See also: IRC log
<artb> Date: 9 August 2007
<artb> Scribe: Art, Anne, Marcos
<artb> ScribeNick: Art
<artb> AB: minutes http://www.w3.org/2007/07/11-waf-minutes.html
AB: minutes: http://www.w3.org/2007/07/11-waf-minutes.html
... any objections to approving these minutes?
[NO Objections]
RESOLUTION: July 11 2007 minutes approved
AB: I propose two in Sept and two
in October
... comments?
MC: OK
AvK: OK
<scribe> ACTION: Barstow schedule 2 conf calls in Sept and 2 in October [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action01]
<trackbot-ng> Created ACTION-113 - Schedule 2 conf calls in Sept and 2 in October [on Arthur Barstow - due 2007-08-16].
AB: any comments?
AvK: we haven't discussed it yet but I'm pretty sure we want WAF to continue beyond November 2007
MC: QUT wants WAF to continue
AB: I presume Nokia wants the WG to continue
<scribe> ACTION: Barstow create a Draft of our new Charter by the end of September [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action02]
<trackbot-ng> Created ACTION-114 - Create a Draft of our new Charter by the end of September [on Arthur Barstow - due 2007-08-16].
AB: who will attend?
AvK: will attend if it doesn't clash with HTML WG
MC: yes I will attend
AB: of course I will be there and I expect Guido to also attend
<scribe> ACTION: Barstow closely reveiw the November TP week schedule for clashes [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action03]
<trackbot-ng> Created ACTION-115 - Closely reveiw the November TP week schedule for clashes [on Arthur Barstow - due 2007-08-16].
AB: if we absolutely had to meet on the weekend before, what are your thoughts?
MC: that would be OK with me
AvK: likely; I may also just send someone else (e.g. Arve)
AB: should we create a PP on
behalf of WAF?
... I would expect a very short intro about what we are
doing
... but no strong recommendation on Ajax
... other than to acknowledge that it can be useful in the
context of creating Web Applications with Widgets and/or
XBL2
MC: we should make statements about scope?
AB: here is the WS home page: http://www.w3.org/2007/06/mobile-ajax/
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/waf/
ABe: we could make a statement
about Widgets on mobile platforms
... Widgets offer an install once value proposition and can
restrict or limit network access there after if so desired
AB: logistically, an issue is the paper is due Aug 15
MC: I can take a 1st cut, send it to Arve and Guido
AB: what are the highest priorities for today: Widgets, XBL2, Read Access or the Workshop Position Paper?
ABe: Widgets is #1 prio
AvK: not sure we can do a lot of in-depth discussions on Read Access without Jonas or Thomas
<scribe> ACTION: Barstow September voice confs need to include Thomas and Jonas [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action04]
<trackbot-ng> Created ACTION-116 - September voice confs need to include Thomas and Jonas [on Arthur Barstow - due 2007-08-16].
MC: Widgets is my highest prio
AB: let's take XBL2 first; Widgets 2nd; Read Access 3rd; Position Paper 4th
AB: XBL2 Spec http://www.w3.org/TR/xbl/
... any status on Implementations to report?
... I wonder if Mikko is available to dial-in and give us a
short status of his impl?
AvK: I don't have anything to report
AB: I also don't have anything to
report
... I haven't from Apple/Webkit team lately. At one point it
was on their "someday pile". Not sure of the
status.
<scribe> ACTION: Barstow check with Webkit team about XBL2 implementation plans [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action05]
<trackbot-ng> Created ACTION-117 - Check with Webkit team about XBL2 implementation plans [on Arthur Barstow - due 2007-08-16].
AB: anyone know about Mozilla's XBL2 plans?
[Silence]
<scribe> ACTION: Barstow check with Mozilla about their XBL2 implementation plans [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action06]
<trackbot-ng> Created ACTION-118 - Check with Mozilla about their XBL2 implementation plans [on Arthur Barstow - due 2007-08-16].
AB: Marcos, do you know Cam's plans wrt XBL2 impl?
MC: you'll have to ask Heycam, I'm not sure
AB: is it likely XBL2 will need to go back to WD?
AvK: yes. For starters it depends on Access Control but HTML5 too. Jonas has also submitted some comments that are not addressed by the Candidate.
MC: the FPWD of the Primer was
published in July
... currently only have received minor Editorial comments and
we have fixed those
... We published an article on Ajaxian
<marcos> http://ajaxian.com/archives/xbl-20-primer
MC: we received over 250
comments
... average rating is ~1.3
ABe: I think some of the examples don't illustrate the main advantages
<mikko> Hi art, I noticed only now you've pinged me.
<mikko> should I dial in?
Mikko: I added support for the
implementation element
... can also add handers
... I created a couple of examples
... I don't have feedback on usage
... what are the group's expectations for my impl?
... I know we wanted to use it with the Primer.
MC: I think it would be useful
for the test suite.
... and of course it will also be helpful for the Primer
Mikko: I cannot implement all of the features.
MC: yes, understood but it can still be useful.
Mikko: should I still try to add something; how should I proceed?
MC: you've got template and binding working right?
Mikko: yes
MC: what about the extends attr?
Mikko: yes, it is implemented
<mikko> status: http://dev.w3.org/cvsweb/~checkout~/2006/waf/XBLImpl/XBL_Status.html?rev=1.5&content-type=text/html;%20charset=iso-8859-1
Mikko: I have a table in CVS that says what's been implemented and what has not been implemented
AvK: do external bindings work?
Mikko: no
MC: can style be implemented?
Mikko: yes it could be
MC: what about the script element?
<anne> http://dev.w3.org/2006/waf/XBLImpl/demo/ is pretty cool!
Mikko: yes I think it could be implemented as well as style and the src attribute of the implementation element
AB: how much time do you have to work on it?
Mikko: I can add the stuff we just talked about but not likely to have time for any more.
MC: we could ask Lachy if he can participate
MC: I met with Cameron last week.
We talked about a model for gathering XBL2 test cases from the
Public
... The W3C has coincidentally opened a Public repository to
gather test cases for some mobile specs.
... I am hopeful to use this repo to collect XBL2 tests.
... We don't have a commitment yet from anyone to create test
cases but we do have a model/framework on paper.
... We created something like 100 test cases on paper.
... Would like the test suite to be very comprehensive.
... Still thinking about whether the XBL should be internal or
external to the host language.
... We will use XHTML as the main hosting language.
... I am expecting to have the repository set up in about one
month.
... Cameron and I are still working on a template
... Mikko do you want to be included in the template
discussions?
Mikko: yes, please.
<scribe> ACTION: Marcos send XBL2 stuff to Mikko and Cameron to review [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action07]
<trackbot-ng> Created ACTION-119 - Send XBL2 stuff to Mikko and Cameron to review [on Marcos Caceres - due 2007-08-16].
MC: In the Ajaxian article I
asked for volunteers for the XBL2 Primer
... so far no takers :-(
... We need to find some people.
<scribe> ACTION: Barstow actively seek some contributors to the XBL2 Primer (e.g. WAI community) [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action08]
<trackbot-ng> Created ACTION-120 - Actively seek some contributors to the XBL2 Primer (e.g. WAI community) [on Arthur Barstow - due 2007-08-16].
<scribe> ACTION: Barstow contact Mark Birbeck to determine his interest in contributing to the Form Controls part of the XBL2 Primer [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action09]
<trackbot-ng> Created ACTION-121 - Contact Mark Birbeck to determine his interest in contributing to the Form Controls part of the XBL2 Primer [on Arthur Barstow - due 2007-08-16].
<artb> ScribeNick: ArtB
AB: latest Editor's version: http://dev.w3.org/cvsweb/2006/waf/access-control/Overview.html
<anne> http://dev.w3.org/2006/waf/access-control/Overview.html
AvK: I've done some re-org
... and added a couple of new sections: Access Request and
Access Check
... URI matching is more aligned with Jonas' comments
<anne> http://dev.w3.org/cvsweb/2006/waf/access-control/Overview.html.diff?r1=1.83&r2=1.84&f=h
<anne> http://dev.w3.org/2006/webapi/XMLHttpRequest-2/Overview.html
AB: open issues - http://www.w3.org/2005/06/tracker/waf/issues/open
AvK: Closed issue #2 - we will
use Access-Control
... Closed issue #7
... Closed issue #8 - I told Matt the reason via an e-mail
AB: during our Sept and October telcos, the Read Access agend will include any Open Issues
AvK: what are the agenda items?
MC: automatic updates
AvK: inter-widget communication
AB: what about Autodiscovery (section 8)?
MC: yes
ABe: the issue is "Find out what
needs changing to make this work for Atom too."
... I think this issue should be remove because it is not
relevant
MC: do we need to point to
HTML5?
... That would put a Normative reference dependency on the
spec.
... Is that OK?
AvK: we could address this at
CR
... HTML5 defines a registry and presumably we would register
"widget"
MC: should we ping the HTML WG?
<anne> rel registry: http://wiki.whatwg.org/wiki/RelExtensions
AvK: not at this point in time.
ABe: I have some issues with the
Security Model (Section 7).
... I raised some issues yesterday
MC: I captured those issues
... in the discussion of the <access> element
MC: I propose we skip this for now.
AvK: author sets updateURI in the
manifest
... engine then does a request to URI with version number
... response would give a new download location and version
number
... if newer version, new one fetched and the old one replaced;
otherwise ignore
MC: concerned about the update URI in the response being hijacked
AvK: if the update location is not secure you're doomed anyway
ABe: a requirement is the update URI is indirect
AvK: if we use this mechanism, may not need a unique ID
MC: so what, specifically, is the
URI in the request?
... marcos.com/update.php?version=1
ABe: need to define the document returned in the response
MC: could be a new format or a
microformat or a HTTP header
... easiest thing to do is an HTTP header
ABe: that could provide a version but no adequate means for structured data
MC: agree an XML format would be more flexible
ABe: use case for more structured data: the update may need to be signed
AvK: I'm not convinced by that use case and the complexity it would introduce
ABe: another use case is trying to get multiple updates with on request
AvK: not sure we want to address
that UC; would mean ID is required
... would also mean all Widgets are updated from the same
place
<scribe> ACTION: Marcos investigate response format for the update (e.g. see what is done in J2ME) [recorded in http://www.w3.org/2007/08/09-waf-minutes.html#action10]
<trackbot-ng> Created ACTION-122 - Investigate response format for the update (e.g. see what is done in J2ME) [on Marcos Caceres - due 2007-08-16].
GE: we had discussed having two
URIs
... could also optimize the protocol by just sending the
version and if nothing newer there is no response or minimal
content response
AvK: the format of the response would contain an absolute minimum of data e.g. version number and URI; possibly some additional data
GE: Keep It Simple
ABe: what is the benefit of returning an empty document?
MC: there would always be some
type of response the options are an HTTP response header or an
XML file
... e.g. Response 204 - No content (response must not include a
message-body)
AvK: could include version info in a template in the URI
ABe: the updateURI must remain
stable across versions
... If the server does not a newer version, server returns
204
... Otherwise a document is returned with URI for newer version
and the version number [minimal data]
<anne> 1. GET request to update URI which contains a template
<anne> 2. GET request to update URI where the GET request has a HTTP header that indicates the version
<anne> 1. example: http://example.org/updates/fancyWidget{%version}
<anne> 2. example: http://example.org/updates/fancyWidget with Widget-Version: "version" HTTP header
<anne> 3. GET request to update URI where the user agent appends ?version=version or &version=version if there is a ? in the URI already
<anne> 3. example: http://example.org/updates/fancyWidget?version=1.0 or http://example.org/updates/fancyWidget?user=anne&version=1.0
<marcos> 0. GET request that does not use 1 or 2
<anne> strike 0
<anne> it's a dupe
<marcos> 0 is awesome
<arve> 3 === 0
<anne> 1. example request would be http://example.org/updates/fancyWidget1.0
<marcos> http://example.org/updates/fancyWidget{%version}&rssFeed=http://bla.com/news.rss
<anne> http://www.ietf.org/internet-drafts/draft-gregorio-uritemplate-01.txt
MC: so the my Use Case i.e. wanting to add some additional parameters to the URI is satisfied by the template proposal
ABe: I am concerned about the
authoring complexity issue
... e.g. what if Version is mis-typed
... I would prefer an optional HTTP header
AB: Marcos, are you OK with option #2
MC: yes that's OK with me because it builds on #1
AvK: options #2 is OK with me too
RESOLUTION: Option #2 Agreed; response header is XML
[Marcos adds format details to the Editor's Draft ...]
AB: which Widget engines have some type of inter-widget comm support?
MC: Yahoo! has an API (tellWidget)
AvK: could use Window.postMessage() [HTML5]
<anne> http://www.whatwg.org/specs/web-apps/current-work/#cross-document
AvK: this is the "cross document" mechanism
ABe: what would we do with the "domain" attribute?
AvK: good question; I'm not sure
ABe: domain may be location
info
... i.e. Window.location object
AB: how long has postMessage been around?
ABe: we've had it for over two years
GE: do need some type of registry/directory to know which Widgets are available
ABe: need to query the Widget
engine for available Widgets
... like Widget.getWidgetWindows
AvK: what Widgets can I talk
to
... what Widget am I going to talk to
MC: I am concerned about the
security issues
... could be addressed other ways e.g. by some type of
mediating server
... I'm also a bit concerned about widening the scope too
much
AB: regarding scope, I'm OK with
us expanding the scope and doing some "innovation".
... We can publish our ideas and seek comments from the
community.
... Ultimately, if we have good ideas the spec will presumably
get implemented; otherwise it just never leaves Candidate
[Marcos summarizes the Issues inline in his Editor's Draft ...]
[Marcos presents his first draft ...]
AB: looks OK
AvK: looks OK with me to
AB: I propose Macros send the
edited copy to the WG ASAP
... and we give WG members until 24:00 UTC August 12 to send
comments.
... If there are no objections, I will submit the PP to the W3C
on Aug 13
AB: we have now covered all of
the agenda items in the original agenda and the ad hoc agenda
we created today
... I don't have anything else.
... anyone else have anything to discuss?
MC: nope
AvK: nothing else
AB: thanks very much to Anne and
Opera!
... everyone should note that Chaals is a wonderful cook and
host!
... he even serves whale meat to his friends :-)
... [tastes just like chicken ...]
... meeting adjourned
This is scribe.perl Revision: 1.128 of Date: 2007/02/23 21:38:13 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/don't a/ don't have a/ Succeeded: s/Respons 204/Response 204/ Succeeded: s/whale meet/whale meat/ Found Scribe: Art, Anne, Marcos WARNING: No scribe lines found matching ScribeNick pattern: <Art\,\ Anne\,\ Marcos> ... Found ScribeNick: Art Found ScribeNick: ArtB ScribeNicks: Art, ArtB Present: Art Anne Arve Marcos Mikko_(XBL2) Gorm_(Widget_Update) Agenda: http://lists.w3.org/Archives/Member/member-appformats/2007Jul/0020.html Found Date: 9 Aug 2007 Guessing minutes URL: http://www.w3.org/2007/08/09-waf-minutes.html People with action items: actively barstow birbeck confs contact contributors marcos mark need seek send september some stuff voice xbl2[End of scribe.perl diagnostic output]