See also: IRC log
<scribe> ScribeNick: ArtB
<scribe> Scribe: Art
Date: 9 July 2009
<darobin> hsivonen: you gonna join?
<hsivonen> I can call in
<darobin> easy to tell
<darobin> I can hear the trolls echoing behind you
<hsivonen> that's me
AB: agenda posted July 8 (
). One change is to drop 3.a. (Francois' comments) and add LC
Comment #2233 from Josh (
). Any other change requests?
... I note Henri is here
<darobin> new TODO for A+E: http://www.w3.org/mid/1B5EE78B-5FA5-472B-9250-AAADE4A0900B@berjon.com
AB: he says he is representing himself and NOT Mozilla
MS: during AOB want to talk about Team Contact going forward
AB: Reminder there will be no widgets call on July 16, July 23 and Aug 6. Any other short announcements?
AB: LC Comment #2233 (
) is from Josh. This is the only comment where the Commentor
has indicated the group's response is not acceptable. We have
an obligation to try to reach consensus so we will start
... Josh, would you please briefly clarify for whom you speak re your objection to the group's proposal?
JS: I can't formally speak for
Nokia and I can't formally speak for Mozilla
... so I can say I speak for myself
... but I have talked to other people that agree with position
AB: I want to clearly understand
the current model and the objections Josh and Henri have to
... Marcos, would you please briefly summarize the issue?
MC: the essence of the
... the current model is too flexible
... my understand is JS and HV think the flexibilily should be restricted
<hsivonen> For the record, I'm not objecting to anything
AB: please, everyone add to the
minutes if something is missing
... Josh, do you agree with MC's summary?
JS: yes, I do agree
... one issue is how to choose the lang
... another is what to do with the pkg
... the pkg fallback flexibility is too much
... and think user will be surprised by results
AB: what is the issue for the pkg creator?
JS: I've had some discussions off
list and some of that is not part of Public record
... but I will illustrate via an example
[ Josh looks for a log of offlist discussions ... ]
<timeless_mbp> ok... a german company creates a package for a company in Switzerland
<timeless_mbp> the package is therefore originally in German (de)
<timeless_mbp> they also then translate it into Italian (as some portion of Switzerland speaks Italian)
<timeless_mbp> and then they translate it into French (... for the same reason)
<timeless_mbp> they also commission for someone to translate it into English
<timeless_mbp> the German company then proceeds to create an updated version which adds additional resources (pictures)
<timeless_mbp> pictures with text
<timeless_mbp> for reference, Nokia uses Flash for its Text :)
<timeless_mbp> sorry... the updated version has another resource for pears and is in turn translated for De, It, Fr
<timeless_mbp> another version later updates to add Bananas but the resource is only provided in De and It
<timeless_mbp> and finally a final version adds yet another image resource which is only available in De
<timeless_mbp> The user preference is Fr, It, En, De
<timeless_mbp> and the result is that they get pictures from a mix of at least three of those
RB: I think this is a contrived
... one company publishes a widget that supports multiple langs
... think it is rare to get images from mixed langs
... may get one or two links for a diff lang
... which is better than no links
JS: some images are added later,
not part of the HTML
... the initial testing looks OK
AB: I would like you Josh to follow your No email with a clear example of the problem
JS: I'm concerned about bad
... it can cause bad usability
... user in one case thinks a feature is missing
AB: will you please follow-up your email Josh?
JS: yes, I can do that; will point to this discussion
<scribe> ACTION: soref send an email to public-webapps that clearly identifies the L10N problem you see with the current L10N model [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action01]
<trackbot> Created ACTION-379 - Send an email to public-webapps that clearly identifies the L10N problem you see with the current L10N model [on Josh Soref - due 2009-07-16].
MC: our L10N people think this is a good model
AB: Henri, any comments from you?
HS: I am not objecting to
... This is about failure modes
... if the L10N is not complete
... can get a bad experience if get multiple langs for one experience
... want to minimize the "ugli-ness" of the failure mode
AB: I see little to no value in
continuing to debate the "perfect model". On the contrary, I
think we need to get real data from both Implementors and
Localizers and that means getting the spec to Candidate and not
continuing to debate and re-debate and re-re-debate, ad nauseam
the perfect L10 model.
... want to know our options here
... I don't see a compromise position here; does anyone?
RB: no I don't think the two models can be merged
JS: what are the exit criteria
... is it 2 impls?
JS: do these impls need to be shipping?
RB: that would be a problem because you'd have to have widely shipped implementations and widely used content to satisfy your criteria, at which point we can't change the spec if it's broken anyway
AB: is voting the next step?
MS: no, not necessarily
... you can record a decision without recording a vote
... can just send an email to list saying "we discussed this and decided ..."
... and tell people to send comments if they have any
... want to avoid the use of "objection" if possible
... need to think about "I can Live With It"
... note that consensus is a goal and it can't always be achieved
... must also move forward
... can document we tried to get consensus but couldn't
... not absolutely required to get a vote
... must clearly describe the issue
... and state the proposal on how to move forward
... and then if someone wants to file a Formal Objection they can do that
... need to record the decision
... and provide an opp for others to comment
AB: thanks Mike
... ICLWI means the group can continue
... my proposal is we mark this comment as Resolved_No
... and that we continue
... I can respond to the mail list
... and point to this discussion
... who supports keepiing the L10N model as is?
MC: I do
RB: I do
Benoit: I do
<abraun> i do
AB: I do
... that is speaking on behalf of Nokia
... Jere is on holiday
<MikeSmith> hsivonen, you are still permitted to express an opinion -- it's not a formal vote
AB: I am hearing quite a bit of support for this
JS: when I talk to Maemo guys, they do not like the current model
RB: this isn't a formal vote
JS: if during the impl phase, serious issues are found, what can we do?
MC: we can fix it if indeed the impl experience dictates that a change is needed
JS: if we can fix it if we show
that we really screwed up, then that's great
... but if I find out that we cannot fix even though we screwed up that is very bad
AB: one of the realities is people are already implementing and shipping the LC version
<scribe> ACTION: barstow respond to Josh's no on 2233 with a pointer to these minutes [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action02]
<trackbot> Created ACTION-380 - Respond to Josh's no on 2233 with a pointer to these minutes [on Arthur Barstow - due 2009-07-16].
AB: all of the LC comments are
logged in the Comment Tracker (
). There are about 50 comments. Besides comment #2233 which is
still marked as Open, all of the others are marked as
"Resolved_Yes" which means we have responded to the comment,
and we have either implemented our response or done
... we are still awaiting responses for 19/46 comments: 16 from Opera's Anne van Kesteren; 1 from Opera's Martin Nilsson (#2230); 1 from Celestial Wake's Jeff Decker (#2228); 1 from Nokia's Jere Kapyaho (#2216).
... Mike, would you please clarify "how long must we wait for a Commentor to respond to our proposed resolution?"?
MS: there is no documented
... but the common practice is 2 weeks
... must consider though that during July and August 2 weeks is probably not realistic
... the fair thing is to wait for people to return from vacation
RB: in the case of Editorial comments, do we need to wait for 2 weeks, even in the Summer?
MS: no, I don't think so
... but if there are Substantive comments then I think waiting for a response is OK
AB: Marcos, what was the date of
your latest reply to Anne?
... or any of the commentors?
... I think it is July 3
... Anne's comments were not Substantive
... we also responded to all of his comments during the f2f meeting
... so he is well aware of our position on his comments
... none of the other comments were sustantive
... we did fix some bugs
Benoit: wondering if the 2 week period has already passed for the other comments
MC: I think we have addressed everyone's comments now
AB: what about #2230 from Martin Nilsson of Opera?
MC: his comments were for
... and non-normative parts of the spec
... AFAIC, we are done with his comments
... I will forward his comments to the list; he may not be subscribed
... just waiting for an OK
AB: re #2228, I asked Jeff Decker
to reply to your response Marcos and he hasn't done so
... but those were clarification?
MC: yes that's right
AB: re #2216 from Jere
... he won't be back for 3-4 weeks
MC: this is not blocking
AB: he was generally OK with the L10N model, right?
AB: it appears that we feel like
we do not have to wait for any responses before making a
decsion about moving to CR
... is that true?
MC: yes, that's correct
AB: draft proposal: We are ready
to publish a Candidate Recomm for the P+C spec
... any objections to that proposal?
[ None ]
RESOLUTION: we are ready to publish a Candidate Recommendation for the P+C spec
<scribe> ACTION: barstow create the Trans Request for the P+C CR [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action03]
<trackbot> Created ACTION-381 - Create the Trans Request for the P+C CR [on Arthur Barstow - due 2009-07-16].
<scribe> ACTION: smith work with Art to schedule the Director's Call for the P+C CR [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action04]
<trackbot> Created ACTION-382 - Work with Art to schedule the Director's Call for the P+C CR [on Michael(tm) Smith - due 2009-07-16].
AB: Robin, Marcos, what is the status of the A&E spec ( http://dev.w3.org/2006/waf/widgets-api/ )?
RB: I have been doing a bunch of
... would be good to link to a published Window Modes spec
... because one attr points to it
... and that is a Blocker
<scribe> ... dropped one attr
UNKNOWN_SPEAKER: clarified some
... updated preferences based on input from Hixie
... updated IDL
... need to fix showNotification
... and need input from Opera
<scribe> ACTION: marcos provide a specification for the showNotification method [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action05]
<trackbot> Created ACTION-383 - Provide a specification for the showNotification method [on Marcos Caceres - due 2009-07-16].
AB: that would be great
... if Charles or a Team Contact wants to push this to LC, that's OK with me - while I'm gone
RB: as soon as I fix these 2 pending issues, I can ask for a LC decision
AB: we can use the CfC mechanism
AB: anything else on A+E spec?
RB: please read it Everyone!
MC: I'll review it after I create the showNotif text
AB: Robin, what is the status of
the WARP spec ( http://dev.w3.org/2006/waf/widgets-access/
... what's the next step? FPWD was published about 4 weeks ago.
RB: we haven't had any
... OMTP people seem to be happy
AB: no feedback can be a concern but can also mean everyone is OK with it
RB: well, we talked about it a
lot before FPWD with TLR, et al.
... a good soln to no feeback is to go to LC
AB: I'm thinking the same
... so, is it ready for LC?
... is it feature complete?
RB: yes, I think so
... I do need a RNG schema but I can do that during CR since it is Informative
AB: does it meet all of the relevant requirements?
RB: the spec includes a list of reqs
AB: so I think we can say yes, it meets them all
MC: yes, I agree
AB: draft resolution: we are
ready to publish a LCWD of the WARP spec
... any objections?
[ None ]
RESOLUTION: we are ready to publish a LCWD of the WARP spec
<scribe> ACTION: barstow submit the request to publish LCWD of the WARP spec [recorded in http://www.w3.org/2009/07/09-wam-minutes.html#action06]
<trackbot> Created ACTION-384 - Submit the request to publish LCWD of the WARP spec [on Arthur Barstow - due 2009-07-16].
AB: there has been some discussion ( http://lists.w3.org/Archives/Public/public-webapps/2009JulSep/0108.html ) about the Widgets Updates spec ( http://dev.w3.org/2006/waf/widgets-updates/ ). Note we can continue to work on the spec while the PAG is still in operation.
MC: the proposal is to drop the checkForUpdate method
AB: are you OK with that?
MC: yes, I am
AB: any concerns or objections to removing that method?
[ None ]
AB: after that change is made, do we want to publish a new WD?
RB: there are a couple of other
minor changes that should be made as well
... I can make the changes if MC is too busy
MC: yes, that would be OK with me
RB: I prefer publishing a new WD as it emphasizes that work has not stopped
AB: I agree with that
RB: we can ask Charles or TC to make a pub request after it is ready
AB: OK, that's fine with me
... anything else on Updates?
[ No ]
AB: the next Widgets call will be July 30
MS: XHTML2 WG charter will not be
renewed and consequently, all of the *HTML work will be done in
the HTML WG
... that means I will spend 100% of my time on HTML, at least for the time being
... another Team Contact will be appointed for Widgets
... by the next call expect the new TC for Widgets to join
AB: thanks for your good work and
good luck in your new work!
... anything else?
[ No ]
AB: Meeting Adjourned
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/if this is about failure modes/This is about failure modes/ Succeeded: s/that is difficult because the impls are shipping and difficult to change/that would be a problem because you'd have to have widely shipped implementations and widely used content to satisfy your criteria, at which point we can't change the spec if it's broken anyway/ Succeeded: s/reda/read/ Succeeded: s/AB;/AB:/ Succeeded: s/AB: I can make the changes/RB: I can make the changes/ Found ScribeNick: ArtB Found Scribe: Art Present: Art Josh Robin Mike Marcos Henri Benoit AndyB Regrets: Marcin Agenda: http://lists.w3.org/Archives/Public/public-webapps/2009JulSep/0144.html Found Date: 09 Jul 2009 Guessing minutes URL: http://www.w3.org/2009/07/09-wam-minutes.html People with action items: an art barstow email marcos respond send smith soref with work[End of scribe.perl diagnostic output]