Dotmobi or not mobi

Under the glassy surface of the mobile application world, there's the heaving and grinding of tectonic plates. They're shifting, buffeted by powerful currents deep in the hot, fluid core of the mobile business. These currents are due to the interminable internal politics of handset manufacturers, network operators, content providers, and increasingly the developers of phone applications.

History has not been kind to the mobile application space. WAP 1.x in 2001 was cwap, and was rejected by the market. With WAP 2.x, many developers of WAP sites tried foolishly to replicate their Web sites, rather than seeing it as a new way for people to access “just the right information” quickly when on the move. It shouldn't have been such a failure - the Japanese equivalent, imode, has been pretty successful.

The next big thing for mobile software was J2ME (Java) and native apps running on mid-market handsets. Only fools, idiots, the wealthy and the uber-geek had smartphones back then, but the mid-market was pretty big. Actually deploying Java applications is a royal pain in the ass, though. Every model handset was its own compatibility circus, and then you still need to persuade users to download and install the .jar.

Third-party app developers wanting to do something more complex had to look to Symbian and Palm OS and (cough) Windows Mobile, which provided “device independent” platforms that could run third-party applications. In theory. Most who tried ended in the same frustrating bog: Which model? Which version of firmware? Sorry, this application is not compatible. It was almost as much of a pain developing for “device independent” platforms as any other.

What if you were just a brand trying to let users see info about you on their phone? Too much PT. Fail.

And this is the kicker. Developing applications or information tools for handsets is a nightmare. Every handset does meet certain requirements, has certain device standards, although most are of no use to beast, man, or software developer. Basically, every single model of phone has some kind of technical gotcha that needs a developer to sit, test, debug. Add to this that most manufacturers have complicated application-signing processes, like Symbian Signed.

Apple is a barely significant rounding error in global mobile handset shipments.

Roger Hislop, contributor, ITWeb

Every time a manufacturer releases a new firmware version, there's a chance your app will fail. Every time a manufacturer releases a new model, there's a chance your app will need extensive redevelopment to work.

Then came dotmobi. More of an intellectual change of gears than anything else (plus a couple of bits of development stuff, such as XHTML-MP and CSS rules for pages. Like no, the poor little mobile CPU does not want to calculate beautiful graduated multi-hued backgrounds.)

The idea with dotmobi was to get around handset idiosyncrasies by doing the processing and even interface rendering on the back end, with the handset browser displaying only what the user needed to see. With some clever design work, a dotmobi service could appear almost indistinguishable from a native application running on the handset. It's even better in many cases than native apps, because dotmobi does not suffer the average handset's pitiful processing power and access to bandwidth.

Basically dotmobi was “software as a service”, it was “cloud computing”, it was “the network is the computer”, long before it became a fad for hip IT directors. .mobi itself is just a TLD with a note to operators to please not reformat pages and break stuff.

At the heart of the dotmobi concept is the very important message: you need a different strategy to address content to mobile users than Web users, because mobile phones are not poor man's computers. They're a whole, different thing.

Dotmobi was “driven” by Nokia as a strategic vision. Things were moving along, albeit sluggishly: there was a direction to go in where device independence would become a reality.

But just as this new philosophy was starting to grow roots, along came iPhone and ruined the party.

Apple has no interest in device independence, no interest in cross-handset functionality, no interest in anyone other than sucking in the buyers of its hardware. It is a strategy solely focused on the tightest possible integration between the lines of code and the silicon chips, and making sure that no one, no one, no one at all ever interferes with it.

Unfortunately for Nokia and its allies in dotmobi... and the end-user, the iPhone was wildly successful at launch, and Apple's genius for taking the blindingly obvious and already existent and repackaging it as an earth-shaking innovation won. Again. The iTunes store for apps was simple, elegant, worked and let app developers make a buck (unlike Nokia Forum and its bastard offspring, the Ovi Store, which is high farce).

The other handset manufacturers were left flailing around helplessly. Application developers glommed onto iPhone, and wide-eyed tech commentators and tech press followed unquestioningly.

Except Apple is a barely significant rounding error in global mobile handset shipments, especially in Africa (IDC numbers and Gartner numbers).

iPhone is not the phone in the hands of the masses of consumers out there. It can never be. It's expensive, complex, fragile. But sexy. Seductive. Easy to gush about in blogs and magazines. Relatively straightforward to write apps for, be published, and find buyers that can pony up a couple of bucks in a few clicks. App development flourished on the App store, endless gallons of ink were spilled about it.

Which is why every half-baked brand manager looking at a mobile strategy is leaping straight into briefing a software house to write them an iPhone app. It's even become cheap now - in Europe you can pay less than EUR1 000 for an app delivered in three weeks.

What brand managers and digital marketers in South Africa somehow manage to remain entirely oblivious to is that the iPhone in South Africa is a negligible audience. And one that can be reached through Web-based dotmobi applications anyway. What a waste of time, money and effort.

It's time for the mobile application industry to give itself a damn good slap, and refocus. Those who want applications on their mobile can get them in the easiest, fastest, most reliable, most efficient way through dotmobi and Web-based services.

A mobile software industry that gets seduced into the dead end of device-based applications will set the industry back as badly as in the WAP fiasco. It's time for the mobile content and mobile applications industry to get their thumbs out and stop this app train before it comes off the track, and again crushes the brand managers and investors who pour their millions into the industry.

Brand managers are about to get that 'Ow! Burned again! Disappointed! In trouble with my boss for getting no return on the investment' feeling. And that's in no one's interests.