It would make sense for Nokia to not put all its eggs in one basket. Windows is not doing well on mobile devices, so Nokia investing in other options would be the safe thing to do.

In a sense, not putting their eggs in one basket was their downfall: http://www.theregister.co.uk/2011/03/10/nokia_ui_saga/

That's not about putting their eggs into separate baskets, it's about Nokia's shortcomings in terms of both management and engineering.

I worked very closely with Nokia during that time period. As closely as one could work without being directly employed (we deployed a product directly to the S60 baseline).

What happened at Nokia was the end result of years of engineering neglect. Symbian was a monster. It was poorly engineered in general, and certain subsystems (like messaging) had a reputation for simply being impossible to work with. It was just a huge mess.

The UI saga was just another symptom of that mess. That touchscreen phone referenced in the article? It existed well before the iPhone in various incarnations. They spent years trying to push that thing over the finish line, and failed. The underlying graphical subsystems gave them problems. The new form factors presented by the emergence of touchscreen slate devices proved at odds with a whole host of assumptions made throughout the (huge) Symbian/S60 codebase.

The whole thing was just a giant hot mess.

It's possible that competing internal frameworks could have worked. The ultimate problem is that they never developed one compelling UI paradigm/library that actually worked. Much less worked in a well integrated way across the whole platform.

Nokia failed at project management. They failed at engineering. They failed at leadership.

Now they're just another OEM.

History has shown us that Nokia was always good at hardware and very bad at software.

I suspect they've learnt from that and now leave the software side to Microsoft, while leaving them free to focus once again on creating good hardware.

And you think that Ubuntu on mobile devices is going to do any better?

Nokia screwed up, but diverting resources to an even more miniscule market is even crazier.

It is not miniscule, it is 0% right now ;) But again, the question is of putting all the eggs in one basket or not. Ubuntu and other non-windows mobile OSes each have different approaches, hard to say which will succeed. It's foolish to bet on just one if you are as big as Nokia.

What I don't understand is why these late-comers continue to try and build their own self-contained ecosystem of apps. Microsoft with all its money still barely has a decent ecosystem and even that has a lot of holes. There's 0 chance that any other is going to be given the time of day. Blackberry hinging on Gingerbread apps is just silly.

Instead they should be betting on the web, completely. Ubuntu, being a traditional Linux distro, has a big library of desktop apps that can be cross-compiled in Emscripten, but instead they split their much smaller resources between several different options.

Web apps just aren't as nice as real apps, and all the existing platforms already support them too. If there's nothing your new platform can do that the existing ones can't, and your message is "just deal with the 80% solution that is the web app world", why would anyone use your new system?

Apps that exist are better than apps that don't exist. Ask WebOS users, the few that still are out there. The only way to make web apps as nice is to put the engineering effort into making them as nice (something only Mozilla is seriously doing at this time). Another proprietary mobile ecosystem isn't going to be adopted.

I don't understand your argument. A new platform has to offer something the existing platforms don't have. Web apps cannot be the differentiator for a new phone OS because the big two already have standards-compliant web browsers. The browser is table stakes; of course any new phone OS must offer one, because you can't be taken seriously without it, but the browser in itself is not a reason to use the platform.

Or are you suggesting that new phone OSes ought to add proprietary non-standard browser extensions in place of a native apps API? Given the ugly history of proprietary browser features it is difficult to imagine anyone taking such a strategy seriously.

Well why cross-compile to Emscripten, when you can cross-compile directly for arm,e.t.c (thus leveraging diractly unix/desktop heritage)

actually there have been talk about one common api for all .nix that is tizen, sailfish, meego, ubuntu, plasma

That's true up to a point, but even Nokia has limited resources and developing, launching and marketing a tablet computer at Nokia's scale is not cheap.

So given that, if they were going to put some eggs in another basket, why would they choose Ubuntu over Android, which has an existing application base and at least some user awareness?

It seems like we're at the same stage right now in the mobile OS world as we were in the personal computer market back in the 80s, with multiple competing platforms such as the Amiga, Atari ST, BBC Micro, Commodore 64, PC, Mac, etc. Each had it's advantages, and each had significant success, but there's actually only room in the market for two or three of them long term.

> So given that, if they were going to put some eggs in another basket, why would they choose Ubuntu over Android, which has an existing application base and at least some user awareness?

I agree, Android would be a better second bet. But they don't necessarily need to limit themselves to 2 - look at how many options Samsung is keeping open for example.

Except Samsung is drowning in profits, while Nokia is just drowning.