An Open conversation about the future of Qt.
-
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 Surely, the APIs are a "little" * cough cough * outdated, but that mostly has to do with the management of Qt making it chase the QML dragon and lose sight on what is really important in this very moment.
It is the outdated-ness of the APIs that make them inapplicable for mobile application, today this doesn't show mostly because the mobile development aspect of Qt is almost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 But instead of working to modernize or replace the outdated APIs, the decision was made to just scrap it all and go all out QML.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 :)
-
it's not the thread I was talking about, but close: "link":http://qt-project.org/forums/viewthread/6751/P15.
-
-
bq. "I commissioned a deeper investigation into the scope of work required for Qt to work fully on Windows 8. I will receive the full review of this investigation early next week. I will make sure to share this detail with everyone as soon as I am able to."
"Chuck Piercey":http://lists.qt-project.org/pipermail/interest/2012-June/002751.html
Director Product Management & Key Accounts -
Well, for one, Qt is open-source so it is impossible to "kill" it. As far as I understand, Qt Project itself is also now not dependent upon Nokia per se for governance. I think the administrative costs and server costs etc. are paid by Nokia but those probably could be taken care of.
The question really is, where do the Qt contributors and maintainers employed by Nokia go now? How do we ensure that they can keep working on Qt full-time? I hope that divison of Nokia is sold to a home(s) where they can still keep working on Qt open-source.
As for mobile platforms, port to WP8 is a non-trivial endeavor. Android is a more likely destination, with some success already by Bogdan and others.
(About me: You have not seen me much around here, but I do love me some Qt and would love to contribute to Qt in my spare time, especially Android port.)
-
There is more to come.
Microsoft is "rumored":http://www.forbes.com/sites/ericsavitz/2012/06/21/microsoft-first-branded-tablets-next-their-own-cell-phone/ to be activly working on their own (non-Nokia) mobile devices; Research In Motion (BlackBerry) is "rumored":http://www.reuters.com/article/2012/06/24/us-rim-split-idUSBRE85N0D320120624?feedType=RSS&feedName=businessNews&utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+reuters/businessNews+(Business+News) to split or sell mobile devices division.
-
My two cents is this: I hope that in an ideal world the Trolls can be sponsored by a foundation like the FSF and also that Nokia would do the right thing and donate Qt to the same foundation and keep it free from commercial interests forever. Why Nokia has gone down the Windows 8 Phone route makes me shake my head in absolute disbelief.
-
OK. So this joint venture wasn't about using M$'s OS, but for Microsoft to put it's OS on Nokia devices. That makes more sense. I didn't even think of it that way.
-
My theory was MS actually wanted for Nokia to fail, so it can grab a nice chunk of Nokia's market share, which was pretty much dominating the mobile market a few years ago. Unfortunately for MS this plan didn't work out all that well, surely, Nokia lost its market, but it was immediately taken over by Apple, Samsung and a few other smaller Android platform players, MS is a big and slow to respond company...
-
I readed all the thread and nobody mentioned nothing about the "KDE Free Qt Foundation":http://www.kde.org/community/whatiskde/kdefreeqtfoundation.php, that means that in the worst, case the KDE proyect will aquire the ownership of Qt, and Qt will be controled by a real FOSS community.
The open source Qt is guaranted, there are no need to make a fork or to worry about the future.
