SVG Working Group Teleconference -- 11 Nov 2010

<trackbot> Date: 11 November 2010

<jwatt> I would if I could figure out how to get the number pad to come up in the new version of skype

<jwatt> for anyone else hitting that problem in future, you need to hide your sidebar to get the number pad to show up

<jwatt> crazy

<anthony> Scribe: anthony

<scribe> ScribeNick: anthony

Telcon Time

CM: Ed we did we resolve what was discussed in terms of telcon time?

ED: What we discussed at TPAC meeting was to have the telcon one hour earlier than before
... as I understood it that was one hour before the actual time
... and not the shifted time
... so essentially it means 20:00 UTC

CM: 2 hours ago was my translated time
... but it's actually 1 hour ago

AG: So 7am for me and 9am for CM

PD: So 9pm for you ED?

ED: It's 10pm here now

PD: Noon for PST
... that does not apply to the task force correct?

ED: As far as I know the task force has not changed the time
... I think that is 20:00 UTC
... I'll double check
... Yes task force 20:00 UTC

CM: That's the time we want to move to for our teclon time right?
... this means that the task force time is at the same time?

ED: Actually I think, 20:00 UTC is the time we have now in the telcon system
... for SVG teclons
... check that now

<ed> http://www.w3.org/Guide/1998/08/teleconference-calendar.html#s_2216

ED: I'm happy with we have now
... so we shouldn't change anything
... as in starting 1 hour ago

AG: That's fine with me
... PD is that fine with you?

PD: That's great

CM: And you also discussed having single telcon a week?

ED: I think everyone at the meeting was ok with that
... and we decided to go with Thursdays

RESOLUTION: We will have 1 telcon per week at 20:00 UTC on Thursdays

Little TPAC Summary

CM: I did skim through the minutes
... but I just wanted to get some broad ideas on what was decided about plans and directions

ED: What I heard we were aiming for some what stable specs for June 2011
... those specs will include public fx Transforms 2D/3D, Filters spec (that apply to HTML), join Animation model spec
... the SVG Integration spec
... and as a lower priority the Advanced Gradients

CM: At what sort of level by June?

ED: Not sure we decided on, but I heard somewhat stable
... but it suggests and agressive schedule for getting drafts out
... I think we can probably have a couple of sepcs ready by then. Not sure if they'll all be at the same level

CM: Might be duable if we all put an effort in

PD: Two additional items
... We were going to put Erik on the SVG DOM
... And thinking about simplifying the DOM, having getters and setters and simpler list API
... only other thing was we identified owners
... Transforms is and has been driven by Anthony
... Animation was going to be driven by Dean
... Filters was going to be driven by Robert
... Advanced Gradients sounded like a partnership by Tav and Anthony

CM: Robert?

PD: ROC

ED: I think someone from Apple said that they were happy to be an editor of Filters
... I'm pretty sure I have an action relating to that
... can't remember off hand who it was

PD: The only other thing we said was that, to do two things around testing
... as we look to make tests, we move to make tests for the new W3C testing harness
... and we deliver tests and specs as much as we could at the same time
... and Doug was going to lead investigation into doing crowed sourcing for test development

CM: Obviously we are not going to convert all our existing tests to that?

PD: No

ED: I think it's probably best to go with the current test suite now
... but for future test development
... we should use the new harness

CM: Does it enable more automated testing like ref tests?

ED: Yes, and I think it makes sense to use ref tests where we can
... not everything can be script automated
... not sure if scripted tests and ref tests will cover everything
... but it will cover a big portion

CM: The drivers for these task force specs are they the editors?

PD: My understanding was that they are at least the owners, as in they are responsible to get it from point A to point B
... And both myself and anyone else should contribute to make that happen

Plan for 1.1 Test Suite

CM: Was 1.1 2nd Edition discussed at the F2F as well?

ED: What we discussed at TPAC was to try and finish the last call issues
... and close them before December 15th
... and around the same time
... we should have a somewhat stable charter document
... because it will take a few weeks for the AC review to happen

CM: That's mostly a task for Doug and Chris

ED: I know Doug has started work on it
... and he showed some work on it

CM: The plan for having thses last call issues by the 15th Dec
... was that we could publish the spec at the next maturity level as well?

ED: I think the plan is to ask for publication some time in december
... and have it move to recommendation some time in January
... if everything goes according to plan
... because the plan is to not have the 1.1 2nd Edition in the charter document
... just new things

CM: Test suite do you have a status of that?

ED: I've been doing updates and I've been working on it today
... generating reference images
... and fixing minor issues
... and I sent an email to the list
... with the issues
... some tests have some unclear or missing pass criteria
... some tests have missing revision numbers
... due to the way they were checked in
... missing 'ko' flag
... some tests are using
... red to indicate pass
... but not to concered about that
... So there are lot of animation tests that are missing written pass criteria which is bad

CM: Do we need to give someone an action to look at some of those

AG: Might have to triage those because there are alot of those

ED: Seems that there are about 30 or so

CM: Can you put that on a wiki page

<ed> http://lists.w3.org/Archives/Public/public-svg-wg/2010OctDec/0130.html

ED: Already emailed the list

AG: If we divide that up between us, could have that done in no time

CM: Time line for tests suite work?

ED: I think that it is ok to have it done by december if we all do the work
... I've run it from start to finish a few times
... and everytime I've run through it I've come across issues
... this list is not complete with all the issues
... so basically it's making sure it's ok for releasing

CM: Is that something you were going to continue on and do?

ED: I think so

CM: So things that need to be done are
... addressing pass criteria
... and fixing red
... and there are some tests which need to be approved by us

ED: I wasn't too concerned about those
... we could put more in
... or we can keep going with the ones we have already

CM: By adding to those?

ED: The ones that I've been reviewing

CM: Ideally we'd be able to approve or not the ones you've reviewed so far

ED: If you have a list of those we could go through those quickly

CM: Don't have the list at the moment

ED: Can we run the test suite status generation script again?

AG: Yes, I can run the test suite status again
... after the telcon, and email out

ED: Just make sure to run an update before you generate

CM: There are still some tests that heven't been reviewed
... I don't know at this point
... if we want to consider not reviewing those

ED: The sooner we close on the final list of tests
... the easier it will be
... we still need to run through the implementation status of each test

AG: Might be worth waiting for the status report

CM: There are a bunch that I reviewed
... of the Microsoft tests
... and some of them have questions or arguments and I've marked those as reviewed
... and there are still some that haven't been responded to

PD: Is that right?

CM: I can go back a check back through my email
... and if there are ones that have outstanding comments I'll email out

PD: I'll also look
... if we've missed anything that would also be great

SVG 2

CM: People are still bringing up issues that might apply to SVG 1.1.
... since we want to get the document out
... don't want to make too many drastic changes to it
... I want a place to address these issues
... that come up
... What is the current plan for SVG 2

ED: I think we touched or discussed topics
... but we didn't talk about it alot
... we are still on the plan on using the modules for parts of the spec

CM: When you were talking before about these core DOM changes
... it sounded more like stuff that effects SVG 2
... rather than a separate document

PD: The way I think about it is SVG 2 is going to be modulised like we talked about
... you're right about the DOM work
... we need to figure out where that goes
... and the SVG integration module
... these two are portion of the SVG 2
... might want to look at or add to or improve things
... which is part of the SVG 2 track

CM: In terms of what SVG 2 the document is going to include
... is say the integration document going to be part of the SVG 2 family of specifications
... or will that text go into the SVG 2 document itself
... I remember talking about a while ago about producing modules
... and bringing them together
... never been quite sure logistically where the text is going to go

ED: We do have the SVG 2 base document there in place
... it is possible to add stuff to it

<ed> http://dev.w3.org/SVG/profiles/2.0/

CM: The question is if I'm going to add some text
... where do I go to add it

PD: I think that for example, I think that Anthony wants me to look at SVG Compositing and maybe that's a module in itself
... should probably make a first pass at what the modules are
... so we don't have to revisit it

CM: We should decide what features are going to be part of SVG 2
... in the broad SVG 2 time frame and then from that and see which are modules
... that are going to be seperate at the moment
... and which ones are not
... and the features that are not modules
... will have to go into the document itself
... and we'll have to work out what the structure is like
... one thing that bugs me about SVG 1.1 it's wordy in some places
... where it doesn't need to be
... and not detailed enough
... and I wonder how much we want to use from the 1.1

AG: I remember Doug was saying use SVG 1.1 but mark it up as text that is unapproved
... then we review it
... so that it goes to a reviewed status

CM: If someone has time it would be good to go through and rewrite whole sections

<jwatt> scribenick: jwatt

<scribe> scribe: Jonathan Watt

rx/ry clamping on <rect>

<heycam> http://www.w3.org/mid/20101101023449.GK28301@wok.mcc.id.au

ED: I'm happy with the suggested wording and don't mind changing tests
... do we want to put it in 1.1 or not

CM: I'm happy to put it in 1.1
... it's a small clarification

ED: I think it's better to put it in than not

PD: if I don't come back to it, assume that I think it's okay
... the cost of changing for us is high right now though

<ed> http://dev.w3.org/SVG/profiles/1.1F2/test/svg/shapes-rect-03-t.svg

ED: if we make the proposed change, that test would need to be changed
... what Firefox is doing right now would then be the correct behavior I believe

PD: I'll comment on the list

gzip MIME-type

ED: people seemed unhappy with the changed wording
... do we want to have it as a supported format for SVG, or is it just intended for transfer?
... I think people generally set the HTTP header
... you tend to end up with .svgz files on your local machine, and it's actually useful
... I think most editors support it

CM: I think the preferred way is to have your server set up to send the appropriate headers, rather than having a separate MIME-type
... I'm not convinced the spec needs to say anything about gzip, when it doesn't say anything about other compression formats

PD: would we potentially change 1.1 for this?

ED: we won't change it if people really dislike the change

PD: we aren't going to support opening .svgz from the local file system in IE9

CM: making it a requirement would be a big change at this point

ED: okay, I'm fine with dropping this for now in that case
... I'd like to spec it out in 2.0 though

Actions and issues in Tracker

CM: I see there are many open actions in Tracker
... and we're probably at the point again when lots of it is now irrelevant, making Tracker less useful
... I think it would help if it got back to the stage when it's an accurate reflection of what has to be done, and when
... I think we need to add a component to the telcons for this

PD: does Tracker support cross-group collaboration

CM: no
... the fx task force are using tracker
... you can get a view of all the issues assigned to you across all Tracker instances, but that's about the limit of the cross-group integration
... "My Tracker"

PD: one thing that concerns me is that people in the CSS WG hadn't heard of the work Antony had done
... and in another case a group left at TPAC because they didn't know there was a meeting

CM: I'd imagine you have reps from groups out in other groups reporting back
... to their group
... which is probably the way to solve this
... minute emails to the lists have a good summary of the actions and resolutions at the top
... maybe an email with just that and a link to the full minutes should go out to other groups lists

ED: fantasai does send out summaries via twitter
... I think pulling out the resolutions and putting them at the top would be an improvement

PD: we don't use resolutions enough

ED: we need to make sure we act on them
... track them in the wiki?

CM: I think tracker has some crude tools for resolution tracking
... I think we should be clearer about desicions and making sure they get turned into resulutions

JW: I'm concerned we don't remember the details of what we talk about and decide, and the details get lost in minutes
... we should really have topic pages in the wiki where we summarize the important details from telcons
... so that we don't have to waste so much time going over the same topics because we forgot the details of when we discussed the topic last time
... or if not a summary, at least topic pages where we add links to relevant minutes

PD: shame our wiki doesn't work very well

ED: I talked to Peter Linss (css wg co-chair) after TPAC who said we might be able to share resources

PD: I'd be willing to look into resources
... I'd like a solution that pushes out info, like to my email box
... and provides good query tools

CM: I worry that we may just swap one set of problems for another, and take up a lot of time switching
... and that changing wouldn't solve the problems

trackbot: end telcon

- DRAFT -

SVG Working Group Teleconference

11 Nov 2010

Attendees

Contents