SVG Working Group Teleconference -- 28 Apr 2011

<trackbot> Date: 28 April 2011

<tbah> +33 is me

<heycam> Scribe: Vincent

<heycam> ScribeNick: vhardy

heycam: Introductions for Rik and Vincent.

cabanier: Our group is mostly interested in the SVG Compositing and SVG Filters spec. and make it ready for use in CSS. The SVG Compositing spec. is the highest on our list.

heycam: later on we will go over the LC comments for the SVG Compositing spec.

vhardy: I am interested in the SVG work in general and like Rik, in the compositing and filter efforts. I am also interested in what is happening in the joint CSS/SVG effects task force.
... I am working on the CSS Regions effort.

heycam: this is related to prior work the SVG working group has done in that area.
... are you going to join the SVG WG meeting?

cabanier: I am trying to get approval for the trip. It is in process.
... hopefully I will be able to attend.

vhardy: likewise.

heycam: we would like that Patrick for Microsoft and Dean from Apple be able to join. However, they have the Apple WWDC conf. in June at the same time. I am not sure about Patrick's ability to attend. We are considering moving the meeting to the Bay area.
... so it is still undecided. Doug is trying to find out what can be done.
... I will send an email to try and find out.
... First agenda topic: 1.1 second edition progress
... I have commited the suggested changes.
... that's for ACTION-2834
... I have made the changes for Brian's animation issues.

ed: I want to mention that I have made minor changes to the pointer-events section.

<ed_tmp> ed: both related to ACTION-2671 and ISSUE-2364

heycam: the animation issues was from Brian Birtles (Mozilla).
... the deadline for PR publication request is May 6th. I am concerned about Chris Lilley's action (ACTION-2910)
... is it feasible to try and grap that action from him?

ed: Chris has done some of it. We could try to get it and then finish the work.

heycam: I sent him an email a while back but did not hear back from him.
... I will send him an email reminder with the desired publication date.

ed: If we could get him to check in what he has done so far, we could take it from there.

anthony: is there a lot of work left?

heycam: not a lot.
... I could complete the work if Chris commits his work.

<scribe> ACTION: heycam to contact Chris about commiting progress on ACTION-2910. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action01]

<trackbot> Created ACTION-3026 - Contact Chris about commiting progress on ACTION-2910. [on Cameron McCormack - due 2011-05-05].

<heycam> http://dev.w3.org/SVG/profiles/1.1F2/test/status/implementation_matrix.html

heycam: at this point, there is some more work to be done on the final bits of the test suite and the implementation report.
... lets look at the implementation report

ed: we are waiting on getting some font resources from Chris.

<ed_tmp> text-dom-04 / 05

ed: that applies to the text-dom-04 & 05 tests.
... there was a suggestion to split fonts-desc-04 in two tests.

heycam: do we have an action for that?

ed: the wiki page does not have an action list.
... checking on tracker.

heycam: I think we had a plan and actions to address all these remaining tests.

<heycam> ACTION-2962?

<trackbot> ACTION-2962 -- Chris Lilley to convert fonts-desc-04-t. and 05-t to use WOFF -- due 2011-03-06 -- OPEN

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/actions/2962

<ed_tmp> ACTION-2965?

<trackbot> ACTION-2965 -- Chris Lilley to make a WOFF font for text-dom-04-f.html -- due 2011-03-06 -- OPEN

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/actions/2965

heycam: for the font tests, I thought that Chris was trying to get the WOF font sorted out first.
... we should try to resolve these issues before publication and drop the corresponding tests if we cannot.
... two implementations are needed to enter PR.

(discussion on whether the tests should be removed or if the spec. should be removed).

tbah: filters-light-03-f has one pass and one partial pass. I can get Inkscape to pass the test.

heycam: I am hopefull that Mozilla could pass. We have a patch but it has not landed yet (not in the public builds).

ed: I have seen it pass in mozilla dev builds (at the last f2f). I am not sure it needs to be public.

heycam: this is reasonnable if we can explain it has been seen and that it will become available publicly in a reasonnable timeframe.
... we could also mark that it works in an internal build.

ed: yes, a disclaimer or note is fine, like we did for the Opera build.

<scribe> ACTION: heycam to try and get the Mozilla patch for filters-light-03-f to become publicly available or, if not possible, add a note to the implementation table. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action02]

<trackbot> Created ACTION-3027 - Try and get the Mozilla patch for filters-light-03-f to become publicly available or, if not possible, add a note to the implementation table. [on Cameron McCormack - due 2011-05-05].

heycam: painting-render-02 is ok, because we also tested it on ASV. So the two implementations are ASV and Abra.
... text-dom-04 and 05. The 05 is checking the SVG strings methods. I am surprised this does not pass.
... text-dom-04.

ed: I wrote it. This was checking SVG DOM methods for measuring text. It would be good to have WOF fonts for the test. Pending on Chris to check in the font.
... I could take the action from him to make it work.

heycam: you think it may work with a WOFF font instead of an SVG font?

ed: yes
... 04 is testing getSubstringLength with different values and checking exceptions.
... 05 is testing a couple different DOM methods, passing an index to retrieve a string.
... I will take the action from Chris to get a WOFF font for text-dom-04.
... I already reassigned ACTION-2965

<scribe> ACTION: heycam to investigate how to fix text-dom-05. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action03]

<trackbot> Created ACTION-3028 - Investigate how to fix text-dom-05. [on Cameron McCormack - due 2011-05-05].

heycam: there is no pass at the moment.
... there is a partial pass on Opera.

<ed_tmp> partial in ie9 too

heycam: I think this test could be dropped if we could not fix it.

ed: I would be ok to drop it too.

heycam: text-dom-05 tests edge cases.
... text-intro-02

ed: multiple implementations have problems on the 3rd line.
... it is pretty close to passing on WebKit

heycam: are you satisfied with the test.

ed: I would like to hear about what WebKit thinks of that test.

<scribe> ACTION: ed to contact the WebKit team about text-intro-02, probably Dirk. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action04]

<trackbot> Created ACTION-3029 - Contact the WebKit team about text-intro-02, probably Dirk. [on Erik Dahlström - due 2011-05-05].

<ed_tmp> Dirk Schulze

heycam: text-intro-09 is similar with a different font, WOFF font and system font.

ed: I'll mention both tests.

heycam: these BiDi tests are important to have.
... it would be good to keep them in the test suite.
... that's all for the test suite.

ed: there are a couple lines that a few 'unknowns'. Please update your implementation report if you can.
... It would make the table a bit nicer.

heycam: I'll make another run through it.

<scribe> ACTION: vhardy to style the implementation report. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action05]

<trackbot> Created ACTION-3030 - Style the implementation report. [on Vincent Hardy - due 2011-05-05].

heycam: In the worst case, if these test cases cannot be made to pass, are there objections to remove them from the report.

vhardy: is it an option to go through the transition call with tests that we know are correct and implementable but do not have a pass.

heycam: we have dropped a few tests already.
... the remaining ones, I think could be dropped without hurting the spec.
... ok, no objection to drop the tests that do not have two implementations if need be.

<scribe> ACTION: heycam to check with Doug and Chris about keeping a few tests that are implementable and correct but do not have two full passing implementations. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action06]

<trackbot> Created ACTION-3031 - Check with Doug and Chris about keeping a few tests that are implementable and correct but do not have two full passing implementations. [on Cameron McCormack - due 2011-05-05].

heycam: SVG Compositing last call comments.

anthony: the LC period for compositing ended on the 14th. So far, we have had two comments come in. One from Tab Atkins at Google and the other from Rik Cabanier.
... I think that Tab was ok with the responses on the email discussion. We did not give an official WG response.
... Rik recommended to split things apart. This would be a bit of work. I do not see any reason to take some drastic action.

heycam: Do you have a link to the discussion.

<anthony> http://lists.w3.org/Archives/Public/www-svg/2011Apr/0041.html

anthony: Tab suggested renaming the clip-to-self property to comp-op-region. I agree with that.
... we would keep the same values for the property.
... this was there to have compatibility for Java2D.

heycam: I am happy to have it renamed.
... which implementations do we have?

anthony: ASV3 had an implementation with a different syntax. I am pretty sure Abra implements this.
... Alex Danillo is ok with the name change.
... as long as the functionallity is correct, he does not have an issue with the name.
... regarding enable-background (another comment from Tab), he is asking to change the example. I agree for the second part.

heycam: yes, more examples would be good.

ed: regarding enable background, do we want to make the x/y/width/height optional?

anthony: is enable-background still in filters?

ed: it may get deprecated.

anthony: for the optional x/y/width/height, I do not have a strong feeling?

heycam: Is that an explicit optimization from the author?

anthony: yes.

heycam: so it makes sense to drop it then.

RESOLUTION: clip-to-self is renamed to comp-op-region in the SVG Compositing specification.
... drop the x/y/width/height parameters from the new value on enable-background in SVG Compositing specification.

cabanier: do we need inherit on enable-background?

heycam: it is needed for all non-inherited properties.

ed: what do we do if people specify x/y/width/height

anthony: it should be ignored.

cabanier: I think that even if the x/y/width/height are ignored, then it should not make any difference.

heycam: the parameters should be marked as ignored if present.

<scribe> ACTION: anthony to officially respond to Tab's comments on SVG Compositing LC. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action07]

<trackbot> Created ACTION-3032 - Officially respond to Tab's comments on SVG Compositing LC. [on Anthony Grasso - due 2011-05-05].

<scribe> ACTION: anthony to rename the clip-to-self property to comp-op-region. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action08]

<trackbot> Created ACTION-3033 - Rename the clip-to-self property to comp-op-region. [on Anthony Grasso - due 2011-05-05].

<scribe> ACTION: anthony to add a note to enable-background new value to say that the x/y/width/height values have no effect. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action09]

<trackbot> Created ACTION-3034 - Add a note to enable-background new value to say that the x/y/width/height values have no effect. [on Anthony Grasso - due 2011-05-05].

RESOLUTION: document that x/y/widht/height in the new value on enable-background have no effect.

<scribe> ACTION: vhardy to check with product team that we are not overlooking issues by dropping the x/y/width/height parameters on enable-background. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action10]

<trackbot> Created ACTION-3035 - Check with product team that we are not overlooking issues by dropping the x/y/width/height parameters on enable-background. [on Vincent Hardy - due 2011-05-05].

<anthony> http://lists.w3.org/Archives/Public/www-svg/2011Apr/0043.html

anthony: Rik sent a response say that we should split Porter Duff and the blend modes. This was in response to Alex's email.

cabanier: my concern is that Porter Duff blending is different from the other blending modes. I am affraid this may generate confusion. For example, the Porter Duff rules are not group invariant.
... If you put the content in a group, then it would render differently.

<cabanier> <circle><g><circle)(circle pd=clear)(/g) is not equal to <circle>(circle)(circle pd=clear)

anthony: yes, this is correct

cabanier: in the blending modes, then you do not have that behavior.
... if the background is accumulating.

anthony: I do not understand how this effects the blending. I thought the blending was just a mathematical operation.

cabanier: it has to do with how enable-background is implemented.
... the default of a group is to accumulate the background. This is why we get a different behavior.

anthony: I agree that knock-out is not very well defined. If Rik could provide better wording, that would be awesome.
... I can work at rewording the knock-out section to clarify the behavior.

cabanier: on the mailing list, I forwarded an article that describes what it does.

anthony: yes, I saw the link.

<scribe> ACTION: anthony to propose new wording for knock-out. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action11]

<trackbot> Created ACTION-3036 - Propose new wording for knock-out. [on Anthony Grasso - due 2011-05-05].

anthony: so the request is to separate the blending and the Porter Duff rules in two separate sections.

RESOLUTION: split the Porter Duff and blending modes into two separate sections in the document.

<scribe> ACTION: anthony to split the comp-op values into two sections, Porter Duff and Blending in Section 5 of the Last Call specification. [recorded in http://www.w3.org/2011/04/28-svg-minutes.html#action12]

<trackbot> Created ACTION-3037 - Split the comp-op values into two sections, Porter Duff and Blending in Section 5 of the Last Call specification. [on Anthony Grasso - due 2011-05-05].

anthony: Rik had another comment about extending the blend modes. We'll have to discuss this next time.

<heycam> trackbot, end telcon

- DRAFT -

SVG Working Group Teleconference

28 Apr 2011

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output