SVG Working Group Teleconference -- 19 Jan 2009

<trackbot> Date: 19 January 2009

<scribe> scribe: erik

<scribe> scribeNick: ed

ISSUE-2201

ISSUE-2201+

ISSUE-2201?

<trackbot> ISSUE-2201 -- Return value of SVGAnimationElement.getStartTime unclear -- OPEN

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/issues/2201

<heycam> http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#getStartTime-return-value

CMC: wasn't sure if we decided if we return a value for the interval if it wasn't active at the moment (so the next interval)

ED: just wondering which is more intuitive to authors

AG: what does batik do?

CMC: think something slightly different, actually returns NaN
... Opera currently will do the current interval whether if it's active or not
... if we returned the start time of the current interval if they're not active scripts could still compare that against the document time to determine if that's active or not
... don't mind either way

AG: do we want to wait for jwatt to have his opinion?

CMC: brian brought up the discussion originally
... he also asked on svg-developers if anyone used this, and what they thought
... no replies yet

ED: it may be that sometimes it's tricky to time the call such that the getStartTime is when the interval is active, and you might expect to get something instead of an exception if the next interval is computed

CMC: what about the duration one?

ED: that's getSimpleDuration, so the name says pretty much what it does

CMC: difficult do decide given that there's other timing things that are not exposed
... given that we can expose a bit more information by having it return the current interval and that you can compare against the documenttime, and that you can call it on beginevent, it should be safe to du the "current interval"

AG: i'm fine with that, opera would have to change?

ED: yes, to raise the exception

<scribe> ACTION: heycam to change the http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#getStartTime-return-value erratum to use "current interval" and to move the item to proposed [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action01]

<trackbot> Created ACTION-2407 - Change the http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#getStartTime-return-value erratum to use \"current interval\" and to move the item to proposed [on Cameron McCormack - due 2009-01-26].

ISSUE-2199

ISSUE-2199?

<trackbot> ISSUE-2199 -- Underlying transform value erratum now conflicts with 1.2T -- RAISED

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/issues/2199

<heycam> http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#cleanup-wording-underlying-value

CMC: i've updated this to bring in the text in 1.2T
... copied the text and changed some links from SMIL to SMIL Animation
... I've brought it back to draft to review it again

AG: doesn't look like a big behavioural change

CMC: it's not really explicit in the text, but this text implies that when you're doing the scale transofrmation animation and the scale is zero (like in a to animation and you need to know the starting value)
... it's not explicit in tiny 1.2 either anymore
... we used to have a table
... but was removed before moving to CR
... based on comments from DOH

AG: wasn't much in the spec before anyway

CMC: yes, this was rather undefined
... no tests in the 1.1 testsuite that rely on the underlaying value

AG: I've no problem changing that to proposed

ED: one thing I was wondering about was if the references affects this at all

CMC: the first link there is just describing the from-to calculation

ED: right, this is probably close enough
... and good to use similar wording

CMC: how do we do tests for errata items?

AG: we should add tests in the 1.1F2/tests/svg directory going forward

CMC: we'll have to release the testsuite again, right?

AG: right

CMC: so maybe we could publish the second edition spec along with a new snapshot of the testsuite

<heycam> http://www.w3.org/2003/01/test-cases/text-intro-06-t.svg

ED: that sounds good yes

CMC: the links should point to the right place though, they don't currently
... we should move the old tests over

<scribe> ACTION: heycam to create a test for http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#cleanup-wording-underlying-value (check if we have a test for it in the 1.2T testsuite) and then move the erratum to proposed [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action02]

<trackbot> Created ACTION-2408 - Create a test for http://dev.w3.org/SVG/profiles/1.1F2/errata/errata.xml#cleanup-wording-underlying-value (check if we have a test for it in the 1.2T testsuite) and then move the erratum to proposed [on Cameron McCormack - due 2009-01-26].

ISSUE-2103?

<trackbot> ISSUE-2103 -- i18n comment 3: Inline -progression typo -- CLOSED

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/issues/2103

ISSUE-2203

ISSUE-2203?

<trackbot> ISSUE-2203 -- color-profile "local" IDL attribute erratum incorrect -- RAISED

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/issues/2203

<heycam> http://www.w3.org/mid/20090116032239.GB3580@arc.mcc.id.au

CMC: this was raised by me a while ago
... I think i was incorrect in raising it
... it was for the IDL and there's an attribute named "_local"
... and "local" is a keyword in IDL
... so the underscore is actually needed
... like the comment said

ED: ok, let's remove the errata item

CMC: yes, that's probably safe to do

<scribe> ACTION: AG to remove the color-profile "local" IDL attribute erratum for SVG 1.1 (addressing ISSUE-2203) [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action03]

<trackbot> Created ACTION-2409 - Remove the color-profile \"local\" IDL attribute erratum for SVG 1.1 (addressing ISSUE-2203) [on Anthony Grasso - due 2009-01-26].

www-svg feedback

<heycam> http://lists.w3.org/Archives/Public/www-svg/2009Jan/0024.html

CMC: for these methods what happens if you call them before the timeline is started

ED: so I think I was able to answer his first question about animationsPaused
... though we could possibly clarify it

CMC: for beginElementAt
... we should add the offset to time zero I think

AG: yes, makes it less confusing I think

CMC: would be consistent to do it that way I think
... my suggestion would be if you did "beginElementAt(5)" it would be five seconds after the document timeline started, not from when it was called if it was called before the timeline started

ED: that would probably be the most intuitive yes

CMC: and then we'd get the timelineBegin from 1.2T if you wanted to start the timeline earlier

http://lists.w3.org/Archives/Public/www-svg/2009Jan/0029.html

CMC: it's a bit weird to fix up the starttime to be some negative time
... that's jwatts suggestion to run the animation before the timeline has started

ED: the ElementTimeControl API is slightly different in 1.2T, something to note
... IMHO it's a bit strange to not look at 1.2T when implementing SVG SMIL support from scratch

<scribe> ACTION: heycam to reply to http://lists.w3.org/Archives/Public/www-svg/2009Jan/0024.html to say what we decided and that we've made erratum for it [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action04]

<trackbot> Created ACTION-2410 - Reply to http://lists.w3.org/Archives/Public/www-svg/2009Jan/0024.html to say what we decided and that we've made erratum for it [on Cameron McCormack - due 2009-01-26].

<heycam> http://lists.w3.org/Archives/Public/www-svg/2009Jan/0026.html

animate-elem-82-t query, http://lists.w3.org/Archives/Public/www-svg/2009Jan/0026.html

CMC: brian points out that this test was copied over from 1.1, but it relied on the behaviour that we changed in 1.2T based on DOH's comments

AG: this test has always had issues related to it
... was it even approved?

CMC: does say so
... presumably we've changed the definition of paced to something we're happy with in tiny
... and this test should be removed and/or changed from the tiny 1.2 testsuite at least
... but should we also look at changing what's in 1.1?
... or just leave it
... opera does the 1.2T way right?

ED: since this issue went back and forth a number of times I'm not sure if we decided to hold off on changing anything

CMC: this test in the 1.2T testsuite, we should eitehr change it or remove it
... only some subtests rely on the pacing, don't know which

<scribe> ACTION: review 1.2T/tests/animate-elem-82-t.svg and make sure the positions are correct for the paced animations, and move the test back to draft/unapproved for now [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action05]

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

<scribe> ACTION: ED to review 1.2T/tests/animate-elem-82-t.svg and make sure the positions are correct for the paced animations, and move the test back to draft/unapproved for now [recorded in http://www.w3.org/2009/01/19-svg-minutes.html#action06]

<trackbot> Created ACTION-2411 - Review 1.2T/tests/animate-elem-82-t.svg and make sure the positions are correct for the paced animations, and move the test back to draft/unapproved for now [on Erik Dahlström - due 2009-01-26].

testsuite

AG: haven't started writing the xslt stylesheet yet

ACTION-2385

AG: most things can be copied across, but some empty fields need to be filled out

CMC: so once that's done the testsuite will be in good shape for adding tests to it

AG: you can do that now anyway, we can make the xsl look for certain elements and only change tose

CMC: we haven't copied over the tests yet have we?

AG: we did
... but they're still using the old template
... i've been organizing the last bits of the f2f, so I'll have more time this week to look at it
... if anyone wants to make a new test just use the new template

<anthony> http://dev.w3.org/SVG/profiles/1.1F2/test/templates/

AG: there are some new fields
... the testdescription field allows references to the spec to be put in
... and which parts of the spec is being tested
... and because multiple parts can be tested you can put multiple references
... tehre's a field for pass criteria
... the approved attribute has been removed, it's only using status now

CMC: does testdescription need to be there?

AG: probably not
... will give an example

<anthony> http://dev.w3.org/SVG/profiles/1.1F2/test/svg/text-tspan-02-b.svg

ED: what about future specs, we might not have a full absolute link?

AG: good point
... we could make an xslt for publication

ED: we already put some spec links into the harness, based on naming convention

AG: that's based on sectionnames, so with xlink:href you can point to a specific place in the spec

- DRAFT -

SVG Working Group Teleconference

19 Jan 2009

Attendees

Contents