and give it our full support so it becomes a better competitor to Lightburn even on Windows — and then watch it lose sales and market share because it’s really hard to compete with free.
Well if you want quality you pay for sw.
Not really, some of the best software at what it does is FOSS. Of course not every FOSS program is going to be as polished or feature rich as you expect the closed source paid for stuff to be, especially early in its life – but then neither was the earlier versions of the paid software as polished as the new ones (at least in general).
when solution to problem is “f*** this”, it tells a lot :D
This is why you need a Windows 7 at home with updates turned off, not all programs will support linux, end of story.
I’ve been a Linux user for twenty years and I have seen many leaps and bounds for Linux over that time. In my line of work doing CCTV and Home automation I have to rely on a VM running Windows because 90% of the software I use for work doesn’t run in Linux. A couple of programs run using Wine but I have to spin up a VM.
The way I see it if a developer doesn’t want to support Linux then either find a alternative that has Linux support or just spin up a VM. Running a VM is no different then running a flatpak or a AppImage pessary.
I have tried linux a few times but always give up because inevitably there is some software that won’t run well on it that I need or want. About half the time that software is a game. I think that is not going to change. I think that is how windows11 will win out over linux in the end.
Other than the handful of games or their publisher that have actively made efforts to stop their games working with proton Valve building on WINE has made virtually every game just work on Linux as a trivial click install and it works. So if you are finding gaming is the one thing holding you back but you otherwise like Linux give it another go (perhaps check with the ProtonDB for your favourite games first just in case – but the Linux gaming experience is now better than I ever remember Windows being most of the time).
I once upon a time used to run a GPU passthrough VM so the windows games and VR in the early days would just work but I never had to deal with Windoze crap in the day to day. Which was great for years, but when that computer died with a mobo failure 4 odd years ago, but since then I’ve never missed it at all anyway.
It’s a kind of closed circle. Users don’t go Linux because there is no software. Developers don’t do software for Linux because there are no Users. And it’s like that for a long time.
To be honest I don’t expect people to go Linux. But I wish developers start to develop more on Android.
This is why I avoid closed source paid for closed source software like Covid in the first place.
This is why I discourage all those I advise to avoid investing time learning and building skill with proprietary software. Whatever agency you’re provided is at the whim of the developers and subject to change at any time. I understand their profit motive and the economics of the decision they’ve made and don’t fault them, but nobody has to be a bad guy to legitimize the anti-proprietary-software position.
I don’t see what difference open source software would make.
The only way you’ll have any agency is by investing time in becoming a software developer yourself, which ultimately costs you many times more. Otherwise you’re still under the whim of people who don’t even know that you exist, because you’re not a paying customer and they don’t need to need to care.
In the world of opensource you can never have the rug pulled and service you rely upon lost entirely…
Even with no development skill you can continue to use an existing version no matter which direction the developer take, and for every time a serious change of direction happens there will almost certainly be a fork or two that matches your preferences anyway – As the source is all there to work on and somebody capable in the userbase will not like the same things as you.
Where with closed source if Autodesk/Adobe/M$ (etc) decide to pull 95% of features of the product/tier you are paying for and put it into some more expensive version. Or simply add some new features and in the process massively disrupt your workflow…
I do think issac is a bit wrong though – there is no point in mastering the art of your open source option beyond personal satisfaction and perhaps future proofing your career if something else is effectively the required program for being employable.
For some time, until you have to update/upgrade your distro for something else, and it stops being compatible, or someone simply takes it down from the repositories and poof, it’s gone for you for all intents and purposes. Then you have to start hacking and doing tricks.
Dead ends that only delay the inevitable. If the majority don’t follow along, the fork will fizzle out in the lack of serious resources and users.
Closed source -> the bin.
As many others have pointed out, I would just release an appimage, state only supported on the latest Ubuntu LTS, and leave it at that.
Snapmaker is facing the same challenge, build both a deb and rpm, then someone comes along and has some distro that’s missing a dependency.
Appimage solves the problem, gives a support statement one can stand behind. Have a one pager FAQ on the website that states that, and point folks to that when they raise a ticket or support query.
Why not flatpack or snap, because it avoids the whole snap Vs flatpack drama that seems to never end. And I’d still be saying that if I was still at Canonical, its just to much faff and timesuck. Hopefully down the road, we can get a level headed agreement to merge those (likely when you can get ice in hell).
Option one, do a appimage, option two make it a flatpack. If Canonical pays for or offers build the snap for no cost, you can opt for that.
So… How exactly does this solve the developers’ dependency conflicts like the LibSSL conflict Jason mentioned a while back? You have to be able to build at all before you can start worrying about the size and shape of your package, but maybe your suggestions account for that (I’m not really developer so I actually don’t know)?
They simply wanted to cut off expensive developers in favor of cheaper ones. Code that runs on windows will run on linux, just as well if not better, and require no additional work to “package”. Unless you purposely make it windows only. Getting packaging options is just a convenience and even then it should be fully automated.
It is true that junior developers may not understand cross platform development so if they are hiring on the cheap it would help explain this mess.
They obviously don’t pay their PR team(person?) very much.
Potentially a suicidal move.
Create a gap in the market and someone will fill it.
Now there’s a strong incentive for the creation of an open source substitute, and they have just let loose into the market the people who would know how to create it…
Watch this space.
Except less than one percent of their users are on linux…
Not if you measure the users by weight.
Not necessarily. The distribution is bimodal; for every one spatially intense you have one scrawny. Which is less visible, so it skews the perception.
When a free alternative fills this gap for Linux then and is ported to Windows, this decision will come back to haunt them.
Just being free doesn’t mean a thing if it’s still almost unusable and lags behind in features.
Without a way to make money out of the software itself, these free community drive projects rarely get off the ground and end up forking and in-fighting, and then abandoned. Their only hope is being picked up by businesses which see synergy benefits in maintaining a free-to-all piece of software to sell something else.
Have you tried VisiCut? It lags in the builti-in design editing side but there’s Inkscape or your favorite SVG editing tool for that. It’s had camera support for over 10 years and same goes for having a materials library but it gets better, a project contains not only the materials selection but also the laser settings for the materials.
https://visicut.org/
That’s why I said alternative, it has to be as good to be a viable alternative.
There are already less than viable options now.
Unless it’s in the server space, abandoning Linux in any form touches no one’s bottom line.
I develop .NET applications on Windows by day, run almost exclusively Linux at home. 1,000 attempts to get into Linux programming have ended for me with “I’m doing too much programming at the day job to focus on this at home”. But the curiosity is there.
I mostly use Gentoo which I am pretty sure is the primary build target of few enough user applications that I could count them on my fingers. But.. apart from a bit of crap that’s offered only on locked down container platforms… everything just works.
So I am really curious… what is this per-distribution support that developers so complain about. Again.. it feels like no one is building “for Gentoo” yet it all just works!
Is it in the packaging? File paths, permissions? Is it because pre-built binaries target specific library versions and those differ? The kind of problem that completely disappears when building from source? I just don’t get it.
I’m pretty sure that if I took a piece of Qt Hello World code and ran it at home, or even at work in WSL it would “just work”. For that matter, would it need to be modified to build it in Windows even? What does a laser cutter controller have to do that is different? Read a file in.. write out to a serial port… The rest is all just crunching numbers right?
What are the per-distro weeds that developers get so stuck in?
You can’t be master of all things.
Linux is a mature OS.
Which means it’s a pig F, just like Windows, only with SystemD.
But it’s users already know it’s issues, hence Stockholm syndrome.
It’s their hog Fing mess and they like it that way.
Some JS programmers claim to like that mess!
So much they use JS on servers! (Hard to believe, but true.)
There is no limit to rationalizations and bad decisions.
A programmers just works, and a consumer’s just works are two very different things.
One bit of context missing from the discussion is that what’s actually happening is they’ve decided to rewrite Lightburn and so they’re dropping Linux support.
Businesses have the right to make bad decisions, I guess. 🤷♂️
It’s missing from the discussion because it’s not true. Sure, we’re doing refactors here and there to try and make a better product, but we have no plans to rewrite it.
What about open sourcing or offloading problem components like camera support?
Before flatpaks & appimages were around, we simply used a program called “Alien” to translate between rpm/deb/tar.gz
A quick search shows it still exists.
This is exactly why our company no longer uses software we don’t have the sourcecode for. It’s not that we’re going to compile ourselves, but if the company providing it goes under or starts behaving like assholes it’s simple to find a fork.
We used Lightburn for a couple years right at the start but it quickly became apparent that the devs were only interested in adding gimmicky features to justify the ridiculous subscription pricing for something that could have easily been open-source with a “professional” option like all the cool kids are doing these days (and it would probably would be more successful too.) For paid software it’s worth $20 IMHO, considering that beyond the gimmicks it does very little above the stock software that comes with laser-cutter controllers other than not being Chinese.
A good plugin for Inkscape can easily replace Lightburn and provide vastly more capabilities. G-Code is already there in an existing plugin, and I easily got an SDK from Ruida (manufacturer of the controller in almost every laser cutter) back when I was thinking of doing this myself before I got distracted by something new and shiny.
Inkscape’s G-Code plug-in finally got some love? That’s amazing! The last time I played with it was so long ago and it was completely incompatible with newer versions of inkscape. I had no idea there was further development.
Don’t forget the VisiCut is open source and has been under active development for over 10 years.
Please make a simple program to send data. I don’t need all the software to set and to reduce objects. I just need something like a slicer. Send and that’s it. Then the projects I already have I can still burn.
Amen. I’ll run Win 11 for design work in my office, but use a cheap/rugged Linux machine in the workshop to send jobs and control the laser.
I got away lightly by this, when I first got my K40 laser cutter I really didn’t like the shonky strippit-and-shippit bastardised Corel Draw thing that came with it, it was Windows only and I wanted to run it off a Raspberry Pi. Looked around for a quick FOSS solution and although I found Lightburn, I also found K40 Whisperer (https://www.scorchworks.com/K40whisperer/k40whisperer.html) which is totally free, but looks a bit clunky. It absolutely worked and did everything I needed.
So I used K40 Whisperer, telling myself that as soon as I needed more I could always stump up and pay for LightBurn, and maybe change out the main board for a nicer one, and a adjustable height table in there, maybe change out the lens and mirrors, give it a nice paint job (that last one isn’t true, but you see where this is going).
None of that happened, so I’ve bumbled along making significantly more things from laser cut ply than 3d printed plastic because it so much faster to iteratively prototype with.
And now here I am, probably close to the end of the unit’s life (the laser tubes only have so many uses in them and the vast proportion of the cost of a K40 laser cutter is the laser tube) and it’s rapidly being solved for me :D
Great decision. Linux in all it’s forms are great and all but the headache isn’t worth it. When a random app can replace a dependency and break your app or cause a bug hunting hell why bother? I doubt even 5% of their user base is exclusively Linux, probably less than 3%. Just not worth it.
Less than 1% of our users use Linux, exclusive or otherwise. Exclusive would be even lower than that.
I’m saddened because I really enjoyed using Lightburn which is a good software, and I’m in the 1% with only Linux machines around. I don’t want to have a windows machine just for that, though the old VM trick would work with some compromises. It’s a shame to see good software leave the Linux ecosystem though.
Thank you to the Lightburn team for the great work you did, I was happy using your software, and will stick with 1.7 until something better comes up.
What are you basing this on? Sales? Download numbers? What?
Trials and activations. Whenever someone uses the software, it does a fingerprint through the licensing tool we use and that allows us to see what operating systems are used. Then we went through and coalesced the Linux, Mac, and Windows versions down. So if someone had 2 Linux computers and 1 Windows computer used on a license, we would count those as such. We also looked at currently active vs total numbers across the board.
Really, this is as close as we can get it to “truth”. Downloads are a little harder, because they’re done through GitHub historically and we’re switching to a different site for various reasons. GitHub does have release stats you can pull, but they’re honestly not as useful for data statistics purposes. For example, the total downloads for Linux for 1.6.03 (released 6/25) across our 3 released packages (self extracting installer, 7zip, and AppImage) sum up to greater than the total number of Linux activations we’ve ever had. The 1.6.04 version (released 7/22) across all release Linux release packages is greater than the currently active licenses for those who run it on a Linux computer. Now part of the discrepancy is people trying out the software, which we can also see with the licensing tool, but I don’t have those numbers in front of me so I won’t get into that.
As for sales, a license is agnostic to your OS. And we really would be more fine with supporting Linux if it wasn’t the combination of low usage + outsized support burden.
People keep mentioning wine and proton.. do you think that’ll be a viable option for Linux folks, moving forward?
Thank you for taking the time to fully answer the question, it’s certainly appreciated.
It’s too bad about Linux support and I hope in the future this changes, I wouldn’t go for the appliance route as I already have dozens of SBCs and wouldn’t want to purchase another for a single application.
I don’t own a Windows PC and haven’t in well over 20 years so my views on Linux are skewed as I’ve used it for so many years. I simply can’t use Windows not because it’s awful but I can’t deal with the update > reboot > update > reboot nonsense and it’s only gotten worse over the years. A VM is viable I suppose but it’s a bit heavy for daily use to spin up.
Russel Brand was picking up momentum again recently, but then got into drugs again, not the chemical type this time but the ‘opium of the people’.
The realization that people who can’t figure out Linux have access to laser cutters just took a high slot on my fear list.
There is a reason for the labels saying “Do not look directly at laser with your remaining eye.”