- From: Henrik Frystyk Nielsen <henrikn@microsoft.com>
- Date: Mon, 9 Dec 2002 14:28:31 -0800
- To: "Jean-Jacques Moreau" <moreau@crf.canon.fr>, "Martin Gudgin" <mgudgin@microsoft.com>, "Marc Hadley" <marc.hadley@sun.com>, "Noah Mendelson" <noah_mendelsohn@us.ibm.com>
- Cc: "W3C Public Archive" <www-archive@w3.org>, "Nilo Mitra" <EUSNILM@am1.ericsson.se>, "David Fallside" <fallside@us.ibm.com>
Marc and I went through the remaining of Gudge's issues for Part 1 and here's what we think we (spec editors) need to do: > 4 Section 2.6 > > Bullet 3 should be reworded to be clearer and use infoset terms. > "The Value of Code" seems ambiguous in the current langauge. First time in each part expand with what it really means. > 9. Section 2.7.4 > > It would be nice to add a note saying that these 18 exceptions > are based on a 'writer makes right' approach and that a canonicalization > algorithm ( a reader-makes-right ) approach would obviate the need for > them. Marc to send out a note to the WG asking for guidance and also regarding whether the sig stuff should go in appendix > 13. Section 5.1.1 > > For consistency the type information for the attribute should > come immediately after the infoset properties Just do it > 15. Section 5.1.1 > > It is implied but not explicitly stated that nested > encodingstyle override ancestor declarations. For clarity, I think there > should be some explicit statement. Do nothing > 16. Section 5.2.1 > > The relay AII is missing from the list of AIIs allowed on a > header block. Just do it > 17. Section 5.2.2, 5.2.3, 5.2.4 > > We use different language in 5.2.2 concerning intermediaries > dropping these attributes than we do in 5.2.3 and 5.2.4. We should be > consistent Just do it > 18. Section 5.2.2, 5.2.3, 5.2.4 > > Concerning what changes intermediaries make to these attributes > I think it would be good to reference 2.7.4 Just do it > 19. Section 5.4.2 > > The constraint on the xml:lang attribute values should be in > section 5.4.2.1 Nope, the constraint is on the parent. Do nothing. > 20. Section 5.4.2.1 > > For consistency the type information for the EII should come > immediately after the infoset properties Just do it > 21. Section 5.4.2.1 > > We need to add that the [prefix] property of the xml:lang > attribute MUST be 'xml' Just do it > 22. Section 5.4.3 > > For consistency the type information for the EII should come > immediately after the infoset properties Just do it > 23. Section 5.4.5.1 > > The language concerning allowable attribute should use the same > style as we use for header and body blocks Just do it Henrik Frystyk Nielsen mailto:henrikn@microsoft.com >-----Original Message----- >From: Jean-Jacques Moreau [mailto:moreau@crf.canon.fr] >Sent: Thursday, December 05, 2002 05:16 >To: Martin Gudgin >Cc: W3C Public Archive; Marc Hadley; Nilo Mitra; Noah >Mendelson; Henrik Frystyk Nielsen > >I'll be fixing some of these since I introduced them in the >first place (*zero*). Plus a few more. Details below. > >Martin Gudgin wrote: >> 1. Section 1.5.1 >> >> The definition of SOAP Module has "*zero*" in the paragraph. >> What is the significance of the asterisks? > >Done. The asterisks in the original email were for empahising >differences with the current text. > >> 2. Section 2 >> >> The third paragraph begins with the word 'It'. What is 'It'? I >> *think* it's 'The SOAP Processing Model' but from context it >could be >> interpreted as 'This section' > >Not in the copy I looked at. > >> 3. Section 2.2 >> >> In the third paragraph after table 2 "role names defined above" >> should read "role names define in Table 2" where Table 2 is >a specref. > >Done. I also had to remove all remaining references to the >former Robustness section, since the spec no longer validated. > >> 4 Section 2.6 >> >> Bullet 3 should be reworded to be clearer and use infoset terms. >> "The Value of Code" seems ambiguous in the current langauge. >> >> 5. Section 2.7.2 >> >> The first paragraph after the numbered list ends with the word >> 'above'. 'above' what? I would suggest we change 'above' to 'by the >> intermediary' > >Done. > >> 6. Section 2.7.4 >> >> In Bullet 1, "can be removed from" should read "can be >removed, by >> that intermediary, from" > >Done. > >> 7. Section 2.7.4 >> >> In Bullet 1 "[ref to header block processing rules]" >should be a >> specref > >Done. > >> 8. Section 2.7.4 >> >> In Bullet 2 "[ref to header block processing rules]" >should be a >> specref > >Done. > >> 9. Section 2.7.4 >> >> It would be nice to add a note saying that these 18 >exceptions are >> based on a 'writer makes right' approach and that a canonicalization >> algorithm ( a reader-makes-right ) approach would obviate >the need for >> them. >> >> 10. Section 3.2 >> >> The two numbered bulleted items seem to be duplicate >information from >> the earlir (unnumbered) bulleted list in this section. > >I can't find this in 3.2 MEPs. > >> 11. Section 3.3 >> >> The first paragraph has "*zero*" in it. What is the >significance of >> the asterisks? > >Done. > >> 12. Section 5 >> >> There is text missing concerning the fact that if we >don't specify a >> value for a given infoset property then any legal value is >acceptable. >> >> 13. Section 5.1.1 >> >> For consistency the type information for the attribute >should come >> immediately after the infoset properties >> >> 14. Section 5.1.1 >> >> I think it would be better to reword the MUST NOT / MAY >for where >> encodingstyle can appear as follows: >> >> The encodingStyle AII MAY appear on the following: >> >> <list as currently in spec> >> >> The encodingstlye AII MUST NOT appear on any other >elements in a SOAP >> Message Infoset > >Done. > >> 15. Section 5.1.1 >> >> It is implied but not explicitly stated that nested >encodingstyle >> override ancestor declarations. For clarity, I think there should be >> some explicit statement. >> >> 16. Section 5.2.1 >> >> The relay AII is missing from the list of AIIs allowed >on a header >> block. >> >> 17. Section 5.2.2, 5.2.3, 5.2.4 >> >> We use different language in 5.2.2 concerning >intermediaries dropping >> these attributes than we do in 5.2.3 and 5.2.4. We should be >> consistent >> >> 18. Section 5.2.2, 5.2.3, 5.2.4 >> >> Concerning what changes intermediaries make to these >attributes I >> think it would be good to reference 2.7.4 >> >> 19. Section 5.4.2 >> >> The constraint on the xml:lang attribute values should >be in section >> 5.4.2.1 >> >> 20. Section 5.4.2.1 >> >> For consistency the type information for the EII should come >> immediately after the infoset properties >> >> 21. Section 5.4.2.1 >> >> We need to add that the [prefix] property of the >xml:lang attribute >> MUST be 'xml' >> >> 22. Section 5.4.3 >> >> For consistency the type information for the EII should come >> immediately after the infoset properties >> >> 23. Section 5.4.5.1 >> >> The language concerning allowable attribute should use >the same style >> as we use for header and body blocks >> >> 24. Section 5.4.7 >> >> SupportedEnvelope should be a sub-section. 'described >below' should >> be a reference to that sub-section > >Done. Also promoted the previous section. Also removed the >diff markup. Also replaced "as described below:" by <specref...>. > >> 25. Section 5.4.7 >> >> qname AII should be a sub-section. > >Done. Copied the text for that section from 5.4.8, which also >has a QName AII. > >> 25. Section 5.4.7 >> >> The note at the end of section 5.4.8 also applies here. > >Done as part of above. > >I also created subsections for 5.4.8, so the two sections are >now symmetric, structure-wise. > >> 26. Section 8. >> >> 'Johnathan Marsh' should be 'Jonathan Marsh' > >Done. > >
Received on Monday, 9 December 2002 17:29:15 UTC