An Open conversation about the future of Qt.
-
It’s time for an Open conversation about the future of Qt.
There have been a lot of private conversations going on and I wanted to create a forum to talk about them publicly. To be clear, I am not in the loop with anything going on within Nokia and have no information beyond what is in the press.
I have great respect for the people who work on Qt, and I am grateful for all of the things that Nokia has done to build Qt and the Qt community. Yet, today we are at a point where Nokia’s corporate interest is diverging from that of the staff, and the community as a whole.
I have no doubt that there are people at Nokia who understand the community issues, and who want to do the right thing, as I’m sure there are others who are motivated by reducing expense and liability.
For Qt to thrive in the future, there are many issues which need to be resolved, including people, legal rights and finances.
I am but one small voice. I don’t claim to speak for the community. But I think if you want to be Open, you need to discuss the important things Openly. I understand that Nokia has owns IP and they have the right to dispose of them the way that they choose. But the staff and the community have a stake in the outcome and we each has our own choices to make.
I could go on for many pages on “What I would do if I were King of Qt,” and I will try to write it and post it. But for now, I want to get the conversation going in public. What do you think are the best and worst case outcomes? What are the obstacles that need to be overcome? Where do we go from here?
Peter Winston
ICS CEO -
Good luck with that Peter, the last week and a half I've been told opinions, voiced in this forum do not reflect the intent of Qt's developer base. Maybe you will have more luck doing this live on the next DevTalk, unless it gets cancelled or something...
Most people here seem to hope for Qt getting purchased by some resourceful and decent company, I've also read about forking Qt the last few months. IMHO the way to go for Qt is to focus on its advantages it has over other SDKs, cross-platformability and nativity to be precise, and build on those instead of trying to copy others, reinvent the wheel and try to turn Qt into something else. That pretty much sums the best and worst case scenario - best case a decent company buys Qt and puts it back on the right track, worst case is things continue to get worse, be that with the current ownership or a new one, and Qt gets forked - a last resort for the sake of saving it.
Me as well as many other users have been "moaning" about the future of Qt for quite a while, but nobody listened, it was "grim pessimism and conspiracy theories", now that naivety can no longer powder reality it gives me no joy to be the one to say "I told you so"...
-
I was offing to be such a home, or work with others to create a home. I don’t have the resources of a Nokia, but I think decent enough. Just as critically, I think I have a business model which will allow Qt to continue as an open project.
Your point about re-inventing Qt is the most important long term question in the post Nokia world, and I don’t have an answer, yet, Right now, I’m focused on the short term. How to align the people, process, and finances, to see what can be done.
The good news is that Qt is still very strong.
The developers using Qt never heard of the project that was killed, and they don’t care.
The world already believes that Nokia is a Microsoft shop, and Qt is on its own. -
We all have been dismayed by the recent events.
To solely bank on Windows Phone (and more or less Series40) is a high-risk bet, and there is hardly any indication so far that it will not be lost. But, as difficult as this is to admit for me due to the fatal consequences this decision might have for Nokia, it also permits Qt to part, which will be, as it is, benefitial to Qt.
On the quiet Qt beeing developed and used by Nokia was always a reason for the scarce commitment to Qt from other manufacturers, be it Google or Samsung. The future of computing is mobile, as well for Qt. And the future of mobile is "Android, iOS and Tizen":http://communities-dominate.blogs.com/.a/6a00e0097e337c8833016306291bad970d-pi. If Windows Phone 8 doesn't change tack (and there is "serious concern":http://communities-dominate.blogs.com/brands/2012/05/how-microsoft-lost-the-future.html it won't) Nokia's market share in smartphones will be "negligible":http://communities-dominate.blogs.com/.a/6a00e0097e337c88330167671c9e48970b-pi by the end of the year.
I understand that the future of Qt is a burning question especially for people with a substantial financial investment, but it is not an immediate question, as "Nokia will support Qt":http://lists.qt-project.org/pipermail/interest/2012-June/002530.html at least until Qt5 is released. I hope there will be an official announcement soon, but I understand that there is none yet, as they have been seemingly "caught off-guard":http://lists.qt-project.org/pipermail/interest/2012-June/002465.html as well.
I think we should be careful with jump to conclusions, because things will get clearer within the next weeks. But what we can do is expressing support for Qt, showing that there is a strong demand for it, and that investing in it will pay off.
The partnership with Nokia was a huge benefit for Qt, be it QtCreator, QtQuick, the LGPL and Open Governance, the certification and ambassador progam or the developers days and the contributor summit. Nokia has brought Qt in a position where it actually is an interesting investment for a large audience
I'm quite sure there will be a solution which is benefitial to Qt and its community.
-
The Qt SDK is open source, so it will be around for ever.
Now, I don't post much here because I prefer eye contact, still my two cents
a) I use Qt every day
b) plan to do so for until we are done.
c) Nokia has issues, yes. It's tough to be number one and lose ground, but I don't see the end. And I base that opinion on real news, like the WSJ, not ehmm say news 'makers'.Best to Nokia! ;)
-
I been thinking why Qt is such a great toolkit, IMHO better than any MS tools or others.
The answer to this question is in the official Qt4 book 2nd edition, in the "Brief History of Qt" chapter. It was developed with passion by two young developers who started their own company. Others companys can have also great developers that money can hire, but sometimes they rather be else where, or are bored working just to get the job done.
I once meet a ex-troll that told me that Trolltech was the best and fun place he worked, but that had changed a little with Nokia takeover.And I read recently in a troll blog that Elop had told that Qt should focus on QML only and deprecate all others funcionalities. So to me is clear that Nokia managemente has no interest in Qt has great cross plataform desktop toolkit.
Nokia management is not to be trusted, is just about MS lobby, why else would they kill symbian (and no symbian was not old and outdated has people say, if you think that, check nextstep or objective C or whatever name they call it, it's older than 30 years), why else would they not put N9 and N950 on the spot light, because they are truly the best phones out there, Qt is the best toolkit outhere, and Nokia had all the tools and ecosystem (whatever that maybe) they need, to conquer the world, before MS come along. I just read in the last days news that Nokia is working with MS to bring WP7 has fast has possible to new feature phones. So what's there on Nokia management about Qt future ? Please tell me because I don't know.
PeterWinston, in IMHO I think to create a new home for Qt, I would start by private talking to Qt greastest players. First Digia, they seem to be the best player right now, has they are developing Qt, and managing Qt license. It's important to figure out what kind of relationship they have with Nokia. Then Kdab, others forum onwers like qtcentre, contact the Trolltech founders they could have insightfull advices. Join a small group, talk first, measure all things including the poison pill, and if all of you together think it's time to act, then act like a fox. For the sake of Qt, and for the sake of the community.
In my personall opinion, I would not mind if Qt would have to loose the LGPL license to became a new profitale independ company again, focusing on desktop cross plataform, and on mobile android, IOS and (why not ?) WP7 development ( perhaps we should start a new pool about it :) ).Joao de Deus
-
I am not so sure about Digia, surely, they contributed with some bugfixes, but their overall input is negligible and it does seem they might be more into using Qt to milk its commercial customers without contributing any more than volunteers do for no pay at all.
I've always suspected Nokia of being the instigator of that complete focus on QML and stealing developer resources from other much needed areas, and while some people claimed QML has nothing to do with Nokia, and its appearance just happened to coincide with the purchase, but so far no one has explicitly stated that Qt would have the same direction and priorities had it not been under Nokia management.
I don't mind losing the LGPL license either, Qt should be free for open source projects, but commercial licenses should be more flexible, keeping income size and development team sizes in mind. For big players those THOUSANDS OF DOLLARS are completely justified, but for indie developers Qt should be more forgiving, with different licenses from 200 to 1000$ depending on the actual usage of the framework. Or why not a percentage of the income too? Like for example 10% of sales of Qt based applications go to Qt, but not in the pockets of some company that simply milks it, but actively invested in the development of the framework.
Something more - developers and companies who actively contribute without pay can get credits which get them discount from commercial licenses, or in the cases of significant contributions even the right to use the framework commercially as long as they contribute to it. It is only fair... -
Thank your for starting this conversation Peter.
I have great respect for the people at Nokia who work very hard to make Qt a success. But their
management can not be trusted at all. Elop takes all steps needed to make sure their is no alternative for Nokia but their current crash course. And unless there is external intervention (shareholders, government, EU, ... ) he will succeed and Nokia will collapse before the year end.So let's investigate the opportunities this brings for the Qt community. If we can make the right
choices this can make Qt a stronger framework.On the technical level, I would prefer Qt to get back to its roots and focus on being a cross-platform C++ toolkit for desktop applications. I've always found that most of the value of
Qt is inside the QtCore and QtGui parts, and that the other parts only take away development
focus and resources (Script, SQL, SVG, Webkit, Xml, Declarative), and would not mind if those
parts were dropped to make sure the Core and Gui parts can thrive.On the financial level, I would not mind paying a reasonable fee to make sure there is continued development of the QtCore and QtGui libraries.
Maybe there are other steps that can be taken to decouple the fate of Qt and Nokia. It would be interesting to get the ideas of the Qt management within Nokia, maybe they can post them in an
anonymous way. -
I am pretty sure the Trolls will appreciate a discussion very much. All the trolls I ever spoke to were very passionate about Qt, much more so than about Nokia phones. I am sure every developer there will be happy to see Qt go on and many will consider staying with whichever organisation will keep Qt going.
I would actually love to see a foundation take over development of Qt where commercial users can join, pooling their resources on a team of developers. Ideally those would share offices like they do now. At least from my experience that does speed up development quite a bit.
Sorry, but I do not want to trust any one company with the future of Qt. Nokia was not all bad, it did put quite a bit of marketing effort behind Qt and it did put open governance into place. But still I do not feel comfortable with the idea of having a new set of upper management muck around!
Something will need to happen fast though: The Trolls are damn good developers, I am sure many of them will find new jobs elsewhere pretty soon if there is nothing on the horizon to give them hope that Qt will go on and they can still have a place in this new future. Once a critical mass left development will be much harder for those that stayed and any new people.
The Qt Contributors summit next week would be a good place to discuss this issue in person. Most important companies in the Qt universe should be have at least some delegates as should Nokia. I would really appreciate getting some public updates here or on twitter or wherever. I hope Peter and others will follow up on this topic with his thoughts on what happens there and will follow twitter closely. Lets all hope for some good news from the summit!
-
I use Qt less than year but I immediately felt its potential. But for the development is need the financial stability . Many companies currently earn (and not bad by the way) on the sales of applications through online stores. According to this as an option of additional income, I propose to create open source projects, from developers around the world, put up for online shopping, sales revenues and donations collected in a single account of the Qt project. First need to do the following:
- Organize a collection of ideas. It will be a page or section in one of the Qt resource where everyone can post ideas and suggestions for future applications;
- Create a list of rules or guidelines for the establishment and maintenance teams who are implementing the application from the center of ideas;
- Create a single cash account for Qt.
I understand that this does not give quick results and may not attract sufficient funds, but it will be a practical step to financial independence of Qt community.
-
[quote author="Taamalus" date="1339821111"]Nokia has issues, yes. It's tough to be number one and lose ground, but I don't see the end. And I base that opinion on real news, like the WSJ, not ehmm say news 'makers'.[/quote]Nokia had an executive management problem before Elop, and it has been multiplied with Elop. This is no information which requires you to read (and agree to) Tomi Ahonen (who is, by the way, the most influential analyst in the mobile industry); it just requires you to read the "annual and quarterly reports":http://www.nokia.com/global/about-nokia/investors/financials/reports/results---reports/ or, if you prefer, the "Wall Street Journal":http://online.wsj.com/article/SB10001424052702303822204577468251129110174.html, which do not show a company under reorganization, but rather a company on "... its way toward irrelevancy", which is now "... primarily [evaluated] on the value of its assets, patents and cash reserves, rather than as an operating business ..." because of the "... zero value for the device [...] business ..." and "... little hope for a turnaround from here even with a refined strategy.".
[quote author="Erik Janssens" date="1339848604"]On the technical level, I would prefer Qt to get back to its roots and focus on being a cross-platform C++ toolkit for desktop applications.[/quote]You don't. The focus on mobile is essential for Qt. The increasingly majority of sold platforms is already mobile, not desktop. The emerging markets are tablets and integrated systems (smart home, including televisions, media and home automation), not desktop.
Whereas Qt was available on a majority of platforms (or basically any platform) in the last decade by supporting Windows, Mac OS and Unix, it is by now only available on a minority of platforms, keeping on falling behind. This will directly affect the appeal of Qt for a large audience; and it is for sure not "code once, deploy everywhere".
In addition, you will discard a large part of what has been contributed to Qt over the last three years, be it functionality or the attention that has been brought to a large audience of developers.
[quote author="tuxedo" date="1339858254"]I would actually love to see a foundation take over development of Qt.[/quote]I think a foundation beeing responsible for Qt is worth considering.
A membership fee funds the Qt project, including the infrastructure, the certification and ambassador program and the development environment. Members are allowed to offer commercial licenses, which are either used to directly generate income by providing support, training or commisional work or indirectly by having a specific platform supported in Qt (and therefore having a larger developer audience).
Non-members retain their option to either use Qt under the terms of the LGPL (an option which has to be preserved under any circumstances, this is the most beneficial that has happened to Qt), the GPL, a commercial license bought from a member or beeing a member on their own.
The important thing is that all contributions are made to the foundation. This allows for bringing different interests together, as members do not depend on each other (we for example know that Qt on Tizen was effectively prevented because Samsung did not want to contribute to Nokia).
Members can contribute to the foundation as it is in their interest, but to the benefit of all.
bq. Many things are indeed rather unclear yet, so please give me and the others Qt developers from Nokia a break. We're working very hard trying to get some clarity here and finding solutions.
But don't forget that Qt is an open source project, and that there is a large interest into it from many different industries. So Qt will live on whatever happens. The contributor summit is coming up in a couple of days in Berlin. I hope we'll know more by then.
"Lars Knoll (Qt Chief Architect/Maintainer)":http://lists.qt-project.org/pipermail/interest/2012-June/002561.html
Qt Chief Architect/Maintainer at Nokia -
Damn, those Nokia results sure look bad. Even more losses than Q4'11. At this speed, Nokia will be broke in another 3-4 quarters... How do they manage to have new product lines and still have lower sales than last quarter when all they had was a dead end platform?
Qt already works well on desktops, supports the BIG and pretty much only THREE in the desktop world. Surely, the APIs are a "little" * cough cough * outdated, but that mostly has to do with the management of Qt making it chase the QML dragon and lose sight on what is really important in this very moment.
It is the outdated-ness of the APIs that make them inapplicable for mobile application, today this doesn't show mostly because the mobile development aspect of Qt is almost non-existent. But instead of working to modernize or replace the outdated APIs, the decision was made to just scrap it all and go all out QML. And for the n-th time, QML is a good technology that has its applications, but it is far less universal then the old native API, despite it being like 7-8 years old. And being forced to go for QML on the mobile, where the last thing CPU and memory starved devices need is more bloatware, virtual machines, interpreting and what not, or alternatively you also have the privilege of being able to stick to a 7-8 year old API if you don't want to use QML.
A MODERN NATIVE API is crucial to the future of Qt, only it can provide the winning hand Qt needs to be a viable option, desired by developers. The fact is QML arrived way TOO LATE to the party, and cannot really compete with basic stuff like HTML5 + JS. I know many people are amazed by QML, and not only newbs, whom find C++ way too hard, but QML doesn't offer that much advantages over its competitors and has plenty of downsides too. HTML5 is supported almost everywhere, with the exception to embedded, where QML2 is a no go either, for 3d graphics WebGL is there too, so currently choosing QML pretty much only grants you platform limitations. QML has only one benefit and it is the possibility to extend it with native code, but if extensive computations are needed there are workarounds to extend an HTML based application as well. The one benefit of QML simply doesn't have the weight to make up for its many downsides, especially its limited platform support.
But enough with the QML rant, there has been enough of it already. QML won't save Qt, in order for Qt to rise in popularity and usage it needs to provide THE ONE THING THAT IS MISSING - and that is A MODERN NATIVE CROSS PLATFORM FRAMEWORK that supports all major mobile platforms on top of its excellent desktop support.
Qt is better than NET. Qt is better than the Android SDK. Haven't used Apple's frameworks, they seem fairly complete and modern, but I am not charmed by Objective-C (SUBJECTIVE-C would be so much more appropriate name), Qt is better than C# MONO...
Point is, there are VERY FEW developers that would turn their backs on a framework that offers the best performance, is easy to use and covers over 99% of the devices on the market. Why would anyone bother writing an application that works on only one platform, with a technology that offers inferior performance and efficiency, if one can write an application that works on all platforms at the peaks of their capacity? Maybe a few devoted, fanatical fanboys, but generally, every creative person out there wants for his creations to be as widely available as possible. And THAT is what Qt needs to be, not chasing and arriving late at trendy fads but offer the one thing that is missing and the one thing other frameworks cannot even afford to offer. Qt doesn't need to try to compete with the rest, it needs to offer what the rest cannot compete with. It is THAT SIMPLE!!!
-
Please, for the sake of this thread, keep this discussion where it "belongs":http://qt-project.org/forums/viewthread/16465/. The thread on the mailing list is already plagued by the usual suspects; we don't need the same here again.
-
@Lukas - I suggest you read this post very carefully and do a little thinking before you rush into yet another of those clichéd response of yours:
This threat is about the Future of Qt, and I only brought QML up as being pretty much it's only present, and I think you should be able to see how the present is related to the future, since the latter is a product of the first. Now, I do realize this threat is more concerned with the financial future of Qt rather than its development direction, but the two are related as I am about to show further down this post.
So if you don't mind, stop regarding me as if I am some pest, frolicking and spreading FUD, for I may as well be more concerned with the future of Qt than you. At least enough to the point I am not making any illusions of it, and of what NEEDS to be done to ensure it.
That is right, it is not about what I want, neither is nor should be about what you or anyone else wants, for what people want is subjective and rarely relevant. There is one thing that is incomparably more important than what people want, and that is WHAT PEOPLE NEED. And what Qt needs while we are on the subject.
It is a fact that today it is Nokia that pays the trolls, and as such Nokia comes as not simply the major, but the predominant Qt contributor, a fact that has pretty much tarnished the hopes of MOBILE success of Qt. Well, I don't think that it should be Nokia or any other company paying the trolls, the FIRST and MOST IMPORTANT thing Qt needs in order to secure its future is INDEPENDENCE, or in short, Qt must pay for the work of its primary development force - the trolls. That means Qt must become more successful commercially, WITHOUT sacrificing the accessibility it got through the LGPL license.
Which brings me to the next point - in order for Qt to become more successful commercially without sacrificing its accessibility, it needs a MUCH LARGER user base and much more flexible licensing fees. There aren't many developers that can afford Qt commercially today, and there is a whole world of developers who will be able to afford more flexible licensing. So on top of the few big commercial clients Qt has today, it can have scores of smaller which will significantly increase the revenues generated by the framework, to the point there won't be a need of any big company, funding it and enforcing its limiting corporate politics on it, Qt will be able to support itself just fine.
BUT Qt has to win those developers first. So in order for Qt to get what it needs, it must give developers what they need. And enough with the illusions, developers don't need QML, it has its merits, it has its future, but focusing on it won't save Qt, not in the short term, and certainly not in the long run.
There is already a cross platform solution, working on every commercially viable platform, and it is called HTML5+JS, and for Qt to become a success it needs to compete with it. That means:
First and foremost - catching up to it when it comes to support, and why not even exceeding it, by developing a fall-back capable API that could bring Qt on embedded, which QtQuick2 can't. And NO, I don't mean QtGui, I mean something still relevant to our time, only not OpenGL based.
Second - offer nativity, which is something the interpreter/VM based HTML5+JS combo cannot offer.
And this is all it would take for Qt to become a solution other development frameworks cannot compete with, which will inevitably grab the attention of every adequate developer out there, giving developers an elegant and effortless way to gain audience on every platform on the market, exceeding what their proprietary frameworks can offer. If Qt provides this much needed and currently TOTALLY absent from the market solution, this will attract many consumers, many code contributors and many funding contributors, this will grant Qt complete independence and long term safety, which, combined with the larger developer base spells out the formula of ROUSING SUCCESS. Why struggling to compete in fear and insecurity of the future when you can simply dominate and be 100% free and independent?
-
No, I just think that this should not be another QtQuick discussion, as we know how such discussions tend to be conducted (also with a view to the mailing list). This has nothing to do with a personal opinion about someone else, but rather that this thread deserves better.
However, I'm not the authority who is to decide what is allowed to be discussed here, or not; I just ask to keep the potential consequences in mind.
-
There are a lot of issues in terms of "what Qt future should be"
I want to keep the fucus of this thread on "How make future possible"
What needs to be done, to keep the development team going.Software developers are not interchangeable. If we want Qt to be able to survive outside of Nokia, we are going to need some action.
I for one want to see the Open Governace process maintined and moving forward.
I am leaving for Berlin now, see some of you there.
-
Just reading an article on the "BBC":http://www.bbc.com/news/technology-18527509 and it mentions Windows Phone 8 having native C/C++ API support. Does this mean a Qt port to WP8 is possible / allowed? If so, it might be an interesting option...
-
I vaguely recollect that this is more like "native" - but it's not full c++ support, just a subset that MS will allow. There was a separate thread for that on DevNet, you might want to look it up there, as I can't guarantee my memory to work well :)
-
Bah... thought it would be too good to be true.
-
it's not the thread I was talking about, but close: "link":http://qt-project.org/forums/viewthread/6751/P15.