minutes, 9 July 2015 SVG WG telcon

The minutes for today's teleconference are available on the web:

   - in HTML: http://www.w3.org/2015/07/09-svg-minutes.html
   - plain text: http://www.w3.org/2015/07/09-svg-minutes.html,text

Copied below for searchable archives.

________________________________


   [1]W3C

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

                               - DRAFT -

                    SVG Working Group Teleconference

09 Jul 2015

   [2]Agenda

      [2] https://lists.w3.org/Archives/Public/www-svg/2015Jul/0004.html

   See also: [3]IRC log

      [3] http://www.w3.org/2015/07/09-svg-irc

Attendees

   Present
          Cameron, AmeliaBR, nikos, stakagi, Tav, shepazu

   Regrets
          Erik

   Chair
          Cameron

   Scribe
          AmeliaBR

Contents

     * [4]Topics
         1. [5]Behavior of overflow:auto
         2. [6]Paris & Sydney F2F meetings
     * [7]Summary of Action Items
     __________________________________________________________

   <trackbot> Date: 09 July 2015

   <scribe> Scribenick: AmeliaBR

Behavior of overflow:auto

   Nikos: In SVG 1.1, auto overflow is treated as visible. In CSS,
   it creates a clipped but scrollable or swipeable region
   ... For SVG 2, I think the plan from previous discussions was
   to make SVG similar to CSS.

   <nikos>
   [8]https://lists.w3.org/Archives/Public/www-svg/2012May/0096.ht
   ml

      [8] https://lists.w3.org/Archives/Public/www-svg/2012May/0096.html

   Nikos: However, I haven't been able to find any clear
   resolutions, just this mailing list post.
   ... In which Cameron mentions that this is unlikely to be a
   breaking change.

   <nikos> [9]http://jsfiddle.net/dodgeyhack/9naLshrm/

      [9] http://jsfiddle.net/dodgeyhack/9naLshrm/

   Nikos: There is some mention of moving it to the Integration
   spec, but that doesn't address elements within SVG such as
   symbol.

   The JSFiddle is just a simple test of current behavior.

   Cam: I'm not sure what Edge will do; I'm pretty sure Firefox
   still treats it as visible.

   Nikos: Everyone does visible for auto as far as I can see.

   <nikos> [10]https://svgwg.org/svg2-draft/render.html

     [10] https://svgwg.org/svg2-draft/render.html

   Tav: Didn't we change the behavior for markers?

   Cam: Yes, but that was just a matter of changing the default UA
   stylesheet.

   Nikos: I've added a table into the relevant section of SVG2
   describing current values for each element.

   AmeliaBR: To be consistent with CSS auto behavior, we need to
   deal with pan&zoom or scrolling. There are already values for
   both visible and hidden, don't need to repeat either.

   Cam: We discussed this at the F2F, but it's really too big a
   topic for now.
   ... It would seem more of a natural transition if browsers
   currently clipped to the rectangle.

   AmeliaBR: For me, auto implies that the content is accessible.
   I wouldn't want to treat it as hidden.

   Cam: It depends which is the relevant feature: that inner
   content is accessible, or that the defined region is a strict
   limit.
   ... I'm inclined to keep the current behavior, since it is well
   supported.

   Nikos: I don't really have a strong opinion. I thought about
   making auto hidden for nested SVG elements, but visible for the
   outer stand-alone SVG.

   Cam: That's interesting, because there might be differences in
   current implementations.

   AmeliaBR: For stand-alone root SVG, you'll get scrollbars I
   think.

   Cam: What about inline SVG?

   <heycam> [11]http://mcc.id.au/temp/ov.html

     [11] http://mcc.id.au/temp/ov.html

   AmeliaBR: Currently there are cross-browser inconsistencies.
   But that's a user stylesheet issue, visible vs hidden, not
   related to auto.

   Cam: This test case shows the effect of overflow: auto on
   inline SVG.

   Results: IE treats auto as visible, everyone else seems to
   treat it as hidden
   ... In contrast, everyone treats auto as visible for content
   inside the SVG:
   [12]http://codepen.io/AmeliaBR/pen/ee899b133d902fb82a2f6357f4bb
   908c

     [12] http://codepen.io/AmeliaBR/pen/ee899b133d902fb82a2f6357f4bb908c

   Nikos: Did we not already agree that overflow: scroll should be
   available, if scrollbars or similar mechanism is possible?

   Cam: So auto could be similar, if scrolling is allowed.
   ... But I think we can get a resolution on the original issue,
   should it be treated as visible or hidden?
   ... I would recommend visible based on current behavior.

   AmeliaBR: Can we add wording that leaves open the possibility
   of panning or scrolling, but makes it visible if
   panning/scrolling is not possible?

   Cam: That would be consistent with the wording we already have
   for scroll.

   Nikos: It would nicely complement scroll, which is display
   scrollbars if you can, but otherwise clip content. Auto would
   be display scrollbars if you can, but always make content
   available one way or the other.

   Cam: I think that would be alright.

   Nikos: Current behavior for scroll is visible in Firefox and
   clipped in IE/Chrome

   RESOLUTION: Overflow:auto should be treated as overflow:visible
   if scrollbars are not possible for the given element;
   overflow:scroll should be treated as overflow:hidden if scroll
   bars are not possible

Paris & Sydney F2F meetings

   Cam: For the Paris meeting, coordinated with CSS, that's 99%
   sure that space is available & other logistics will work. The
   CSSWG is supportive of some overlap for FX work.

   [13]https://www.w3.org/Graphics/SVG/WG/wiki/F2F/Paris_2015

     [13] https://www.w3.org/Graphics/SVG/WG/wiki/F2F/Paris_2015

   ^^^ Not much there ^^^^

   <nikos>
   [14]http://www.w3.org/2015/06/12-svg-minutes.html#item14

     [14] http://www.w3.org/2015/06/12-svg-minutes.html#item14

   <heycam> [15]https://wiki.csswg.org/planning/paris-2015

     [15] https://wiki.csswg.org/planning/paris-2015

   Cam: The currently decided dates are the previous week. CSS is
   August 25-27, we were going to meet at the same time. I'll
   update the wiki.
   ... I also wanted to mention the Sydney F2F. I sent an email to
   the list. Unfortunately, logistics don't work for having the
   meeting at the same time as CSS, as we'd wanted.
   ... I suggested going back to having it after, but we will need
   to confirm & decide relatively soon.
   ... This would be similar to the Feb2015 schedule. Any concerns
   should be sent to the mailing list.

   Nikos: Would that be the next week, or immediately after?

   Cam: Immediately after, with an overlap day for FX.

   trackbot, end telcon

Summary of Action Items

   [End of minutes]