I’m busy with my real-world job trying to piece together a portal using Oracle’s Portal product. It feels very bloated and overly complicated but there’s solid support for the product which is at least part of the reason why the decision makers probably went with it. It just takes some time to get my head around all the little nuances. Although it leaves me wondering when I come across some portlets released by Oracle themselves which seem lacking, such as the Exchange 2000/2003 portlets not supporting SSL (they use OWA to access user’s information, which is almost always done over SSL … wtf?) or a discussion forum which has the word “Product” hard coded into everything. You’re not adding a discussion, you’re adding a “product”. You’d think there’d be a little more customization built into these things but.. nope.
Anyways, that’s what I’m up to.
Yes, Virginia, there is an HTML 5. But it’s going to be known as “Web Applications 1.0”. I have not had the chance to fully read the spec but only glance through it. It looks like there’s going to be a lot of new things to play with, but I’m worried pages are going to get very complicated very quickly.
That’s one of the things I really like about HTML 4, it really isn’t complicated. It provides only basic data structure but allows you to build up complex structure from those basic elements. HTML 4 is made of individual blocks and it is up to you to build a castle. With HTML 5 I think we’re seeing some parts of our castles are now prefab which would seem to make our lives much easier. But I’m a little worried that things are getting overly complicated. I see things like the bitmap canvas and wonder if some sort of descrimination isn’t being built-in (or further enhanced) against non-visual clients (visually impared users, applications which can only process text, text-only interfaces). It’s as if HTML 5 is trying to mimic the features Flash or JAVA can already provide.
I’d like to see an HTML 5 where HTML 4 syntax was tightened up. Where we had a much larger definition on how browsers should handle/render specific tags. Try and relieve any abiguity left over from HTML 4. Try to settle down on a simple, easy to use, easy to implement standard that can be supported for years to come. Rather than develop a spec that will put current browsers into obsolescence in only a few years. It will be Netscape 4 all over again.
And I really wish these specs didn’t lay down those features clients ought to try and implement now with the assumption they will become “law” when the w3 signs off on the spec. We ran into problems with this sort of approach back in the days of Netscape 4 and CSS. Let’s hold off a bit and wait for a final spec before we all rush out to develop new browsers. Maybe then we can get some consistency between platforms.
But we’ll never have consistency.
Because instead of tightening the already existing spec of small blocks with which we can build whatever we want, we’re being given prefab parts that will fit everyone’s castle differently. We’ll have inconsistent support across browsing platforms. User experiences will differ and us web developers will either have to not use the new HTML 5 features or work around bugs in browsers as needed.
It’s CSS all over again, or at least it appears to be. And that worries me a bit.