draft minutes for the TAG teleconference of 26 january 2012

Draft minutes can be found here:


and below in plain text:


       [1] http://www.w3.org/

                                - DRAFT -

               Technical Architecture Group Teleconference

26 Jan 2012


       [2] http://www.w3.org/2001/tag/2012/01/26-agenda

    See also: [3]IRC log

       [3] http://www.w3.org/2012/01/26-tagmem-irc


           Peter Linss, Ashok Malhotra, Jonathan Rees, Henry Thompson,
           Noah Mendelsohn, Jeni Tennison, Yves Lafon, Dan Appelquist

           Dan Appelquist (partial), Larry Masinter, Robin Berjon, Tim

           Noah Mendelsohn

           Yves Lafon


      * [4]Topics
          1. [5]ISSUE-50 (URNsAndRegistries-50): Persistent references
          2. [6]ISSUE-58 (namespaceDocument-8): XML Schema Namespace
          3. [7]ACTION-560: Polyglot document review
          4. [8]Web Application Storage
      * [9]Summary of Action Items

    Jeni to scribe next week, if Larry not present

    for next week, regrets from Ashok and Peter

    approval of minutes:

    RESOLUTION: Minutes of 4-6 January 2012 F2F
    [10]http://www.w3.org/2001/tag/2012/01/04-agenda are approved

      [10] http://www.w3.org/2001/tag/2012/01/04-agenda

    Yves to create a local arrangements page the next f2f meeting

    Noah: I will probably miss the IETF meeting prior TAG f2f

    ht: the point was for us to learn about IETF more than being sure
    TAG is represented

    approval of january 19th minutes postponed

ISSUE-50 (URNsAndRegistries-50): Persistent references


      [11] http://www.w3.org/2001/tag/products/persistence.html

    Dated version:

      [12] http://www.w3.org/2001/tag/products/persistence-2012-01-23.html

    ht: there are high level goals, despite focusing mainly on http uris
    (but not only). What is lacks crucially is next steps

    the only pending thing was the workshop report, which is linked now
    and was circulated widely.

    Key deliverables and schedules are empty as we are missing next

    jar: if we had a clear view of who would be the customers, it would

    jar: maybe a community group?

    ht: community group are about what W3C can do.

    not sure what we want out of this is a REC track doc

    <noah> I've tried to push us to have goals and success critera that
    aren't document; documents should be deliverables in support of some

    Yves: CGs are not only to do pre-REC track work, it can be there to
    clarify things on a topic

    Noah: how about saying that our work is done, we helped.

    ht: we need to do something, as if things break here, we wouldn't
    have done our job

    HT: Success would be coming into existence of a mechanism for
    creating persistent domains

    jar: we are not empowered or able to achieve this kind of success.
    We can produce document, recipes to get persistent domain names, but
    other will have to implement

    <noah> I note that we pretty much aren't empowered to do anything. I
    still think it sometimes make sense to point out that success can be
    defined at more than one level. Sometimes, I think it's OK to set
    success criteria even when the means of achieve that success is not
    entirely within our control

    real success would be creation of such domains, but it's not
    something we can achieve

    <jar> noah says 2 goals: #1 persistent domains, #2 publishing a
    recipe for causing persistent domains to come about

    <noah> I thought I said two goals: 1) real success is causing the
    community to converge on a practical means of maintaining persistent
    domains -- that's a goal we can't achieve by ourselves in support of
    which 2) publish a document, a goal we can definitely achieve

    <noah> ACTION-528?

    <trackbot> ACTION-528 -- Henry Thompson to create and get consensus
    on a product page and tracker product page for persistence of names
    -- due 2012-01-24 -- PENDINGREVIEW

    <trackbot> [13]http://www.w3.org/2001/tag/group/track/actions/528

      [13] http://www.w3.org/2001/tag/group/track/actions/528

    <noah> ACTION-653?

    <trackbot> ACTION-653 -- Noah Mendelsohn to schedule telcon
    discussion of Persistence product page (which was drafted for but
    not reviewed at F2F -- due 2012-01-17 -- PENDINGREVIEW

    <trackbot> [14]http://www.w3.org/2001/tag/group/track/actions/653

      [14] http://www.w3.org/2001/tag/group/track/actions/653

    Noah: is the question "are there people lining up to create that CG"
    or "are there people to produce work in that CG" ?

    ht: both, first work would be to get people engaged

    <noah> From: [15]http://www.w3.org/2001/tag/products/

      [15] http://www.w3.org/2001/tag/products/

    <noah> Draft of product plan was discussed 4-6 January 2012 TAG F2F:
    this version was proposed by Henry Thompons, but more discussion is
    needed to agree on the plan.

    <JeniT> jar, you need to propose it and a some minimum number of
    people need to express interest

    <noah> [16]http://www.w3.org/2001/tag/products/

      [16] http://www.w3.org/2001/tag/products/

    <noah> close ACTION-528

    <trackbot> ACTION-528 Create and get consensus on a product page and
    tracker product page for persistence of names closed

    <noah> . ACTION: Henry to update
    to reflect agreement on 26 January telcon regarding settling at F2F

      [17] http://www.w3.org/2001/tag/products/persistence-2012-01-23.html

    <noah> ACTION: Henry to update
    to reflect agreement on 26 January telcon regarding settling at F2F
    [recorded in

      [18] http://www.w3.org/2001/tag/products/persistence-2012-01-23.html
      [19] http://www.w3.org/2012/01/26-tagmem-minutes.html#action01

    <trackbot> Created ACTION-666 - Update
    to reflect agreement on 26 January telcon regarding settling at F2F
    [on Henry Thompson - due 2012-02-02].

      [20] http://www.w3.org/2001/tag/products/persistence-2012-01-23.html

    <noah> ACTION-653?

    <trackbot> ACTION-653 -- Noah Mendelsohn to schedule telcon
    discussion of Persistence product page (which was drafted for but
    not reviewed at F2F -- due 2012-01-17 -- PENDINGREVIEW

    <trackbot> [21]http://www.w3.org/2001/tag/group/track/actions/653

      [21] http://www.w3.org/2001/tag/group/track/actions/653

    <noah> close ACTION-653

    <trackbot> ACTION-653 Schedule telcon discussion of Persistence
    product page (which was drafted for but not reviewed at F2F closed

    <noah> ISSUE-58 (namespaceDocument-8): XML Schema Namespace Document

ISSUE-58 (namespaceDocument-8): XML Schema Namespace Document

    <noah> ACTION-23?

    <trackbot> ACTION-23 -- Henry Thompson to track progress of #int bug
    1974 in the XML Schema namespace document in the XML Schema WG --
    due 2012-01-24 -- PENDINGREVIEW

    <trackbot> [22]http://www.w3.org/2001/tag/group/track/actions/23

      [22] http://www.w3.org/2001/tag/group/track/actions/23

    ht: the schema datatype spec has a small section about namespace

    issue is that the ns document is missing an anchor

    HT: For XSD datatypes, it suggests that to talk about e.g. any uri,
    you use XMLSchemaNamespace#anyURI. Problem is, you can get a
    namespace document, and it has no such anchor.

    HT: I filed a bug a very long time ago

    the bug is still open, while schema language is updating its version

    <jar> are we talking about [23]http://www.w3.org/2001/XMLSchema ?

      [23] http://www.w3.org/2001/XMLSchema

    <JeniT> yes, I believe so

    <jar> isn't the same true of the XHTML ns doc?


      [24] http://www.w3.org/TR/xmlschema-2/#namespaces

    <jar> oh no it's not.

    ht: I notified the chair and it will be going forward

    <noah> The rec is at: [25]http://www.w3.org/TR/xmlschema-2/ which
    defines namespaces [26]http://www.w3.org/2001/XMLSchema and
    [27]http://www.w3.org/2001/XMLSchema-datatypes (the latter being
    rarely used, I think)

      [25] http://www.w3.org/TR/xmlschema-2/
      [26] http://www.w3.org/2001/XMLSchema
      [27] http://www.w3.org/2001/XMLSchema-datatypes

    <jar> xhtml doesn't specify use of #-concatenated URIs, xml schema
    does (IIUC)

    <ht> So, e.g. [28]http://www.w3.org/2001/XMLSchema#anyURI

      [28] http://www.w3.org/2001/XMLSchema#anyURI

    Jeni: do you need to do conneg to be able to resolve the fragment?

    ht: if you ask for application/xml first you'll get the xml schema,
    not the html

    <jar> it would be easy enough to kludge up some RDFa

    Jeni: if you have anchor in the HTML, it will mean something you may
    not mean
    ... what does [29]http://www.w3.org/2001/XMLSchema#foo mean?

      [29] http://www.w3.org/2001/XMLSchema#foo

    <noah> FYI, [30]http://www.w3.org/2001/XMLSchema-datatypes.xsd

      [30] http://www.w3.org/2001/XMLSchema-datatypes.xsd

    <noah> <simpleType name="anyURI"><restriction

    <jar> I also think it is meant to refer to the datatype, not to an
    HTML element… but maybe not clear... this is a good exercise to
    force the question

    <ht> "Each built-in datatype defined in this specification can be
    uniquely addressed via a URI Reference constructed as follows:

    <ht> the base URI is the URI of the XML Schema namespace

    <ht> the fragment identifier is the name of the datatype

    <ht> For example, to address the int datatype, the URI is:

    <ht> [31]http://www.w3.org/2001/XMLSchema#int"

      [31] http://www.w3.org/2001/XMLSchema#int

    Ashok: you don't want to refer to a place in the document, you want
    to refer to a datatype. So a different use of a fragid

    noah: when talking about 3023bis, we tripped over the same issue
    with rdf+xml, saying that the fragment is not identifying what is
    meant by 3023bis. Now there are two more exceptions, xpath functions
    and now xml schema

    <ht> <xs:simpleType name="int" id="int">

    <ht> [from the schema-doc for schemas --- so there are anchors there
    already, the conneg problem is that there are no such anchors in the
    other leg of the conneg]

    jar: no need to talk about 3023bis. Only read 3023 which is what is
    in effect right now

    <jar> application/xhtml+xml already specifies rdf-style fragids…
    we could just assume that and put rdfa in the ns doc.. don't talk
    about 3023bis now

    noah: our goal with 3023bis was to help move out, so it's relevant

    jar: relevant to 3023bis, but not to this discussion

    ht: the wording in the schema spec is using "

    Each built-in datatype in this specification (both "primitive" and
    "derived") can be uniquely addressed via a URI Reference constructed
    as follows:


    if we conneg, we have a document where the anchor is defined, one
    where the anchor is not

    <noah> noah: I thought our tentative agreement with RFC 3023bis
    authors (Chris Lilley and Murata Makoto) was that
    application/rdf+xml would be called out as the one exception to the
    rule that someURIreturningXML#fragid would in all cases be
    interpreted per XPointer

    ht: note that this issue was before http-range-14

    <noah> noah: I'm pointing out that we're hearing that XSD datatype
    namespace documents and XQuery function namespace documents may be
    two more exceptions. I'm suggesting this should be discussed with
    Murata-san and Chris

    <jar> FOAF spec sets good example, I think? (for the rdf-style
    fragids at least)

    <noah> noah: I hear Henry saying "yeah, I have a broader action to
    get back to the 3023bis folks"

    Jeni: the other way of resolving it would be to remove the ids from
    the schema doc

    <noah> ACTION-564?

    <trackbot> ACTION-564 -- Henry Thompson to track fragid issues in
    3023bis, report to TAG and/or communicate with 3023bis editors as
    appropriate -- due 2012-02-14 -- OPEN

    <trackbot> [32]http://www.w3.org/2001/tag/group/track/actions/564

      [32] http://www.w3.org/2001/tag/group/track/actions/564

    ht: the group will probably think that it will breaks backward

    <jar> ht: "can be addressed" in schema spec is webarch-anomalous

    <JeniT> it's been a useful example I think on the fragid work

    <noah> XSD 1.1 datatypes document:

      [33] http://www.w3.org/2009/XMLSchema/datatypes.xsd

    <noah> Oops, that's for markup like <xs:facet>. Sorry.

    Ashok: let's turn the question around: you got a collection of data
    types, you want to talk about data types, and collection of data
    types. What mechanism can be used?

    jar: RDF semantics tells you what those URIs means

    <jar> google rdf-mt

    <JeniT> ashok, you could use RDF that said xsd:int
    rdfs:isDescribedBy <schema.xsd#int-anchor>


      [34] http://www.w3.org/TR/rdf-mt/

ACTION-560: Polyglot document review

    <noah> ACTION-560?

    <trackbot> ACTION-560 -- Henry Thompson to review HTML polyglot last
    call Due 2011-06-06 -- due 2012-01-24 -- OPEN

    <trackbot> [35]http://www.w3.org/2001/tag/group/track/actions/560

      [35] http://www.w3.org/2001/tag/group/track/actions/560

    <Ashok> Yes, Jeni, we could now ... there was no RDF when we decided
    on that convention

    <JeniT> ashok, sure :)

    <noah> Referring to HTML issue 120

    ht: this has to do with my concern with HTML issue 120, allowing to
    remove all ns definitions from html5

    <noah> [36]http://www.w3.org/html/wg/tracker/issues/120

      [36] http://www.w3.org/html/wg/tracker/issues/120

    <noah> Henry please clarify: are you speaking specifically of the
    text/html serialization? I assume application/xhtml+xml supports
    namespaces and puts them in the DOM in any case, right?

    the status quo seemed to mean that ns definition was allowed in
    html5, but it is still unclea


    <noah> RDF or RDFa?

    <JeniT> RDFa

    <Zakim> noah, you wanted to get clarification and ask about the DOM

    noah: by html5, you wanted to say the html serialization or the xml

    ht: it's about the use of ns declaration with the html parser (as
    opposed to the xml parser)

    <JeniT> the RDFa-in-HTML doc that talks about interpreting namespace
    declarations is

      [37] http://www.w3.org/TR/rdfa-in-html/#backwards-compatibility

    ht: RDFa talks about the DOM and about the infoset. There is
    something in the html spec about how to get an infoset out of the
    html parser

    <noah> ACTION-560?

    <trackbot> ACTION-560 -- Henry Thompson to review HTML polyglot last
    call Due 2011-06-06 -- due 2012-01-24 -- OPEN

    <trackbot> [38]http://www.w3.org/2001/tag/group/track/actions/560

      [38] http://www.w3.org/2001/tag/group/track/actions/560

    <noah> noah: My point was, we should focus not so much on whether
    HTML5 claims to drop or not drop namespaces, but rather on what's in
    the DOM, and whether downsteam specs can build on that to get
    namespaces when they need them

    <noah> ACTION-560 Due 2012-02-28

    <trackbot> ACTION-560 Review HTML polyglot last call Due 2011-06-06
    due date now 2012-02-28

Web Application Storage

    <noah> ACTION-632?

    <trackbot> ACTION-632 -- Ashok Malhotra to frame issues around
    client-side storage work -- due 2012-02-07 -- OPEN

    <trackbot> [39]http://www.w3.org/2001/tag/group/track/actions/632

      [39] http://www.w3.org/2001/tag/group/track/actions/632

    <noah> ACTION-647?

    <trackbot> ACTION-647 -- Ashok Malhotra to draft product page on
    client-side storage focusing on specific goals and success criteria
    Due: 2012-01-17 -- due 2012-02-07 -- OPEN

    <trackbot> [40]http://www.w3.org/2001/tag/group/track/actions/647

      [40] http://www.w3.org/2001/tag/group/track/actions/647

    <noah> ACTION-572?

    <trackbot> ACTION-572 -- Yves Lafon to look at appcache in HTML5 --
    due 2011-11-29 -- OPEN

    <trackbot> [41]http://www.w3.org/2001/tag/group/track/actions/572

      [41] http://www.w3.org/2001/tag/group/track/actions/572

    Noah: draft product page on this

    we got feedback from Robin on the ML

    <noah> Draft product page:

      [42] http://www.w3.org/2001/tag/products/clientsidestorage-2012-01-24.html

    Ashok: Robin comments were very useful. I don't think that he
    objecting to the product page

    noah: I'm proposing we get as far as we can with those of us on the
    call, and we will address any remaining concerns Robin may have when
    he's available

    Ashok: we need to clarify what the issues are.

    <noah> Goals

    <noah> The goal of this work is to document for the community good
    practices relating to the use of client-side local storage by Web

    <noah> YL: Seems to be about use of current technologies....

    <noah> AM: You are right, was going to get to that, but will do so

    <noah> AM: One of the things you want are apps that can run
    connected or disconnected

    <noah> AM: Should be easy to switch modes

    <noah> AM: If you actually look at appcache, which helps you with
    offline, you can't do that. The app is either online or offline.
    Your files are the resources that you have stored locally. You
    cannot update, etc.

    <noah> AM: There's no synchronization, etc. If we start with "this
    is what client-side storage should enable", you want up with a bunch
    of requirements, many of which are not yet addressed in current
    technology. So, we're speaking about requirements for a client-side
    storage facility.

    <noah> AM: This is different in spirit than what we usually do.

    <noah> YL: Clarification about app cache. It's more about having
    consistent state of resources belonging to web apps. If you're
    offline, you'll have in your cache all the data you need to run.
    it's consistent, and it's all there. WHen connected, you might get

    <noah> YL: That's a bit different from local storage, or a
    comprehensive approach to disconnected operation that might include
    PUT/POST, etc.

    <noah> YL: So appcache and local storage are different

    <Zakim> noah, you wanted to say we want architectural principles.

    noah: we should emphasize on architectural issues. The success
    criteria starts with "what are the requirements", but it may not be
    TAG's role to take the lead here. TAG's role might be more in the
    use of URI, the interaction with the HTTP cache, etc...

    Ashok: we also need to say what we want local storage to do for us

    noah: we want to understand the use cases, but the main thing is
    about the architectural principles

    like linking

    ex: if you are creating something locally and it will be exported,
    what is the linking story there?

    let's continue in two weeks, when Ashok and Robin will be there


Summary of Action Items

    Action Items [NEW] ACTION: Henry to update
    to reflect agreement on 26 January telcon regarding settling at F2F
    [recorded in
    [End of minutes]

      [43] http://www.w3.org/2001/tag/products/persistence-2012-01-23.html
      [44] http://www.w3.org/2012/01/26-tagmem-minutes.html#action01

     Minutes formatted by David Booth's [45]scribe.perl version 1.136
     ([46]CVS log)
     $Date: 2012/01/31 18:34:21 $

      [45] http://dev.w3.org/cvsweb/%7Echeckout%7E/2002/scribe/scribedoc.htm
      [46] http://dev.w3.org/cvsweb/2002/scribe/

Baroula que barouleras, au tiu toujou t'entourneras.


Received on Tuesday, 31 January 2012 19:02:19 UTC