Widgets Voice Conference -- 11 Sep 2008

Date: 11 September 2008

<scribe> Scribe: Art

<scribe> ScribeNick: ArtB

Agenda Review

AB: agenda posted yesterday: http://lists.w3.org/Archives/Public/public-webapps/2008JulSep/0617.html
... regarding V2/NG features for Widgets, Claudio won't be here today and asked me to postpone the discussion until next week
... any problems with that?

ABe: no

MC: no

Announcements

AB: I was hoping for a BONDI Widgets update but Marcos said they are meeting this week
... I'll add that to next week's agenda
... Marcos was contacted by the Mobile Web Test Suites WG regarding Widgets test suite

MC: MWTS WG would like to help other WGs with their test suites
... they have identified Widgets as a potential candidate
... Discussions are still preliminary
... For the first step, I've asked Dom to look at the spec from a "test-ability" viewpoint
... If there is anything we can do in the spec itself to facilitate testing, we want to know that now and reflect it in the spec

AB: this is great Marcos
... I told Dom we are very interested in engaging MWTS WG

MC: Arve, how does this sound to you?

ABe: it would be best if I was able to get someone from our QA team involved

AB: I'll do the same

Requirement Document

AB: what is the status on the LC comments, Marcos?

MC: waiting for confirmation from Kryztof
... also Josh
... also I18N WG
... also Bryan and MWBP WG

AB: so that is quite a few loops to close

MC: I gave the MWBP WG a deadline (think it was end of this week)

AB: Bryan submitted some comments about the Closing of Issue #17
... To me, proxy support could be a candidate for the V2/NG feature list
... Would that make sense?

MC: yes that would make sense to me, especially if it will keep the doc from being blocked

AB: in general, I don't want to re-open Closed Issues unless there is large consensus among the people that closed the issue that new input/evidence suggests we re-open

<MikeSmith> (I agree about not re-opening closed issues.. it's one of the worst things that a WG can decide to do.)

MS: I agree with Art's comment about not re-opening Closed Issues
... it has caused large problems for some WGs
... Re-opening issues will delay our specs and hence implementations
... This inevitably will result in some people not being happy.
... Unfortuantely that will happen but we also need to be cognizant about schedule.
... That said, we do need to keep track of all feature requests
... and such requests to the V2/NG list.

<Zakim> MikeSmith, you wanted to weigh in on V2 feature-list idea

MC: I agree with what Mike said
... in the case of the proxy, I don't view it as a feature.
... I don't think we can spec the proxy input as Bryan proposed it.

<timelyx> "oops"

AB: other than chasing the commentors, is there any other work that needs to be done?

MC: no; I think the document is ready for a new publication

is next publication of the Requirements doc another Last Call or a _plain_ WD?

MC: I want to go straight to another LC

AB: my take on the Process Document is we need to publish another WD before a LC doc
... http://www.w3.org/2005/10/Process-20051014/tr.html#return-to-wg

BS: what are the major diffs?
... between the 1st LC and what we now have in the latest ED

MC: the changes are "pretty significant"
... we could do another WD and then a LC but I don't think we'll get much feedback during that time frame
... People won't submit comments until another LC is published

BS: I agree

AB: I agree as well but I wonder if the process people will raise an objection

JS: could we just have a short review period for the WD?

BS: yes, I think we can do that
... Is there a minimum review period?

MS: I don't think so but 2-3 weeks is typical
... most people wait until the last day
... to submit comments
... It is very important to stick to the deadline for comments

MC: what if we publish a new WD ASAP and have just a 2-3 week review period
... and then on October 2, publish the LC

BS: so a WD would have just a 2-week review period

AB: In Turin we said we wanted the LC to end on Oct 13
... If the LC has a 3-wk review period it would then have to published on Sept 22
... That would then mean we could ony have a 1-week review period for the WD, assuming it was published on Sep 15
... having a 1-wk review period seems a bit odd

BS: agree, but I think Marcos has done a good job of answering all questions

<marcos> :)

AB: Mike, what do you think?

MS: I think this plan is OK
... We should do whatever it takes to make the Mandelieu f2f meeting as productive as possible

MC: what is the point of the 2nd Last Call?

BS: I assume those that submitted comments

MC: but they have already received confirmation from them that we have addressed their comments

AB: another proposal would be to just have no _plain_ WD and a LC with a 4-week review period
... Mike, can we do that?

MS: we would need to provide some rationale

AB: the PD says "In the case of substantive changes, the Working Group MUST republish the technical report as a Working Draft."
... in section 7.4.6 http://www.w3.org/2005/10/Process-20051014/tr.html#return-to-wg

MC: we want to be as productive as possible

AB: Mike what do you recommend?

<marcos> "A Working Group's Last Call announcement is a signal that:

<marcos> * the Working Group believes that it has satisfied its relevant technical requirements (e.g., of the charter or requirements document) in the Working Draft;

<marcos> * the Working Group believes that it has satisfied significant dependencies with other groups;

<marcos> * other groups SHOULD review the document to confirm that these dependencies have been satisfied.

<marcos> "

MS: we can review the PD here as a guide and not be overly restrictive
... I think we can just publish a new 2nd LC WD
... without a new _plain_ WD
... But we must make sure the LC addresses all of the comments that were submitted by the deadline for LC #1

AB: propose we do not publish a _plain_ WD and make the next publication the 2nd LC
... any objections?

BS: no

MC: no

ABe: no

MS: no

JS: no

RESOLUTION: we will not publish a _plain_ WD; our next publication will be LC #2

AB: what date are we shooting for?

MC: September 15

<MikeSmith> (ideally, anybody who submitted comments on the LC draft should be directly contacted -- with a CC to the list -- that we have published an updated draft)

AB: actually, I will neded to send a notification to the Chairs list

Core API and Events spec

AB: what's the status

ABe: I think the Sep 15 deadline I provided in Turin is still mostly do-able
... I am adding some input
... have a question about all of the properties

MC: I am wondering if rather than a list of props we use some type of Get and Set methods

ABe: regarding the Window interface, we have an open issue

<marcos> MC: I will spec that linkage between the default properties and the Widget interface (hence bridging the two specs so that the correct properties from the correct config doc are loaded).

ABe: using Web IDL, I don't know how to associate the relationship between the Window interface and the Widget interface

<arve> I'm not seeing any way in WebIDL where I can express that WindoWidget extends the Window interface, rather than replacing it

MC: does Web IDL have a notion of extends?

ABe: not really

<marcos> interface WindowWidget extends Window{ .... }

ABe: for the purposes of what we need
... I will contact Cam about this

JS: do we really need to clearly specify this?
... could prose be sufficient?

ABe: we could do that
... but in practice, I think it should be more explicit, especially for implementors

AB: how do we get consenus here?

<timelyx> I think we don't want interface declarations to bind an interface to a specific other interface

<timelyx> because it excludes the ability to bind it to some other interface

<timelyx> what we should want is an independent declarative statement:

<timelyx> Window supports WindowWidget;

<timelyx> so that someone elsewhere could write:

<timelyx> MyObject supports WindowWidget;

<timelyx> and this is really independent of the definition of WindowWidget, enabling others to later make similar statements for other objects.

ABe: this is probably good
... but should we ask Cam to express this in Web IDL
... It would be better to express this in Web IDL than in prose
... I will follow-up with Cam

Automatic Update Status

AB: what's the pub plan?

MC: I want to publish this ASAP

AB: propose we publish FPWD of the Automatic Update spec as soon as it is pub ready
... any objections?

ABe: I have plenty of issues with the doc but no objections to publishing it as is as the FPWD

BS: no

JS: no objections to FPWD

RESOLUTION: we approve the FPWD of the Automatic Updates spec as is

ABe: when is the publication moratorium?

AB: not sure but I'll let you know
... Meeting Adjourned

- DRAFT -

Widgets Voice Conference

11 Sep 2008

Attendees

Contents