Doc Searls on XMPP and SIMPLE
The SIMPLE case for XMPP. Eric Norlin points to XMPP rises to face SIMPLE standard ? Vendor coalition challenges standard used by Microsoft and IBM, by Cathleen Moore in Infoworld. He reminds us that XMPP is the initialism for the more familiar Jabber protocol, which Peter Saint-Andre suggests that we distinguish a bit by "proposing that XMPP is to Jabber as HTTP is to the Web." As both Cathleen and Eric point out, SIMPLE is a big developer (IBM and Microsoft) project, while Jabber is an independent developer (Jabber, Inc.) and open source community (Jabber Foundation) project that has attracted the involvement of some big companies on the demand side of the market (Hitachi, France Telecom, Sony, Hewlett-Packard). Last I heard, IBM and Apple were also making use of the Jabber protocol. (Here's a case where IBM is using Jabber to help build a wireless emergency network in Washington, D.C.) XMPP and Jabber are taking off because they are (truly) simple infrastructural building materials that are highly compliant with the NEA nature of the Net. SIMPLE is being pushed by IBM and Microsoft for a variety of reasons; but that very variety may be the core of a problem. Here's Cathleen Moore: Moreover, because the SIMPLE protocol is still incomplete, IBM's and Microsoft's implementations have required the addition of proprietary extensions to make their offerings work. Admitting that its forthcoming Microsoft Real-Time Communications Server 2003 contains propriety extensions to fill out the SIMPLE protocol, Microsoft steadfastly maintains its commitment to the SIMPLE standard as it matures. "We are absolutely committed to being SIMPLE compliant," said Ed Simnett , l ead product manager of RTC Server at Redmond, Wash.-based Microsoft. "SIMPLE is not a mature protocol at this point, so we've taken a snapshot and implemented against that," Simnett said. "As we make subsequent releases, we will be looking to make sure we are absolutely compliant to the standard." In fact, Microsoft views SIMPLE as a superior technology choice because of its extensibility into other media types such as audio, video, file transfer, and whiteboarding, according to Simnett. "One of the reasons we picked SIP and SIMPLE is because it allows more than presence and IM. We think that one of the most exciting long-term vision areas here is making sure IM and presence become base building blocks in the broader real-time communications arena," Simnett said. "Other protocols focused too much just on instant messaging and not enough on the other media types we think will be important." Having just shipped a SIP gateway for its Sametime IM platform in November, IBM Lotus also maintains its commitment to SIMPLE. The standard is young and vendors are implementing it differently, but interoperability is the goal, according to Kevin McLellan , marketing manager of workplace collaboration products at IBM Lotus in Cambridge ,Mass. "The fact that the largest industry players, [IBM] and Microsoft, are supporting SIP and SIMPLE suggests we will work to make sure that the standard evolves to achieve interoperability. That is the goal. We are also working hard to make sure it evolves in the right way." IBM intends to add native SIMPLE support to Sametime in a subsequent release, McLellan added.
Meanwhile, IBM's SameTime and Microsoft's MSN Messenger barely interoperate. Both companies have done little in the past to demonstrate a willingness to meet market demand for interoperable IM. Last month Microsoft announced a product that will bridge SameTime and MSN Messenger, but just for corporate intranets. Meanwhile, any company, any customer, can develop its own Jabber-based IM system. Countless numbers have done exactly that, including IBM. SIMPLE may be a fine protocol when it's done; but it's not. And from the sound of what Microsoft says above, it risks violating POGE: the Principle of Good Enough. Without POGE we would have no TCP/IP, no HTTP, no HTML, no SMTP. In other words, no Net, no Web, no Net-based email. POGE also accounts for the success of XML and Linux. It's why XML-RPC moved faster than SOAP. There are two questions you need to ask about protocol efforts like SIMPLE and XMPP: Who's pushing it? and Why? Then you look for pudding proof. So far I don't see it for SIMPLE. I see lots of it for XMPP. Disclaimer: I'm on the Jabber, Inc. advisory board. One of the reasons I'm on that board, however, is because I want a world with wide-open and ownerless IM protocols. [The Doc Searls Weblog]
9:33:46 AM 
|
