Instant messaging protocol hits speed bump
THE IETF (INTERNET Engineering Task Force) this week put its instant messaging (IM) standard on the back burner, choosing to overhaul and revamp the protocol through a conglomeration of proposals submitted by various IM companies before June 15.
Although the IETF had been working on IMPP (Instant Messaging and Presence Protocol) since last year, the proposed protocol was tangled up in details and "abstractions," which necessitated a change, said IMPP working group co-chair Dave Marvit with Fujitsu Laboratories America.
"It turns out, the more you think about IM and try to get a really good standard that's both secure and scalable and easy to write to and all that good stuff, it's hard," explained Marvit, adding that the group has been "banging around discussing the issues kind of in abstract" until realizing about two months ago that writing a protocol of this scope could not be done correctly by a committee.
"For a while, they weren't taking any more drafts and I think they were getting into the weeds a little too much - it got to a level where maybe not as much was getting accomplished as could be," said Brian Park, senior producer for YahooMessenger. "Now that they know what a lot of the issues are, this new proposal will take existing issues into account- I think this is a good step in the right direction."
Marvit said the IETF will be collecting standards proposals from IM companies and creating a subset or one large proposal from parts of each submission in hopes of finding a proposal a majority of companies can agree on.
"We're not abandoning anything; we're saying let's put the working group on hold. We will not use that as the process for this next step," Marvit added. "When we get those protocol proposals, if they all stink, we'll say 'OK, back to the drawing board.' If they're good, if some are good and some aren't, we'll ditch the ones that aren't. We'll take those and say, 'Here are the good proposals of pieces of the good proposals. Now we have something concrete to talk about.' There's a big difference between discussing a protocol in the abstract and having a protocol to look at and take good ideas from and adjust them."
Another IETF meeting near the end of July would allow those who have strong opinions on a piece of the proposal to voice their concerns and figure out the next step for the IM standard. However, the timeline for a finalized, official IM interoperability standard is still up in the air and depends on the proposals the IETF receives June 15.
Park stressed that for many companies, especially Yahoo, speeding the protocol approval process toward implementation is of the utmost priority.
"I expect and hope that this time around, there are not going to be a lot of people in the way of making this happen," Park said. "I mean, I can't speak for what AOL is going to do or say, but I think most of the other companies really want this standard to happen and believe it's important. As long as the majority of those companies agree on something, we'll be able to move forward quickly."
As for grumblings that AOL is not helping the standards process along despite a pledge last year to work closely on an IM protocol, Marvit said he believes them to be false.
"I'm not as worried about the political intrigues, I'm really not. I'm focused on the protocol side," Marvit said. "Any contribution [AOL] wants to make to the group is most welcome, and they know that. Certainly, there isn't anybody oobstructing the process in any way . I don't know how you would do that."
Should the standards process become bogged down again for whatever companies, IM companies involved in the proposal process, such as TribalVoice and Microsoft as well as Yahoo, could attempt to raise public interest by spreading the word about the benefits of IM interoperability among their users. A second option, Park said, would be to prove interoperability's usefulness by banding together with other IM companies and linking their IM systems.
"Coming together and showing it's a possibility, even if it's not a standard that people can talk to another community, opens [users'] eyes up to say 'OK, if this is possible, why can't I talk to this community or that community?'" said Park, who did not comment on companies with which Yahoo might consider joining forces. "There definitely can be arguments made that showing customers what can be done will help speed the process of interoperability because people start asking for it more and more once they see it as a reality."
However, this forming of partnerships is less desirable than an open standard, because it does not benefit the entire industry.
"Open standards are really the ideal way to do [interoperability]," explained Park. "The reason it's ideal is because there's one standard that every developer in the world can build a service without worrying about what all the other companies are doing: They just build a cool product and put it out there knowing it's going to work because there's a standard they can apply their code against."