One of the most interesting topics of iPhone speculation is the choice of interpreted, web technologies as the development platform. I greeted the news with a big smile, and a sigh of obviousness. Having spent a few frustrating years preaching the potential of agile mobile development platforms, it sits near and dear to me to here that Apple is paying attention to a bigger market.
Of course, the old-school, “Mobile 1.0″ crowd’s reaction is just as I would expect. Some of the claims make me laugh, so I felt motivated to chime in on the topic. Let’s break down the big three that I’m hearing:
“No SDK means no killer apps.” There are two issues here: (1) That there are ‘killer’ mobile apps that aren’t already included in the iPhone; and (2) That killer apps can’t be built with web technologies. For the first bit, ask yourself what the killer mobile apps are? Number One is voice… Number Two is SMS… Number Three varies, but support for syncing PIM data, taking pictures, listening to music, checking email, and browsing the web, pretty much covers it. For the second part, to assume that killer apps can’t be built with web technologies would require denying the last ten years of Internet development. The Web has changed everything — and it was built with web technologies ;-) Besides, Apple hasn’t commented yet on whether they’re exposing select native API’s via JavaScript.
“No clear revenue stream (for developers and operators) means no developers.” Stop thinking Mobile 1.0. Stop thinking traditional channels. Stop thinking about the Operators and Manufacturers “owning a customer”. Drop all this telcom baggage and start looking at the Web. There are plenty of companies making significant revenue simply because a large number of people have a browser and a data connection to their PCs. If anything, the mobile market becomes more interesting (and potentially more lucrative) when application development is cheap and the legacy mobile bureaucracy is out of the way.
“Developers need low-level access to the hardware.” This actually came up in a recent conversation, and I just about walked away at that point. Are you kidding? Do you have any idea how much of a PITA (and HUGE waste of time) it is to develop high-quality, reliable, usable, native applications on embedded hardware? I do. And I can assure you that you want no part of it. I appreciate the occasional need, and I’m sure Apple can give the John Carmack’s and Google’s of the world a l33t SDK; but if you’re looking to develop innovative, profitable mobile applications, there’s no reason for you to be tracking down memory leaks and hardware bugs. The less time you waste fighting the hardware, the more time you’ll have to launch new software. (If you don’t believe me, compare the rates of software and business model innovation that happens on the Web vs. on mobile phones. Mobile phones have done wonders for flattening the world, but they can’t compare to the Web as an environment for cheap, rapid innovation.)
C. Enrique Ortiz 9:39 pm on June 22, 2007 Permalink |
Hey Erik….
Killer applications can be developed using web technologies, no doubt about. But thin/web clients they are. There are things that just can’t be done using the web model… Access to native is about taking advantage of the handset’s full capabilities, vs. just purely thin clients; and for now only Apple will be creating such apps. As a person who makes a living out of mobile software, I need both, the ability to write killer web apps, as well as killer local apps. There is room for both, and requirements for both. I know know all that…
ceo
erik 12:42 pm on June 24, 2007 Permalink |
Hi Enrique!
I hear what you’re saying, but the argument assumes that Apple will hide native API’s from their widget engine. From the keynote, it sorta sounds like they will, but we don’t know that for sure. I entirely expect Apple to offer device-specific API’s via JavaScript, thus solving this thin vs. native issue.