W3C

- DRAFT -

TAG Weekly

27 Mar 2008

Agenda

See also: IRC log

Attendees

Present
Stuart_Williams, Tim_Berners-Lee, Ashok_Malhotra, Henry_Thompson, Noah_Mendelsohn, Norman_Walsh, Jonathan_Rees, Dave_Orchard, T.V._Raman
Regrets
Dan_Connolly
Chair
Stuart Williams
Scribe
Ashok Malhotra

Contents


 

 

<Stuart> Scribe: Ashok Malhotra

<Stuart> scribenick: Ashok

Convene

Noah: AOB item -- could you send travel directions for f2f?

Stuart: Any objections to approving minutes from last week

No objections.

RESOLUTION: Minutes from 20th March 2008 are approved

Next meeting April 3 -- regrets from HT

Regrets from Tim and Stuart for 4/24

Norm will scribe next week

SW: We now have all the minutes from the f2f

DO: Problem about minutes being in two places

SW: I am willing to accept the places where the minutes have landed
... Propose we accept minutes subject to Jonathan's changes being included

<Stuart> Jonathan's corrections are at: http://lists.w3.org/Archives/Member/tag/2008Mar/0139

HT: No objection

RESOLUTION: f2f minutes approved subject to Jonathan's changes being incorporated - i.e. http://lists.w3.org/Archives/Member/tag/2008Mar/0139

<scribe> ACTION: Henry S, Norm, Dave -- revise minutes from f2f with Jonathan's changes [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action01]

<trackbot-ng> Created ACTION-127 - Henry, Norm, Dave -- revise minutes from f2f with Jonathan's changes [on Henry S. Thompson - due 2008-04-03].

abbreviatedURI-56 (ISSUE-56)

HT: They have been waiting for this

Draft from HT ... subsequently revised

<ht> http://lists.w3.org/Archives/Member/tag/2008Mar/0093.html

Original wording

<noah> Second para:

<noah> In this connection we find the prose about CURIEs the current RDFA

<noah> spec. [2] troubling.

<noah> Is there a typo in that?

<noah> Should that be >in the current RDFa spec<?

<ht> yes

<noah> tnx

<timbl> " If you really think you must allow for not only the use of CURIEs,

<timbl> but of the union of [-form CURIEs with URIs, in new specs etc.,

<timbl> then you should at least note that this involves a hostage to

<timbl> fortune with respect to future versions of the URI spec, which

<timbl> might conceivably introduce URIs with initial square brackets.

<timbl> "

Tim: URI's will not start with [ ... no need to say this

Noah: I agree to removing it

HT: Happy to remove

<timbl> "We would _much_ prefer a proposal which made clear that it was only

<timbl> addressing the need for an abbreviated URI format in non-XML

<timbl> languages, or new XML languages, or new contexts within old XML

<timbl> languages, where _only_ such abbreviated forms are allowed."

Tim: I'm worried about pushing them down the path of only non-XML or new XML languages

<Norm> Practically, this means *requiring* the square brackets, right?

<timbl> no IRI consensus ... ah

<Zakim> ht, you wanted to query the IRI parallel

TVR: Agrees

HT: Then we need to go to IETF and say "revise URI spec"
... But I don't want to do it ... URI's are just fine. Case has not been made to change the way HTML is written
... There is a case for folks who are authoring N3 by hand ...

<Zakim> Norm, you wanted to observe that CURIEs don't rely on XML Namespaces either

NW: It's a new kind of abbreviation for URIs

<Zakim> noah, you wanted to say IRIs and CURIE's are different because of prefix bindings

<ht> HTML _has_ a 'prefix' mechanism, which is the base URI mechanism -- works in HTML and XHTML

<Zakim> timbl, you wanted to point out there is a serious requirement for RDFa .. and this turns out to have implications on HTML atributes

<ht> I look forward to hearing from Timbl what the RDFa requirement is

Tim: The reinvention of namespaces is a separate bug

<timbl> TimBL: this is one of the many cases in which we have a tradeoff between the eventual cleanliness of the system abnd the compatibility with history. For the sake of posterity, we should have one type of attribute, one routine to parse it which will in future accept [Curie] . For history, we will ban the use of these in any attributesd defined today. neither situation is ideal.

HT: The other impt thing you said, people are going to write RDFa in the same way as they write N3 now ...

<timbl> TimBL: if people are going to write RDFa like they write RDF/XML or N3 today, then they will use many namespaces (10-20) and use each many times, so they ...

<Zakim> ht, you wanted to agree with the way TimBL started

HT: by hand using URIs from multiple domains which require multiple prefixes

<timbl> really need namespaces, or the document becomes large and unreadable.

TimBl: Not only hand authoring -- also affects readability and length of file

<dorchard> Maybe use EXI would solve the length problem for machines :-)

HT: Remembering when to type [ ] and when not will be a huge pain

<timbl> Having to remember when you can use [] and when you can't , in other words when a given attribute was historically introduced, is a real pain.

<dorchard> I'm a bit worried about which optimizations this is really for.

<timbl> It is very tempting to design other solutions in this call.

<timbl> When was the last time that people looked at mount point type use of virtual schemes as abbreviations for URIs, in other words declare that the prefix and the scheme name are in the same space?

<Zakim> noah, you wanted to say, this makes me wonder how hard to push back

HT: We agree RDF languages need this ... it shd be allowed to proceed

<timbl> @prefix doi: <http://h.doi.org/lookup/>.

Noah: I don't see a way to do this for the scope they are going for
... If it was only for XML language I would have some trepidation ... N3 is not an XML languages
... Can we say "use it only where there is a need ... not a universal solution"

<Stuart> http://lists.w3.org/Archives/Member/tag/2008Mar/0069

SW: Can we agree on everything except the 3rd item?

Tim: Instead of removing, can we point out the problem in both directions

SW: We need some text we can say yea or nay to

Noah: When you try to generalize it you get into real trouble
... If you apply generally you will get more breakage

<timbl> "One can imagine an alternative proposal which made clear that it was only

<timbl> addressing the need for an abbreviated URI format in non-XML

<timbl> languages, or new XML languages, or new contexts within old XML

<timbl> languages, where _only_ such abbreviated forms are allowed. That

<timbl> is, a position taken _against_ any

<timbl> possibility the CURIEs might be used where URIs are called for in

<timbl> XML languages today. It would though have to acknowledge the possible negative

<timbl> consequences of success in going down this path, namely that

<timbl> ordinary users will not understand that the [-CURIE is not a

<timbl> universal alternative to URIs, and will start using them in"

<ht> How about this: "The TAG has not converged in its discussion about the pros and cons of the [-CURIE and its proposed use. We will working on input for you on that issue -- in the meantime please do not assume that silence gives consent with respect to that aspect of the draft."

SW: Raman, could you support the msg except item 3

<Stuart> http://lists.w3.org/Archives/Member/tag/2008Mar/0069

<timbl> How about " acknowledge the possible negative consequences of success in going down this path, namely that ordinary users will not understand that the [-CURIE is not a universal alternative to URIs, and will start using them in existing languages where URIs are expected, causing tools to break and users to be frustrated."

<Norm> Say nothing?

SW: The question is -- can we agree on the rest of the draft except item 3?

TVR: I wd object to item 1 because it seems to say we have not had a conversation with them.

Discussion

TVR: I can live with item 1
... I can live with item 2
... I think item 4 is a good thing to say

<Norm> I'm sensitive to Raman's comment on point 1, I'd be in favor of making it clear that we aren't asking for *our* clarification but for clarification *in the spec*.

HT: I believe the remaining items are editorial. But I will check.

<Zakim> timbl, you wanted to propose a replacement for 3 in http://www.w3.org/2008/04/curie.html

<jar> +1 re placeholder for section 3, everything else can be sent

Tim reads out above

TVR: I can live with that.

Raman leaves the call.

<Zakim> Norm, you wanted to say I think the TAG should make points 1, 2, and 4

SW: Any more discussion of Tim's revision?

<jar> +1 re Tim's version

SW: Proposes "TAG have reached consensus to adopt HT and JAR's original proposal with item 3 changed as per Tim's revision"

<timbl> Tim's revision -- http://www.w3.org/2008/04/curie.html?

<jar> "[-CURIE" -> "safe curie"

No objection

<Askok> No abstentions

RESOLUTION: TAG have reached consensus to adopt HT and JAR's original proposal, http://lists.w3.org/Archives/Member/tag/2008Mar/0069, with item 3 replaced with as per Tim's revison, http://www.w3.org/2008/04/curie.html, as TAG feedback to the XHTML2 WG on http://www.w3.org/TR/2007/WD-curie-20071126/.

<scribe> ACTION: Stuart to send final text HTML WG [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action02]

<trackbot-ng> Created ACTION-128 - Send final text HTML WG [on Stuart Williams - due 2008-04-03].

RDFinXHTML-35 (ISSUE-35)

Norm: I had some high-level questions and I wanted to ask if TAG wanted to discuss them

<timbl> I feel that there should be a policy of trying to guide the (x)HTML family of languages so that they can be managed as one in the future, with an adapter layter

Norm: What will the world do if RDfa spec by XHTML2 does not beocome part of HTML5

How do we feel abt redefinition of attributes like rel and rev?

<timbl> rel= and rev=

Norm: Attributes declared to accept a URI or CURIE but we have discussed this so I'm happy to drop it

<Stuart> http://lists.w3.org/Archives/Member/tag/2008Mar/0072

SW: Is the comment to make is to ask the 2 HTML WGs to coordinate?
... Norm, could you draft a partial comment?

JAR: I sent my comments out personally ... I have nothing to add

<Stuart> ACTION: Norman to draft comment for feedback of RDFa review [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action03]

<trackbot-ng> Created ACTION-129 - Draft comment for feedback of RDFa review [on Norman Walsh - due 2008-04-03].

<jar> "virtual URI schemes" !?

Tim: Shd we reinvestigate the idea that we shd not distinguish between prefix's and scheme names?

TimBL: Move QNames and URIs into same space with one fell swoop
... Getting compatibility will take work ... you want it as an extension to URI spec ... don't want languages inventing indepently

SW: Do we have all the comments we want to make on the RDFa call?
... We shd review media-type, frag-id interaction wrt HTML and RDFa

TimBL: Is there an ambiguity there?

SW: I shd go look at this more closely

<timbl> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Mar/0297

Bristol travel arrangements

SW: I will put up a logistics page

Noah: If you could just send mail with hotels

<timbl> SW: HP labs is on the N edge of Bristol, but you might not want to stay there ... downtown Briston is more interesting, especially for walks in the evening.

<timbl> ... We would need a taxi to pick you up in the morning.

<timbl> 100 miles from LHR

<timbl> 1.5-2hs drive

<timbl> Or train to Paddington (15mins every 15ins) then train to Bristol.

<timbl> £140 for a pickup by car

<timbl> "Rail air link is bus to Reading then mainline to Brisol parkwy/Temple Meads"

<timbl> 1.5 hrs train, every hour

Summary of Action Items

[NEW] ACTION: Henry S, Norm, Dave -- revise minutes from f2f with Jonathan's changes [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action01]
[NEW] ACTION: Norman to draft comment for feedback of RDFa review [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action03]
[NEW] ACTION: Stuart to send final text HTML WG [recorded in http://www.w3.org/2001/tag/2008/03/27-minutes.html#action02]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2008/03/31 17:09:40 $