Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Open Source GUI KDE

Open Source Advocates Hope They Don't Have to Fork Qt (phoronix.com) 137

Posted by EditorDavid from the last-resorts dept.
Phoronix reports on a new concern about Qt, the free and open-source widget toolkit for creating GUIs and cross-platform applications: Wednesday a KDE developer who serves on the board of the KDE Free Qt Foundation commented that The Qt Company is evaluating restricting new releases to paying customers for 12 months. That was said to be under consideration due to COVID19 / coronavirus impacting their finances and needing to boost short-term revenues... [Slashdot editor's note: the comment also claims the Qt Company "says that they are willing to reconsider the approach only if we offer them concessions in other areas."] This comes months after The Qt Company already shifted to make Qt long-term support releases customer-only, among other steps to boost their commercial business at the beginning of the year.

Following all the speculation and concerns from the statement by KDE's Olaf Schmidt-Wischhöfer, The Qt Company released this very brief statement:

There have been discussions on various internet forums about the future of Qt open source in the last two days. The contents do not reflect the views or plans of The Qt Company.

The Qt Company is proud to be committed to its customers, open source, and the Qt governance model.
But in the event of a one-year freeze on free releases, Phoronix now reports, "several individuals and projects are already expressing interest in a Qt fork should it come to it." The hope is first and foremost that The Qt Company and KDE / KDE Free Qt Foundation can reach a mutual agreement without this embargo on future releases, which would effectively close up its development... Among those backing the concept of forking Qt as a last resort if necessary has been developers from consulting firm KDAB, the Qute browser developer, and the QGIS project as one of the leading geographic information system software packages, among many KDE developers themselves.

The mailing list thread is quite active in talking about the possible fork if necessary, including aspects like web-hosting down to what such a fork should be called ("Kt" seems to be a popular choice so far with several different members in the community).
This discussion has been archived. No new comments can be posted.

Open Source Advocates Hope They Don't Have to Fork Qt More

Open Source Advocates Hope They Don't Have to Fork Qt

Comments Filter:

  • Don't forget your competitive advantage. (Score:5, Insightful)

    by Brain-Fu ( 1274756 ) on Sunday April 12, 2020 @02:47PM (#59937776) Homepage Journal

    Java already exists as a cross-platform toolkit for UI development, and Microsoft is really pushing to get there with the .net core.

    The "free" aspect of QT is half of why people use it. The other half is direct integration with high-performing C++ code.

    Take away the "free," and people will start looking for alternatives.

    • Take away the "free," and people will start looking for alternatives.

      They aren't taking anything away. The non-free license would apply only to new code and only temporarily.

      My company uses Qt and PyQt. I hope it stays free, but we won't switch if it doesn't. The existing code works well so a 12-month delay on upgrades is no big deal.

      But I don't see why they should care what I think, since in all the years we have been using Qt we have paid them exactly $0.

      • Re: (Score:2)

        by Luckyo ( 1726890 )

        That raises an interesting question. Who are their paying customers?

        • That raises an interesting question. Who are their paying customers?

          Most likely PHBs who think a phone call is a more reliable source of support than Stackoverflow.

        • Re: Don't forget your competitive advantage. (Score:5, Informative)

          by mr.morbo ( 6346556 ) on Sunday April 12, 2020 @09:22PM (#59939070)

          Company I worked for about 10-15 years ago was a paying customer. It did buy pick up the phone support and direct access to their engineers.

          The library was incredibly buggy at the time so we opened a lot of tickets with them.

          Their engineers seemed quite clued up and bugs were closed really quickly. Patches were delivered to us to get us working ahead of the next release containing the fix. They were also pretty quick to suggest alternate - sometimes better - ways of doing things to avoid the bug behaviour.

          Way back then the support was worth the license cost. Can't speak to whether that remains the case.

        • Re: (Score:3)

          by darronb ( 217897 )

          I wish I could have been a paying customer.

          I did quick and dirt ports of Qt to the original Kindle and PalmOS... used it for years after the switch to LGPL... always intended to buy a license when it made sense.

          Finally, when I stopped contracting (where requiring a paid license would be a competitive disadvantage), I thought "ah ha! I can finally buy a license".

          Nope.

          Not only did they price themselves into car-payment territory, the real killer is the license detail that once I switch to commercial all my

          • What is the problem with iOS?

    • Java already exists as a cross-platform toolkit for UI development

      WTF? Java is a turd that works BEST when it is kept away from a user interface

      and Microsoft is really pushing to get there with the .net core.

      WTF are you on about? There is no UI library for .net core at all.

      You literally posted nonsense for the early karma points

      • I was thinking of the inclusion of WPF, as listed here [microsoft.com]. But I had it in my mind that this was cross-platform. I didn't realize that it was windows only. Be that as it may, your statement that there is "no UI library for .net core at all" is incorrect.

        And your statements that "Java is a turd" and that it "works BEST when it is kept away from a user interface" are obviously just matters of opinion, not facts.

        So your accusation that I "literally posted nonsense for the karma points" seems pretty unfounded.

        • > And your statements that "Java is a turd" and that it "works BEST when it is kept away from a user interface" are obviously just matters of opinion, not facts.

          Try to make an argument that Java is not a turd, without bringing in sever-side Java.

          Yeah, it's a turd client-side.

        • Be that as it may, your statement that there is "no UI library for .net core at all" is incorrect.

          No, it is correct. .Net core is the "back end part" of .Net and *contains* no GUI libraries.
          However under windows you have .Net Gui libraries which obviously use .Net core.

          So on linux and Macs you can only write CLI programs, aka servers or command line tools.

      • Java already exists as a cross-platform toolkit for UI development WTF? Java is a turd that works BEST when it is kept away from a user interface

        About all I find Java good for IS the user interface, even if it is junky clunky and weird for most UI work, at least it's portable. Using Java for back end data processing is stupid if you ask me. It's a resource hog which is slow in intermittently goes off and does it's garbage collection thing without so much as laving a "I'll be back from lunch in 15 ms" sign in the window: when it's taking out the garbage, you wait. I don't like that. Yea, it's fine for UI's where the human can barely perceive the d

      • WTF? Java is a turd that works BEST when it is kept away from a user interface
        And why would that be the case?
        Swing and JavaFX are both excellent. Hard to imagine anything better.

    • Java does not run on iOS - Qt does.

  • and do a code cleanup, get rid of all those examples and other cruft & kludge that make it unnecessarily bigger package after its compile, just the library and plugins, just enough to make the qt based apps run good, anything more is just bloat

    • "Bloat" is what people call features that others use.

      • "Bloat" is what people call features that others use.

        Also, Qt has a download engine that lets you pick exactly what you want to download. If you just want the C++ library, it is fairly lean. PyQt doesn't add much. The mobile (iOS and Android) libraries, web stuff, and GUI "creators" are where the real bloat is.

  • Open source project tries to turn the corner to profitable, then last open version gets forked. Sorry, no profit for the originator!

  • "There is no master but Master, and QT-1 is His prophet."

  • Fracturing the Linux desktop (Score:5, Informative)

    by guacamole ( 24270 ) on Sunday April 12, 2020 @03:29PM (#59937892)

    These asshats were already responsible for fracturing the Linux desktop. If people don't recall the whole story, and that was a long time ago, in the late 1990s the QT developers and their allies built the first version of KDE WITHOUT open sourcing the QT library. This brought the Linux community to brink of war. Some distributions already started shipping KDE. Eventually, GNU and it's allies developed and released an early super-alpha version of GNOME desktop some time in 1998. Only after thar, Trolltech (that's literally the name of QT company) released QT under GPL license, leaving the Linux community fractured forever. Thank you very much, trolls.

    • I'm wondering why gtk/gnome still survives after qt going gpl? Probably there are other reasons for this? Most likely it's the programming language since there is more people skilled in C among free software devs than C++, so Qt/KDE remain relatively shortstaffed.
      • trust! or in the QT sense the lack there of.

      • Re:Fracturing the Linux desktop (Score:4, Interesting)

        by weilawei ( 897823 ) on Sunday April 12, 2020 @04:15PM (#59938058)

        I'm still using MATE, however many years later it's been since the debacle that was GNOME 3. Works great. It doesn't get in my way or try to look pretty, and it's slowly picking up features that prove themselves useful over many years in other desktop environments.

        Still works, ain't broke, they haven't gone on some religious crusade with their codebase or the UI.

        Bottom line is is works, and it works very well after all these years, so why would I switch?

        Wouldn't you rather your cars interface to the steering and throttle was relatively conservative? Or would you prefer a redesign every 6 weeks where they randomly decide you shouldn't be using your feet to control the throttle and brake, but using voice control for it?

        • From the average user's point of view there is no real difference between KDE and GNOME. Most humans just point and click on buttons. Drag and drop works between most Qt and GTK+ apps, so what do they care? They both implement essentially the same steering wheel and pedals, and what do users care if the headlight switch is different in every car brand (as it is today) if it's essentially similar and located in more or less the same place? They're just going to leave it on automatic anyway.

          P.S. I also use MA

      • Mostly it's inertia, there's too much code base for a ground up rewrite of GNOME. Outside of GNOME apps you almost never see GTK used as a goto cross platform toolkit.

        As for Qt/KDE being short staffed, they've kept a consistent well regarded desktop with features GNOME still doesn't have, with only 1 fork to make the people who couldn't bear to move past v3.5. GNOME seems to get forked every other week just to appease the egos of various distro maintainers.

      • Re: (Score:2)

        by Octorian ( 14086 )

        I think programming language choice is the reason. Sure, the license spat is the official reason that gtk/gnome really exists. However, at the time, I got a very strong impression that the "real" reason was that many F/OSS developers hated C++ with a passion. (and thus were perfectly comfortable recreating everything C++ does in C, using conventions and macros).

        Of course C++ dynamic linker performance used to be really bad back in those days, and GCC wasn't terrific at it back then either. Things have chang

        • Sure beat when Guac/MS developed .Nyet and then pushed that garbage on GNOME. The last thing I wanted to see was a horrible C#.
      • Guac and his buddies worked with MS to create .Nyet on GNOME and then had huge funding from MS to help keep the Linux desktop fractured. It really was more about GNOME and MS than it was about KDE.

      • The real reason for GNOME/GTK survival is that KDE/QT owners waited until GNOME was developed into a full-fledged and a usable desktop before finally making the decision to open-source QT. If QT was open source to begin with, there would be no GNOME. But once GNOME and GTK were out, clearly they became unstoppable. Another reason is the difference between GPL and LGPL (used by GNOME libraries). This means that if you want to develop closed-source commercial software, you are free to do it with GNOME. But

    • Re:Fracturing the Linux desktop (Score:5, Informative)

      by Daemonik ( 171801 ) on Sunday April 12, 2020 @04:04PM (#59938016) Homepage

      Pfft, the Linux desktop was already fractured, or are you ignoring NeXTStep, Afterstep, CDE, Enlightenment, TWM & OpenLook? Everyone transitioning from Unix to Linux brought their favorite desktop environments along with them.

      Furthermore, while Qt wasn't GPL while KDE was in beta, it WAS free for non-commercial use, which suited the majority of Linux users at the time. Regardless, the KDE Free Qt Foundation secured Qt as open source prior to KDE 1's release.

      Also, it's hilarious that KDE has remained under control of the same foundation since it's creation 20 years ago, with only a fork for the people who wanted to stick with the 3.x base while GNOME has been forked a bazillion times.

      • Also, it's hilarious that KDE has remained under control of the same foundation since it's creation 20 years ago, with only a fork for the people who wanted to stick with the 3.x base while GNOME has been forked a bazillion times.

        The lack of interest in KDE is hilarious?

      • "or are you ignoring NeXTStep, Afterstep, CDE, Enlightenment, TWM & OpenLook?"

        Sun abandoned openlook and went to CDE (Motif) so that fracture was healed before the Linux desktop was ever really a thing. twm was never a desktop, only a window manager. Enlightenment only tried to be a desktop for a brief moment, and it never tried very hard. By NeXTStep perhaps you mean OpenStep, that was a thing. And there was never a CDE for Linux back in the day, although Caldera did have a Motif-based desktop. So real

      • Most of the things you mention are not a desktop environment. Afterstep, Enlightenment, and TWM were just window managers. Nextstep was a desktop environment, and a very good one, but it was neither open source, nor available on Linux. I believe there was an effort to develop an clone of Nextstep for Linux, it was called Openstep, but it remained in a perpetual alpha state, and now apparently dead. Openlook is a joke from 1980s or early 90s. Nobody bothered with free version of openlook because it's like ad

        • I was saddened when Sun abandoned Openlook in favor of CDE.Openlook was clean and beautiful. CDE was cluttered and ugly.

          From a development perspective they were more-or-less the same, but the Openlook UI was nicer.

        • Nope OpenStep is something different, it is NeXTStep that also runs on hardware that is not from NeXT. What you are thinking of is GNUStep. One always felt if the effort that went into GNOME and KDE had gone into GNUStep the Linux desktop could be just like a Mac these days. That might or might not be a good thing. Would certainly have opened up the door to ports of comercial apps to Linux from OSX. The problem was the cathedral approach of GNUStep compared to the bazaar approach of GNOME and KDE. Oh well w

    • > fractured forever. Thank you very much, trolls.

      Competition is a feature, not a bug.

    • Re: (Score:3)

      by mridoni ( 228377 )

      Except for a few components, Qt is LGPL, not GPL. That makes it even harder for the Qt Company to profit from it, and the fact that their pricing model is rather insane (US$ 5,000 per seat/year for companies over US$ 250K of revenue, US$ 500 for startups) doesn’t really help.

      • Those components are essential, and those are GPL. This means closed source commercial software can't be developed using the free QT library.

    • These asshats were already responsible for fracturing the Linux desktop.

      You're being rather biased here, as Trolltech was responsible for making Linux GUI development saner than it had ever been before.

      Eventually, GNU and it's allies developed and released an early super-alpha version of GNOME desktop some time in 1998.

      "Captain, there be asshats here!" (sorry, Scotty).

      Only after thar, Trolltech (that's literally the name of QT company) released QT under GPL license, leaving the Linux community fractured forever.

      Qt was bound to be GPL'd at some point, regardless of the existence of GNOME (which is really where the desktop schism made its worst marks). Having said that, Qt's original license was pretty close to GPL anyway. The licensing concerns, while valid, were bound to be worked out at some point. After all, KDE was the poster-child

    • "Fracturing"

      You seem to think this is Windows, and miss *the entire point* of Linux.

      Choice! Individualization! Freedom!

      I wonder what you would think, if you heard how Linux is developed:
      *Everyone* has a fork.
      Every developer. Every distribution. Every advanced user.
      It is inherent to the whole model.

      Yeah, they were asshats. But giving us more than a monoculture like with Apple or Microsoft, is not the reason.

    • Re: (Score:2)

      by Kjella ( 173770 )

      At least Trolltech had a business model so they could pay people to work on it and not be complete junk like GTK. That's what really lit the fuse here, they were going for a dual license model where you'd have to pay them to write proprietary Linux apps or use it as a cross-platform library since only the Linux version was free.

      They updated the license from FreeQt to QPL to GPL but to make your code part of the official project it still required a contributor's license so they could use it in their commerci

      • At least Trolltech had a business model so they could pay people to work on it and not be complete junk like GTK.

        If you believe GTK is junk, that's fine, but that has nothing to do with the license. The GNU software foundation insisted on a specific model for GTK, so it was a technical decision, not because of the license.

    • No, I would argue that it was the GNOME world that fractured things.

      • If GNOME did not happen, then there would be on open source desktop for Linux. GNOME fractured nothing. It's the greedy Trolltech folks who waited until the last moment to finally open source the QT. If QT was open source to begin with, then KDE would probably the only full fledged and viable free desktop to this day.

    • Re: (Score:2)

      by HiThere ( 15173 )

      Actually, that fracture has been very beneficial. Whenever the Gnome folks decide to make the code useless, KDE is available. I've had to use it several time. Gnome3 still isn't as good as Gnome2 was, and KDE (whatever the current one is) isn't as good as KDE3 was. At the moment I've decided to let someone else deal with the problem, so I'm using Wx (which depends on Gtk).

      But the purpose behind it sure wasn't benevolent.

    • These asshats were already responsible for fracturing the Linux desktop.

      What? Are you saying having many choices isn't a good thing? That is sacrilege and you can be burned at the FOSS stake for such things.

    • Sorry, that summary is completely wrong.

      KDE had a special license to use QT. If there was any concern about it then it is the GPL maniacs about no one really cared. There is no "forever fragmentation".

      If one is so arsine that he prefers the gnome toolkit over a real framework, it is his own fault.

      Thank you very much, trolls.
      The troll is you ...

  • They do not care about a 1:1 relationship between work given and money paid. All they care about is "What the market will bear.". Whar you will bear. If it means walking over dead bodies, they ask "how much will I make". Always "make". Never actually "earn", except as a side-effect.

    That is why all lisences of all your works should be GPL3 at the very least. And actively imaginary-"property"-crime /destroying/, ideally.

    Only rest, when everyone who sides with stealing and racketeering amd terrorizing like tha

  • I have not been able to develop with Qt/KDE, but hopefully, they will not have to fork. This is not good for anybody.

  • Qt is a bloated heap. Start over. (Score:4, Interesting)

    by OneSmartFellow ( 716217 ) on Sunday April 12, 2020 @06:49PM (#59938542)

    It's an OS hiding inside a development library.

    It relies upon layer after layer of macros and duplicates (poorly) functionality in the STD c++ library.

    I used to be a fan, until Qt5, when it was split into a million pieces, requiring inconsistent project depends manipulation, and causing a headache for users trying to install the equivalent runtimes

    Yeah, hate to say it, but Qt is a dead horse.

    I guess the higher-ups recognize that and want to try to cash out before it all implodes. Can't blame them really. A shit ton of hard work went into it
        It used to be awesome.

    Sad story.

    • Thin variant of Qt would be really nice. Drop custom containers and rely on STL instead. Drop Moc and rely on modern C++ and template metaprogramming.

      • I would love that, a little piece of me dies every time I have to convert a QString to a std::string or vice versa, have to do a clean & rebuild because because the compiler decided I didn't need to re-moc, or have to explain to PFYs the joys of forcing all UI object manipulation onto a single thread.

    • Your computer is an operating system that is running a windowing system that is running a browser that is acting like an OS and windowing system and you are complaining about QT? You seem to be a wee bit short sighted.

      • I don't need a "graphics library" to provide strings, collections, threading, asio, file handling, syncronization objects, state machine templates, etc, etc, etc.

        I want it to provide a convenience wraper for the shit-show that I get invoved with when trying to develop cross platform Windowing apps. Hell, it doesn't even need to provide OpenGL. That's pretty straight forward too, and pretty damned platform agnostic.

        It only has to be relatively up to date with the C++ standard - let's say, at this point C++

    • Re: (Score:2)

      by Megane ( 129182 )

      My problem has been how they aggressively deprecate older versions of OS X for no particularly good reason except the usual millennial reason of "old thing bad". Then when someone makes a build of some open source thingy, it always requires a relatively recent OS version. I even had one such app that instead of saying "sorry, this won't work on your OS", it would simply crash on launch.

      Now I see mention in these comments that their LTS versions are non-free? I guess that might explain it.

  • What about the "poison pill"? (Score:3)

    by GuB-42 ( 2483988 ) on Sunday April 12, 2020 @06:54PM (#59938564)

    The Foundation has license agreements with The Qt Company, Digia and Nokia. The agreements ensure that the Qt will continue to be available as Free Software. Should The Qt Company discontinue the development of the Qt Free Edition under the required licenses, then the Foundation has the right to release Qt under a BSD-style license or under other open source licenses. The agreements stay valid in case of a buy-out, a merger or bankruptcy.

    https://kde.org/community/what... [kde.org]

    • You're going to need to lawyer the crap out of that. Unless the time frame is also specifically called out. From what I can tell none of that precludes a 12 month delay in release to the general public under the open license.

  • Pay for software (Score:3)

    by DidgetMaster ( 2739009 ) on Sunday April 12, 2020 @06:58PM (#59938574) Homepage
    Back in the old days, a decent software developer could build version 1.0 of something useful and get a few thousand people to pay $10 or $20 for it. That was enough to prime the pump and get enough revenue to build additional features for version 2.0. A good software startup could actually 'bootstrap' their business this way. Nowadays everyone expects tons of features in a first version, expects it to be fully tested in an enterprise environment, and get it all for free! That is a tough environment for any startup to tackle without a ton of upfront funding.

    I am building a product using Qt right now. If I can get some revenues (I am a startup facing those challenges too), I would love to send some of it to Qt because I want the software to get better and it adds real value to my product. I just want it to be a reasonable amount. Right now, there seems to be only two choices...free or very expensive. Something in-between would be nice.
    • It's consequence of economics. If you treat software as a product based on copyright license then market forces will eventually move the price to 0 since this is the actual cost of passively allowing someone to copy your software. And earnings before it hits 0 are subject to variety of random factors, unfortunately nowadays software quality and effort put into development aren't most important ones of them. Alternative monetization model is required which directly exchanges money for development effort, rat

    • Re: (Score:2)

      by dacut ( 243842 )

      I just noticed that Qt does have a startup edition that's available for $500 (instead of the usual $5000). It's the same as the regular commercial edition, but only sold to businesses with under $250k in yearly revenue.

      It's a bit screwy if it's an embedded (non-mobile) device, though; seems they want a per-seat fee on top of that.

  • Kill the golden goose.

  • Serious question. My understanding is that because they release code under the GPL, and open-source contributions by others (also released under the GPL) have made their way back into the official code base, The Qt Company is now bound by the GPL.

    Wouldn't releasing a derivative work, without releasing the code, be a violation of the license?

  • Copperspice (Score:4, Insightful)

    by The Evil Atheist ( 2484676 ) on Monday April 13, 2020 @01:12AM (#59939688)
    I'm not affiliated with them in any way, but Copperspice is not only a fork of Qt, but it removes the horrible Qt moc crap. I think they've also improved performance a bit by removing old interfaces that uses lists and what-not.
  • but I don't see how kde could not be further developed with qt not changing for a year.

Slashdot Top Deals

If all else fails, lower your standards.

Close