I mentioned earlier how disappointed I am to see that the iPhone will not support installation of third-party applications unless those apps are approved by Apple, and presumably distributed by Apple or Cingular. Merlin Mann continues to speak for me on the impact of this; I heartily endorse both his Digg-inspired assessment that this “does indeed suck monkey butt” and also his prediction that the decision will have no appreciable effect on sales — in the short term, at least. (More on that later.) Will I never buy an iPhone because of this decision? Never is a long time, and as Merlin says, who knows how this could evolve. Every Dashboard widget I’ve ever installed has come from Apple’s Dashboard site, and that doesn’t seem to have made anything worse about Dashboard for me; and I’d seriously consider paying hundreds of dollars just for the video/iPod combo. But I wouldn’t pay $1,936.00 (link via Consumerist) for the iPhone unless they reverse their call on third party apps.
Why does this matter so much to me? And why should it matter to anyone other than supergeek Radar/Digg/Slashdot readers, as one of the commenters on my earlier post asked? Randall Stross makes one argument for why in today’s New York Times, and Cory Doctorow takes that argument further at Boing Boing. I’d like to add a few stories, though, from my four years as an owner of three different Treo phones — not theory, not politics, not necessarily supergeekdom, just practical reasons why the open PalmOS software has mattered dramatically to my use of the Treo.
- When the Treo 300 first launched on the Sprint network, Sprint did not provide text messaging for it. A third-party developer called PDAapps came up with Treo300SMS, an installable application that acted as a gateway to Sprint’s web form for sending and receiving text messages. A total hack, and something Sprint would never have approved had they been given that control. Text messaging is a natural application for a Treo since it has a decent, usable keyboard, and Sprint charges people to receive text messages, so it benefitted them to have the capability — but they hadn’t gotten to it yet. As a result of having an open platform, PDAapps made the Treo 300 more useful and appealing than it would have been until Sprint and Handspring got around to making text messaging work. Of course, now every Treo comes with an excellent SMS application, but for the months that passed between the 300′s launch and SMS support, third-party development filled in the gaps, to both Sprint and Handpring’s benefit. An open platform allows developers to implement functionality the platform provider hasn’t gotten around to yet.
- The Treo has always come with an email application, and quite frankly that application has always been pretty sucky. The usability was poor, and at least at first (I haven’t checked recently), the mail app didn’t support protocols and security systems that some mail servers require. Fortunately, a third-party developer, SnapperMail, implemented a fantastic email client that supports all sorts of systems (POP3, IMAP, SSL, and so on), and includes applications to read image attachments, open zip files, and more. (Don’t take my word for it — Walt Mossberg of the Wall St. Journal calls it, “the cleverest and most capable hand-held e-mail program I’ve seen [...] the closest thing on a hand-held to the kind of full-featured e-mail programs people use on their PCs.”) Because the Treo lets me install third-party apps, I’m not stuck with whatever crappy email program Palm decides to give me, and probably frees them from customer demands they otherwise would have to rush to fulfill — again, to their benefit. An open platform allows developers to reimplement and replace functionality the platform provider has gotten around to, but has failed to do well.
- When the Treo 300 came out, it did not have Bluetooth; that didn’t arrive until a later model. It did, though, have a USB cable for connecting to your computer, and the clever developers at JuneFabrics came up with PdaNet, which allowed you to use your Treo as a laptop modem through the USB cable. When I found this, I was overjoyed, since it meant I could get around paying outrageous fees to hotels for connecting to the Internet while on the road. I know from people at Palm that Sprint absolutely flipped out when they heard about PdaNet, and can easily surmise they’d never have approved its use. Later, when Palm added Bluetooth to the Treo, Sprint shipped it with Bluetooth networking (which allows you to do the same thing PdaNet does, without the cable) disabled. I and many other Treo fans flipped out and promised to leave the platform, and Sprint relented, contacting many of the bloggers who had complained and saying, “This sort of use only represents a pretty small fraction of Sprint Vision customers, so we see it as one of those areas where if this is how customers choose to connect, we don’t stand in the way.” (June Fabrics also went ahead and implemented Bluetooth networking anyways.) In this case, an application Sprint immediately hated and feared wound up in the wild, and in the end Sprint determined that it was causing them little or no harm. Like the movie industry reacting to VHS, what they thought might be the death of them instead turned out to increase the utility of their network, and made people like me happier with their products. An open platform allows developers to meet needs that scare the platform provider, and allows consumers to have those needs met where otherwise the platform provider would block a capability.
- PalmOS has over 20,000 applications available for it, developed over the ten years the platform has been around. On my Treo, I use or have used the Oxford English dictionary, Google Maps, Scrabble, Monopoly, Bejeweled, a diet and exercise tracker, a BART schedule, a crossword puzzle program (which lets me download and play the New York Times crossword puzzle anywhere I am), a shopping list keeper, a currency converter, various flight trackers, as well as the email, SMS, and networking apps I mentioned above, and many more. My wife and I played Scrabble on BART last night after looking up the BART schedule and checking times for the movie we were going to see — all on the Treo. (Geeky though this is, one of the first things I wondered about the iPhone was, how long until it supports crossword puzzles?) The result of this is that I am far more bonded to the Treo than I ever would be without those applications. As a phone, calendar, and contact manager, the Treo is nice, but with that list above, it becomes much more. I don’t know how many applications Windows Mobile has, but their site claims “hundreds”; I assume the count is really in the thousands, though also thousands less than for PalmOS. Why does Palm continue to hold its own against Microsoft in this realm? I’m sure that some of it has to do with simplicity, but some of it, I bet, has to do with how much more you can do with PalmOS. An open platform allows its users to get far more done, and latches them to that platform far more tightly as a result.
It doesn’t surprise me at all that Apple would look at these arguments and reject them — as with the original Macintosh, the NeXT, and the iMac, expandability is not a 1.0 feature for Steve Jobs. Apple also has a significant amount of history competing with its developers (Audion/iTunes, Watson/Sherlock, Konfabulator/Dashboard, and so on), which says to me that they don’t consider a strong developer community a requirement of success. You could also easily look at this decision and compare it to the original Macintosh as an integrated, single-source platform, competing against the PC compatible world, which would also suggest they’re repeating past mistakes. Beyond just Apple being Apple, though, Cingular might very well have pointed to PdaNet as exactly the reason they would want control over applications on the iPhone, and perhaps they found a sympathetic ear with their partner. I easily can see how this came to happen.
It will be interesting to see what happens now. The Macintosh eventually got expansion slots; iTunes did not. Maybe you can look at the iPod and see the lesson learned as, if the design is good enough, Apple can keep control of the whole stack — the Macintosh didn’t try to take too much, it just lost the lickability that strategy requires. I don’t believe you can take the one device everyone carries with them and lock it up as a closed platform, and still succeed. I would rather take two devices, where the one with the antenna is open, than just the one, closed. You can’t line up against every other company in the world and say, we get all of this, without those companies deciding to fight you. It would be smarter for Apple to figure out how they can make others (and not just Cingular) successful on top of what they build, rather than trying to own and control everything. For me that strategy is a deal-breaker, and I think it should be for you, too.