Phone calls/SMS
Instant messages (AIM etc.)
Web browsing (Safari is always running)
Email (Mail is always running)
Music/podcast playback (with popup controls, no less)
Screenshot capture
Clock alarms and timers
Calendar notifications
Push notifications from third party apps to do just about anything–with no system resources wasted

Not half bad.

Anyway, I really think the whole situation is overblown. I’ll concede that there are a few apps that could legitimately benefit from running in the background (namely Pandora), but I think the vast majority have no need for it. AIM is the other main one, but I think the push notification system works well enough as a substitute – you still get notified when you have a new message, which is all you’d need a background process for anyway.

And as I said, this is not a technical limitation. It’s entirely possible that a future iPhone will have sufficient enough battery life for Apple to allow third-party apps to run in the background. I believe they will once the hardware gets to that point. They’re just not going to give people a sub-par experience in the meantime.

The problem with folks comparing normal desktop computing with mobile phone computing is the problems are different. By this I mean that mobile computing is “real-time” computing, where as on the desktop if email takes 1 second extra to do a task, no biggie, but if your phone call is out for that long voice becomes choppy or calls are dropped. Predictability is more important than raw performance. Most of the multitasking people are complaining about are human level multitasking (i.e. send an email while on phone, or the person who wants to listen to VM while taking a note). Yes these do require some “true” multitasking, but there is a mission-critical task and a secondary task.

Someone above commented : “it was noted that a CPU is mostly idle”, and for desktop computing, this is true, and definitely a usable resource. In mission critical apps, this is accepted so that when the critical task comes in, there is no (or at least predictable) latency to handle the task within task parameters.

A handheld computer and a phone are 2 different devices. This would be totally unacceptable as the first, but something required in the second. I need my phone to ALWAYS work, and my email/calendar to be a close second. Everything else is nice, but if there are stutters/delays in these (I guess media can’t have stutters, but my call to the nurse as I run down the hall can’t getting dropped no matter what) I can accept that.

Actually Windows is better at multitasking than Leopard because of a faster implementation of the context switch.
With snow Leopard Apple has fixed this, and this results in a 10% overall performance gain.

Also creating a new process is inherently slow due to the nature ‘of the beast’. So all flavors of Unix share this ‘problem’.
It is a common solution for Windows and Unix programmers alike to create as little as possible processes and create lots of threads to handle the parallel aspects of an application.

So the key difference in speed for most parallel applications isn’t a comparison between process start-up speed but a comparison between thread start-up speed and thread handling (context switches mainly). And as I said – and that’s a fact – Leopard is slower in context switching than Windows and that’s the main issue in multithreading. (I also know this from practice because I test heavily threaded applications on linux and Windows XP side by side. My observation is that Windows XP is equal in performance compared to linux. And I know that Vista/Windows 7 is faster than XP and linux is faster Leopard in this respect.)

So as Mac users we have to wait for Snow Leopard to have equal performance, and that’s nothing to brag about.

In the US we have two large carriers. One is ATT, the other Verizon. Verizon being CDMA. It’s a huge chunk of the market. I am not saying VZ should have gone first, just that after ATT, maybe 6 months later, plan to have an offering. Personally I would rather see an offering for every network. It would mean some decrease in margins but much larger unit sales. My ideal is to see the iPhone just sell thru the iPod channel and blow them out. One sku for GSM, another for CDMA and enable for whatever carrier you like. That would have been changing the rules of the game and a big benefit to consumers.

I don’t see GSM v CDMA as an either or. I see support both as part of the game. You support the big networks and that is just part of the cost of doing business. Now we are seeing the what happens when you don’t support carriers, you get one iphone killer hyped after another, mainly because you have 3 carriers left out of the game. They have to respond someway, and their answer is FUD and distortion that also hurts the brand. I would rather see them plugging their version of the iPhone and competing on price for data, etc.