The next web - beyond the browser!!

Hi…

Today it seems that everybody puts web and internet into the same bucket. At least from a technical perspective there is a strong distinction.  What is called “the web” is only a type of communication which could use the internet as a media. The standards to mentioned are HTTP and HTML, right? The internet itself is much broader in its abilities. All of us now other types of communication like SMTP or SIP or whatever.

For the end user – thank goodness – this should be transparent. She or he simply use the service. But under the covers there are quite severe things to look at. For example HTTP is not transaction oriented nor does it easily support status. Today we know it was not the best idea to disregard this at first (There was no idea of a DOM in the browser upfront. Today we generate the HTML in the browser using a DOM and JavaScript).

What we did over the years was to use layers on top of this levels (like the Web Service standards cope with drawbacks of the HTTP protocol) or we tweaked what we had. For me HTML is – at least what we see as HTML today – tweaked to an extend which makes it incomparable to its early ancestor.

So looking at all this hype around AJAX it really kills me. I did a session last week where a guy in the audience asked me about the architecture behind AJAX. So I drew the simple version (or to be honest I copied the one once published in the IEEE magazine). He was really astonished and asked about the place where the state is kept, how the used is identified, where does the business logic lay, how do you debug this and so on. Every answer I gave started with “I know it is ugly but…”

OK, IT industry accept it: Users will start to expect their user interface to use the nice AJAX things. Because from a user perspective it is nice and desirable. So we as an industry must react. What Microsoft does by presenting a framework called ATLAS is certainly the best way I saw so far to flee the trap. But what a trap? Developing AJAX applications from scratch is all but easy. So it will be more necessary than ever to relay on a proved, extensible framework to help.

But sorry is this the end? Hopefully not. If we are honest to ourself: The web had a huge success regarding its size and user acceptance. Regarding the technical baselines we are way behind what can be taken for granted in a desktop app.

So should we accept it and find new ways around it by tweaking? Sorry, I am not willing to support here.

We can do much better and we should. For our own and for the sake of our users. So please stop racing behind the browser and start to rethink. There are applications well suited inside a browser window. And there are (maybe the majority) which simply aren’t. I’m typing this from within a train coach, see what I mean ??

CU

0xff