SVG Working Group Teleconference -- 15 Jan 2015

<trackbot> Date: 15 January 2015

<ed> scribeNick: Tav

<ed> https://www.w3.org/2002/09/wbs/19480/Sydney2015/

ed, Reminder to sign up for Sydney.

ed, Also if you need to call in,...

<richardschwerdtfeger> http://rawgit.com/w3m/aria/master/svg-aam/svg-aam.html

update from the svg accessibility TF

richardschwerdtfeger, Trying to get HTML5 to follow say strategy.

richardschwerdtfeger, Core mapping guide for all specs.

richardschwerdtfeger, Not in core spec: name computation. How to handle text API.

richardschwerdtfeger, SVG ahead of HTML.

richardschwerdtfeger, Reviewed one taxonomy. Had demos from Doug on screen reader.

richardschwerdtfeger, Change in SVG spec in video/audio for Linux.

shepazu, Accessibility API mappings... most in SVG WG familiar with ARIA.

shepazu, SVG will be publishing SVG2 accessibility mapping... we should have a dedicated call. Doug wants to get opinion of other members.

<ChrisL> agree is better to understand it

ed, heycam : We should review.

<ChrisL> will there be anything to review in time for the australia f2f meeting?

ed, We can dial into Friday's meeting.

richardschwerdtfeger, Does Friday next week work?

shepazu, Rich want to get it out soon.

richardschwerdtfeger, Hope to get it out in February.

shepazu, We should put out a poll if the time next week works for them.

<ChrisL> friday next week is good for me; the week after is not

Tav, works for me.

shepazu, Is Adobe interested? I'll put out a poll.

<ChrisL> +1 based on the email sent in today

richardschwerdtfeger, I can set agenda next week but need adequate lead time.

shepazu, I think we're making good progress...

new invited expert to svgwg

<Smailus> First one from Oreilly was 2002

shepazu, SVG Essentials co-author active on SVG mailing list, accessibility task force. Interested in becoming an invited expert.

shepazu, Co-chairs approve, PLH says its OK. Chris are you OK?

ChrisL, No objections.

ChrisL, W3C has a five day period to complain.

shepazu, I've already approved her. I don't anticipate any problem.

<TabAtkins> Ah, Amelia as IE? Heck yeah.

data-* in SVG

ChrisL, Nobody has objected to proposal. Question is how to do it.

<TabAtkins> Suggest moving the accessor API to Element, and just define the data-* attributes as being global attrs for all SVG elements.

ChrisL, Might be useful in other XML but nobody seems to care. This is really an non-name-space alternative to name spaces.

heycam, Tab said something which Tab has already mentioned above.

heycam, Next step is to try to get it to happen.

ChrisL, If done on element who does the work?

<TabAtkins> That's the DOM spec.

heycam, Anna needs to do work?

<TabAtkins> https://dom.spec.whatwg.org/

<ed> s/anna/anne

<shepazu> http://www.w3.org/TR/dom/

ChrisL, Action is back on Anne.

shepazu, If Anne doesn't have time we could ask Robin Berjon

ed, Do we need to reserve those attributes in SVG?

heycam, Yes.

ed. Does anybody object to Tab's suggestion?

<shepazu> +1

RESOLUTION: "We will reserve "data-*" attributes to be used in SVG content. The API for handling them is on element.

ed: Who get the action?

<scribe> ACTION: heycam to add "data-*" attributes notes to spec. [recorded in http://www.w3.org/2015/01/15-svg-minutes.html#action01]

<trackbot> Created ACTION-3694 - Add "data-*" attributes notes to spec. [on Cameron McCormack - due 2015-01-22].

<ChrisL> scribenick: ChrisL

<length> referenced to CSS is incorrect for SVG

Tav: length goes to css definition which is not ours, svg differs on percent, unit identifiers etc

svg 1.1 already did that right

Tav: we need a two step definition for attrs and properties

ChrisL: this is different for svg attrs compared to properties. it was right in svg 1.1

Rossen_: so it was right in 1.1

ChrisL: assume we want to refer to css3 values and units

heycam: i made some edits and split percentages from length for comnsistency with css
... the other part for unitless lenghts, not sure but may be a mistake
... to link directly, we have something in types that supports unitless lengths
... but css3 supports parsing of lengths without units in some contexts so we should define it that way
... invoke parser with that flag set

ChrisL: sounds like you know what spewc changes are needed

Tav: odd to click on length on rect and get that. its a property. don't see that you dont need units

heycam: keep in types chapter, and then link to css spec there whle talking about the unitless flag

<scribe> ACTION: heycam to fix length to point to types chapter and flag unitless lengths [recorded in http://www.w3.org/2015/01/15-svg-minutes.html#action02]

<trackbot> Created ACTION-3695 - Fix length to point to types chapter and flag unitless lengths [on Cameron McCormack - due 2015-01-22].

'context-fill', 'context-color' should apply to <pattern> and <hatch>

Tav: context-fill and stroke inpmemented in inkscape for markers
... missing for pattern and hatch. state explicitly that they can be used
... easier in impl to apply everywhere
... swapping stroke and fil inside a group for example
... so owe should allow on pattern and hatch

<heycam> +1 for explicitly saying these apply to hatch, pattern, etc.

ChrisL: is there somewhere we want to exclude it?

heycam: allow anywhere with referencable graphics

Tav: gradients?
... for a stop

heycam: ok and that is not referenceable

<TabAtkins> Heh.

heycam: define how it is interpreted in html context and also outside

ed: hatches, dont they take current color or do you need to specifiy the?

Tav: you need to specify them, like patterns

nikos: the examples show that but the spec text does not

ed: what is the lacuna value?

Tav: like patterns

<ed> <rect fill="url(#hatch)"> <--- what color is the hatch?

heycam: fill and stroke inherit. but what defines color of hatch line?

Tav: properties on hatch path

<heycam> stroke properties are used

ChrisL: if spec and examples contradict we should fix

nikos: attr list is incomplete

Tav: spec styling is left off

nikos: needs fixed eventually

ed: okay, shuld apply to all these context properties

<scribe> ACTION: tav to look at applying context-* outside of referenced things. [recorded in http://www.w3.org/2015/01/15-svg-minutes.html#action03]

<trackbot> Created ACTION-3696 - Look at applying context-* outside of referenced things. [on Tavmjong Bah - due 2015-01-22].

Tav: need to keep track of the ancestors anyway

heycam: but, inside a pattern, cant distinguish between parent value and context value

Tav: aha

heycam: or refactoring graphics into a definition, changes the rendering

Tav: needs more thought

feTurbulence

Tav: rendering gives solid patches because of a divide by zero

<nikos> For anyone that hasn't seen it: http://jsfiddle.net/dodgeyhack/mo7x85zw/

Tav: dealt with in inkscape using jaspers solution, throw out that value and do it again
... presto defines the vector to be zero which is incorrect

ChrisL: can we agree on jaspers solution?

nikos: that solution looks good

<scribe> ACTION: erik to fix feTurbulence for zero length vectors [recorded in http://www.w3.org/2015/01/15-svg-minutes.html#action04]

<trackbot> Created ACTION-3697 - Fix feturbulence for zero length vectors [on Erik Dahlström - due 2015-01-22].

progress on svg2 readiness

Rossen_: doc linked from agenda, public now

<Bogdan> https://www.w3.org/Graphics/SVG/WG/wiki/SVG2_readiness_assessment

bogdan: normally we take unready things off to force them to be fixed
... more than 60% of the spec would fall into that
... so define responsible people for each chapter to fix this

heycam: will do painting chapters

Tav: working in structure and text

ed: working on structure

bogdan: focus on svg element
... cpap 5 is huge

ed: shadow dom also a big remaining thing

I can take intro and concepts, since those seem to also have issues

Rossen_: how do we track these? actions, wiki?

I will take styling as well

ChrisL: I will do chapters 1,2,6

ed: chapter5 is huge with a lot of issues, need help there

Tav: have been liberal in adding issues

Rossen_: some issues may be easy to clear off

bogdan: count is subjective, tried to count but

Rossen_: definately a conversation starter
... like css 2.1 where we prioritized discussion of sections with highest readiness impact

heycam: what does readiness and severity mean

bogdan: (explains but scribe missed due to local noise)

severity goes from 0=good to 3=terrible

heycam: so look at the table and see if you agree for your chapters
... erik, can take svgsvgelement section which is mostly dom stuff
... easy to come up with a proposed behaviour
... thanks for putting that analysis together

birtles: mark turner is doing an svg meetup in sydney

<birtles> http://www.meetup.com/SVG-AU/events/219775517/

birtles: event set up
... asked tospeak but suggest others speak as well
... it is on the thursday night

(adjourned)

- DRAFT -

SVG Working Group Teleconference

15 Jan 2015

Attendees

Contents