We’ve got some breaking news, folks: Intel and Nokia have just announced a partnership in which they will merge Maemo and Moblin into a new product called “MeeGo“, aimed at just about any type of device you can imagine. This is one of the rare (but very welcome!) moments of convergence in the mobile Linux space.
MeeGo will build upon the Moblin core operating system, with Qt being the application development environment. It will be hosted in a completely open fashion by the Linux Foundation, so everybody can join in on the fun. It will run on pocketable mobile computers, netbooks, tablets, mediaphones, connected TVs, in-vehicle infotainment systems, and more.
MeeGo using Qt has a major advantage: Qt now also runs on Symbian S60 and Symbian^3, and Nokia stressed that applications written for MeeGo only have to be recompiled in order to run on Symbian. The power of Qt at work, people.
“Our vision for seamlessly communicating between computing devices from the home, auto, office or your pocket is taking a big step forward today with the introduction of MeeGo,” said Intel President and CEO Paul Otellini, “This is a foundational step in our evolving relationship with Nokia. The merging of these two important assets into an open source platform is critical toward providing a terrific experience across a variety of devices and gaining cross- industry support.”
“MeeGo will drive an even wider range of Internet computing and communication experiences for consumers, on new types of mobile devices,” said Olli-Pekka Kallasvuo, CEO, Nokia, “Through open innovation, MeeGo will create an ecosystem that is second to none, drawing in players from different industries. It will support a range of business models across the value chain, building on the experience and expertise of Nokia, Intel and all those who will join us. Simply put, MeeGo heralds a new era of mobile computing.”
The first release is expected in the second quarter of 2010, with the first devices shipping later this year.
If it is really open source, and open to everybody on all devices as claimed, then it will run on ARM just as well as on Atom.
Interesting.
I wonder how easy it will be to port existing Qt applications.
I wonder how easy it will be to port existing Qt applications.
Just as easy as porting Qt apps to Maemo has been so far, ie. rather easy. Qt runs already on Maemo just fine, including the most recent version, and there’s no special tricks you need to do to compile a Qt app for it. You just have to take into account the screen size and device’s restrictions and you should be good to go.
From http://meego.com/about/faq
yeah and AMD too… currently it doesn’t boot on amd cpus even though they support the required instructions (sse2 and sse3 iirc)
Moblin requires SSSE3, which is different from SSE3, and only available on Core2 or Atom (not sure) and later Intel CPUs.
Even the CoreDuo board I have can’t boot Moblin.
Good point. Moblin currently do not run on my eeePC 701 because it is a Celeron, not an Aton processor.
ANother thing, is that I want to see how the UI comes out. If they can create an easy, qt-based, windows manager and desktop, I’ll go bye-bye KDE from my desktop, cause desktops currently are just becoming more and more fat, when I just want to run apps and compile my little programs
second quarter of 2009? I think this is a mistake, probably 2011
According to the FAQ, it’s 2010.
So they ditch clutter? Very good move in my opinion. Clutter does not have any traction anyway and QT is way more advanced.
Unless I didn’t understand and they keep clutter?
After reading a little more, I’ve found this:
http://meego.com/developers/meego-architecture
Looks like both QT and clutter stay. Even better. More options is always better than less.
Looks like MeeGo will allow Clutter/GTK applications, but toolkit development will be focused on Qt.
Yeahhh….
Clutter is a Scenegraph engine for the Linux desktop. If you don’t know what ‘scenegraph’ is the A)look it up in Wikipedia, B) your obviously not a programmer.
As far as QT vs Clutter…
http://www.clutter-project.org/sources/clutter-qt/1.0/
I know what Clutter is and I also know that Qt itself has similar functionality built in.
Why replicate something in YADL (Yet Another Damn Library) that replicates what Qt does? Except Qt does it with a nice set of exposed classes and an API that is programmer friendly.
This is ultimately what is costing GTK and the ‘development environment’ around it traction and developers – it’s a massive panopoly of dependencies where developers aren’t sure what to use, everything is programmed in a different way, nothing is really standardised and deploying and maintaining those dependencies on a distribution is a nightmare.
Not to mention that hacking an object-oriented system onto plain old C means that programming with plain GTK+ is ugly and necessitates tons of boilerplate code.
Maemo was good, Mobin not bad, but MeeGo? What a babyish sounding name.
Bizarre. I don’t get modern marketing, not at all.
So, Nokia are doing it again. Screwing with Maemo, making things incompatible and upgrade paths murky. Nokia, you suck.
Who said ANYTHING about incompatibility? Nokia have already stated that existing Maemo apps will run on MeeGo.
In fact, from what we have seen so far MeeGo is essentially Maemo 6 plus x86 support and some Intel widgets..
And the switch from Debian to Fedora as OS base.
Which makes me so sad
Moving from Debian to Fedora is a true step back for me.
I hate to see a Debian based mobile OS go. I’ll be moving all my efforts from Maemo towards MER.
Just like already with Moblin, there will surely be Linux distributors picking up the GUI layer and put their own disto below that. Eg. openSUSE’s and Ubuntu’s Moblin derivates do that.
With Qt supporting symbian also, I highly doubt anything will be backward incompatible. I would say apps will be so backward compatible, it will even work on Windows mobile and windows xp
.
I hope it’s as good and true as that video clip claims. This is an open source wet dream is it not? Imagine being able to integrate all your computing devices together in an open fashion with no restriction? The possibilities are endless
He, did anyone say ‘Plasma’?
(for the uninitiated: having (network)sharable widgets scaling to any screensize on any device was part of the initial design idea behind plasma…)
They won’t use Plasma or any KDE technologies in the base OS. Nokia is cautious to be fully compatible with lower end Symbian phones an that means pure Qt.
That doesn’t mean that there won’t be KDE tech-based applications, because Nokia already does that with the office viewer app, but that one won’t be part of the developer-targeted “tech stack”.
However, the KDE project people are working on their own technology stack and they have currently no intentions to support Symbian — only Linux (Maemo/MeeGo).
Links? References? I’ve seen nothing about MeeGo needing to be compatible with Qt on Symbian.
What is ‘pure Qt’ anyway? As soon as you write anything with the Qt toolkit, like KDE software you are saying it becomes ‘impure Qt’?
Good question.
Perhaps kdelibs brings a bit too much stuff of it’s own to the table. It also used to be that once a program went with kde libs, it became Linux-only, but that should no longer be the case.
KDE community has some work to do to convince people that KDE libs are ok even if you are not building a KDE app. The “KDE SC” rebranding might help here.
You seriously want references?
Then read anything Nokia has published in the last two years. It was the whole point of the Trolltech takeover to get a unified development platform, for god’s sake.
Have you been living under a rock????
For applications then yes, they’ll probably be keen to have people writing applications that will run on a purely Qt platform to support a wide array of devices.
However, that isn’t enough by itself, and it certainly isn’t enough for devices like the N900. A lot of infrastructure needs to be built and developed to create a usefully usable device with the functionality needed. You aren’t going to get that by Qt alone, or Qt will have to be expanded in a big way. I’d be surprised if Plasma and some of the new stuff doesn’t find it’s way on to devices like N900s. They’ll have missed a trick if it doesn’t. In that sense, no matter what you do you’re not going to have the same stuff ported to Symbian regardless.
Edited 2010-02-16 01:35 UTC
Yeah right. Since Symbian phones will have lot lower resolution, weaker processors, less memory, etc this statement is bullshit. It’s like Microsoft saying that .NET can be runned on different platforms, technically true but false in reality.
Yeah right. Since Symbian phones will have lot lower resolution, weaker processors, less memory, etc this statement is bullshit. It’s like Microsoft saying that .NET can be runned on different platforms, technically true but false in reality.
First of all, if the application in question isn’t a game then the amount of memory and slow processor has no bearing whatsoever. Secondly, screen resolution again has no bearing if the application is developed in a way that it can scale well from small screens to large screens.
Now, how about your claim about Symbian-powered hardware? Well, where do you base your claims?
Come down from fairyland. Why you think Microsoft failed on phone business. Too large scale of different devices. But in Nokia case it is even bigger. We are talking screens from 320×280 to 800×600, processors from 300Mhz to 1Ghz and so on. Hell if look some WinMo software that was made for 640×480 screens look crap on new HD2. So scaling doesn’t work. Nokia has 3! platforms without chassis specification on ANY! Try scale your applications on all those without making crappy user experience. Major reason why so many does iPhone software is because you have 1 platform and 2 chassis to aim.
Why you think Microsoft failed on phone business
Because, well, they’re Microsoft.
Hell if look some WinMo software that was made for 640×480 screens look crap on new HD2. So scaling doesn’t work
You’d be an idiot if you did just pixel scaling…. You may have not realized it, but in computer software “scaling” can refer to several different things, and in this case I mean element scaling, not pixel scaling; you can have the application’s elements scale up, down, and rearrange them when needed. Thus the same application can both fit larger screens and smaller screens. Also, font can be scaled, with a larger, crispier font used on large screens, a smaller one on small screens.
As I said, if the application’s interface is programmed properly it can easily be made to fit several different screen sizes without modifications to the application’s source itself.
I really shouldn’t feed troll since your Microsoft comment was way stupid. But if you have ever done any, I mean any UI desing, either in form of websites or applications, you know that you can’t just shrink all components. You can’t just shrink content that was optimized to certain size to 75% of that without ruining experience, this is FACT. You think Qt is some miracle that nobody figured before, plah same as Java, .NET, hell even HTML suffers same problems when scaling down. It’s hard even on desktop level where you have UI full of information and you need to downscale it, it doesn’t happen without rethinking how you represent the information.
HTML does not have this problem. It is not supposed to describe any rendering an all. It just describes content. It’s CSS for rendering. QT also uses CSS. You can design your interface independently from how it will be presented and just adapt the CSS for how you want your application to look on each device. The user could even do it himself if he doesn’t like how the widgets are arranged and look.
Edited 2010-02-15 16:16 UTC
People who do N900 software have 1 platform and 1 chassis to aim. People who do N97 software have 1 platform and 1 chassis to aim. People who do software for the eeePC 1000 or whatever have 1 platform and 1 chassis to aim. People who do software for the iPhone 3Gs Have one platform and one chassis to aim.
People who do software for many platforms have many platforms to support. What is special about the iPhone? It’s just another phone you have to target and it doesn’t have QT, which makes it a pain to port software to.
Most Symbian phones on sale today has at least one ARM11 cpu (clocked at 434-600 MHz), but coming models feature ARM Cortex-A8 instead, memory at least 128MB or 256, and resolution is often 360 x 640. Qt works like they say!
Edited 2010-02-15 14:33 UTC
GTK support from Intel and Nokia will now die faster than you can say “financial constraint”, but I think nobody who has really looked at both toolkits will deny that Qt is more complete and modern(CSS, QML,etc). And application developers still are a C++ crowd (when it comes to efficency and speed).
Nokia and Intel have spend a lot of cash on GTK devs the last few years .. so I think this move will have its consequences.
Which major GTK backers are left?
Red Hat is still a major contributor. Novell covers mostly the GTK# side of things. Oracle is also cutting back .. Canonical is mostly polishing the rough edges ..
That said I hope this won’t delay Gnome3 any further.
Edited 2010-02-15 14:31 UTC
http://www.gtk.org/development.html#Team
Canonical never developed GTK.
Maybe the Intel guy will go, maybe Intel keeps him just in case. It won’t affect GTK much.
GNOME 3.0 will use GTK 2.
Not so fast, if you look who pays http://www.lanedo.com/ for its GTK work you will find that it is mostly Nokia.
That is going to stop .. so 4 from 9 on that list are likely to get into trouble money-wise.(Just my guess, but I think it is not that far fetched)
Edited 2010-02-15 14:55 UTC
Canonical never developed GTK.
wrong: Cody ‘bratsche’ Russell is working on gtk+ and he’s paid by Canonical. he’s not on the team, but he should very well be by now.
Maybe the Intel guy will go, maybe Intel keeps him just in case.
being “the intel guy” I can already assure you: I’m not going anywhere. by the way, I’m also the Clutter maintainer, and for that too I’m not going anywhere.
GNOME 3.0 will use GTK 2.
wrong. Gnome 2.30 will use gtk+ 2.20; Gnome 3.0 will use gtk+ 3.0.
Hey cool, thanks for the clarification. Maybe it is not doom and gloom for GTK after all.
What I really want to know is will Meego use Clutter? Does not look like it will (Qt has similar functionality) .. but maybe for the Netbook side of things?
Edit: OK, GTK/Clutter will be supported (http://meego.com/developers/meego-architecture), but is that lecacy support?
Edited 2010-02-15 15:29 UTC
Edit: OK, GTK/Clutter will be supported (http://meego.com/developers/meego-architecture), but is that lecacy support?
I may be talking out of my ass here, but as far as I can see Clutter will only be “supported”; it will not be recommended, the interface of MeeGo will always be Qt, all the internal apps will be Qt, and Clutter is only supported for backwards compatibility. It could mean that Clutter will eventually be phased out altogether in a few years’ time.
Yeah, common sense dictates that GTK and Clutter will be dropped at some point. That would reduce the footprint quite a lot and make the platform much simpler.
Let’s see if ebassi is still being paid by Intel for GTK and Clutter work in 2 years time. At the moment there isn’t that much clutter/gtk lecacy software for Maemo/Moblin that can’t be replaced with Qt counterparts in 2 years time.
One thing is sure though, Nokia and Intel won’t be spending heaps on cash on new features for GTK and Clutter. They are lecacy now. Googles ChromeOS relies heavily on Clutter, so we don’t have to worry about ebassis future that much I guess.
Well, since GTK support is there, I don’t think they will drop it anytime soon. There are way too many applications that run on GTK right now and for the decade to come for them to consider dropping GTK support. Those are mostly desktop applications, but many of them are still usable on netbooks and phones with big screens and many are better than the alternatives. Clutter is not so much needed than GTK though. I don’t think it cost that much to continue to support it anyway. It does not need to be installed on devices that don’t use it to be useful.
I don’t think that is true. For MIDs and phones there isn’t that much. Fennec maybe .. but QtWebkit-based browsers will be just fine.
For netbooks OK, there are a few good GTK apps, but let us see in 2012 GTK and clutter will be gone or at least majorly deprecated for all but the biggest most powerful netbooks.
Considering that you guys don’t even have a concrete roadmap for GTK 3.0 (just a rough draft without any dates; see http://live.gnome.org/GTK%2B/3.0/Roadmap ), you can’t promise that.
I suspect that’s not going to be up to you.
Anyway I’m sure there will always be devices running clutter as a more lightweight solution. But these companies are looking at the big picture, and thinking economies of scale between disparate devices.
Well, I wish I could share your optimism. As far as I see there is still no concrete plan in place for GTK 3, nor any features that Gnome 3 currently requires of it to drive development.
If there is a GTK 3.0 it will merely be a version bumped 2.30. There just isn’t the plan, time, developers or resources to develop anything new in a major way, and certainly not for the Gnome 3 timeframe.
Edited 2010-02-16 01:43 UTC
To be fair, although GTK has had a known shortage of developers and resources over the years, it existed before Nokia and Intel got involved.
One thing that kind of surprised me on the MeeGo website is the many consumer platforms this OS is trying to target: your phone, car, TV, netbook, etc. That’s a lot of platforms to tackle! I feel like this is going to be the new ‘battle-for-the-desktop’ (except not really desktop) for OS makers. ChromeOS is already trying to do this, and LocusOS (mentioned in the osnews article, “Concept: Stream Adaptive Computer System, Locus OS” 8th Feb 2010) has the potential, as well as MeeGo.
NOW, having said that I think it would be awesome if the new MeeGo and the aforementioned LocusOS would join forces. However, that would probably be unlikely since MeeGo is Linux driven and LocusOS appears to be more Microsoft driven. However, it would be really cool to see an OS that would run on many platforms (car, tv, netbook, etc.) like MeeGo but have an UI perspective like LocusOS.
with so many platforms/screensizes etc to support, they should have a serious look at Plasma. Plasma has already proven to be able to build flexible interfaces for a variety of devices (ok, desktop and netbook released, mobile under development). Widgets sharable over a network, written in about any language you want, scalable, easy to write… Seems like the way to go. Better than fragmenting things even further, they should take advantage of what is already there and work with the KDE community!
Of course I think that, as member of that community – but even if i wasn’t, it seems to make sense…
Well, the goal here is to go seemlessly from platform to platform for application developers.
At the moment Qt offers that, KDE (Plasma) only kinda with a lot of pitfalls.(especially when you include the major OSes and Symbian)
And don’t tell me KDE is already there, it isn’t. It is once Amarok, KOffice and Marble run flawlessly on all those devices with just a recompile.
I surely wouldnt claim plasma is ready for that now – but neither is most of their infrastructure. It wouldnt cost them much in time and resources to work with the very active plasma dev team and come up with something superior to what they are aiming at now… of course if they dont, the plasma ppl will do it themselves eventually – its on the agenda for 4.6 or something anyway.
Well, Aaron Seigo works for Nokia and is the top Plasma contributor .. so if Plasma would suit their needs it would have been upto him to tell them, wouldn’t it?
No, Aaron is sponsored by them to do whatever he likes to do – which is of course plasma stuff. But he’s not an employee in the technical sense of the word, where they tell him what to do and he reports back. And even if he was, getting that word out might be hard if you’re way below in the management stack. Furthermore, the different parts of the company don’t always communicate that well either.
Well, paradoxically those applications are written with Qt, so if you’re saying that they can’t be made to run flawlessly with a simple recompile then the Nokia/Intel Qt premise of cross-platform development has failed.
However, you can’t port to a wide range of devices with a simple recompile. That has never been possible and porting has always and will always be required. What Qt does is make any differences a hell of a lot easier to deal with.
Edited 2010-02-16 01:49 UTC
Talking about platform convergence… Wow…
A linux based platform with qt on top supported by both Intel and Nokia and the linux foundation in general. No strings attached…
We will look back on this announcement as a seminal turning point for the evolution of software that works across a myriad of devices.
This has been a-coming for a long time but it is finally crystallizing into something concrete with a real roadmap.
It^A's exciting to be a witness to real industry-wide innovation and collaboration.
Quite possibly. I’ve always maintained over the years that Qt was the only option because, well, you need developers, developers, developers, developers on your side. Without that bugger all happens. Took a bit longer than I thought.
However, there is a ton of stuff in the application stack of your average Linux based installation that needs serious sorting out. You can’t just dump Qt over the top of any OS and mask deficiencies.
Edited 2010-02-16 01:59 UTC
What is with the dearth of decent naming schemes the last few years?
Argh.
Wii
Moblin
MeeGo
ipad
Android
Ugh.
Nokia + Intel + Nintendo -> MeeGoWii
Qt is the better cross platform toolkit, glad to see they recognized that.
The problem is that Qt is overrated. It is way better than GTK but is far from a state of the art toolkit to build a decent application with.
The underlying problem is that C++ is still not the ideal language for efficient GUI development. Objective-C is much better suited.
It would have been a better decision for Nokia (and Google and maybe even Microsoft) to use Objective-C + GnuStep and invest in it.
This way they will forever struggle to catch up with Apple.
WTF are you talking about? QT is a toolkit that can be used in python, C++, java, Ruby, Ada, PHP, Lua, tcl and any decent language. They can just write a binding for Objective-C, but it looks like nobody is interested right now, maybe because nobody want to use it? Just like nobody want to use GNUStep?
Edited 2010-02-16 10:56 UTC
Yes, good point. Although I work with Qt at the moment, I was an Objective-C programmer for 10 years, and I really like it. One of my first projects for KDE was attempting to write Objective-C language bindings for Qt/KDE in fact.
I did go back and try some iPhone development last year, and I felt Qt programming with C++ had a lot in common with Cocoa Objective-C programming after being used to mostly coding in Ruby.
But Qt is better for language bindings (other than Smalltalk) because the object model and method calling style maps better onto more languages. Then again I still prefer Interface Builder to Qt Designer.
Overall, I would rate Qt and Cocoa as equally good. They are about the best two toolkits out there as far as I’m concerned. I don’t think Cocoa gives iPhone developers a killer advantage over a multi-touch enabled, QGraphicsView based widget oriented version of Qt.
as a nokia n900 owner i am happy to see the future of the platform be as plural as possible.
as a KDE user i am also delighted by the emphasis on QT for app development.
as an opensuse user i am happy to see the switch to .rpm format too.
all round good news day.
Is there anyone else as excited as me? This makes me soooooo excited.
For years I watched corporate folks (Sun, Nokia’s Maemo, Novell, Red Hat, etc )latch on to GTK and now things are swinging the other way.
I had been disenchanted, as I prefer KDE. When the time came to make a choice for a GUI toolkit, I chose to learn Qt for Ruby apps because it was so easy to do despite the trend towards using Gtk (in my mind). Now, however, the equilibrium seems to be shifting in the other direction!
I’m a Ruby not a C++ programmer, but Qt4 bindings are really easy to use and find documentation on, so it is great to no longer feel like a second class citizen. I can no feel confident that the apps I have written will be multiplatform with less muss and fuss!
My only question is, how will this will shake out for netbooks when KDE4 netbook edition is non-beta? It seems pretty exciting though, no matter what…
Good to see that Qt is getting the love it deserves. Somewhat of a shame about Clutter, of course. I wonder if it will continue to live on in Gnome 3, or die a slow, sad death? Considering it was an “Intel project”, I think maybe the latter…. Last time I tried using it I was hampered by a severe lack of up-to-date bindings and documentation (only C and Python were being kept current), so IMHO this probably spells the death sentence.
Still, going with Qt makes sense, since it’s clearly the more mature platform, has a much more complete API and more active developers, not to mention *much* better documentation. I imagine Intel’s also happy about gaining access to Nokia’s mobile software stack (e.g. power-management, chipset and radio drivers and the like), while Nokia of course wants to push Qt, partly because it is vital to their Symbian comeback strategy.
For developers meanwhile this can only be seen as positive, since it means there are three mobile OSes (Symbian, MeeGo and WinMo), one netbook platform and any number of desktop platforms that can all be easily developed for using one toolkit. Good stuff.
This MeeGo thing sounds like a good thing, but I fear that they may be a little too late to the market.
The problem is that in today’s world, marketing seam to be much more important than the quality of the actual software. This means that Apple and Google with all their marketing power probably will eat the lunch of these systems. WinMo could have had a chance given Microsoft would have had a lot of money to spend on marketing, but they are turning WinMo into WinPhone7 where the place for QT is less clear.
It could work well as a get out of Symbian strategy for Nokia though, if they should change their minds on Symbian
In the US, maybe. In the rest of the world (the GSM part), Nokia is everywhere. Their marketing power is hard to match. Android is almost unheard of and the iPhone is just another phone. Poeple usually don’t know what an operating system is. They don’t buy Android or Maemo. They buy Nokia, Samsung or LG. They care more about the look, the size and the number of megapixels in the camera than what it runs.
In the rest of the world, I think it is the other way around. Apple and Google will have a hard time to catch up some market share. Google may get some if they manage to have Nokia use Android in some models.
That may be true for feature phones, but when it comes to smart phones people usually know, and care what OS they run.
When did you last see somebody use a N900 in the street.
Edited 2010-02-17 08:17 UTC
The N900 is not yet sold in my country. Well, yes it is sold, but not by the service providers yet. I’ve asked my service provider and they told me they would probably sell it later this year.
Anyway, the N97 mini is very popular. It’s everywhere. And very few customers actually know what OS it runs. If they put MeeGo on it, nobody will notice. They will just see that the phone looks better and that’s the only thing that matters. They couldn’t care less that it’s MeeGo, Android or the iPhone. Neither Nokia nor Apple talk about the OS in adverts. They just talk about the features like email, GPS, touch or pinching. Granted, Apple also talks about the apps but I don’t think it is a way to gain customers. They do it in order to lower their marketing cost. They share the cost of advertising with the app vendors.
It looks like LG might be going with Nokia/Intel MeeGo rather than Google Android.
http://www.gearlog.com/2010/02/mwc_lg_gw990_will_become_an_in.php
http://www.engadget.com/2010/02/16/lg-gw990-to-be-among-first-meego…
http://www.liliputing.com/2010/02/lg-gw990-smartphone-to-run-meego-…
If there is a Nokia/LG axis running Meego, that would become serious competition for Google/Android and Apple/iPhone in most of the world.
Not exactly on topic, but am I *really* the only H. P. Lovecraft fan here?
Brain cases, anyone?
No, if the compiler or SDK is called Yuggoth, that would rock SO much….
…but have that lurking fear that the project support will peter out over the next year leaving it up to the open software developer pool to try and pick up the pieces and make it all happen without support from major companies.
Then again I have lurking fears that I am being watched by my government.