Friday, September 21, 2007

Developing enterprise applications for mobile devices remains way too hard

A logjam exists between developers and their ability to productively deliver enterprise applications and data to mobile devices, such as cell phones, PDAs, and so-called converged devices like the Apple iPhone.

The logjam is complexity and too many obnoxious variables. To develop applications that reach even a small number of major handset environments means big-time custom plumbing, from the various data sources, to the mixture of networks, to the choices on synchronization, to the various security needs, to the many user interfaces and mobile client operating systems. Managing all these variables requires a high degree of skill across many different skill sets. There are not many developers that fit this bill in your average enterprise.

And this all means a lot of time and money is required to bring just a few basic applications to just a few basic mobile clients. No wonder enterprise mobility stubbornly remains below the radar for IT leadership. Mobile remains relegated to the crowded back burner of IT imperatives.

And given all the variables and high degree of required customization, few ISVs have emerged to try and make a living at producing mass-market mobile applications. The subsequent lack of killer applications, other than standards-based email and text messaging, reduces the appetite to take on the infrastructure complexity for taking the corporate datacenter out to the mobile client across commercial mobile networks.

Also impacting the complexity is the diversity in how mobile devices work from geographic and regulatory market to market. It's nearly impossible to envision a global approach to mobile wireless computing, as we've seen for desktop and web computing. Designing for one mobile market does not give you much of a leg-up in reaching many others.

The "inclusive platform" approach of dictating the exact device and/or runtime environment up and down the stack that mobile applications play in is itself stifling given the inability to take advantage of the low-cost devices and services available via commercial mobile service carriers. What's more, aligning the back-end and front-end infrastructure does not necessarily align with how the mobile telecommunications carriers and handset operators, well ... operate. They don't like the idea of losing control of what their clients do on their networks.

Clearly, wireless handheld delivery of enterprise data and applications has yet to reach its untapped and vast potential.

Microsoft for years has been grappling with these issues, with many fits and starts. There have been some impressive successes with Windows Mobile, but Microsoft has by no means sewn up the field of mobile enterprise applications design and delivery. Microsoft with its Windows Mobile approach has not yet achieved a critical mass for how business applications and data can be driven out to a field-based workforce. And it's likely that today's widely heterogeneous environment for end points and devices will be with us -- at least in the U.S. -- for a long time to come.

These seemingly intractable roadblocks to wider mobile business use are one reason we're seeing what amounts to appliances for the client devices. The Apple iPhone, which launched with great fanfare in late June, is a prime example. Apple has fused hardware, software and applications -- as well as a few critical APIs -- and has picked one opening-inning carrier, AT&T, as well as Wi-Fi generally, for connectivity. And that combination of attributes, along with what should soon be more APIs, makes the converged device/appliance approach not just a consumer affair -- it begins to help reduce the complexity for enterprise mobility too.

Microsoft in the summer said that it likes the idea of such fused mobile clients -- the appliance on the client -- so much that it has hinted it will produce such integrated devices too. Rumors persists that Google also has its sights on a mobile handheld appliance of some sort. Things are clearly heating up.

These nifty clients should go quite a way to making enterprise mobility far easier by allowing developers to exploit them with fewer interface and connectivity variables to manage. But the converged client still needs a back-end or middleware counterpart to help coordinate an enterprise's data, logic, and security needs. Sybase is hard at work on what may prove to be a game changer for such enterprise mobility middleware -- especially when coupled with a converged device such as the iPhone.

Sybase has bet its future growth on mobility. And while the Dublin, Calif. company has not yet announced the details of its full stack, the Sybase vision makes a lot of sense. Expect in the next year to see what amounts to a distributed middleware system approach to enterprise mobility from Sybase that may break the current logjam in enterprise mobile development and deployment. If Sybase can shake up the enterprise mobility infrastructure industry -- and partner effectively with the likes of Apple's iPhone -- then Microsoft's response will need to be swift and significant.

Dare I say it? What the heck: Look to 2008 to be the year that enterprise mobility finally gets some legs.

No comments:

Post a Comment