What is the future of Qt? | Hacker News Reader

What is the future of Qt?

blog.qt.nokia.com

82 pointssathyabhat23 comments

This is a sad story... Qt was gaining huge momentum under Nokia's acquisition. QML, their new technology for mobile interfaces was just pure joy to use. The platform was moving fast, and it seemed that nothing was impossible, including ports of Qt to Android or iOS.

Now, it just seems that it's slowly going to die.

As an enthusiastic Qt / QML developer, I'm sad...

I'm sure Qt still have a place on the desktop as a multiplatform GUI toolkit, but it does seem like the end of the road for Qt's mobile stuff.

I think the best that could be hoped for it would be a role akin to MonoTouch, but I think the amount of work to get to that stage would be too great, especially considering Nokia would have no interest in it.

I really hope Qt doesn't die. As I'm sure you know, Trolltech nourished Qt for more than a decade until they were bought by Nokia a couple of years back, and Qt is an awesome toolkit that has wide support on Linux et al. I can't fathom Qt just dying out -- even if what used to be Trolltech fades, people will at least continue to hack on the codebase, and hopefully professional support is absorbed by the remaining Linux companies.

Qt is the widget toolkit for many first-class Linux applications, including VLC and all of KDE. KDE and Qt specifically are very closely linked, so KDE-based distros should be especially anxious to absorb any Qt refugees from Nokia.

I'm sure the KDE project would absorb as many people from Qt as possible. But who is going to pay them? Which distros a) still use KDE exclusively and b) have the resources to pay for a number of Qt refugees?

1. Trolltech was profitable when Nokia bought it. 2. Nokia could sell Trolltech/Qt. They justify their actions on the basis of needing some money. Selling Troll/Qt could bring in some money (let's exclude the small detail that MS wants QT dead, not sold...) 3. If KDE demanded the Qt source out of escrow, it would be one more black eye on the already nasty look Nokia, so Nokia might just let the Trolls go...

As an enthusiastic Qt/Embedded developer, I'm hopeful.

Qt/Embedded suffered under Nokia. QML, nice as it is, is a dog on anything less than 700 MHz of core. Strings flying everywhere. The Trolls were told to put everything into QML and Mobility, to the point where I've filed embedded bugs and been told "Sorry, it's not a priority issue".

Add this to the word I got through the backchannels that Nokia was discouraging (and not even testing) static builds of Qt libraries and I could see the writing that Qt was pretty much completed for non-QML embedded devs.

I'm hopeful that wherever the code lands next that we'll see some of the floors get swept and we can all move on. Lighthouse is a nice gateway to better things, for example.

Yes, especially since Mobility really is a weird beast. See http://doc.qt.nokia.com/qtmobility-1.1.0/gallery.html . Maybe it's just me but this screams "over-engineered", I still haven't figured how you're supposed to use it.

I agree. My take on Mobility is that they needed a higher level of abstraction to make QML attractive to Javascript/Cocoa Touch/Android XML developers.

The way I read it is, they will maintain Qt to maintain the past ... but not to build their future

which I think is bad for anyone who wanted to invest in Qt, we all want to develop for platform of the future not the past

This seems to support my hunch that far from planning to bet irrevocably on WinPho, Nokia is actually trying to keep its options open: pick up a sweet deal to be Microsoft's premium partner in smartphones for a couple of years, use that time to shuck off Symbian, keep MeeGo alive as a tablet OS. Then after that couple of years, when the MS smartphone exclusive expires, take a fresh look around and decide whether to go with one or more of WinPho, Android, MeeGo or whatever may be around at that point. Doesn't seem like a self-evidently terrible strategy, though there's an obvious risk that WinPho will continue to founder and that Nokia won't have the time or money or brand loyalty to tough it out through another couple of years of failure in smartphones.

Another risk for Nokia is that this deal doesn't so much change MeeGo's status as keep it the same. I get the impression that MeeGo can remain a "promising", half-finished project indefinitely, for as long as Nokia keeps it in development. The technically accurate name for this state is not 'vaporware' but 'GNOME stew'. :) Perhaps only the raw terror of betting Nokia on MeeGo smartphones could get it to coalesce into a really competitive platform. (Competitive in technical/UI terms, leaving aside other problems like market share.) Or fail to coalesce and be cancelled.

MeeGo's apparently going to see commercial use as a tablet OS, but unless Nokia shakes its apparent attitude that a tablet is a little thing you do with your Linux-based platform when you don't dare to put it on your smartphones then the tablets won't be the crucible that MeeGo apparently needs. (And they aren't likely to do well in the market either.)

It's sad, but not unexpected, to see Qt in trouble. Qt is an amazing development environment, and I was getting rather psyched for the prospect of being able to use it to develop for phones, tablets, netbooks, and the desktop. I only hope that if Nokia is done with it, they spin it off or drop it completely (thus activating the FreeQt agreement) rather than defunding it so they maintain control but starve it to death.

However, spinning it off (as some are suggesting in the comment thread) has major difficulties. In particular, since buying Trolltech, Nokia has LGPLed Qt. That seems to me to drastically change the situation for prospective commercial licensing, as companies can ship commercial, proprietary products with Qt without paying licensing fees. How would Trolltech 2.0 make money? Sure, there would be a number of companies that still pay the commercial licensing fee, particularly for embedded systems or for support, but would there be enough to continue funding Qt development?

> However, spinning it off (as some are suggesting in the comment thread) has major difficulties. In particular, since buying Trolltech, Nokia has LGPLed Qt. That seems to me to drastically change the situation for prospective commercial licensing, as companies can ship commercial, proprietary products with Qt without paying licensing fees. How would Trolltech 2.0 make money?

The business model would have to be different, that much is true. But where does the funding come from now for GTK+ for instance? Or the Linux kernel?

There are enough Qt-using companies out there that maintaining and improving Qt should not be an issue if it is divested from Nokia. The concern would be making sure there is a neutral platform for the companies that contribute.

Keith Rusler, from comments section of the article: "I have to say, Nokia made a bad decision jumping to WP7 knowing that Qt wouldn’t be on it. Now that Nokia did this, they basically went from Qt “Code once, run everywhere” to “Code once, run nowhere”."

Nokia needs to sell Trolltech NOW. The apparent conflict of interest is too great regardless of what the real conflict of interest might be.

And I, like most Qt programmers, don't know what the real level of conflict-of-interest level is. But if I've learned one thing in the corporate world, it's that appearances matter. The appearance that Nokia wants to kill Qt is enough to screw-up a lot of decisions.

Otherwise, KDE should take steps to force an Apache release of QT source. Better now than later. See: http://www.kde.org/community/whatiskde/kdefreeqtfoundation.p...

The big change for Nokia is not the adoption of WP7 as an OS,it is the adoption of .NET as the development platform of choice - not to mention that Nokia's Symbian emulators already used Windows as the development platform. In .NET Nokia is adopting a development platform which provides tools and standards for integrating handsets, slates, and desktops into the back office and cloud while reducing issues of fragmentation. While Qt held that promise, it was a long way from the maturity (10+ years) that .NET provides.

Adoption of .NET will still allow C++ development for people with those skills and inclinations while not adding any more complexity than exists. More importantly .NET will allow more flexibility going forward - even reusing existing VB code or writing new code in F# or IronPython. Qt was not the future for mobile app development because of the relatively high barrier to entry imposed by C++ for casual app developers and enterprise IT departments.

"While Qt held that promise, it was a long way from the maturity (10+ years) that .NET provides."

What? Qt 3.0 is even older than .NET.

And .NET 1.0 came out in 2002, which isn't even 10 years ago yet.

Funny.

I'm frequently saying that .NET 1.1 (and java) is just VCL(Delphi) with virtual machine.

My friend corrects me: In his opinion delegates came from QT.

Nokia will drop QT like a rock if they'll run out of money. Say what you want but Nokia is NOT a software company.

Say "have dropped" and I agree...

One of Nokia's problems has been lack of focus. I am a huge fan of Qt so I hate to say it, but keeping Qt, MeeGo, and especially Symbian alive (even with reduced investment) makes it sound like this problem has only gotten worse now that WP7 is thrown in the mix.

Actually great new - but the should support the Android-Port of Qt. This is very important - make it easy for Qt developers to stick their head into other revenue sources.

> but the should support the Android-Port of Qt

That makes absolutely no sense for Nokia, at any level of resolution.

Reply on news.ycombinator.com