SVG Working Group Teleconference -- 21 Aug 2008

<trackbot> Date: 21 August 2008

<heycam> Meeting: SVG Nuremberg F2F Day 1

<heycam> Scribe: Cameron

<heycam> ScribeNick: heycam

Publication schedule

DS: It's important that we get the 1.2 Tiny spec to Rec by the end of the year
... because we have several standards bodies that have deps on this spec that need it to become Rec
... it's important for the mobile use
... i think it's important that SVG start focussing more on desktop browser uses soon
... in order to do this, we need to go to LC by the end of next month at the latest

ED: what do we need to get done for LC by the end of next month?
... the changes appendix will need to be updated

CM: a lot of work there
... what level of changes should we include there?

<ed> http://www.w3.org/2007/10/htmldiff

<ed> http://www.w3.org/TR/SVGMobile12/http://dev.w3.org/SVG/profiles/1.2T/publish/changes.html

<ed> http://dev.w3.org/SVG/profiles/1.2T/publish/changes.html

<ed> http://www.w3.org/Graphics/SVG/WG/track/products/2

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/

<ed> http://www.w3.org/2007/10/htmldiff?doc1=http%3A%2F%2Fwww.w3.org%2FTR%2FSVGMobile12%2Fintro.html&doc2=http%3A%2F%2Fdev.w3.org%2FSVG%2Fprofiles%2F1.2T%2Fpublish%2Fintro.html

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/1

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/15

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/258

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/268

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/274

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/issues/333

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1720

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1865

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1932

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1942

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1883

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1440

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1787

<shepazu> http://www.w3.org/Graphics/SVG/Group/track/actions/1904

<shepazu> http://lists.w3.org/Archives/Public/public-svg-wg/2008JulSep/0131.html

Action/issue triage

CM: All of the 1.2T relevant actions/issues have been ported across to the new tracker
... one i'm not sure about is andreas' http://www.w3.org/Graphics/SVG/Group/track/actions/1787

DS: we could e-mail him

CM: probably best
... it's an action to check to make sure edits have been finished, i think
... andreas hasn't replied to kalle yet (which is the second part of that action)

http://www.w3.org/Graphics/SVG/WG/track/products/2

http://www.w3.org/Graphics/SVG/WG/track/issues/2003

CM: action is done, so i'll close the issue

http://www.w3.org/Graphics/SVG/WG/track/issues/2014

CM: what is the wording in the spec about restricting the spaces?

<ed> http://www.w3.org/Graphics/SVG/Group/repository/spec/mobile/1.2/1.2NG/publish/linking.html#IntroFragmentsViews

<anthony> http://lists.w3.org/Archives/Public/www-svg/2008Jun/0023.html

<anthony> http://gbiv.com/protocols/uri/rfc/rfc3986.htmlhttp://gbiv.com/protocols/uri/rfc/rfc3986.html

<ed> Replace "Spaces shall not be allowed in fragment specifications and commas must be used to separate numeric values within an SVG view specification."

<scribe> ACTION: Anthony to allow spaces in viewbox fragids and to suggest that percent-encoded spaces be used if spaces are used, and to reply to kalle [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action01]

<trackbot> Created ACTION-2142 - Allow spaces in viewbox fragids and to suggest that percent-encoded spaces be used if spaces are used, and to reply to kalle [on Anthony Grasso - due 2008-08-28].

ACTION-2142?

<trackbot> ACTION-2142 -- Anthony Grasso to allow spaces in viewbox fragids and to suggest that percent-encoded spaces be used if spaces are used, and to reply to kalle -- due 2008-08-28 -- OPEN

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/actions/2142

trackbot, close ISSUE-2014

<trackbot> Sorry, heycam, I don't understand 'trackbot, close ISSUE-2014'. Please refer to http://www.w3.org/2005/06/tracker/irc for help

http://www.w3.org/Graphics/SVG/WG/track/issues/2021

We decide to punt ISSUE-2021 to SVG Core 2.0.

ISSUE-2021: ASV returns (0,0,400,400) for the test case.

<trackbot> ISSUE-2021 Bounding box of <image> subject to aspect-ratio preservation undefined notes added

http://www.w3.org/Graphics/SVG/WG/track/issues/2022

We decide to move ISSUE-2021 to Core.

CM: i don't think it should be unsupported or in error
... it should be valid, and then the processing due to eRR just happens if the resource is 404 or whatever

ED: eRR would mean that the rendering would stop there

<ed> so if you have an svg with three images in it, the svg has eRR="true" and then the second image is 404

<ed> then no image would be visible because of eRR

<ed> it's not in error, and the IRI isn't unsupported

<ed> it's just an effect of eRR handling

DS: unsupported values should be shown in the error console
... we can have a suggestion in the spec to say that if the UA has some notion of an error console or log, that it reports unsupported values and how they were remedied (the lacuna value used, etc.)

<scribe> ACTION: Doug to add the suggestion to report unsupported values in the error console [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action02]

<trackbot> Created ACTION-2143 - Add the suggestion to report unsupported values in the error console [on Doug Schepers - due 2008-08-28].

DS: in Linking it says "a circular references represents an error", i don't like this

we decide to leave that to be an error

<ed> http://dev.w3.org/SVG/profiles/1.2T/publish/struct.html#ExternalResourcesRequiredAttribute

<ed> Indicates that resources external to the current document are required. If an external resource is not available (for example the request for the required resource times out), progressive rendering is suspended, the load event is not fired for the element, and the document goes into an error state (see Error processing). The document remains in an error state until all required resources become available.

<ed> so...we should sync the wording in linking with this, saying "error state" but not "in error"

<shepazu> hi, aemmons!!

<ed> From linking: "When attribute externalResourcesRequired has been set to true on the referencing element or one of its ancestors, then an unresolved external IRI reference (i.e., a resource that cannot be located) shall represent an error (see Error processing)."

[We discuss the newly introduced fallback behaviour in the Linking chapter, and decide it's too much of a change to introduce this late, and we'll move it to Core.]

We need an action to make the following changes in this section:

- move the first dot point about circular IRI references to a paragraph of its own (so it doesn't just apply to invalid IRI references, but any circular references

<ed> Replace:

<ed> When attribute externalResourcesRequired has been set to true on the referencing element or one of its ancestors, then an unresolved external IRI reference (i.e., a resource that cannot be located) shall represent an error (see Error processing).

<ed> With:

<ed> Note that when the attribute externalResourcesRequired has been set to true on the referencing element or one of its ancestors, then an unresolved external IRI reference (i.e., a resource that cannot be located) will result in special handling, see [definition of eRR].

- Add an "invalid IRI reference" definition to intro.html

- Check that processing for invalid IRI references is defined for each attribute that can have an IRI reference

- Move the third bullet point to Core.

<scribe> ACTION: Cameron to rework the invalid IRI reference section (14.1.4) as in the minutes here [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action03]

<trackbot> Created ACTION-2144 - Rework the invalid IRI reference section (14.1.4) as in the minutes here [on Cameron McCormack - due 2008-08-28].

http://www.w3.org/Graphics/SVG/WG/track/issues/2027

ED: there were some files that referenced from <animation> elements something that had overflow
... i recall fixing some of those or all of those cases i found so that they contained the graphics within their area, so clipping wasn't an issue

AG: and when i investigated julien's email, i found that opera and bitflash did the same thing and produced the reference image we had
... so i suspect julien looked at the old version of the test suite

<ed> file://localhost/Users/ed2/w3.org/SVG/profiles/1.2T/publish/coords.html#EstablishingANewViewport

<scribe> ACTION: Anthony to reply to Julien to say that the tests have been fixed in the new release and that there's something in the spec that warns against this clipping issue (in that URL above) [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action04]

<trackbot> Created ACTION-2145 - Reply to Julien to say that the tests have been fixed in the new release and that there's something in the spec that warns against this clipping issue (in that URL above) [on Anthony Grasso - due 2008-08-28].

<scribe> ACTION: Cameron to take a quick look at 16.2.18 to see how to make it slightly better [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action05]

<trackbot> Created ACTION-2146 - Take a quick look at 16.2.18 to see how to make it slightly better [on Cameron McCormack - due 2008-08-28].

http://www.w3.org/Graphics/SVG/WG/track/issues/2029

<anthony> scribe: anthony

ED: Looks like everyone is in agreement to allow spaces

<scribe> ACTION: Erik to change stroke-dasharray to allow space separated values [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action06]

<trackbot> Created ACTION-2147 - Change stroke-dasharray to allow space separated values [on Erik Dahlström - due 2008-08-28].

ISSUE-2030

http://www.w3.org/Graphics/SVG/WG/track/issues/2030

ED: Dino said he had an action to clean up feature string
... Chris says that Mobile 1.1 had incorrect feature strings
... compared to 1.1 Full
... but does Tiny 1.2 have new feature strings?

CM: Dunno
... so in Tiny 1.2 does it have any of the 1.1?

ED: Tiny doesn't have any of 1.1 feature strings
... not sure if this was deliberate (Tiny 1.2 new ones)

CM: In the minutes from 2005 there was suggestion from Scott to change them and issue an errata for 1.1
... so this issue is now on Mobile 1.1
... and the decision was they we not going to change the 1.2 feature strings

<ed> http://www.w3.org/TR/SVGMobile/Tiny/feature

ED: SVG Tiny 1.1 feature strings there are 3 of them, I think

<ed> http://www.w3.org/TR/SVGMobile/Basic/feature

ED: Not sure if there is anything we need to fix for 1.1 Tiny or Full

CM: It does list 1.0 feature strings
... so it's a bit of a mess
... I guess we are not going to change anything for Tiny 1.2 either then

ED: What exactly is the issue about?
... seems to have been changed
... but not sure
... I think it could be a problem that there is no hasFeatureMethod Tiny
... I wouldn't mind seeing a new Issue raised to address the problem of missing hasFeature method

CM: What it says in the original mail doesn't seem to be problem anymore

<heycam> ScribeNick: anthony

RESOLUTION: We will close the ISSUE-2030 as it does not effect Tiny 1.2

ED: I'll raise a new issue about hasFeature to replace ISSUE-2030

ISSUE-2031

http://www.w3.org/Graphics/SVG/WG/track/issues/2031

CM: [Reviews minutes]

ED: Is this something that would work in Batik?

CM: Not implemented

<heycam> "When an element is not displayed (i.e., when the 'display' property on that element or one of its ancestors has a value of none), that element must not be the target of pointer events."

<heycam> -- 13.7

ED: I think this wording is a bit unclear

<ed> so, if the element has display=none then I don't expect any user interaction to create events that has that element as its original target

<ed> ...but on the other hand I do expect the element (if it's in the document) to still be able to receive events that are dispatched to it

<ed> ...with dispatchEvent, or by other means

<aemmons> I agree with ed 100%

<ed> so if an element that has focus is made invisible with a display=none, what is expected to happen with the current focus?

<ed> is it still focused?

<aemmons> nope, focus is re-evaluated

<aemmons> lots of content does this

<aemmons> a few groups of 'menus'

<ed> that's undefined behaviour as I read the spec

<aemmons> enable display only on one

<aemmons> move display through various groups for a different menu

<ed> do you have any pointers for where it says the focus must be moved on display=none?

<aemmons> sure, sec..

<aemmons> http://www.w3.org/TR/SVGMobile12/interact.html#focus

<aemmons> The SVG User Agent must not navigate to an element which has display='none'. (An element which has display='none' is not focusable.)

<aemmons> The SVG User Agent must allow navigation to elements which are not visible (i.e. which has a 100% transparency or which is hidden by another element).

<ed> but it wasn't display=none when it became focused

<aemmons> Yes, Dos not explicitly say if it is changed that a new focus is chosen

<aemmons> But, early on in 1.2 content and implementation, it was implied

<aemmons> by us, but also content tools and developers

<aemmons> and also I beleive other impleemntations

<aemmons> I guess if it is not obvious something has to be added to clarify either way

<aemmons> But I believe the usefulness of content would be reduced - how in content then would you force a change in focus without resorting to ECMAScript?

<ed> I'll send an html example to the publicwg list

<aemmons> ok, my example still stands in the original issue - this type of construct is very useful for making UIs with SVG

<ed> the thing is though, focus handling is really tricky stuff, and I don't want HTML and SVG to be incompatible because that makes it even more complex

<ed> see http://lists.w3.org/Archives/Public/public-svg-wg/2008JulSep/0171.html

<ed> firefox and opera keeps the first field focused, even after it was display=none:d, and after redisplaying it you can continue to type into it

<ed> safari doesn't move the focus anywhere, but doesn't allow typing more characters

<ed> (well, the field still has a focusrect in safari)

<ed> btw, apologies for the broken html...

DS: If that's not the specified behavior then we should examine the behavior
... we should probably coordinate with HTML on this

<scribe> ACTION: Doug to take up ISSUE-2031 with the HTML WG [recorded in http://www.w3.org/2008/08/21-svg-minutes.html#action07]

<trackbot> Created ACTION-2148 - Take up ISSUE-2031 with the HTML WG [on Doug Schepers - due 2008-08-28].

<aemmons> Ok Erik, I see where you are coming from. This is an interesting problem but please we should not do anyhitng hastly - the whole mobile content that I can see from both us and other relies on focus a given way

<aemmons> to the point where if it were deiced the other way we'd have to be non-compliant or else tick off a lot of people

<aemmons> I am sure this is not just us as well but others