Jan 21, 2008 13:36
Blog sites: Get it, got one. These babies take entries I've written and display them to the public. Cool.
Social Networking sites: Get it, got one. These puppies take bits if information I choose to provide, and allow me to see similar information about others. They provide various types of social interaction. Cool.
Webmail sites: Get it, got one. These sites give me an email address, and let me send and receive emails. Cool.
Portal sites: Okay. I get it, sort of, and I don't have one. Why? Well...
Portal sites are about a 'customized webspace', a 'personalized foyer' for the internet, and various other bits of marketroidal ideas. In a nutshell, they take all sorts of pieces of information from the web, and give it to me in a sort of 'online dashboard'. I could, for example, have the number of unread emails in my inbox in one corner, and a list of topics from my favourite news source along the bottom. Blah blah blah.
Hey, that's cool, but this 'portal provider' wants me to provide all sorts of information. My email address and password, for a start, so they can check that for me. And how useful is this beast, really? What does it streamline? What becomes easier, or faster, or more convenient?
Moreover, why should I fork over details to some company for this sort of thing? I already HAVE an 'internet dashboard', sort of. My browser. My tabs, if correctly organised, can show me these things. My bookmark toolbar makes lists I use a lot accessible at a moment's notice. BUT...
It still isn't quite the same, is it?
Why shouldn't it be?
And why is GreaseMonkey so difficult to bother getting into?
UPDATE:
More to the point, why is AJAX cool because it 'enhances the end-user experience'? Shouldn't it be cool because it allows for a cleaner separation between the presentation and data layers?
And finally, my question:
Why are web sites obfuscating the interface of their 'AJAX' application? Shouldn't they be publishing it? Obscured calls don't add security, although I suspect that's the usual motivation.
UPDATE:
In fact, does this relegate the web server to an rpc server with an additional 'getFrontEndInterface' call, some session tracking, and a few http headers?
Sorry if this doesn't make sense even to those who know what all this stuff means, I'm just musing and don't have anywhere else to make notes right now.