I wasn't so afraid about the scaling as much as the duplication of intent.
Granted I've never used their chat feature (I haven't used Facebook in
ages really) but if it does what it says on the box why would you have two
distinct protocols to perform the same job? It seems costly in the sense
that you have to put specific resources for each when you could use them
in a single, non competitive, team.
But in effect their current web based client is not really any different from using something like BOSH or the gmail gtalk client, they are different front end protocols yes but there is no reason they cant just be a simple wrapper for a common backend, I don't see why you think that's so costly, there is no reason that it would be if designed properly, how do you know they didn't think about XMPP when they designed the web client infrastructure? Their backend might already be perfectly designed for easily bolting an XMPP client protocol interface onto it without masses of work.

I'd say they meant for IM purpose. We'll see when they start pushing a
client that offers some of the features you can find on the website
(xmpp:facebook:poke extension anyone? ;)) and work only in their own
client that, oh by the way, displays some advertisement too (I'm being
negative here I know but one may still wonder considering Beacon).
Possibly, although I doubt they will, but we will just have to wait and see.

But yeah you're right, at this stage it's all hypothetical anyway. I'm
excited from a developer point of view and yet careful on the way Facebook
will use it business wise. If Facebook was using XMPP to push ads for
instance that would hurt Jabber in people's mind I'm afraid.
I don't think them having their own client that shows ads in it would hurt Jabber/XMPP, you don't have to use a client that shows ads, that's the beauty of XMPP. The only way it would hurt it is if they locked it down so you could only use their client with its ads, which would pretty much defeat the point of even using XMPP in the first place.

Lets just wait and see what happens.

Richard


Reply via email to