Having only two is great for developers, though, at least as long as native apps are preferred by users.
There's a limit to the number of different platforms one can support, especially if they have their preferred programming languages like Android with Java and Apple with ObjectiveC, Microsoft with C#... and even worse, they all have their own UI metaphors, so just porting the code isn't enough. And of course they all have their own app store with weird rules and specific payment options.
Edit: and then I'm forgetting the OpenGL versus DirectX debacle that we'll also get on mobile if MS gets their way. At least now both big platforms use OpenGL ES. Sometimes I wish that standards actually worked and not everyone tried to invent their own :/
I agree that native development across platforms is a pain, but I think the answer is to make cross-platform development easier, not remove platforms. We're getting there- MonoTouch+MonoDroid+Windows Phone mean that you can iOS, Android and WP with just C#- the front-end UI will need to change each time, but you can save a lot of backend reprogramming. Mono-X might not the answer- maybe JS will spread to native as well, or something like that. But let's at least try before we cut people off.
"There's a limit to the number of different platforms one can support"
I keep hearing that. However, I don't think customers care as long as comparable products from developers who have decided that a platform is part of their market develop products for it. That is, not every ISV can, or should, cover every platform. Some should specialize in only one platform. Some should shoot for market entry on "less crowded" platforms.
I create several apps on iOS, Android, Blackberry, Windows 8, J2ME and Brew. The idea that it is to hard to do this is simply false.
It is NOT in your best interest to lock yourself into 2 platforms who will have increasingly less interest in giving you good terms on your iron-clad developer agreements. I can't believe you'd even think this was a good idea.
I'm sorry but I've professionally worked in the mobile space for years and the idea that it is easy is simply false. I've done BREW and J2ME you know how hard it was. Devices weren't "spec compliants" but "examples compliants" (only a few basic examples would work). Any non-trivial app targetting J2ME devices required a gigantic QA team and the money needed alone to acquire the various devices to be able to test your app was sufficient to drive most small players out of that market.
It was hard on J2ME / Brew because of fragmentation. Then for a few months/years it was "easy" because one iPhone was the only player in town and ruling the entire app market. No fragmentation. Things were great.
But now fragmentation is here again: various Android versions, various iPhones / iPads / etc. and dealing with all these different devices is a complicated things.
You cannot say that it is "easy" because you shipped a "todo app" (or whatever) on these devices.
There are a lot of companies with succesful mobile apps out there stuggling to solve the fragmentation issue.
Of course you can do it but it's not anywhere near "easy".
So saying "The idea that it is too hard to do is simply false" is quite misleading.
Oh I agree fragmentation within an OS is TERRIBLE. I dont however equate that with multiple OS's. I hope we continue to have at least 3 if not 4 or 5 healthy players in the smartphone mobile space.
There's a limit to the number of different platforms one can support, especially if they have their preferred programming languages like Android with Java and Apple with ObjectiveC, Microsoft with C#... and even worse, they all have their own UI metaphors, so just porting the code isn't enough. And of course they all have their own app store with weird rules and specific payment options.
Edit: and then I'm forgetting the OpenGL versus DirectX debacle that we'll also get on mobile if MS gets their way. At least now both big platforms use OpenGL ES. Sometimes I wish that standards actually worked and not everyone tried to invent their own :/