minutes, 20 November 2014 SVG WG telcon

http://www.w3.org/2014/11/20-svg-minutes.html




    [1]W3C

       [1] http://www.w3.org/

                                - DRAFT -

                     SVG Working Group Teleconference

20 Nov 2014

    [2]Agenda

       [2] 
http://lists.w3.org/Archives/Public/public-svg-wg/2014OctDec/0066.html

    See also: [3]IRC log

       [3] http://www.w3.org/2014/11/20-svg-irc

Attendees

    Present
           Dirk, Leonie, Brian, Thomas, Satoru, Cameron, Erik,
           Nikos, Bogdan, Doug, Tav, Tab

    Regrets
    Chair
           Cameron

    Scribe
           ed, heycam

Contents

      * [4]Topics
          1. [5]How to write tests for SVG 2
          2. [6]status update
          3. [7]focus and navigation
          4. [8]mailinglist policy
          5. [9]paint server rendering when objectBoundingBox is
             zero
      * [10]Summary of Action Items
      __________________________________________________________

    <trackbot> Date: 20 November 2014

    <Tav> ??P9 is me

    <ed> scribeNick: ed

How to write tests for SVG 2

    TS: I'd written a test for tspans, based on svg 1.1 tests
    ... checked into old repo, not sure it was carried over to the
    new github repo

    CM: the tests were not moved over to github
    ... we decided that we should move our tests to the w3c
    web-platform-tests repository
    ... I was given an action to create a few simple examples for
    scripted test, a reftest

    TS: the old repo had an svg2-tests repository

    CM: right, the new github repo only contains the specs, not
    tests
    ... once we have sorted out how to put tests on
    web-platform-tests we will move the tests from svg2-tests

    DS: so, heycam, you'll ping james graham for how to get that
    going, right?

    CM: yes

    TS: james is the guy behind web-platform-tests?

    DS: yes

    <Tav>
    [11]http://www.w3.org/Graphics/SVG/WG/wiki/SVG2/Testing_Require
    ments

      [11] http://www.w3.org/Graphics/SVG/WG/wiki/SVG2/Testing_Requirements

    DS: ok, so we will get some beginners documentation

    Tav: the page I just pasted in needs to be updated

    DS: most instructions should still apply

    TS: where in the directory structure tests should go, naming
    scheme etc, that sort of thing is missing atm

    CM: I'll do that

    <scribe> ACTION: heycam to write up documentation for directory
    structure, test naming scheme for the svg tests on
    web-platform-tests [recorded in
    [12]http://www.w3.org/2014/11/20-svg-minutes.html#action01]

    <trackbot> Created ACTION-3688 - Write up documentation for
    directory structure, test naming scheme for the svg tests on
    web-platform-tests [on Cameron McCormack - due 2014-11-27].

status update

    CM: what is the status of svg2 plan?
    ... based on the TPAC discussions
    ... what are the most urgent parts?

    krit: do as many edits as possible until february, things that
    aren't done by then get removed

    CM: so, should we remove things that are half-written?

    krit: we should try to publish more often
    ... we can delay things until the next revision

    <BogdanBrinza> and my mic appears to not be working and I had a
    plan to drive the investigation of stable vs moving parts

    TS: yes, iterating on what we have, to publish 2.1, 2.2 etc

    CM: faster turnaround and fewer features sounds good to me

    DS: ppl being disappointed about missing features, we need to
    improve the communication on how we intend to adress this
    (faster iterations)

    krit: we should publish every six weeks

    CM: what were the next steps?

    <BogdanBrinza> Yes, I'm putting the plan together now

    <BogdanBrinza> and would update the action and put this on
    agenda for next week to figure out what can we do, yes

    krit: [something]... by february
    ... the plan was to publish the LC in feb

    <smailus> Asthetically it may be ugly, but communicating that
    the published version in March 2015 is SVG 2.0 is important to
    communicate that this isn't the whole "SVG 2" spec.

    <BogdanBrinza> LOL I'll make sure to check Hawaii plan as well

    CM: I'll wait until next week and pick things off that list
    ... thank you for the update

focus and navigation

    DS: I was comparing focus and navigation from svg 1, 1.2 Tiny
    and 2.0
    ... there's been a lot more focus stuff added in 2.0
    ... it's scattered over the spec, we need to focus it more
    ... I gave a webinar about svg a11y yesterday

    <shepazu> [13]http://schepers.cc/svg/accessibility/focus/

      [13] http://schepers.cc/svg/accessibility/focus/

    <shepazu>
    [14]http://schepers.cc/svg/accessibility/focus/focus-tabindex.h
    tml

      [14] http://schepers.cc/svg/accessibility/focus/focus-tabindex.html

    krit: it does work in safari

    CM: reason for it not working in firefox is I need to review
    some patches for tabindex in svg

    DS: another issue around tabindex...
    ... if you look at the first one on the left, tab and it goes
    in order
    ... the same on the circles

    <BogdanBrinza> it somewhat works in IE..

    DS: the ones that light up green have a link
    ... tried to get it to work, based on an old test i wrote
    ... worked in chrome
    ... wasn't satisfied with the wording in the spec
    ... there's also some bit that defers to html5, I think we
    should resolve on those things
    ... out of order tab navigation, when you set the tabindex to a
    positive number
    ... the first instance I numbered 0-12 in that order
    ... in the second one I set them to tabindex=0
    ... which means order is document order, not in visual order
    ... apparently there's some probelems with setting to positive
    values
    ... current recommendation is for ppl to use tabindex=0
    ... based on an issue that Florian raised for CSS UI level 3,
    they're going to move nav-index to level 4

    CM: nav-index is similar to tab-index, but in css?

    DS: don't know honestly, but I think that's right

    CM: I saw the post on the public-svg-wg list about deferring it
    to level 4

    DS: want to know more about it, how should we handle making
    things focusable
    ... and navigation
    ... looking for help to work on focus and navigation, anyone
    interested?

    ED, CM, LJWatson: yes

    DS: [...] not exposing the focused element in voice
    ... we should try to make sure svg2 has these things exposed to
    a11y apis in SVG
    ... would like to see that across browsers
    ... and tests
    ... public-svg-a11y is the mailinglist
    ... you can subscribe, or ping me and I'll add you
    ... the svg-a11y-tf has launched

    krit: the work in webkit [...] don't have that much time
    anymore

    DS: we'll figure it out

mailinglist policy

    DS: svg was one of the first WGs to do all work in the public
    ... we wasnt sure how this was going to work at the time, so we
    setup a parallell mailinglist (public-svg-wg)
    ... not clear to me that we should use that one
    ... it's not really public, just publicly readable
    ... I propose that we close that mailinglist
    ... and that we do all our work on www-svg

    nikos: we already resolved to do that
    ... we jsut havent executed it

    <nikos>
    [15]http://www.w3.org/2013/09/19-svg-minutes.html#item03

      [15] http://www.w3.org/2013/09/19-svg-minutes.html#item03

    CM: sgtm, we may need to upadet some things first
    ... the repository notification emails
    ... and [missed]

    DS: other groups have tracker go to the public list

    CM: that could get noisy

    DS: so we could send those to the public-svg-wg list then?

    CM: ok, sounds good

    RESOLUTION: we will deactivate public-svg-wg, keeping the
    archives, repository changes will go to the w3c-svg-wg and
    tracker + technical discussions to www-svg

    DS: agenda items should also go to www-svg

    <heycam> ScribeNick: heycam

    <ed> [16]http://www.w3.org/Graphics/SVG/WG/wiki/Agenda

      [16] http://www.w3.org/Graphics/SVG/WG/wiki/Agenda

paint server rendering when objectBoundingBox is zero

    ISSUE-2416

    <trackbot> ISSUE-2416 -- Consider allowing zero height or width
    bbox gradients, and/or adjusting the bbox for stroke to be the
    actual stroke region -- raised

    <trackbot>
    [17]http://www.w3.org/Graphics/SVG/WG/track/issues/2416

      [17] http://www.w3.org/Graphics/SVG/WG/track/issues/2416

    ed: this is an old issue we've had open for a while
    ... I recently came across this while browsing bug reports for
    Chrome and Opera
    ... I created a test

    <ed> [18]http://jsfiddle.net/e7o46dgL/

      [18] http://jsfiddle.net/e7o46dgL/

    ed: this is to show that current implementations render paint
    servers even if the shape element has an empty bounding box
    ... the spec doesn't agree with this

    krit: we did have a test for it?
    ... a line with a gradient or stroke, and it was black
    ... I remember having to implement it for Cairo because it
    could not render it

    ed: that's the only browser I haven't tested
    ... if you have Safari there?
    ... I think we should update the spec to say that it should
    render

    <ed>
    [19]https://svgwg.org/svg2-draft/coords.html#ObjectBoundingBoxU
    nits

      [19] https://svgwg.org/svg2-draft/coords.html#ObjectBoundingBoxUnits

    ed: the last para in that section

    krit: did you test with gradients?

    ed: should be the same thing

    krit: not necessarily

    ed: they're both in terms of the bounding box

    krit: but they can be implemented differently
    ... there's a difference between how graphics libraries handle
    gradients and patterns
    ... with the assumption that it has the same behaviour for
    gradients, I agree with your change

    ed: and if not?

    krit: go with the majority

    ed: for animation, it doesn't make sense that fill or stroke
    goes missing if it's zero sized bbox

    heycam: was this one of the 1.1 errata issues?

    ed: don't know that it was
    ... the spec text doesn't call out patterns explicitly

    <ed> "When the geometry of the applicable element has no width
    or height and objectBoundingBox is specified, then the given
    effect (e.g., a gradient or a filter) will be ignored."

    <ed>
    [20]http://www.w3.org/TR/SVG11/coords.html#ObjectBoundingBoxUni
    ts

      [20] http://www.w3.org/TR/SVG11/coords.html#ObjectBoundingBoxUnits

    ed: there's an additional sentence in SVG 2

    heycam: I agree with your proposal

    ed: if you animate a straight line, rotating, you don't expect
    the paint server to stop applying at a certain point
    ... maybe not the transform itself, but adjusting x1/y1/etc.
    ... I'm fine with creating a few more test cases for gradients,
    and filters as well

    <scribe> ACTION: Erik to make some tests for gradients on
    zero-sized bbox [recorded in
    [21]http://www.w3.org/2014/11/20-svg-minutes.html#action02]

    <trackbot> Created ACTION-3689 - Make some tests for gradients
    on zero-sized bbox [on Erik Dahlström - due 2014-11-27].

    thanks for calling in stakagi and birtles, sorry about the new
    telcon time! :)

Received on Thursday, 20 November 2014 21:29:55 UTC