Weekly Forms WG Teleconference -- 14 Jan 2009

<John_Boyer> Scribe: Leigh

<John_Boyer> scribenick: klotz

<markbirbeck> previous call overrunning...

Thank you Paul

<John_Boyer> Scribe: Paul

<John_Boyer> scribenick: prb

Upcoming Telecons

Steven: chairing next week

John: should be just same as current agenda, sans anything done
... precede important topics with *

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2009Jan/0017.html

next FtF

John: travel difficult for some, worth raising subject of virtualisation
... difficult to get more than 8 hours in a day, for all timezones
... virtual FtF should be split, 2 days in one week, then 2 day s in next
... whichever FtF this is done for, if any,

Charlie: understands travel difficulties
... pessimistic about approval

klotz: budget OK for one trip this year

<ebruchez> same here, the Mountain View meeting doesn't imply any travel for me

<Steven> I can make the Feb meeting

<Steven> Nick, please phone in again

nick: (problems on line) already have tickets for Google, Amsterdam is close, less problems

John: easier question is: do we want Amsterdam as virtual, or should we proceed with actual face to face

<Steven> (Me neither ;-) )

<nick> same for me Amsterdam isn't a problem for me

Mark: I abstain, as Amsterdam is easy for me, so people with more issues should say

klotz: If we decide now that we have virtual ftf, and it works, we may never meet again
... we should postpone the decision until summer, don't see what giving up buys us

Mark: when we talked about doing it London, I was planning other things, such as XForms Day School, around it, as we have all world's XForms experts in one place at a time, we can do that sort of thing

Klotz: that's not just location, but planning,

mark: another way to look at it is: how many can say now that they can go?
... all europeans can go

Steven: moved to amsterdam as I couldn't get funding to go to london, but if Mark organises other things, I can get funding from other sources to give presentations
... not much difference between Amsterdam and London, if other things were happening, then I can get to London, and happy to do so

John: interesting, as it amounts to Charlie organising an IBM location, and Mark organising other things

Mark: roughly, I'd do the running around and evaluating places and dinner etc.
... IBM have big centre near the river near my office, so I can check it out in the weeks beforehand , and I'm open to doing so

John: we don't know if Charlie will get approval to go yet

Charlie: I can coordinate rooms etc.

Mark: I pictured a Ubiquity afternoon, or similar, around the WG meeting

Charlie: if john and I can get approval for one trip, we need to consider whether London or San Jose

John: I don't have approval yet, but I didn't get it last time, and I was cancelled before, and I have outstanding airline credit enough for San Jose, but not Europe
... more likely I'll get approval, but I don't know what happens in June

Mark: didn't mean to divert conversation from next FtF, just following Leigh's statement

John: assuming most people can make a live meeting ...

Klotz: Amsterdam is good, as the PTB are used to seeing it,

Mark: maybe we can do extras in Amsterdam, did you mention about CWI possibilities

Steven: w3c NL is based here, and the new man there may wish to make his mark

John: do we want to leave it in Amsterdam

klotz: I suggest we agree some parameters now,

Charlie: I can't book until three months before,

klotz: revisit in early march, do we need to gather any new information by then?

charlie: what is economic situation?

John: the only people who will have real issues, are charlie and I

Klotz: actually, I have approval, but it could be taken away

John: I can't think of anything else we'd be waiting on

<unl> @john: so lets a ftf in victoria ;-)

Charlie: consensus sounds like go ahead with Google then decide in early march about European one

XForms 1.1 situation

<scribe> ACTION: John to publish new firefox implementation reports from Keith [recorded in http://www.w3.org/2009/01/14-forms-minutes.html#action01]

<trackbot> Sorry, amibiguous username (more than one match) - John

<trackbot> Try using a different identifier, such as family name or username (eg. jkugelma, jboyer)

Keith: Firefox 3 they have done a bit of work in, and official release is behind development cycle, it was hard to get development version

I got official release, and tested the various failures, and corrected some of the issues in Chapter 10, and submitted back to group, >80% now pass

wellsk: It was part of triage

John: maybe someone can take a copy of our so called loan form and make it work
... it brings up issues of test suite maintenance

wellsk: I want to see if anyone was able to take over responsibility of test suite in the WG

John: that's important, as we go on, there may be tweaks to the suite

wellsk: I don't mind taking responsibility until someone else can take over

john: are we looking for someone to take over after 1.1

wellsk: sooner if possible, but if I have to take 1.1 to PR, I don't mind. The level of commitment is not too great

John: there may be some cases that rely on some features that are difficult to implement in some processors, and this is the first time we are exercising such a large test suite.
... for example. There are some issues that rely on replace="all", and in ubiquity, the best we can do is to defer to the browser,

klotz: is the issue getting the replace done, or getting the submit-done event
... if it's simply getting the submit-done, we can make a case for that
... if it is the whole package of this kind of submission, that's a different story
... If it's just the event, then we could credibly change the spec

John: which would ripple into the test suite.
... there are tests that are not testing replace=all, but use it to test something else in some way
... If we make a change to the test, we need to ensure that it is in a way that makes it more likely that a test that has already passed will still pass
... as the report from Uli showed us, we had 2 implementation reports roll in that had failures that showed the tests to be the problem
... test suite maintenance after XForms 1.1 will certainly be an issue
... is there anyone on the call that might be able to take that responsibility. It will come hand in hand with implementation report maintenance
... what is the list of tests that only has one implementation that passes now

wellsk: I'll take that on and see what I can come up with

<scribe> ACTION: wellsk produce a report of tests that are only passed by one implementation [recorded in http://www.w3.org/2009/01/14-forms-minutes.html#action02]

<trackbot> Sorry, couldn't find user - wellsk

John: we need at least one more report, because we have a number of test that are only passed by one implementation in the two reports we have so far
... possibly some people on the phone could contribute to Ubiquity, which would in turn require permissions from your managers
... an implementation report from ubiquity is a few months out, and we're having a long delay here

wellsk: could we use the current status of ubiquity, and compare it to the other 2, would that be legitimate?
... I say this because of work I did with Markmc, that has it on a wiki page, we could use it as a start point

John: whichever ones pass now, could be counted against the feature list where we don't have two passing implmentations, but our conformance level is below 50% right now
... in terms of keeping score, that would be safe, but we need to discuss it, as it is a signal that people should look at it, we can probably discuss it on ubiquity call tomorrow
... are you OK looking at that for now to see where we need further implementation reports outside ubiquity

wellsk: fails won't be counted at all at this point,

John: it may be hard to track, as this will be changing over the coming weeks
... you are producing a list of tests with less than two passes, and ubiquity as is can be used to compile that list

wellsk: I'm ok with that

John: is there a way to get a report from chiba

nick: I don't think I'll be able to get a test report

markbirbeck: it's difficult for us to make both a formsPlayer report, and one for ubiquity
... if we can use this list, we can look at features and produce a partial report that fills some of the gaps

<wiecha> or it could drive the priorities in doing the ubiquity items too

john: good. there have been some reports with pass/fail/unknown
... we can say that formsPlayer , as a long standing implmentation could help out, without committing to running all the tests

wiecha: this will help drive stategy in ubiquity

John: is that feasible for you mark

markbirbeck: is there a list already, or one in production

john: he is producing two lists, one being the most urgent, with no implementations, there are about 30 of those, I think
... A starting report, where formsPlayer passes any of those tests, that would be most important

mark: then we need one 100% conformant implementation?

John: actually, we got rid of that requirement

<w3cllsk> triage, 2 failures list : http://lists.w3.org/Archives/Public/public-forms/2009Jan/0012.html

John: any "MUST" feature must have two implementations, then any other, just one
... Keith has one where neither Firefox nor EMC pass already
... then he'll produce one where there is only one out of EMC, Firefox, or ubiquity that implement them
... for example, mailto and file submissions, which ubiquity probably won't pass anyway
... if this is a bad idea, speak up, the implementation report for ubiquity under IE7 is different to FF3. An implementation is browser+processor

<wiecha> you could add webkit to that list too

John: the world looks at us for cross-browser compliance, I'd like to test the waters, by submitting two implementation reports for ubiquity
... it will have a big impact on xforms for html implementation

wiecha: three ways, for the webkit implementation

john: yes, when we have them for other browsers, submit them

wiecha: we have that now

John: but not an implementation report
... "more than one" is what I'm getting at, and if anyone objects, I'll provide these arguments, as it does reflect the spirit of the w3c

charlie: just making sure that any discussion includes the full set
... we spend a fair amount in the team, dealing with this sort of thing
... testing on different browsers

John: I can say right now, we do have a limited report, but we do have deviations between the two browsers we have reports for, and things like repeat don't work on Safari
... Rounding off, we need to get 1.1 out of the door, it is a classic .1, being 50% bigger than .0
... Can we get a limited Orbeon implementation report? Can you take that to management Erik

Erik: not much has changed, we are very busy, and it's difficult to get this into a plan, we can try to see if any of our user community might be able to help with all or part of the report, but we haven't yet put that call out.
... but other than that, it is likely to get pushed further into the future

John: Great suggestion perhaps we can point them to Keith's work
... The thing you have just suggested, Erik. Nick, could you take that back to Joren

Nick: We can focus on the tests that others fail

Keith: Is Firefox 2 a different implementation to FF3

John: I haven't thought so up to now. It's one thing to say we have an implmentation running on two completely separate browsers, but different versions of the same browser, (arbitrarily, I know) is different.
... trying to spin different versions of the same branded browser, I can see objections

Keith: I just thought the question needed to be asked, I agree with you

John: I couldn't defend it, but I could defend totally different browser support

EMC joining Forms WG

Steven: I have a "done" action item, to contact EMC, to invite them to the group, the sender says it's difficult, because of lawyers to get approval, but give me time, I think it would be good to do it

John: that's great news

Steven: they didn't create a report, just have an implementation?

John: they did send us a report

Relax schemas for XForms

Steven: the other thing was that XHTML2 WG was asking for RelaxNG XForms schemata if anyone has them

klotz: there were the ones that Micah did, and I had a go at modularising them

Steven: I think there is an expert who wants to modularise XHTML RelaxNG

klotz: Nick and I have separate copies
... shall I send it to you

Steven: Shane is the best person to send it to

John: under Action Items on the Agenda, there was a problem reported on copy and delete that we needed analysis done on

klotz: I'll continue to promis to look at it. I'll try in the next two weeks

Actions

John: It would be a good idea to make a quick pass at the action item list. There have been a number done by a number of people, but haven't been reported
... I'm looking at the latest list. Starting at Mark's list - have you looked yet?

Mark: not yet, I meant to do so this morning

Steven: can you paste a link, John?

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2009Jan/0023.html

<Steven> for the minutes

Mark, can you lookk now, and paste into IRC any that are no longer relevant or are done

<Steven> http://lists.w3.org/Archives/Public/public-forms/2009Jan/att-0023/actions-2009-01-12.html

<wiecha> This is done: Charlie Wiecha to review http://www.w3.org/TR/2007/WD-xml-events-20070216/ and present straw set of comments for discussion and approval at F2F.

Mine are up to date, so skipping through to Erik, none look like they are top priority

<wiecha> This is done: Charlie to respond to http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0025.html

John: Paul I think you've done 1 and 3

<wiecha> This is done but we might recombine depending on modularization discussion: Charlie Wiecha to refactor http://lists.w3.org/Archives/Public/public-forms/2008Jul/att-0001/index-all.html a data island and the setvalue/insert/delete.

<wiecha> [oops] Paul -- your substitute probably got targeted at my URL,...sorry

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2009Jan/0023.html

<wiecha> my other actions related to work underway for current modules

John: Is Uli still on the line?

<markbirbeck> I would say that Action 2006-03-02.5 is no longer relevant.

Uli: I'm just looking on my list, something is quite old, "context everywhere"

John: I think that's one's done

<markbirbeck> I think that Action 2007-05-02.4 was eventually completed by Paul.

Uli: I had a look at the Editor's Draft, but I can't see it

<klotz> Action 2008-08-20.1 is done; I integrated it into the editor's draft and sent the results to John Boyer; I don't know if it was merged in

<trackbot> Sorry, couldn't find user - 2008-08-20.1

<Steven> I have done many of the actions, I'll give myself an action to tell Nick which ones

John: I think we decided against it in 1.1, it's 1.2 and ended up going into the binding attributes module, so it's no longer an action item

<klotz> I completed Action 2008-07-23.4

<klotz> I completed Action 2008-01-09.4

John: something about preparing a discussion about submission module. That has slipped by on the agenda, is that something that can be done for next week
... hopefully, steven, that can be on next weeks agenda, otherwise, Uli, send a reminder

<klotz> I completed Action 2007-05-16.3

Uli: I think we had discussion about xforms-serialize event,

John: perhaps you can post to the email list, and get some discussion going there

<klotz> Kenneth and John completed Action 2007-03-07.1

<markbirbeck> I've not been doing anything special with the XML Schemas for XForms, so I think that Action 2008-04-09.1, Action 2008-02-28.2 and Action 2007-06-13.1 are no longer relevant.

Steven: I've given myself an action item to tick off those that I have actually done

John: I did some triage on this yesterday, there are some shorter action item lists for some of the people not coming regularly, but they are not high priority

<John_Boyer> Two from Kenneth have been done.

John: there are two from Kenneth that have been done,

<markbirbeck> I also don't think that Action 2007-09-26.1 is relevant anymore.

John: Nick, as maintainer, I presume that your list is up to date

Leigh: I have gotten one again today, is to seek expert help on the RelaxNG schema

John: one last easy topic - specifications with multiple modules,
... last week, there was some discussion around Nick's bind module, and it seemed that it might be easier to modularise if we don't go the whole hog on cutting it down into one spec per module straight away
... perhaps some of us could pool our editorial resources and combine MIPS, binding attributes - it would be easier for Nick to talk about custom MIPS if he didn't have to reference different specs
... on Charlie's side, there's a lower level module, and a higher data processing module with delete etc in it, and it would be easier to refer to them within the same spec - how do people feel?

Wiecha: That makes pragmatic sense

Leigh: How is it different to what we are doing

<Steven> +1

John: We can discuss on the list if this is the wrong direction, Nick and I can exchange emails on collaboration,

Nick: I'm leaving on vacation next week, so there will be delay

John: I have to be at LotusSphere next week, but after that, perhaps we'll have greater velocity

- DRAFT -

Weekly Forms WG Teleconference

14 Jan 2009

Attendees

Contents