It is time we stopped talking about Rich Internet Applications, the term coined by Adobe to describe Flash-based Web applications, and borrowed by everyone else to apply to their latest efforts to improve
HTML. The discussion is not really about rich applications versus, I guess, plain applications; rather, it is about the next generation of the client. Although it is too soon to know what that is, it strikes me that we already have some strong clues:
- It will not involve wrestling with the Windows Installer every time you need to update an application
- The same code will run on Windows, Mac, and probably Linux and mobile devices
- Local data will exist mainly as an offline or performance-enhancing cache of server or internet-hosted data
- It will most likely run in a web browser
- It will most likely be written in JavaScript
-
What fascinates me is the way in which our most influential technology companies are putting their weight behind different approaches to these common goals.
Apple would really like you to write native code for OS X or the iPhone, but for web applications it is backing Safari and JavaScript libraries like
SproutCore.
Google and Mozilla are also stretching browser technology, with Google adding offline data support through
Gears, and Mozilla determined to out-Flash Adobe by supporting
SVG, Canvas, and
HTML 5. Microsoft has
Silverlight, one of its most interesting developer products for years, which brings
XAML and
.NET together in a neat cross-platform browser plug-in and Adobe has Flash,
Flex and
AIR - all variations on using the Flash runtime as an application platform.
I realise there are other approaches. Perhaps Sun will yet come good with
Java FX. But it seems to me that these are the key players jostling for position as developers try to figure out where to go next for client applications.
I also realise that it is early days. A quick search for vacancies on
CWJobs reveals just 106 for Flex and 79 for
Silverlight, versus 3417 for
C# and 2116 for
Java at the time of writing. Still, I am betting that figures for both Flex and Silverlight will increase markedly in the next year or so. I also believe that Adobe will be hard to catch, because designers already know and love
Flash, and because developers will prefer to target a relatively predictable runtime, even though it is proprietary, rather than trusting browser vendors to stick closely enough to standards to let their tangle of HTML, CSS and JavaScript work reliably across all the platforms.
Silverlight is attractive too, but Microsoft has a lot of catching up to do, especially in runtime deployment, and has not yet come up with anything to match
Adobe AIR which takes Flash onto the desktop.
AIR still needs a bit of work - it is silly that the runtime can write all over your hard drive, but cannot launch a document in Word or Excel - but Adobe has time to improve it. Flex works well with both Java and .NET on the server. Finally, ActionScript 3.0 is close enough to C# or Java that it is not difficult to make the transition.
Could Flash really be at the heart of the next-generation client? It is certainly possible.
