- From: Box, Don <dbox@develop.com>
- Date: Fri, 3 Mar 2000 01:17:56 -0800
- To: "'SOAP@discuss.develop.com'" <SOAP@discuss.develop.com>
- Cc: "Tim O'Reilly (E-mail)" <tim@oreilly.com>, "'timbl@w3.org'" <timbl@w3.org>, "'tbray@textuality.com'" <tbray@textuality.com>, "'ken@bitsko.slc.ut.us'" <ken@bitsko.slc.ut.us>, "'dave@userland.com'" <dave@userland.com>, "'Daniel.Veillard@w3.org'" <Daniel.Veillard@w3.org>, "'connolly@w3.org'" <connolly@w3.org>, "'eric@w3.org'" <eric@w3.org>, "'xml-dist-app@w3.org'" <xml-dist-app@w3.org>, "Henrik Frystyk Nielsen (E-mail)" <henrikn@microsoft.com>, "Steve Vinoski (E-mail)" <vinoski@iona.com>
I spent the entire week at XTech and had an amazing experience. As many of you know, there was a "BOF" of sorts thrown by the W3C on March 1. Three of the five SOAP authors were there (GopalK, DaveW, and me). Several IBM-ers were there. Several Sun and Netscape people were there. A lot of W3C folk were there, including several members of the W3C Schemas WG. A lot of people whose affiliations were not known to me were there as well. Here's my summary of what happened (some of which came from the conference at large). No one advocated taking a year or more to develop a new standard. The world cannot and will not wait. A standardization effort that takes a year to produce a stable artifact is irrelevant to most developers working in this area. Everyone acknowledged the need for a standard in this area to avoid the fracturing of XML and to get application protocols to stop reinventing the wheel every time they need to send a compound type (e.g., a struct). No one had concrete criticisms of the SOAP serialization format. That does not mean it is perfect. However, it means that no one who has looked at the spec has any credible technical show-stoppers that they are willing to share in an open forum. Dave Winer felt XML-RPC was perfectly adequate (and on March 2 advocated the world use it instead of SOAP). No one at the BOF seemed to agree, at least no one verbalized their agreement. See Ken MacLeod's response to this at (http://www.monkeyfist.com/?id=293). Several people (although no one from DM or MS) advocated the W3C rubber-stamping SOAP. I believe I made a compelling case for STOPPING the "messaging vs. RPC" debate which has been beaten to death too many times here and elsewhere. No one at the BOF made the argument that SOAP was MS-centric or Windows-centric or COM-centric. Some people are concerned about the RPC-like nature of the way the SOAP spec reads. The plan is to change that both through modularization of the spec and a better choice of wording for certain SOAP concepts. Look at my XML.com article on SOAP for more on this viewpoint. Everyone seemed to agree that there is value in modularizing SOAP by decoupling it from HTTP (in essence, turning the HTTP mapping into an appendix or auxiliary spec). The authors knew this already and have been working towards that prior to this BOF. Several talks at the conference mentioned their use of SOAP. Most notably, David Orchard from IBM has a working SOAP implementation. I interpreted Jon Bosak's keynote the day before the BOF as making a great case for why an infrastructure protocol like SOAP (or its functional equivalent) should come from the W3C or IETF rather than OASIS and that ebXML should then adopt it (and XML Schemas) as the substrate for defining industry-specific XML interfaces. According to Bosak, W3C defines infrastructure/platform and OASIS (the home of ebXML) is where application standards get layered. This sounds very reasonable to me. Tim Bray (co-editor of the XML 1.0 rec) attended the BOF and was a great help and inspiration. He told of how XML 1.0 came in "low and fast under the radar" and became a standard before every standards wonk in the industry got a chance to rack up frequent flyer miles arguing over where the semi-colons go. There seemed to be consensus around the room that now is a time for another such effort. After the meeting on March 1, I am actually very optimistic about things. That stated, I ask for the support of the SOAP community at large to stop bickering for a week or two while I (and others) try to plot a strategy that gives everyone an open, platform/application/language-neutral solution we all can live with in a time frame we can all live with. In particular, unless you feel we need to wait several years or that we should bake in dependencies on Visual Basic, your silence (on this list at least) will be interpreted as a general agreement on the end-goal of a timely agreement on a universally implementable XML-based protocol framework. As many of you know, I've dedicated a lot of pro-bono time and energy since rejoining SOAP last summer. I have no plans on stopping now, however, I ask for a wee bit of support while I try to gently corral the entire industry behind a unifying XML protocol. Thanks, DB http://www.develop.com/dbox
Received on Friday, 3 March 2000 04:12:01 UTC