by John Browne, on Feb 26, 2013 4:00:00 AM
Now that Microsoft is on the scene so to speak I think you will see a big push to get developers to write apps for the Windows store. Admittedly the smart thinking is that they will only obtain about a 10% market share over the next 4-5 years but I wouldn't discount their ability to enlist/encourage/extoll developers to get onto the platform.
With RIM back in the picture again (although they may be fighting for the crumbs under the table) we're facing the Balkanization of mobile devices. Objective C (iOS), Java (Android), or C# (Windows 8)? What about apps with roots back to VB, COBOL, .NET, or even Win32 API? What about the browser?
It can be a real trial to figure out how to code against all these targets. You could write one version in, say, C# for Windows, then subset it down for Windows 8 RT and the phone, then rewrite it in Java for Android tablets and phones, then rewrite it again in Objective C for iPads and iPhones, then finally build a version using HTML5 and Javascript to run on the browser. A ton of work to cover all those platforms. And every bug fix/update/improvement requires doing it again and again.
Ugh.
So this is why I think HTML5 will conquer all (kind of like love) except that I don't think pure HTML5 apps will be as popular as hybrid approaches, which I also think will make more sense than multiple native versions.
Mobile application development is hot and is going to consume a lot of resources. More and more CIOs are facing not the Dawn of the Dead but the Day of the iPad (or Samsung Galaxy or Microsoft Surface). It may not be high noon yet but the sun is rising quickly and it's shining on a world of BYOD. Can you imagine next year beginning a new line of business app development project that doesn't include a plan for tablets and smartphones? Me neither.
I still think there are compelling cases for the pure native application. The Facebook experience notwithstanding, if I had to write such a critical (and revenue generating) application for a iPhone it would be native all the way. But for a LOB app that the IT department has to migrate from VB or PowerBuilder or C, and has to support 2, 3, or 4 platforms, I think using HTML5 with calls into native services like GPS makes a lot of sense. HTML5 is a rich language and is just getting richer; it offers a widely-supported framework to abstract most of your application code (combined with Javascript and back-end web services to do the heavy lifting); and it's easy to modify with cascading style sheets (CSS).
That way if Marissa Mayer comes out with the yPhone we'll all be ready.
