W3C Tag Teleconference of 27 September 2007

27 Sep 2007



Tim Berners-Lee, T.V. Raman, Stuart Williams, Norm Walsh, Dan Connolly, David Orchard, Noah Mendelsohn
Henry Thompson, Rhys Lewis
Stuart Williams
Noah Mendelsohn


Review of F2F Minutes

SW: F2F minutes will be reviewed on next week's call.

Next teleconference

SW: Propose next call for 4 Oct 2007. Will solicit Dave as scribe.

NW: Regrets for 25th of Oct and 8th Nov.

<DanC> (TVR gave regrets for "October 4, October 18, and October 25.")

NM: Probable regrets for 4 and 11 Oct.

Planning for the Technical Plenary

SW: Proposed two meetings with other groups. 1) Web Security Context to make progress on pwds in the clear and 2) Health Care and Life Sciences for issues related to httpRange-14.

DC: I feel I should stay in touch with security issues.

<DanC> I'd like to be involved in the security discussion, and on Friday I plan to chair an HTML WG meeting.

NW: Will come if I can, but no need to schedule around me.

SW: I think they're meeting Thurs-Sat.

TBL: I'd like to be involved in the Health Care one. I'll be around Sat.

NW: I leave Friday.

SW: Will try and focus on Thurs/Fri.

TVR: I leave Thurs evening.

SW: Looks like Thurs afternoon is possible. I think I leave around 4PM.

NM: I can be around on Sat. if necessary.

SW: Are there any other liaisons I should be looking into?


SW: I will suggest we hook up with MEZ's group (security) either Monday afternoon or Tues.

NW: I'm not sure whether we should meet with XProc working group or not. Part of the charter was influenced by TAG input, but we haven't started on that. Still, there's a connection to XML Functions.

NM: This also relates to the self-describing Web issue. I'm neutral as to whether we need to meet, but very interested in participating if we do.

NW: XProc is meeting Thurs&Friday.

SW: Tag meeting is half day 9-12 on Monday and half day 9-12 Friday.

<Stuart> http://www.w3.org/2007/11/TPAC/

DC: AC meeting is Tuesday evening, and Thurs morning.

NM: Wed?

<timbl_> DRAFT: member-only http://www.w3.org/2007/11/06-AC-agenda.html

<DanC> yes, AC meeting is Tuesday evening and Thursday morning, per http://www.w3.org/2007/11/TPAC/overview.html 2007/09/26 18:06:11

??: Wed. is plenary.

NM: Ah yes, but AC is encouraged to attend if I remember correctly.

<Norm> Sorry, I need to step away for a minute

SW: Back to possible XProc meeting. I have the impression that Tim might like to. Will pursue with their chair (Norm)

<scribe> ACTION: Stuart to talk to XProc chair regarding possible joint meeting at plenary

<trackbot-ng> Created ACTION-59 - Talk to XProc chair regarding possible joint meeting at plenary [on Stuart Williams - due 2007-10-04].

<Norm> I don't feel strongly about it; I just wanted to offer with my chair's hat on

Issue binaryXML-30

<DanC> http://lists.w3.org/Archives/Public/www-tag/2007Jul/0197

<DanC> http://www.w3.org/TR/2007/WD-exi-measurements-20070725/

DC: We had commented on binary XML some time ago, but put our efforts "to sleep" pending their publishing. Well, they've published.

<DanC> http://www.w3.org/TR/2007/WD-exi-20070716/

DC: Measurements seems most relevant to me, because it helps me understand what success is. I think I've heard that Core may be commenting that the extensibility stuff is a concern. That's one of the main issues for me.

<DanC> ... my advice to the TAG is to watch the XML Core review of EXI, and not necessarily to study the EXI specs directly at this time.

DC: I don't think it's urgent for the TAG to comment now.

NM: There was some interesting public comment, perhaps it was on xml-dev, after the specs came out. I think that covered many of the issues we'd raise. So, I agree with DC. No urgency for TAG to comment officially at this point.

<Noah_> (Dan sings us a song waiting to see if Norm comes back.)

<Noah_> (No joy.)

<Zakim_> +Norm

<Noah_> (Joy)

DC: Norm, am I right that the core WG is doing a fairly thorough review of the XML Binary work, and that being the case, the TAG might do well to hang back for now?

NW: I think that's fine.

DC: Any guess on how long we sleep?

NW: I'm guessing at least through beginning of Nov. I think the core WG may have asked John Cowan to pull together a review, and I'm not sure if it's finished.

<Stuart> http://www.w3.org/2007/11/07-TechPlenAgenda.html

SW: Did this make the cut for plenary discussion?

DC: Don't think so.

NM: Too bad, I think it would have been a great topic.

DC: We could push to put it on the Wed. agenda (but I'm conflicted).

<Stuart> zae

TBL: This does seems like a good tech plenary topic, and this seems a good time. There has been a lot of discussion at other venues. Steve's in a meeting just now, so I haven't been able to find out if the agenda is actually malleable.

<scribe> ACTION: Tim to suggest to Steve that Binary XML be put on plenary agenda

<trackbot-ng> Created ACTION-60 - Suggest to Steve that Binary XML be put on plenary agenda [on Tim Berners-Lee - due 2007-10-04].

<DanC> [minor correction to the agenda: my comment on the choice of header name was _not_ on behalf of the TAG]

Responses to TAG review of "Enabling Read Access for Web Resources

SW: I'm not quite sure whether the work that was done was sufficiently responsive to our concerns.

DC: We said "do something on security", and they did something. Not enough?

SW: (scribe missed something about sandboxes)

DC: Well, they added something. Not sure what else we'd want them to say.

<DanC> "For instance, this prevents a malicious site from reading information from the user's intranet using a technology such as XMLHttpRequest." is responsive to our request. I suppose one could say more, but I can't think of anything critical.

TBL: I think the HTTP header could be shorter.

DC: I would think that nothing would work, I.e. no header.

TBL: But the whole point is permission.

<timbl_> Maybe should be "acces-permission not 'control'

<timbl_> it only *allows*

TBL: If you have any web site with any public data, you have to put this.
... Eating our own dogfood suggests w3c put this on all our public data?

DC: Needed on all pages?

TBL: Yes, if it's to be accessible by untrusted scripts.
... I suppose we could do it only on RDF, etc.

DC: Reminding myself. Today, if I have a script on fedex.com with reference to data on buy.com, brower will go "no way".

NM: Yes for data, no for script. I believe browsers will retrieve scripts cross domain. Indeed, I think that was one of the (questionable IMO) rationale's for JSON.

DC: So, returning to our story, with this stuff, the browser will do an HTTP connection and check the header?

TBL: They will do a HEAD.
... The worry is getting something from behind a firewall.

NM: Really only firewalls?

TBL: Also cookies.

NM: Ah, so the story is that my browser has some kind of access control credential, perhaps a cookie, and a malicious site uses my browser to pull data using that credential.

TBL: I think it would be good if the first example was the public stuff. Anything that's intended to be public will have to have this on it.

SW: We also raised the POWDER issue. I don't think they've responded formally, but there are signs things are headed in a good direction.

<Stuart> http://lists.w3.org/Archives/Public/public-appformats/2007Sep/0048

DC: Clarify?

SW: The signs are good that WAF and POWDER will align syntaxes. I haven't been closely involved, just noting that they seem to be talking.

<Stuart> http://dev.w3.org/2006/waf/access-control/Overview.html

TBL: We're looking at 18 June 2007 draft, which still has content access control.

SW: There is also an editors' copy.

<DanC> the editor has since agreed to shorten "content-access-control" to "access-control"

TBL: OK, now it's become "access control".

SW: I don't know whether this is close to Last Call or not. Presumably there will be more revisions. Anyone feel we should provide feedback other than "thank you for the changes already made".

<timbl_> Access-Control:

<timbl_> Access-Control: *

TBL: I want to find out what access control header with nothing on it does.

DC: First example uses "*"

TBL: Do I then not have to say method=GET, etc.

<DanC> Access-Control: allow * <- seems to be "share this with all comers"

<DanC> rule-type ::= "allow" | "deny"

<?access-control allow="*"?>

TBL: I don't think I know of cases where things aren't totally public. Maybe there is need to grant access for particular sites.

DC: I thought that was the main case. Some particular airline company and Fedex. I don't remember the exact details.

TBL: The shoe company wants to quote shipping, so needs to pull the Fedex shipping algorithm.

SW: Do we want to do another review of the document as it is?

TBL: We probably have more urgent things to do.
... Shall we get the system team to do it?

DC: How do we test it. Is there code at the other end?

TBL: Tabulator browser extension?
... Right now, if we had a safe mode, not much would work, but if we could get DBPedia to do this, that would be more interesting.

<raman> I have to leave in about 10 minutes, i.e. half an hour before end of call. I'll quitely drop off, rather than disrupt the conversation flow

TVR: There was some rumbling about Firefox starting on this, but there's hearsay to the effect that the work was done without a careful reading of the spec. I'm getting conflicting information on the status of Firefox efforts.

<DanC> (hmm... is this backwards? "For instance, allow read access from example.org and its subdomains with the exception of public.example.org." )

TBL: Having the systems team do this is a good test of whether people will go through the pain.

NM: Well, the real test will be when people start building software that let you get more value out of the pages that are so labeled. In the short run, that's not there.

SW: Shall we just move to next agenda item?

DC: Stuart, are you content with the current response we've gotten?

SW: Will check again. It seemed a bit "light".
... The concern is that it didn't go into enough depth about the reasons for the current sandbox rules, and what the dangers would be of changing the rules.
... I think I should take another pass at reading it.

Issue XMLVersioning-41

SW: Reviewing action items
... ACTION-16 on David Orchard to Incorporate the NVDL text into the findings. - due 2007-09-27, open

DO: Not done. Propose new new date of 31 Oct

SW: ACTION-28 on Noah Mendelsohn to draft a blog item for review and, pending creation of a TAG blog mechanism, post it. - due 2007-09-06, pending review

NM: On hold until blog ready.

SW: ACTION-38 on Norman Walsh to review the XML part again - due 2007-09-17, open

NW: Making progress. Schedule for next week (4 Oct.)

SW: ACTION-48 on Henry S. Thompson to Produce an exemplary implementation of XHTML Modularization using substitution groups for both bottom-up extensibility and top-down modularity - due 2007-09-24, open

DC: I haven't seen progress on that.

SW: ACTION-51 on David Orchard to And Dan work together to articulate the story that the TAG wants to tell. - due 2007-09-25, open

DC: Dave did some stuff, I owe him an answer. Let's say 4 October for next week's call.

SW: Dave, you sent a note with new drafts.

DO: There was a rename of strategies. What I attempted to do was take the strategies document and focus most narrowly on forwards compatibility, which seemed the most important. I got rid of text on things like "questions language designers should think about". I thought that was useful, but in the interest of focussing more narrowly, I cut it.
... I was trying to focus on "if there's something small we are going to approve, what would it be?"
... Focussed on XXX?, must accept unknowns, planning for version ids. Somewhat opposite of what Noah asked for, which was to agree on a table of contents for the final result. I wanted at least some concrete progress in some areas.

SW: Any more on versioning for now? I think most of us are waiting for Dave and Dan to do more work.

Other business

SW: Any progress on the blog? Have we had a response on running Movable Type in /blog space?

DC: No.

<DanC> Cc: Dan Connolly <connolly@w3.org>, tag@w3.org

<DanC> Subject: Re: TAG blog technical possibility

<DanC> Date: Tue, 18 Sep 2007 14:41:45 +0100 (08:41 CDT)

SW: Anything else to discuss?

DC: How about review of Cool URIs.

Cool URIs Document

DC: Did the editor say "thanks for your comments and we will address them", or "we have already addressed them"?

SW: Which comments of ours?

DC: From the F2F

<DanC> msg from Leo http://lists.w3.org/Archives/Public/www-tag/2007Sep/0109.html

SW: Some of the responses from Pat seemed to have the tone that we the TAG had written the document.

DC: Well, we suggested text and he commented on that.

SW: Yes, but we wrote in the tone of the original.

TBL: I thought the discussion of identity and resource is ground we'd covered before, and that we'd made more progress than current discussion would suggest.

<timbl_> I actualy think 'resource' is a terrible word for daml:Thing

<timbl_> I think 'resource' is fine in 'Infromation resource'.

<DanC> yes, and we pay repeatedly for it...

<DanC> ... though it's not easy to see what the cost of pursuing alternatives would be.

<timbl_> The use of 'resource' coems from RDF, which was made by folks who were thinking of using them for just information resources.....

NM: Is this any more than a disagreement over whether people building computer systems can use words like "resource" in a stylized way.

SW: The word resource are deeply embedded in Web culture.

<DanC> (this is _all about_ the tension in the culture around the R in URL. )

TBL: RDF was originally about online resources which were information resources. The word resource was used, because we're talking about things that are of value.
... Resource is in fact a crummy word. Just not worth going back over all the specs to say "thing".

<Noah_> (UTI = Universal Thing Identifier ?)

<timbl_> (UI)

TBL: When we expanded RDF to talk about a broader range of things, we neglected to change the word.

<timbl_> Universal Identifier

SW: When did you start thinking of giving people URIs?

DC: Think I saw it first on a slide in 1990.

<DanC> (sorry, it was 1994 in Geneva)

TBL: That was pre-RDF.

<DanC> ( http://www.w3.org/Talks/WWW94Tim/ )

NM: Aren't you and Pat both agreeing that the word resource isn't right, but it's too late to change?

TBL: No, I'm actually toying with proposing that we change the word resource to something else.

<DanC> (there's also the possibility that we leave /TR/webarch as is, but introduce a lay term in this "Cool URIs..." tutorial)

SW: Are you talking about the Cool URIs document or Web Arch?

TBL: They need to be consistent.

NM: We're sticking with the decision that 200 is only appropriate for information "resources"?

TBL: Yes. It's just the terminology that would change, not the technical mechanisms.

<DanC> http://lists.w3.org/Archives/Public/www-tag/2007Sep/0109.html

<DanC> "A question of understanding:" -- Leo

<DanC> and "A question of practice: ..." ibid

<dorchard> Separate topic for the chair to deal with when he wants, I'd like to find out of the TAG will be meeting in China around the AC and WWW2008 conference in April/May

<Noah_> I will almost surely not be able to make China.

ftf planning, May 2008

<Noah_> I will almost surely not be able to make China.

DO: I'm concerned about WWW in China in May, and then TAG in Bristol or Venice same month. That's a lot of travel.

TBL: I currently plan to go to China.

SW: Don't know.

<DanC> "20-22 April 2008, Beijing, China [under consideration*]" -- http://www.w3.org/Member/Meeting/

TBL: Starts with intro to W3C on 20 April. AC is April 21 & 22. WWW conference is 23-25.

<Stuart> http://www.www2008.org/

<DanC> apr 21-25 http://www2008.org/

NW: I can try for expense approval.

<DanC> 19-20 Apr in Beijing <- an idea

<DanC> "Refereed paper submissions are due by 11:59pm (Hawaii time) on November 1, 2007."

<Stuart> http://www.www2008.org/CFP/index.html

NM: Need to be back for 24th-25th. That's a long way to go for a short meeting like this. The only possibility for me is to go the week before, but honestly I'd prefer not. I'm not sure I can get funding, and that's a very long way to go, unless there's real synergy with being in China. Maybe or maybe not I could add some IBM business.

DO: I'm suggesting that if we are doing Europe 1st half, it might make sense to swap and put Vancouver in the spring and Bristol in the Winter.
... Do you really want to go to China in April, then Bristol in May?

TBL: Not a problem for me.

DC: Let's not try to decide today.

NM: Maybe the message is, don't put it too early in May. 19 May is 4 weeks after China, which seems borderline OK.

<DanC> (how many hosts do we have in the air? DO in Vancouver, Stuart in Bristol, and... ? [I think the Venice option is unlikely])

TBL: Snag with later May is kids activities, etc.

NM: Is that a problem on, say, the 19th?

<scribe> ACTION: Stuart to follow up with Amy on Tim's schedule for 2008

<trackbot-ng> Created ACTION-61 - Follow up with Amy on Tim's schedule for 2008 [on Stuart Williams - due 2007-10-04].

Summary of Action Items

[NEW] ACTION: Stuart to follow up with Amy on Tim's schedule for 2008
[NEW] ACTION: Stuart to talk to XProc chair regarding possible joint meeting at plenary
[NEW] ACTION: Tim to suggest to Steve that Binary XML be put on plenary agenda
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2007/10/02 19:19:37 $