Well that is actually going to be a problem with almost any infrastructure once you separate the hardware manufacturers from the OS developers.

The better comparison is Android. Where "theoretically" you can write an android app on any Android.. and just deploy it across the market and everyone with Android can now use it right?

Not quite. Anyone that uses Android will know when looking through the Market comments they get people *all the time* reporting "Force Closes Device X", "Won't launch on Device Y", "Keyboard doesn't work on Device Z".. etc. Because even an Android app that is portable still has to adhere to the hardware capabilities (and the different OS versions and manufacturer customized tweaks) across platforms.

Switching this to QT and your first sentence:
That's not entirely true. For the most part... you can take any QT app and pretty much just compile it for whatever you want. The problem comes in to will it work? Take for example the N900. There was a few Mail programs written in QT that people have ported over, and I myself recompiled one for the N900. It compiled fine, launched on the N900 fine. The problem was the UI was not optimized for a tiny screen, and all the options menu's and stuff go off the screen. Making it unusable.

But, the app itself "worked", and was cross-platform compatible, but because the original developer had no intention of a Mobile device using their product - they did not code it to be usable on them.

Apple maintains both their hardware and their software. So they can ensure that every app written for a specific OS version will work the same on their hardware because all of their hardware is very similar in performance. They don't have to play a mix and matching game of some hardware vs others.

So anyway, the point is, the disadvantage here is the same as the advantage here: Multiple peoples hands in the Pot. MeeGo (should be) no different than Android in this case.

I'm sure this has been answered and I either missed it or can't get my head around it, but...

Symbian and MeeGo are open source. Use Qt to develop a UI/UX for each. Use Qt to develop apps that run on either.

Now... how are they differentiated?

EDIT: from end user standpoint, for marketability.

(I think you are asking about development so I will answer as such, if I missed your point let me know..)

By Hardware capabilities.

Theoretically I could write a QT app that uses hardware keys to navigate things or drive a car.

Problem: Half our MeeGo devices don't have a hardware keyboard! They can't play my game! (I believe this actually happened a bit in Androids infancy when the G1 was the only device. When the MyTouch came around I seem to recall people complaining that hardware keyboards were required to use the app and they didn't have one.)

Or, QT is also on my desktop and as dov said earlier: I could code my app to be a Desktop app in QT with a Desktop-style UI (or require left, right and middle mouse buttons for example) and then someone tries to port it to MeeGo or Symbian (because it's QT). Unfortunately the UI will be unintuitive at best, unusable at worse on the smaller devices.

Also, there is likely to be at least some internal differences between Symbian and MeeGo.. so if I code my QT app to rely on something that only exists within Symbian, then porting it over to MeeGo is not a straightforward process. It will need to be fixed/worked around to get it to work.

Etc.

QT (or any platform that I've seen) is not 100% "Code once, deploy anywhere" when you actually talk about 'usability'. Sure, if it's in QT we can likely make it run on our devices... but whether we can use the App or not is a different story.

Actually asking from an end user (ie, marketing) standpoint, sorry.

That is not exactly what I was talking about The point is that, naturally, you will have to adapt to a mobile use-case. BUT. You will do that in Qt. You're not doing it 'in/for MeeGo'. The same stuff is applicable to Symbian, WinMo, UMPCs and even Android/webOS (should you want to use the unofficial ports). In those terms there is nothing specific in MeeGo (unlike when you develop for iOS or Android, where your tools and code are only good for THAT particular platform). Thus, the bottom line for a developer is "I'm making apps for Qt compatible platforms (of a given form factor)", and not "I'm making apps for MeeGo".

Fatalsaint is right, there is no(t much) difference from a user standpoint apart from more high-end hardware coming with MeeGo and 'modest' ones coming with Symbian (and their UIs of course tweaked accordingly). This part of Nokia strategy I don't quite get. Personally, I think that it's this Qt strategy that needs to be pushed, that's the thing that can and should be made cool (I actually blogged about this). Samsung got this right - what they call Bada is in essence their (puny ) equivalent of Qt with all the extra APIs/libs (even though they got two different OSs under the hood depending on device class). Symbian as a brand value is just in freefall. MeeGo is pretty much an unknown. What is your supercool secret weapon that actually has a pretty good reputation and solid (and marketable !) name ? Qt. Well, what's the holdup then ?

Good points attila77 (and I read and liked your article). Qt makes it easier for Nokia to flat out drop the OS that winds up lacking bang for the buck, and no average end user need be the wiser.

Dont agree, We dont need a appstore cause it stops development if company takes control the way apple do it.

We have already seen this kind of problem in nokia OVI store who doesnt grant access for those developing in python and thats a "apple way of doing it" and not a solution.

Meego will not fit for all people but for those who prefer openess and core linux and alot programminglanguage alternatives, Meego is the winner :-)

After reading through the bulk of this thread I am encouraged. What could have rapidly devolved into a "chuck rocks at Nokia for how they have abandonned us" sort of thread it actually has some clear foward looking thinking. Not all the clear thinkers think alike, but that is another reason I love being in this community.

What will make Meego succeed? Limiting my remarks strictly to the telephone end of the OS I ask "What made Symbian60 such an 800 pound gorilla?"

The "Gotta Haves"
1- Consistency across devices
2- Reliability on each device
3- A wide range of price points for devices
4- One click installation of applications
5- Emergency reset procedure
6- Full suite of PC/Mac/Linux tools to manage device and information on it

These, excepting the last point which had only part of that going for it, are what made Symbian60 my UI of choice since I bought a Nokia 6290 in 2007. It is what cemented me as a person who looked at Nokia first, then all others, when considering a handset upgrade. Five of the six requirements are device oriented. The last is external to the hardware but as important as, oh, let's say, antenna design, when it comes to givng the end user the "gee-whiz" experience. Get these six correct and the end user will not be thinking about what is running their phone, they will be using it instead.

I understand the uses of Meego will extend far beyond the smartphone/pda with phone realm. But for it to be at hand and in hand for the average user it has to meet the six items I list above.

Leaving behind the S60 comparison there are other areas where Meego will have to shine to succeed.

Support
I see a lot of mentions of supporting the devices after launch. I see support = new releases of firmware in some people's minds. My 6290 got all the way to v3.xx in firmware and was never heard from again. This was okay because it ran like a top on that firmware.

I think more than wanting support people do not want to have to need support. They want their devices optimised and running smoothly from product #001 to the end of production.

Applications
If you build it they will come. We've seen that over and over again with the Nokia tablets, the iPhone, Android devices, Java, etc. Having Qt as the basis is supposed to make device specific applications easier to create from the "Proto-Application" developed in Qt.

If a clear roadmap of the changes required for specific devices arrives, hopefully before the device pre-order period, then Qt developers will be able to deliver. If there is murkiness as to the correct path to take to make your application device specific then there is going to be a slow walk to having the applications people crave and need.

I'm hoping Meego succeeds. I get a great deal of enjoyment from my N900. It won't last forever. I hope when I get a different device it is an upgrade from the N900, not only in specifications but in OS as well. Meego could be that OS.

Good question. My points:

- Great multitasking implementation (better than Android and WP7). UI as smooth as iOS (hopefully).
- No buttons on the front :-)
- Good development tools with Qt SDK.
- Limited to high-end hardware. Means less fragmentation.
- Nokia likely to be the only manufacturer in the beginning (similar to apple).
Cited as a disadvantage by many, but also limits the amount of different hardware used and therefore fragmentation, which is good.
- Nokia Hardware with nice design and great cameras.
- Nokia brand.
- Ovi Maps.

Why I am still skeptical:

Lack of commitment by nokia. Qt everywhere sounds nice (and will work for most things) but the UI and hardware requirements still have to be tweaked for symbian/maemo seperately.
Nokia needs to sell a lot of meego devices, which they will if they want to.