Summary of 27 May 2002 TAG teleconference

Note: There was no IRC log of this meeting.

Previous meeting: 20 May 2002 [Minutes approved at this meeting]
Next meeting: 3 June 2002. See TAG home for more meeting information. Regrets from SW for next meeting.

Participants: Tim Bray (TB), Paul Cotton (PC), Chris Lilley (CL), David Orchard (DO), Stuart Williams (SW), Ian Jacobs (IJ, Scribe)

Regrets: Dan Connolly, Norm Walsh

Absent: Tim Berners-Lee, Roy Fielding


See initial agenda:

  1. Review of action items
  2. Accept issue raised by Rob Lanphier?
  3. Progress on charmodReview-17
  4. Review of findings in progress
  5. Progress on architecture document
  6. Progress on URIEquivalence-15

Summary of resolutions

  1. Do not accept Rob Lanphier's requested issue at this time. See discussion..

Action items


  1. TBL: Take uriMediaType-9 finding to IETF and IANA. TBL to contact Eastlake and Masinter, copy www-tag. Done.
  2. CL, NW: Review Character Model for the Web (last call). CL Done. NW done to TAG (expected to be made public).
  3. Revise and publish Media Types finding


  1. IJ: Integrate/combine one-page summaries into arch doc. [Partly done in arch doc draft] Summaries from chapters 1 and 2 started to be integrated, but not 3 or 4.
  2. TBL: Negotiate more of IJ time for arch doc
  3. CL: Add concern regarding non-western characters to the POST scenario (issue whenToUseGet-7). Progress has been made through discussions with Martin Duerst.
  4. IJ: revise and publish whenToUseGet-7 finding. Some revisions made. IJ to take into account new comments from TAG.
  5. NW: Draft a finding for formattingProperties-19; find out source of issue from CSS WG.

On hold:

  1. DO/DC/CL: Polish up DO's .1-level draft and find out what's going on with XForms. Partly done; some discussions with XForms reps at AC meeting. DO reports that the XMLP WG is working on a new proposal to address GET bindings in SOAP. The TAG would like the XMLP WG to share the proposal with the TAG as early as is convenient for the WG.

Note: The TAG discussed the impact of www-tag discussions on other Working Groups. The TAG reaffirmed its intention to carry out its technical discussions on www-tag. To avoid confusion between the personal views of TAG participants and formal TAG positions, the TAG feels it is sufficient that emails begin with text such as "in my personal opinion", or similar.

Accept issue raised by Rob Lanphier?

See issue raised by Rob Lanphier asking "What is appropriate error resilience/recovery/"second guessing" in web software?".

<Ian> TB: I think there is no new issue here? Seems too general. Not sure what the recommendation would be. There are issues in RL's email, but as framed, I'm not sure what you could make a finding about.

<Ian> IJ: What about asking QA to answer this in its guidelines?

<Ian> TB: I think there are issues here that are not just QA issues, but as written too much conflated.

<Ian> Action TB: Respond to RL on the list asking for more detail.

<Ian> Resolved: Do not accept this as a TAG issue (but do not decline).

Progress on charmodReview-17

See issue charmodeReview-17. Please note Charmod last call review form.

<Ian> TB: I think we should apologize to the I18N WG and say we think that the TAG will have some material comments to make, but request a 1-week extension.

<Ian> CL: I suggest that people review NW and CL comments for 3 June. Approve some version on 3 June and send to I18N WG. I think we can commit to them having full comments by 4 June.

<Ian> TB straw poll: the basic comment on charmod spec is good reason to split document into three - things well-established, things that might require a CR period, things not yet well-established. Some great explanatory discussion and advice. Should be published. Shouldn't be held up while we fight out other parts of the spec.

<Ian> PC: I agree with that. My WG is likely to lament the early normalization bits.

<Ian> Summarizing:

  1. Action CL: Send Charmod spec comments to www-tag (done).
  2. Action IJ: Ask NW to make his I18N comments public.
  3. Action CL: Inform I18N WG that we think we will comment that the doc should be improved by splitting into different pieces (those well-known, those requiring CR period, and those much less developed). Apologize to I18N for delay, and ask to extend deadline until 4 June. We will review comments from CL and NW at the 3 June teleconf.

<Ian> TB: Is there a study done on the actual cost/size/performance of code required to do early normalization? From an engineering perspective, how big of a deal is this?

<Ian> PC: Charmod doc talks of an algorithm for doing this. In the Query WG, the algorithm was compelling. Most people agreed that this would make normalization an order of mag. faster than they had imagined. I'm not sure the charmod spec says anything about the footprint.

<Ian> CL: Yes, charmod spec does speak to this.

<Ian> TB: I think people will agree that normalization is a good thing. The question is how much you pay for this benefit.

<Ian> PC: Lots of people worried about backwards compatibility cost.

Review of findings in progress

The TAG did not approve any draft findings. The TAG will have as homework for the next meeting to review the following draft findings:

  1. TAG Finding: Internet Media Type registration, consistency of use

    Tentative approval from CL, TB, SW. See email from Tantek and comments from Arjun Ray.

  2. TAG Finding: Using QNames as Identifiers
  3. TAG Finding: When to use GET to make Resources Addressable

    Action IJ: Incorporate comments from SW (TAG only).

Progress on architecture document

<Ian> IJ: I met with RF and TBL last week. Made good progress on a stripped-down version of chapter one. I hope to publish a new draft with this more terse form soon.

IJ shows the TAG some notes. There was general agreement from those present that this terse format is the right direction.

Progress on URIEquivalence-15

See issue URIEquivalence-15.

<Ian> TB: Good stuff on www-tag on this one, in particular email from Martin.

<Ian> Homework for next week: Read Martin's email. (High on agenda for next week)

<Ian> PC: can URIEquivalence-15 and charmodReview-17 be separated? We should be prepared to discuss their connection/relation next week.

<Ian> SW: I agree. Can we say that the succesful production of a character model will go a long way in resolving URIEquivalence-15?

<Ian> TB: I don't think so. But charmod will help us discuss this more clearly.

<Ian> PC: Should we invite an I18N person to the teleconf?

<Ian> CL, TB: Yes.

<Ian> IJ: Martin may not be able to make this teleconf in Japan. Misha might be able to.

<Ian> TB: Let's try to resolve this ourselves first. I will commit to reading Martin's email and the IRI draft. I will try and summarize the key things for the TAG. TB: Not sure how to answer: Do IRIs have any traction anywhere (e.g., IETF, among developers)?

<Ian> CL: Depends on which part. IE, there's an option "Put into UTF-8, then hexify"; that's part of IRI document.

<Ian> TB: Yes, we'd like to see - "Put in UTF-8, then hexify"