Qt LTS Releases To Be Restricted To Commercial Customers, Other Commercial Changes

[cesarcafe]

[...very well detailed explanation snipped...]

To be fair, you have a point on commercial licensing of the GPL source. That being said, how is that any different than the risk taken with any other proprietary, licensed software package? [...]

First of all, thanks a lot for your detailed explanation. It's very instructive. Thanks a lot.

Regarding your question about the risk difference between commercial licenses of GPL sources vs proprietary software, of course it's the same risk. That's what I said from the beginning, although I don't used your example, but I really believe it's a good example: If you use a commercial license of a GPL source, you are taking the same risk as if you are using a proprietary library/system for development... a high risk.

In your rationale, I see you are implying that permissively-licensed code lacks a community willing to develop it further (not true: just consider the community pushing to develop Solaris further after it went closed-source --current OpenIndiana release is just 2 months old), while you are also implying that copyleft code has a community that will never cease development: That's not real, at least not from my experience. You can find thousands and thousands of copyleft projects that were abandoned at some point.

The difference between permissive and copyleft is not that one is going to be abandoned and the other won't (that doesn't depend on the license, but on the size of the community using the code). The difference is that if a copyleft code is abandoned, you are left with code that has a viral license, while if a non-copyleft code is abandoned, you are left with code that doesn't have a viral license. Both situations are bad (abandoned code means you need to put a costly and great effort if you want to continue the project, as you said). But if a project is going to be abandoned, I obviously prefer that it's non-copyleft, so that at least I'm not left with abandoned code that also has a viral license.

[madscientist159]

First of all, thanks a lot for your detailed explanation. It's very instructive. Thanks a lot.

No problem. We've been around for a long time, long enough to see a lot of code and companies come and go, so I'm always happy to share what we've had to learn the hard way in some cases.

Regarding your question about the risk difference between commercial licenses of GPL sources vs proprietary software, of course it's the same risk. That's what I said from the beginning, although I don't used your example, but I really believe it's a good example: If you use a commercial license of a GPL source, you are taking the same risk as if you are using a proprietary library/system for development... a high risk.

In your rationale, I see you are implying that permissively-licensed code lacks a community willing to develop it further (not true: just consider the community pushing to develop Solaris further after it went closed-source --current OpenIndiana release is just 2 months old), while you are also implying that copyleft code has a community that will never cease development: That's not real, at least not from my experience. You can find thousands and thousands of copyleft projects that were abandoned at some point.

My apologies, that was not what I meant to convey. Yes, it is true that there is often enough investment from the overall community to keep smaller projects, or larger ones in maintenance mode, alive. However, from where I sit things start to change rapidly when you hit the scale of Clang/LLVM, Google Chromium, etc. that are under continuous active development -- at that scale, if you are not prepared to fund a similar army of developers, you will effectively be outspent and your fork will fall by the wayside as people go to the tool with more features (a surprising number of developers will only see the improved tooling, not the concessions that have to be made to use it, unfortunately).

The difference between permissive and copyleft is not that one is going to be abandoned and the other won't (that doesn't depend on the license, but on the size of the community using the code). The difference is that if a copyleft code is abandoned, you are left with code that has a viral license, while if a non-copyleft code is abandoned, you are left with code that doesn't have a viral license. Both situations are bad (abandoned code means you need to put a costly and great effort if you want to continue the project, as you said). But if a project is going to be abandoned, I obviously prefer that it's non-copyleft, so that at least I'm not left with abandoned code that also has a viral license.

Now I think we come to the fundamental philosophical difference between our viewpoints.

From my side, having seen so much work locked up under copyright, and having seen rather greedy people try to take things out of the public domain (which is sort of what the BSD like licenses are trying to emulate) and put them back under copyright, I wouldn't want to invest my time and effort in a project just to enrich someone who wants to take my work and stick it back in a proprietary tool without paying me for that work. Especially if the goal is to make money off of it without contributing back. A copyleft license is a simple and effective way to make sure that my donated effort isn't ripped off, that I get something back (specifically, the improvements another person might make to the same project I helped fix up or improve) -- without it, I tend to either make sure I'm paid up front to contribute to the project or just keep my improvements to myself. Basically, I've learned over my lifetime that truly altruistic people are very rare, and people that would treat others as sources of free labor are sadly very common. Share with me your effort, and I'll share mine with you. Make me pay for your effort, and I'll insist you do the same with mine.

From where I sit GPL helps preserve the work until the copyright is spent. BSD-like licenses do not (nor do they actively harm preservation, they are simply neutral much as the public domain is). I'm a particular fan of LGPL licenses for key tools used by a lot of people (such as compilers and platform toolkits), since they're basically a guarantee you can't stop someone from developing software in whatever way and for whatever platform they want. Note that while the this is effectively a limitation of some rights in the software, it stops a different, and IMO more dangerous, potential limitation of rights of people using the software or of developers wishing to modify the software in the future. At its core, this is a true zero sum game -- either one party must lose some rights or the other does. Either the code must remain open by license (restricting the developers) or those same developers have every right to stop releasing source code (restricting both all users and all developers that don't have access to the source code any more).

Yes, we can pretend that this fundamental bistability doesn't exist, and quite successfully in some cases (look at OpenBSD, FreeBSD, etc.) but that's because there is a third element in the system (e.g. the project leadership) enforcing a copyleft philosophy within the primary active (i.e. largest) part of the relevant developer community. It's an unstable state; very hard to maintain, very easy to upset if you have even the slightest amount of resources and inclination to do so. And I have no solutions for it that don't involve radical reforms to IP law, and even then I'm having difficulty imagining just what would fix that inherent bistability of rights short of completely unworkable "solutions" along the lines of either no copyright at all or a mandate for source release for all computer programs.

[moilami]

Investigated a bit and found that Qt is owned by Digia nowadays, which is a Finnish corporation. OS is not a bit of their area of expertise and the company makes most of its money on state funded establishments, meaning they only exist as is because "friends will be friends". Major hostility can be expected by them towards OS assets, of which this was the first step.

[Morty]

If a GPL project is developed by a business that also exploits the project commercially, like in the case of Qt (by selling commercial licenses to users who want to use the project for commercial applications while forcing the rest to develop only GPL projects if they don't buy a license),

And in the case of Qt, this is not true. Qt is also licensed under the LGPL and have been so for years(Since they were owned by Nokia). So there are no issues making a closed source(or commercial as you term it) with Qt without buying a license, you are not forced to develop GPL only projects.

And generally there are no problem developing commercial GPL applications either. For making custom software for large or medium companies it makes lot of sense, you have access to lot of free libraries to speed up your development and it's only required to distribute the source to that company.(Then you can easily resell it to their competitors, as your customer will not share:-)). And for such software you negotiate pay based on development cost and support, not site licenses or installs.

then if you invest money in developing commercial applications with that project, you are taking a very high risk: If at some point they cease business, you are left with a commercial license that can be used with an older unmaintained version only, while future developments from the community will be GPL'ed, so you won't be able to use it.

With Qt you take practically no such risk, due to this.
If they cease business, you will get the whole thing as BSD licensed within a year, giving everyone the ability to take it commercial if they wishes.
And added to that, it will make no sense for the community to not keep developing it under the existing LGPL license. So everyone will still be able to use it.

[dragon321]

Really? The GTK are buggy-designed gui libs.

What desing flaws can you show?

...but freaking CSD.

Totally optional and not default feature. Yeah, so terrible.


Why do peoples still look at GTK 3 through the prism of GNOME 3? Shell isn't even written in GTK 3 at all.

[skeevy420]

What desing flaws can you show?




Totally optional and not default feature. Yeah, so terrible.


Why do peoples still look at GTK 3 through the prism of GNOME 3? Shell isn't even written in GTK 3 at all.

Because a lot of GTK3 programs look like Gnome 3 programs regardless of the desktop environment they're used on.

[carewolf]

Investigated a bit and found that Qt is owned by Digia nowadays, which is a Finnish corporation. OS is not a bit of their area of expertise and the company makes most of its money on state funded establishments, meaning they only exist as is because "friends will be friends". Major hostility can be expected by them towards OS assets, of which this was the first step.

Qt was split off from Digia a few years back to its own company called "The Qt Company". I guess there just weren't any synergies in having Digia's consultancy and Qt's product development in the same company (plus apparently the stock holders had a good day as TQC ended up with a higher price on its than under Digia)

[ddriver]

Qt was split off from Digia a few years back to its own company called "The Qt Company". I guess there just weren't any synergies in having Digia's consultancy and Qt's product development in the same company (plus apparently the stock holders had a good day as TQC ended up with a higher price on its than under Digia)

It was more of a rebranding really, in an attempt to disassociate with the bad rep

[Nth_man]

More than a year has passed, but some things haven't changed for "ddriver":

I think the only troll here is you.
-- Upvoted 6 times at least: https://www.phoronix.com/forums/foru...86#post1063086

[Guest]

This is going to negatively affect some projects or distro releases that relied on it.

Each distribution backporting their own QT LTS will be a nightmare from manpower to compatibility between distributions.

The options:
1. Each distro adopting the LTS support burden.
2. KDE Project absorbing QT LTS maintenance.
3. Distributions dropping KDE/QT for LTS releases.
4. While porting for QT6, code a home brew replacement.

#2 will be the more stable and friendly imo. KDE Community to take the burden and tie QT with KDE and released closely tied together at least for plasma LTS releases. # 1 will bring too much headache for distributions, many of them will do #3 and not think twice and drop KDE from LTS spins just for the work involved.

For a long term solution, #4. But please no forking of QT. With MySQL and MariaDB “drop in” compatibility B crap I have enough.

It's a natural fact. Every time a commercial company gains some kind of success after the someone else work, they decide for a whatever reason to make pay they products. It's natural and happens continuously. The mistake is thinking this never can happen to me. Kde, Xfce and all the rest should think to a free (as in freedom) and open source solution. A nonprofit organization for a better and free gui libs is necessary in my opinion. And the clauses should not permit to sell all the main component for the community that all distros use to commercial companies.
(I know my English is so so, I'm Italian)