Thousands of words have been written here about how Apple’s locked-down treatment of the iPhone has been detrimental to itself, its users, its platform and the state of the entire industry. Seemingly, I’ve been hooted down by the actual events — the iPhone and the App Store are great products and have been tremendously successful — but these weren’t the points I argued. On the points I have argued, there has been fairly broad agreement that’s grown vocal at times, and Apple have made some concessions which I choose to characterize as symbolic, like the recent special-casing of an application using a private API following a plea to Steve.
What I think has been missing has been a look at just how a few steps could dramatically change the landscape, even when selecting the steps specifically to reinforce Apple’s control over the platform. I aim to provide one such a scenario now, in which both of the below steps would come true simultaneously. (They are both pretty much essential.)
Apple allows for code to run that is signed using a certificate with a different root authority than their own live and development ones.
The benefit to the developer: They can run code without needing to register applications with Apple in various stages of the process, like for allowing shared access to the keychain from multiple applications. They can redistribute applications without Apple’s permission process and the need to pass a single, necessarily (up to this point) worldwide App Store content policy. The business risk of not being approved is eliminated. Large beta operations and quicker device turnarounds are allowed, as the 100 device limit (accumulative over the course of a year, not concurrently) is circumvented; debug builds to people not involved in the regular beta operation experiencing weirdness are made possible.
The benefit to the customer: More choice and increased stability. Companies below the Enterprise tier can create in-house applications with a wider spread. Open source and other hacking can be readily performed and encouraged since the spread won’t be capped.
The benefit to Apple: The pressure on the App Store approval process is lifted. One more impetus for jailbreaking, and the risks that come with it in the hands of those who are not suitably equipped, is removed while the proven security benefits of code signing are retained. Increased interest in small company applications and home development drives sales.
Apple splits the App Store in two: Today’s venue (App Store proper), and support for ad-hoc sources, with optional support for developers hooking some elements of their source into an automatic queue for App Store proper approval.
The benefit to the developer: Startups or new applications can build an infrastructure and gradually surface as they include new people into the launch process. Easy installation is retained without the delays associated with the App Store process. Developers of companion applications to a web site, desktop application or hardware appliance can offer up easy, native access just to their target audience. Adding a new source need not be more complicated than tapping a button on a web site. Using the marketing leverage of the App Store proper truly becomes a powerful opt-in choice for its intended uses and needs not stay a bridge to be crossed for every other use.
The benefit to the customer: The App Store proper becomes easier to navigate as Apple may choose higher standards for the showcased applications without fear of monopolizing the platform. Ease of use is maintained while choice is expanded. The device stands to become much more useful as applications that have previously been untenable are allowed entry.
The benefit to Apple: Since Apple may be notified on an opt-in basis of new sources, they can streamline the App Store approval process; since apps are able to be released and distributed more frequently and easily for developers, more stable versions are declared production code, and less snags are hit during approval testing requiring resubmitting bug fixes. Apple can be stricter in their guidelines and distance itself from “fart apps” or $10 000 “I am rich” installation apps should it so choose to without fear of alienating customers for which Apple’s image and reputation are not primary concerns in choosing how to use their own phones. Apple could maybe even reuse their iTunes podcast directory technology in learning how to scour thousands of external sources for relevant information.
This would be eminently doable and it would stand to benefit everyone involved. It would create a freer platform and a freer market and it would release the pressure on Apple and developers to make the absolute right choices the first time and in large scale or risk impacting customers. It’s maybe a lot of work, but if Apple started today, they could no doubt make it in time for the March unveiling of 4.0 and its beta and the inevitable launch of a new iPhone this coming summer. This is not primarily an engineering problem; this is a question of attitude.
There are negatives involved: not every single app will remain in the App Store. But the positive aspects, for all I can see (but I’m terribly subjective) far outweigh this negative aspect. The App Store today scales terribly in all axes but the technical content delivery bandwidth: scrolling through everything relevant in one genre in a store with over 100 000 applications is a horrible experience.
And let’s say the iPhone does take over the market as it seems poised to, and the US Department of Commerce decides to pay a visit in a few years’ time. Please raise your hand if you believe that they won’t be forced to implement some amount of similar measurements.
Apple has nothing to lose but its sickening addiction to control that infests the whole iPhone experience like a bad rash. And it really won’t have to give up a lot of it to gain a lot from it.
Preemptively: “What about the carriers?” Yes, what about them? Apple has said, under danger of essentially perjury, that AT&T didn’t have a role in approving apps, and I imagine that of all the carriers, AT&T has the firmest grip, being the first taker. That said, I wouldn’t think they would let any carrier be an issue.
“What about security?” Look up “phishing filter”. Apple could provide an updated list of dangerous certificates, and the ability to amend that list or replace it entirely (probably most attractively for enterprise use, but in any case evil if the opportunity was missing).
By Jesper · 2009.12.07 02:08
By far the simplest and most plausible change would be Apple removing the time/quantity limits on ad hoc distribution.
They would no longer have to approve an application before it could be sold and installed, but they’d still have the ability to revoke the ad hoc certificate of, or even remote wipe, an app that they found truly objectionable.
By Bergamot · 2009.12.07 08:12
Pretty much exactly what I’d like to see from them – the App store should simply concentrate on offering value in both directions (security / confidence / ease to users, and customers to developers) rather than control.
The phone could simply provide tiered levels of security and annoyance based on certificates and sandbox policies – i.e. installing unsigned software throws up a lovely big warning to the user, particularly if it wants to do anything that would normally be secured.
I suspect the thing with AT&T is probably one of those cases where what was said was technically true – AT&T probably have no role in approving apps. But they probably made their position clear to Apple (about the scope of VOIP and P2P apps) before the product launched. Not in a contractual way, of course.
Nor should we forget that AT&T were only chosen because Verizon refused to accept Apple’s conditions.
The ideal situation is something a bit like the iPod and audio DRM – the product becoming important enough that companies adapted to the iPod market – and the iPhone is probably the first phone to threaten that in the mobile market, because – apart from China – Apple haven’t crippled it on a per-carrier basis.
By JulesLt · 2009.12.07 11:57