Seeing as how I'm being referred to as "some", I feel it my duty to revise my earlier statement, given that Apple announced the 2.0 release on Mar 8, 2008.
I now say, for anyone (anyone!) to quote or refute: "this seems late…"
-- yofal via InvestorVillage
I agree: it does seem late :-)
On the other hand, whether it's late or early depends what's in the announcement.
The 2.0 announcement was the announcement of two radical changes: first, a real SDK for native iPhone applications, meaning that in order for the applications to be ready at OS release developers needed large lead time to get cracking on dev and test; second, Apple announced an exclusive delivery apparatus that required Apple to screen potential applications and issue keys to registering developers, meaning that new types of non-dev administrative overhead needed to be understood and digested before applications could be placed in the hands of users.
I expect the 3.0 announcement will be less earth-shattering, even if it does have valuable features and significant bug fixes and performance enhancements. Apple might, for example, push tech intended to make future hardware more easily accommodated: reminding developers about the advantages of vector graphics over bitmaps for handling resolutions dramatically different over time, etc. Apple could play chicken with the Osborne Effect and pre-announce hardware that would take advantage of v. 3.0 features inherited from Snow Leopard, or simply point out that Snow Leopard's refactoring and efficiencies will be enhanced by improved OpenCL capacity to load CPU demand onto DSPs, graphics hardware, and so forth so that everybody will benefit from v. 3.0 -- a reason to announce that since iPhone v.1 purchasers' first two years (and free software rights) have completed, they should be excited and privileged to shell out $24.99 for iPhone OS v. 3.0.
If Apple is really announcing a bunch of new stuff for iPhones -- including hardware -- the marketing folks might prefer the Powers That Be hold off announcement (i.e., announce later) so as not to kill immediate-term sales. On the other hand, iPhone 2.0's early announcement created an opportunity to clear the channel (thus protect vendors) and to build excitement about an anticipated but not-yet-available product.
I think that Apple's investment in the App Store -- to make it a better place to shop, and a more effective place to market applications -- is likely to become increasingly serious as Apple works to improve the value of its plaform, and I expect Apple to spend some time waxing about the range of things one can do as a result of the third-party applications on the iPhone.
If Apple could build a Delicious Monster-like bar-code scanner on the basis of its integrated video hardware, Apple could phase out the WinCE handheld POS systems and simply use store employees' iPhones to conduct POS transactions. Is video imprint of a credit card plausible? Would Apple need to support Universal Dock hardware extensions for card swiping? This kind of thing could be marketed to other shops together with POS back-end systems for additional software revenue. Remember how Apple over time became a power in film by focusing on achieving an integrated suite of software solutions (video editing, audio mixing, video archiving, graphics manipulation, etc.) that ran on its hardware? The field of POS is less specialized but much broader, and as Apple deploys less and less expensive hardware, its ability to sell software becomes more important. Apple could use its in-house solutions, if they are good and don't depend on outrageously costly things like Apple's massive SAP deployment.
But this last bit tends toward rampant speculation :-)