Minutes of RPC Task Force telcon, July 25, 2001

Attendance

Agenda topics (unordered)

  1. List of RPC issues
  2. Section 5/7 dependencies
  3. Proposal for resolution of issues 16 and 78
  4. Martin Gudgin's evaluation of Infoset description in sec. 7
  5. Feedback from dist-app on sec. 7.2
  6. Evaluate Chris Ferris's CorrelationID extension proposal

Discussion.

Re. #5

Frank- believes it is inappropriate to reference hdrs from the body because, for example, an RPC request is like a method signature which is self-contained, and resides in the body.

Various- believe that legislating a constraint that prevents hdr references from the body is artificial because, for example, the particular serialisation of an RPC request by a client is the business of the client and the receiving application.

Much other discussion on this topic. As a result, we decided that (a) the meaning of "additional information" (first 2 words in sec 7.2) is unclear and leads people to different conclusions (implied suggestion that this para should be reworded), and (b) the header/body distinction issue (issue #101) is the proper vehicle for the discussion of questions regarding sec 7.2

Action: Frank to send an email to dist-app framing the "7.2 concern" as a facet of the hdr/body distinction issue

Re. #6

Discussion centered around whether the proposal should be part of the 'core' spec, should it be required for conforming processors, etc. Identified a couple of levels of "conformance" for extensions in general, (i) features within an extension are identified as MUST, MAY, etc, (ii) the extension itself is identified as MUST, MAY, etc.

Action: David to put the "core" vs "other" issue on the WG's telcon agenda for its next telcon

Re. #3

Frank noted that 78 is the primary issue, after whose resolution, 16 should resolve quickly.

Frank has sent email describing a proposed resolution [1], and another email listing pointers to background sources [2].

The RPCTF plans to take up this resolution at its next telcon (monday July 30), and take the result forward to the next WG telcon (wednesday Aug 1)

Re. #2

A notable dependency between sections 5 & 7 is the reliance in sec 7 on struct which is defined in sec 5. The TF thinks this dependency should not exist because, for example, it does not permit an RPC encodings other than that described in sec 5. Two proposal for removing this dependency were discussed:

Action: Jacek to investigate what other issues would be affected by the rewrite proposed in (ii)

Action: Frank to ask Martin Gudgin's opinion on the rewrite proposed in (ii)

[1] http://lists.w3.org/Archives/Public/xml-dist-app/2001Jun/0110.html [2] http://lists.w3.org/Archives/Public/xml-dist-app/2001Jun/0174.html
Valid CSS! Valid HTML 4.0!
Revision: $Date: 2001/08/02 09:42:24 $