There is also Transports that makes it possible to talk to other IM protocolls from Jabber, but not the other way around. yes it's a Jabber feature and you can't do it the other way around. Talk to your IM messager provider. Myself only uses Jabber, with Transports to ICQ, MSN and IRC.

So the easy sollution for IM-providers is what Google done, implement XMPP in/out from your service. Then you don't need to change your clients, but still your users can chat with other IM users, without changeing clients.

You have also other services combined with Jabber, like OpenID, computer-2-computer communication etc. And there are lots of clients and servers out there using Jabber.

Please check out http://jabber.org/ and/or http://xmpp.org/

And all multi protocoll IM clients I used is not as good as the singele protocoll IM clients. And as I want to chat with all my friends, XMPP is the only choise which is good at both parts.

But major renders, like AOL/Microsoft/Yahoo, only care how to lock in users/customers, therefore they are reluctant to implement their IM client with an open protocol.

I mean, only we the users are the driven force for the grand Im interoperation.

--
Let's play my favorite game at http://www.xbill.org/

@Tim: yeah theres already a open distributed im protocol and it meets all of your requirements, its called XMPP. its the basis for Google Talk.

it's a lame hack.

I think the popularity of multi-protocol clients prove there's a considerable demand for interoperability (to say nothing of the less sophisticated users running more than one client).

The problem is that Yahoo, Microsoft, et al want to lock you into their network, and have you use their client, which integrates with their suite of web properties and thus drive traffic and ad revenue. If IM was completely interoperable, the IM clients would be no more valuable than email clients.

That you can communicate between networks is a total joke. I'm European so correct me if I I'm wrong. But can you imagine an AT&T customer not being able to ring a Sprint customer? Of course not. That is the way it should be in IM land as well.

The Masked Millionaire

I know I sound like a an extremist zealot. But this is an important battle. What we have today exists only because TCP/IP, SMTP and other standards prevailed and made the open Internet. The next generation of services depends on an IM system. It may be XMPP, it may be SIP based IMS infrastructures - whatever. But please don't let it be a bunch of closed systems.

I think there's a deal between AOL and Google for pseudo-interoperability but it doesn't seem to use XMPP, and I suspect it's going to be the same between Yahoo! and Google.

BTW, the MT OpenID server doesn't import names and others data when sign in, which is a bit a pity.

Either, the large market players figure out a way to improve ad revenue from combining IM audiences under one ID. This might be through improved cross-IM ad targeting, that allows them to target users more effectively and therefore charge advertisers more money.

OR, one of the big companies simply bites the bullet and figures out if they don't open up they will lose market share (as noted above), and therefore risk losing audience, and then revenue accordingly.

Ultimately, I think if they can't figure out how it will make them more money, or at least protect what they are making now, it won't happen.

Also, if they do open up ID in this way, I think it will be across a full range o

Also, if they do open up ID in this way, I think it will be across a full range of services e.g. IM, email, social network, mobile access, rather than just IM.

- Open protocol, no centralised server
- DNS based, with a new type of DNS record perhaps? (or uses TXT records?)
- Handles both short and long messages, instant (like chat) and delayed if not available (like email)
- People or businesses could set up email2 servers, just like we do mail servers.
- More efficient file transfers.
- Point to point security, transparent and easy to use.
- White lists by default. This would get rid of 99% of spam. (simple!). Want to send someone a message? Get approval first, perhaps by phone, perhaps by web forms with captchas. A new way of thinking, but a realistic way to stop spam.
- Approved contacts stored on server, not on client.
- Would need a new format of email address to differentiate from the old email system. perhaps myname%mycompany.com or myname#mycompany.com.

A radical idea, but that's what's needed. Obviously no complete solutions here, just a broad concept ideas.

http://code.google.com/apis/talk/open_communications.html

AOL is moving towards XMPP too:

http://florianjensen.com/2008/01/17/aol-adopting-xmpp-aka-jabber/

Yahoo seems to have been playing with XMPP earlier in the year:

http://www.process-one.net/en/blogs/article/after_aol_yahoo_is_also_experimenting_with_xmpp/

Open standards over proprietary services FTW.

Considering that Trillian has been downloaded almost 36 million times from CNet alone, I'd say users do have great interest in interoperability. Absent the networks taking steps, they've taken their own. (Does anyone remember when AOL was upset with anybody trying to access their network?) Besides, AIM is now available via Google Talk, at least within Gmail.

I haven't seen numbers as to which interoperable IM client is tops, but I would guess Trillian. Still, support for the newer social networking IMs, such as MySpace and Facebook (read: Digsby), could be big. I'm thinking things may eventually move towards XMPP instead of separate proprietary networks, though.

AOL's comment makes perfect sense. Solutions like pidgin (formerly known as Gaim on Linux to me) have been the norm for a while.

As far as Digsby, it looks like an awesome client with a lot of useful features. It's just too much for my humble laptop to swallow.