Open-source contributors say they'll pull out of Qt as LTS release goes commercial-only The Qt Company has followed up on its plan to make long-term support releases commercial-only by closing the source for 5.15 today, earning protests from open-source contributors who say that the 6.0 release, which remains open, is not yet usable. Qt is a cross-platform application framework available both under open-source …
This post has been deleted by its author
But forking a now obsolete old branch just to support existing non-commercial users doesn't sound like the sort of exciting new project which will attract may contributors.
It would make more sense to fork 6.0 and say *BEEP* to the Qt company and go on your own - especially since they have shown that they aren't planning to play nicely in future.
Why not fork this one? It's under full support.
$ rpm -qi qt
Name : qt
Epoch : 1
Version : 4.8.7
Release : 9.el7_9
Architecture: x86_64
Install Date: Thu 12 Nov 2020 12:55:44 PM CST
Group : System Environment/Libraries
Size : 17895063
License : (LGPLv2 with exceptions or GPLv3 with exceptions) and ASL 2.0 and BSD and FTL and MIT
Signature : RSA/SHA256, Tue 10 Nov 2020 11:39:49 AM CST, Key ID 72f97b74ec551f03
Source RPM : qt-4.8.7-9.el7_9.src.rpm
Build Date : Tue 10 Nov 2020 11:32:07 AM CST
Build Host : jenkins-172-17-0-2-3664c536-7d7d-4ac4-8b0e-26767e19daa3.blddevtest1iad.osdevelopmeniad.oraclevcn.com
Relocations : (not relocatable)
Vendor : Oracle America
URL : http://qt-project.org/
Summary : Qt toolkit
Description :
Qt is a software toolkit for developing applications.
Or the label on the box warned of a long QT interval
Qt is GPL-licensed
actually Qt is a proper mish-mash of licenses: https://doc.qt.io/qt-5/licenses-used-in-qt.html
i've used Qt on and off for many years now, so this change is all a bit sad
does anyone here have any experience of CopperSpice (https://www.copperspice.com/)? been thinking about making the jump myself for a year or two now, curious to hear of anyone else's experience.
it's not impossible to have dual licensing, a GPL license for any open source project that distributes it, and a private non-GPL license for people who want to ship binary-only versions. since the creators own the code they can do what they want with it, pretty much, even if the two licenses conflict.
Seriously I like to offer both a BSD-like or MIT-like license along with GPL for stuff i put "out there" as open source, and THEN give whoever distributes it the choice of which open source license to use. Wanting to control how people use something (you can't play with MY toys unless you do it MY way) isn't very "free", In My Bombastic Opinion.
It's sorta like: if you give a gift and then dictate (too many) terms on its use, it's not a gift, it's more like a lease.
Shame really, I actually prefer KDE to Gnome these days. The QT mouthpiece commented that he expected people wouldn't be doing anymore free work, I don't think he realized that was for 6+, not just the recently closed 5.15 version. The company has just lost a lot of developers, reviewers, etc.
Qt has always been KDE's Achilles heel. In fact I would go so far as to say that Qt is the main reason Gnome, XFCE, etc. are (as a group) more widely used -- KDE tends to finally reach a stable and just about usable point right as Qt forces some major shift (API breakage, licensing changes) that throws KDE back into complete chaos for yet another decade.
Wonder if the KDE developers will ever cotton on to this fact and start maintaining their own open-source version of Qt? Or will KDE continue on as a mere tech demo of old Qt technology?
I was _EXTREMELY_ disappointed when KDE appeared to go "all 2D FLATSO" like Win-10-nic and the chrome browser and "Austrails". Is KDE's 2D FLATTY-ness a direct result of changes to Qt? Because if that's the case, what's the point of anything 3D (like best-case use of 3D acceleration) in future Qt versions, if it's *IRONICALLY* 2D... [and OpenGL is still "a thing"]
Absolutely right that you can fairly easily install another KDE theme (it's actually not quite that easy, because in KDE's infinitely configurable way, you have to choose different widget styles, different window borders, and *at least* one other thing somewhere else in the settings that I have forgotten, before you can expunge most/all traces of flatso), but I do somewhat agree with Bob that to ship as default a theme that is clearly in the style of W10 and only very slightly less fugly than W10 itself (which really is so eye-searingly ugly it makes twm almost look good) is a really poor aesthetic decision.
The impetus to create the GNOME desktop environment was KDE's reliance on Qt from Trolltech, as well as a preference for C over C++[1]. At the time Qt was available in source form, but modifications could not be distributed. GNOME adopted GTK (or GTK+ as it was then) from the GIMP project, and this eventually encouraged Trolltech to re-license Qt to allow distribution of modifications. And now twenty years later Qt goes closed source :-(
[1] It's far easier to create bindings for other languages if the library is in C rather than C++.
if you make shared libs with C++, at least make sure no symbols are exported that aren't declared 'extern "C"' especially if you want 100% compatibility between, let's say, both CLANG and GCC applications using it...
what you do INSIDE the library should be abstracted and encapsulated, anyway. Anything ELSE would be bad programming habits.
then other languages (Python, Perl, etc.) could have bindings to your library without too many hoops to jump through.
I wonder how long it will be before there is a rethink about this? I reckon hours not days as this sort of scenario seldom works in favour the the company adding the restrictions. It does not pay to play not-nice with the Open Source community.
When Oracle stirred the pot with MySQL it spawned Mariadb pretty quickly.
KDE are too heavily invested in QT to go down without a fight.
Watch this space - tomorrow!
This would be the same Nokia that reported Net Sales of 5.3 billion Euro in third quarter of 2020, with a operating margin of 9.2% and cash and investment assets of 7.6 billion Euro?
I wish my finances were run into the ground to the same extent.
The only part of Nokia that was run into the ground was the cellphone division which they sold to Microsoft.
The company called Nokia nowadays has little to do with the one that was run into the ground. The phone division is what made Nokia a household name and is gone. The networks division doesn't have much left after the Siemens, Alcatel and Lucent mess.
So, I stand by the GP: Nokia was run into the ground. And, yes, it looks like the Qt people are going the same way (quite ironic given the history of Qt).
I don’t think you realise how terrible those balance sheet figures are.
Operating margin of 9% is horrific for a tech company.
A more normal operating margin figure is 30-40% - e.g, ARM runs 50%, Microsoft runs 32%.
For starters, operating margin excludes bank and bond interest cost-of-capital, so the real take-home for the company is more like 3%, which is only a bit better than break even in the short term. This assumes you can continue earning at the same level, with the same products, indefinitely, and that’s where the problem is particularly for a tech company.
In the longer-term, you need to allow for depreciation of the business as a whole. With a company like Nokia, if they just decided to stick to 4G and ignore 5G, the entire value of the company would be gone in 8- 10 years. I’m not saying they will, or have, but the inverse of that (10-15%) is the amount of continuous investment they need to maintain their earnings at current levels.
The figures tell us two things:
#1 Their real earnings are less than their investment needs, which means ultimately an investor who held “forever” will lose money in the long term from here - taking income, but losing their stake finally whenever the company dies in 10-20 years.
#2 While they are investing *something*, Nokia investment *needs* are clearly greater than what they have (from operating margin). So, they can’t be investing sufficient to maintain their income. So they will shrink as a company, only this doesn’t tell us how fast.
Note that I’m *not* saying their strategy is wrong - this might be the least-worst from where they are today.
In the long-term, it’s only even break-even can continue earning at the same level, with the same products, for the next 35 years! Think about it - I have £7bn today, and if my business limps along without anything new for 35 years until my products are finally toast, at the end of that my business is worthless. That’s a business depreciation of 3% annually, which needs to be compensated by short-term profit.
The Problem is that operating margin is the day-to-day profit of selling what you make, it doesn’t allow for anything new. There’s no real R&D in there, to produce genuinely new income streams. This company is a cash cow with fast eroding revenue as new technologies slash into it. Finance people often misunderstand this, because the definition of operating margin includes “R&D headcount”. But that’s R&D in name only to gain tax credits. It does include engineers to coding maintenance features - e.g. to add a Release 18 feature
We'll pretty soon see all open-source projects move over to wxWidgets as the relation with Qt becomes untenable. I've always though the relation to be fragile and fraught with danger since the move towards closing off the free Community versions has had a long history.
I've always been a user of wxWidgets, which has time and again affirmed its open-source commitment and is used in many projects, although invisibly to most users.
I do wonder where this will leave the KDE project, which has relied on Qt for over 20 years now.
Wx is awesome, although its nowhere near as pretty as QTs almost infinite skinability (specifically with the QML stuff. However with some work a professional looking app is possible although it might make the graphic designers amongst us cringe.
The problem is QT is more than just the widget its an entire VCL/dotNet style framework with a superb networking stack , great IPC apis, and so much more.
If you need skinniblity you can always use the wxUniversal version, which is "self-drawn", all the other versions use the native widgets of the platform it's running on.
Although I agree QT is much further along in this field and specifically designed for embedded. There's even a Boot-to-QT version which has its own Linux kernel and boots straight into your QT app.
There are also newer OpenGL based UI's which are coming along nicely but haven't yet reached the maturity of Qt.
In the end it will be a trade-off between price and time-to-market. If you're willing to spend the money and want to be quick to market then Qt is the way to go. If you're making large number of devices where licensing matters for your profitability you can use Linux with some newer OpenGL-based UI frameworks.
I've watched this discussion rise up periodically for about 20 years. I'm not on my desktop but I'm fairly sure it is using Qt 5.24 so I'm not sure I understand all this blah about 5.15 or 6.0
It all feels a bit tabloid
For the licencing, it might be worthwhile going to original sources. The following is from the Trolltech FAQ...
The KDE Free Qt Foundation is an organization with the purpose of securing the availability of the Qt toolkit for the development of Free Software and in particular for the development of KDE software. The Foundation was originally founded by Trolltech and the KDE e.V. (the legal non-profit organization behind KDE) in 1998 and it has a license agreement that ensures the availability of Qt under LGPLv3 and GPLv3 licenses for major desktop and mobile platforms. The license agreement has been updated couple of times over the years, mainly because of acquisitions around Qt or updates to licenses and platforms.
It's also worth looking at KDE.org (funnily enough) and their page KDE Free Qt Foundation.
While it is true that Gnome was started because of licensing concerns about Qt it is also true that the foundation was specifically set up to address those concerns.
Everything looks fairly clearly explained and covered off to me. Nothing I've read here indicates that anyone has bothered to check.
This has been a long time in coming but I believe it's finally about to happen: the ditching of the free Community version and the transition to a closed-source paid-for model.
There's always some bean counter inside the company who'll argue that by shutting down the Community version they'll get more paying customers which will increase profits. And management is too daft to see the repercussions of this or only thinking of their bonuses and how this will enable them to buy that new Ferrari.
Let's face it: greed has been the downfall of many, many companies.
One of the best things I like about wxWidgets is that it's possible to port an MFC application to one that uses wxWidgets if you understand the differences well enough. Other than names of functions, which could be a set of 'sed' lines in a shell script, you have to alter how windows messages are handled as 'events'. it's similar but not the same, and requires actual though to re-write it, but I've done it a couple of times and I like the results.
As a result, if software had been written in C++ using MFC for Windows, chances are a Linux version or a portable version that uses wxWidgets for both windows _AND_ "everything else" could be practical.
"That's the point of capitalism evil capitalists
fixed it for you.
the point of capitalism is for people to earn something of value based on the value and quality of their work, and then use that 'something of value' (like money) to purchase goods and services, etc. the way that human societies have worked since prehistoric times. It has NOTHING to do with exploitation. Evil, on the other hand, has EVERYTHING to do with exploitation. And that's the point.
but whether the people behind Qt's heading-towards-closed-source maneuver are evil capitalists... that will most likely become obvious at some point.
" the point of capitalism is for people to earn something of value based on the value and quality of their work.."
No , capitalism is a system where the profits are returned to the providers of capital - i.e. owners, shareholder, investment bankers.
Its objective is the exact opposite of the above statement.
Workers do not provide capital, and don't have any right to a share of the profit.
'Closed source: support is a cost, so the financial incentive is to write good code with clear documentation.
Open source: support is a profit centre, so the financial incentive is to write bad code with obscure or no documentation.'
In a binary universe the above might be true, but this is not a binary universe.
That is why us pinko commie "don't understand business" license purists insist on free software (free as in freedom) instead of open source which is a much wider term including free software, technology lock-in trap licenses and "you can contribute but only we can distribute" licenses.
It still hasn't got through that much, probably most, open source code is developed by companies. The usual incentive is that it makes sense for companies to be able to collaborate on something mutually beneficial.
Qt may be a bit of an exception here in that it's always been primarily developed by a single company and hence its Achilles' heel, but I don't suppose that fits your argument too well weither.
QT is a deeply frusturating thing from a dev point of view. Its hands down one of the most stress free C++ frameworks out there. The QML can create beautiful UIs, the infrastructural stuff is extremely well designed, even a C++ beginner ought have no troubles building software in it.
But the company itself has apparently no interest in the open source world that gave birth to the very thing they are profiteering from. They've removed small/one-man-band developer licenses and have priced things in a way that absolutely no rational business case could be made for using it for mobile dev (an area where the market is incredibly dicey). I've recently abandoned a big project just because the business modelling I did simply could not make the sums add up once QT licensing fees are introduced into the equation.
Nokia and Microsoft did something horrifying by allowing QT to be purchased by the people that purchased it.
The principle is that the source is free but people pay for maintenance. So LTS version are 100% meant to be paid for; they are maintained! I'm using 5.15.2 at present, there is absolutely no reason why that, or, indeed, any pre-LTS rev shouldn't continue in open source development. It's just numbers - 5.15.2.1 etc ad nauseum.
Indeed. For years, OSS advocates have been saying the income source for software should be the maintenance side of things. But as soon as someone implements it, it all goes to shite.
So where are the companies that run these projects going to get the money needed to stay afloat? And we really need over arching control, because otherwise the devs just go off and do what they want, not what needs to be done - because maintenance is so "BORING".
"But as soon as someone implements it, it all goes to shite."
Yes, Red Hat went to shite in a big way. Billion dollar annual turnover since 2012, two billion since 2015, three billion since 2018. They went to shite so badly that IBM had to rescue them by buying them up for $34 billion. It's a business model that really fails big.
> But as soon as someone implements it, it all goes to shite.
Because that's all that they're complaining about, the closing of the source?
Not in fact that the source is being closed with an "oh you can just version 6" when version 6 isn't fully working *and* there are breaking changes coming?
I have developed with Qt since version 4.7 (~2011), both in commercial and personal projects. During the 4.x days, Qt seemed to be mostly getting better. It was still a bit rough around the edges, but it had a lot of good stuff which the alternate solutions (mostly WxWidgets and GTK+) could not match.
Come the 5.x releases, and I feel like Qt is somehow focusing more on the flashy bits, and less on the technical side, including fixing of bugs. Back in 2013 I was pretty miffed to find that networking on Windows was somewhat buggy, leading me to write my own drop-in class (NNetworkSocket) to make my code work: https://github.com/MayaPosch/Nt
Has anything gotten better since? Qt proclaims to allow for easy portability to mobile platforms like Android, but this mostly a delusion, as I found out: https://bugreports.qt.io/browse/QTBUG-83372
For me being able to look at the code isn't so much the issue. I have looked at the Qt code, and I think it's a horror show that I would not want to touch voluntarily, unless I got paid a lot to do so. This seems to be sadly reflected in the overall quality of the end product, as well as the documentation. I think it's still tragic that the blog article I wrote on using QThreads back in 2011 is still one of the most popular references on the topic: https://mayaposch.wordpress.com/2011/11/01/how-to-really-truly-use-qthreads-the-full-explanation/
So what's changing with Qt? More flashy from what I have seen. Also more CMake instead of QMake, when all that many of us devs wanted (and were expecting) was the demise of QMake with the arrival of C++11 so no preprocessor would be needed any more for the Qt side. None of that seems to be happening, and CMake is unlikely to be ever part of my development flow.
All of this seems to lead to me to look at alternatives to Qt again. Over the years I have already switched everything but GUI stuff from Qt to LibPoco (POCO project). My experiences with WxWidgets in KiCad have made me pass on that framework. So what's left? GTK+? Going native? Write Yet Another Framework?
No easy answers here, it seems.
Too bad! Qt has always had free and commercial releases, but I know I'm not interested in contributing patches to a project that is going to wait until the code is in pretty good shape (due to contributed patches), then close that branch off, expecting people to help fix up the next unstable branch (which will then be closed off once it's in good shape.)
That said... Qt has a big market in in-car entertainment systems, embedded systems, and so on, so I'm sure they'll do fine on their own.
I think that a per ent of open and closed QT developers / users will make the move to flutter.
The two are both OS widget independent and I used both for many years and moved to flutter and it works well on web, desktop, mobile and rasp pi and other small boards.
You can use all the c code legacy you have by using the FFI features of Flutter.
I tend to write all core code in golang and use FFI to expose that to flutter
Long time reader, first time poster. I just couldn't resist given the negativity.
There seems to be a very basic misunderstanding expressed in many of the comments above. While Qt is open source (for some definition of open source) it has never been a community project. Almost all work on Qt over the years has been done by paid engineers working for whichever company happened to own Qt at the time. Many people have made contributions to it (myself included), but that does not make Qt the result of a joint effort of volunteers, who are now being cheated by a big, bad, greedy corporation.
If you don't believe me take a look at the commit log for any of the major source files in the Qt code base.
I regret this step by Qt (I have used it since 1998), but I understand it. It is not easy to survive as a software company these days (assuming you don't have an ad-revenue-generating model).
> The Qt owners want a PERCENTAGE OF REVENUE and you must negotiate it with them.
That is a killer, I agree.
> MSFT is laughing all the way to the bank.
I'm not sure about that. MS has pretty much abandoned the market segments where I expect Qt is making most of its money (embedded, automotive). I can't think of any contender to Qt in these areas. HTML5, perhaps, but that means that you need a modern browser on your device, and that comes with its own set of problems.
I once worked for an Ex Nokia manager as a developer. The guy did not really have a clue about the complexity of the work packages he built.
So experienced people woild be assigned trivial tasks while inexperienced engineers got tasks which should have been split in ten subtasks.
He would start a discussion about some issue and then basically tell you to shut up, because he did not like what he heard.
So, Nokia seems to be one of these places of Euro incompetence.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
