@AC 7/7/2010 22:57GMT: Fragmentation
That is indeed one of the good things about OSS. But most people are probably not willing to tinker like that in case they brick their phone. Justified or not, it is an understandable point of view.
I can understand the motivations of the vendors in seeking to customise their GUIs. Otherwise they merely become a hardware manufacturer with no killer distinguishing characteristics - a sure way to price wars and low profits. But the fragmentation is surely a bad thing for us consumers and the developers. Just look at how the fragmentation in the linux desktop arena has ensured a minimal mass uptake.
Android is just another Linux, but there's Meegoo in the wings ready to enter stage right. So which one of those will prove to be the one to go for? Already there's many different versions and spins on Android - nigthmare! And whilst I think Linux isn't too bad, it really really wasn't designed with mobiles in mind. So I can't see how it can ever compete technologically, especially on battery life, with Symbian, which was designed specifically for low powered operation by good old Psion. Another off putting thing - a friend with an Android phone was seriously worried about taking it abroad. Not because it might not work, but because of the large number of apps which were notourious for just switching on the 3G data feed and doing some stuff without ever asking, despite a system wide setting to the contrary. That's a really good way of racking up a large bill for data roaming.
Apple's proposition to the punter and developer is attractive in some regards - less anarchy for a start. It's just a pity that they get almost everything else (DRM, restrictive lock down, antennae, price etc) wrong. If they fixed those problems, hell, even I might buy one!
Symbian historically had fragmentation by design. Terrible! A fundamentally good operating system with the best possible technical characteristics for a mobile, but ruined by fragmentation at the GUI level and a necessarily complex program environment which guaranteed no developer ever really developed software for it! Who knows where it's going now, Symbian^3 and ^4 may be worthwhile, but only if it is the same on everything. Again, not what the vendors want. And the technological benefits aren't going to be realised if that complex program environment is sidestepped by developers writing apps ignoring the power saving aspects of the OS. Writing really efficient apps is really hard on any platform. Whilst Symbian apparently offers a developer using C++ some tremendous power saving things, I hear that exploiting them is supposed to be really hard. There's clearly not many developers out there with the skills or the patience, otherwise there would be a large supply of fantastic Symbian native apps.
Perhaps Windows Mobile 7? Who knows. Microsoft do have the power to stop vendors tinkering with GUIs, etc. and they could use that to drive a rapid update cycle out to all users no matter what handset they've got. If they did that then it wouldn't matter if it were perfect or not at the beginning; they would be able to improve rapidly across the whole market in a controlled manner. But that would definitely reduce the vendors to being just hardware manufacturers, with low profits.
Currently there's no phone out there that I would be prepared to invest large sums of money in. I've conluded that there's no widespread stability in the ones that aren't locked down by Apple. And Apples are too badly restricted, and bling-factor alone doesn't appeal. That means a free handset on a short contract is the way to go. If it turns out to be a dead duck, who cares. So a HTC desire is a good choice, so perhaps is a N8 when it comes out. Not because I think they'll be the ultimate mobiles, but because I can get them free on a contract. Or, just to really annoy Apple, Symbian^3 ported on to an iPhone.