The dream: Qt on Android
 

All hail Qt on Android from Bogdan Vatra. He has released the alpha version of "necessities"

Re: The dream: Qt on Android
 

Wow this will probably keep the N9 alive :)

Re: The dream: Qt on Android
 

I don't know if this will keep the n9 alive - I do wonder however if this news had arrived any sooner if it would have influenced the decision to join Google instead of Microsoft.
Regardless, what is more interesting is that application developers can now target yet another platform and this platform is still in it's rising star phase.

Re: The dream: Qt on Android
 

It might take a development team like CyanogenMOD to be able to integrate it nicely into the kernels/etc.

But besides that, this article is:
http://www.ffenril.info/wp-content/u...ull_of_win.jpg

Re: The dream: Qt on Android
 

I've come to the conclusion that technology had little if anything to do with Nokia's decision to "partner" Microsoft or with their hiring of ms-elop to do the deed.

Nokia used to be driven by design and engineering.

Then, probably due to Big Investors' pressure, Nokia's management and board switched their priorities on monetizing their leading position instead.

Well, they fumbled their big ticket Navteq and Ovi integration (Symbian may not be the easiest of jungles to live in...) while competitors A and later G managed theirs better and most importantly got devices and new technologies out to the marketplace.

Qt on Android - indeed Qt everywhere (*) - sounds great from an engineering and developers' point of view, but does it directly increase Nokia's revenues? Hmm, I think we have the answer.

I would guess that Nokia's short-lived negotiations with Google revolved around, besides money and level of exclusivity, on whether Google could and would adopt Qt as a top-level component of Android (Google does use Qt elsewhere already). But primarily how Nokia could into that wonderful toll-gate/commission position they now covet.

MS-Ballmer gave Nokia, at least in the short term, a place for their Maps and a small cut of commission revenues. Who cares about technology, their own developer eco-system or employees when there's a last chance to try to validate Nokia bosses vision of their "own" revenue-generating services!

So Qt was more of a sacrifial lamb than a serious driver in this quest for a commission-generating "partner".

Qt of course can still "go places" and it will be interesting to see who will be driving its growth.


(*) Qt everywhere... would've been great for rebuilding secure and throughly integrated IT and communications infrastructure for businesses and governmental sevices while replacing the existing joke sold (actually "leased") by Microsoft. So everybody except MS and elop will end up losers in the deal.

Re: The dream: Qt on Android
 

What if Android 2.3 was released in its current state (and with NFC support) but there was no Nexus S/Nexus 2.

And then OUT-OF-FRIGGEN-NOWHERE they release the Nexus S (Feb 2011) and it comes with Android 3.0 but the "hosting" OEM is NOKIA. And Nokia has cut a deal with Google to spread Android everywhere (due to the deep reaches of Nokia) but in turn Google has to adopt Qt in Honeycomb 3.0 and the tablet (also by Nokia) and smartphone versions are exactly the same except the UI and a new dedicated Appstore for tablets.

So developers get 2 tools for android:
-Java (which has great compatibility, easy to develop, but sacrifices speed, and is bound in legals by Oracle)
-Qt (has good compatibility, good to develop, runs decently speedy, completely open source)

Nokia gets a powerful launchpad into the North American market, stops losing (infact potentially gains) investors and now kills off S40 to be replaced by Symbianv3, stops developing for MeeGo (but not Qt) and replaces MeeGo with an earlier released Honeycomb.

Is that a better solution than NOKIA Phone 7?

Re: The dream: Qt on Android
 

No, because now NOKIA is dependent on a Google-controlled OS layer. MeeGo gets around the problems of Android and WP7 by placing the whole of the OS beneath the UI in the hands of an independent third party that pulls it all together from existing, independent projects.

Much harder for it to be taken in unilateral directions by someone with a vested interest. It's a better use of resources to contribute to existing open source projects instead of adapting something wholly divorced from existing projects and hoping you aren't saddled with an incompatible fork (for any reason) down the line.

Re: The dream: Qt on Android
 

Now, now, you make it sound as if innovation (even on the OS level) is a bad thing.

Re: The dream: Qt on Android
 

Innovation on the OS level isn't a bad thing. The problem is that Android and its course are totally controlled by Google. Not that I see large amounts of innovation in Android, mostly I see things reimplemented and reinvented (see the vendor at MWC boasting their windowing system for Android.) The big thing going for it is that Google is behind it, which is also its biggest downside (Google is the only primary company behind it, and calls all the shots internally.)

Re: The dream: Qt on Android
 

Hahaha.. Some things re-implemented, but there is innovation nonetheless.

And the MWC company isn't really indicative of what Google is doing. Just a company trying to fill a niche in the tablet space by adding windows to Android. BTW, they aren't the first to accomplish this 'feat'.

And Android is forkable, but like many big projects, unless there is vision or fundamental differences in ideology, there's little reason to. At this point, it may be best to scavenge it for parts -- eg. Dalvik is something that more companies should pay attention to.

For the record, I think QT is cool in that it adds a much needed layer of abstraction above the OS but I think QT could be better if the was yet another layer that separated UI concerns from lower level stuff. I know the API segments these things, but isn't the full QT needed to run QT apps? More layers, man. Layers.

Re: The dream: Qt on Android
 

Yes and that is in the process of change. I recall seeing a bunch of Qt Troll blogs basically moving in the direction of modularity where a distinct separation of modules is being done so that you can actually make use of the smallest component of Qt that your application/appliance needs. I think it is targeted towards the embedded market, but what works for embedded almost always is good for phones.

Re: The dream: Qt on Android
 

Did you never heard about the Android-lighthouse project?
http://code.google.com/p/android-lighthouse/
The Qt-Creator integration is nice :)

Re: The dream: Qt on Android
 

Huh? That's what this entire thread is about.

Re: The dream: Qt on Android
 

i have my doubts.

if i'm a developer for Qt, i would target this platform. huge market. but how many Qt developers are there?

if i'm a developer on Android, why would i pickup this skillset when i already work on Android natively?

unlessssssssssssss Qt is so much simpler to developer on Android than that weird voodoo magic thing Google is touting as the Android development experience. however, if this happens, why would Google allow Qt to run? it would be undermining their own effort.

thoughts?

Re: The dream: Qt on Android
 

Because if you do you might, just might be able to target more than just one platform?

Re: The dream: Qt on Android
 

So you would rather NOKIA have no control over their operating system, rather than some??

Besides, I much prefer Google's influence on Nokia, rather than Microsoft's (I guess the lesser of two evils).

And even if the N900 successor had stock android user interface its an okay ux, compared to WP7 which is is quite poor (although buttery smooth).
And don't forget with Android you could personally customize it, something like MIUI ROM is killer.

Re: The dream: Qt on Android
 

Exactly. Until recently Nokia - famed for their logistics - ruled over all other major handset makers (except Apple with their private garden) when they were all running the same "commodity platform" Symbian (which was originally closed but later opened).

Now they're using that 'commodization' as an argument against Android?

Does. not. compute.

Android is open enough so what prevents the major manufacturers from either forking Android or at least forcing Google to play ball with them (e.g. in terms of competitive/alternative components for certain features)??

Now I don't think Google are becoming evil yet, but every sentient player in the mobile food chain feels that they are becoming increasingly arrogant and frankly too powerful.

So why not revive the Symbian partnership but this time around a snapshot of Android code and then start hacking away... If the largest handset makers unite, Google is quite likely to suddenly start listening.

If Qt is a better framework than Dalvik/Java and both can be used side by side Nokia doesn't need to abandon years worth of development and control of their platform. And as a bonus they can let that elop person go and enjoy his loot elsewhere.

Re: The dream: Qt on Android
 

Huh? When the OS is controlled by a neutral 3rd party it's a lot easier for Nokia to direct where it goes on their handsets. Google has a vested interest in pushing the platform in ways they want it to go, and since they're the only major force putting work into the layers below the GUI, if they quit then suddenly Nokia (and anyone else using it) has to scramble to find people to maintain it.

Sure "customize" it, but I'm not after mere customization.

Because Google controls the trunk, and is the only company doing primary development. MeeGo dodges that as the entire OS is already being worked on for other purposes. Whereas forking Android would require investing in resources capable of maintaining a divergent fork.

Well, going with Android you do abandon years worth of development and (thus far) control of the platform. Essentially, you toss all of the FOSS that makes your average Linux distro on to the fire and go for something unique to Google, which gains you effectively nothing except a reduced development effort.

Re: The dream: Qt on Android
 

In 1991, this would have made a wonderful argument against the system that you're advocating; A full 3 years before the Linux kernel's inception (v0.01), a FOSS BSD was already at version 4, was established, and moving aggressively towards POSIX compliance.

Re: The dream: Qt on Android
 

The GNU utilities and GCC were already widespread by that time, and both BSD and GNU had very different licensing schemes. They've also traded ideas back and forth and BOTH are more open than Android in terms of who controls it. Doesn't help that BSD also got beset by copyright lawsuits (which is what really helped GNU/Linux take off.)

Going between FreeBSD and Linux, for instance, is way easier than going from a regular Linux system to anything running Android, and there's lots of crosstalk and compatibility. See Debian GNU/kFreeBSD.

You could have made the argument, but it would have been a weak one. With Android you're burning 30+ years of development and forsaking a much larger development "team" and not gaining much.

Re: The dream: Qt on Android
 

No, no, no. To say that you're not gaining much is to use hindsight as a predictor of what is going to happen in the future. Again, one could have said the VERY SAME for early Linux.

'Not gaining much' is a topic for historians. Small changes in license, focus, ideology, management, etc, can mean large changes in outcome. For instance through some combination of features, Android has garnered widespread adoption in a very short amount of time, and may continue to blossom still.

Innovation should not be stifled due to tribalisms.

As a result of these combination of 'small-gain' differences Android has accomplished in the mobile market what no other GNU/Linux system has accomplished -- and not for lack of effort. IIRC its adoption has also surpassed Linux on the desktop -- a long sought after goal, even by titans of industry (eg. IBM, HP, Dell, Novell).

These 'small gains' may seem rather significant 10 years on.

Re: The dream: Qt on Android
 

No one knew where Linux would go, but since BSD was a minefield people had few places to turn. The FSF was already planning an GPL'd kernel, and Linux came out of nowhere and filled the role. Linux, though, did not throw the efforts of the BSD developers by the wayside and the software for each platform transfers easily.

And a huge amount of this is due to the fact that Google has thrown their weight behind it. I don't doubt they could have achieved the same with a native Linux system.

You are implicitly assuming that this is innovation without acknowledging that it essentially throws previous gains on the fire, and reduces the possible contributors down to those employed by Google and their favored partners.

As for a "tribalism," an insular tribe that hides everything and does everything on its own isn't very friendly. If there is innovation to be had in what Android does, only Google and Android benefit from it.

I can go compile daily developmental images of MeeGo, and run early betas of various distributions and packages step by step through development. Android is totally closed until it is released to the AOSP except to the companies that are working with Google.

Not for lack of effort, but for lack of profile.

Indeed, as actual open source software is pushed by the wayside for Google's pseudo-open system of software.

Re: The dream: Qt on Android
 

Yes, it should be well known, especially today, that an ecosystem extends beyond its codebase. But I disagree with the codebase counts for that little. I cannot say with the same degree of ease that Google's weight would have been solely adequate to lift GNU/Linux to the status of Android on mobiles.

Afterall, a system's structure has huge implications on how it's used. For an anecdote, see Java.


But it doesn't. This is entirely too extreme a statement to be taken seriously. You yourself alluded to the increased ease of interoperability between BSD and Linux, which implies that there is a degree of interoperability (which BTW has been demonstrated) between Linux and Android. That it is less practical is another story, but it's not as though Android is firewalled from '30 years of development'.



You know I'm not talking about Google as a tribe, but referring to the staunch fanboism that, like clockwork, rears its head when something challenges the status quo with something different.

For kicks, consider release of the N900, perhaps more accurately titled 'The cellular radio that drove the fanbois mad'.


No argument there.


Lack of profile? Ahem, Nokia.

I don't think that linux's adoption failings can be trivialized quite so simply.


The source is open. I think you're talking about the release methodology. Certainly even the basement OSS developer source remains 'closed' until it's committed to the tree.

Re: The dream: Qt on Android
 

It is not, inherently, walled off. It is practically, however.

When that challenger is, as a whole, less open and less compatible with what exists I can see why people might be annoyed at the newcomer claiming that they're "open."

I know you're accusing me of being a "fanboi," but there's a massive, massive world of open source full of innovation that existed before Android came around that doesn't work with it due to the decisions of a closed source company that Google bought.

Bionic isn't an innovation over glibc, for instance, but it was proprietary to start. And that's why it exists. Same for the GUI and tied rendering engine, and for Dalvik.

They can. Nokia never gave Maemo the profile it needed to take off, nor did they give the development teams the freedom or support required. A base GNU/Linux system could easily be coupled with a good UI, there is nothing inherent about such a system that holds it back.

Release methodology, development methodology. They are very closed and secretive about things going forward, unlike most major open source projects. Just because the source is open means nothing, I can get source code for operating systems pretty much anywhere. Whether the project is open enough that participation doesn't require being a Google partner is another question entirely (as end-users can participate in the AOSP, but not Android.)

Well the "basement" developer might not have anyone watching what they do or work with anyone else. But major projects tend to hold discussions over mailing lists, maintain public svn/git repos and make available changes on a daily basis (instead of monster dumps after devices with the hardware hit the market.)

A lot of my problem with Android is that it stems from a closed source project that adopted the Linux kernel out of convenience, and Google said "we're open source" but added a whole bunch of conditions to that "openness" that are simply alien to pretty much every project.

No one has issues with innovation, and yes there is silly tribalism. But whether Android is truly innovative (since the whole point of using Dalvik seems to have failed to deliver) remains to be seen, and its lack of openness compared to pretty much everything else that is open is plain as day. Saying "it's innovative and your opposition is just fanboyism" isn't a way to sell your argument.

Re: The dream: Qt on Android
 

Ok, I understand your perspective. I fear any further argument will degrade into a battle of semantics.

I had a lot of fun, though! :)