W3C home > Mailing lists > Public > public-svg-wg@w3.org > January to March 2009

[minutes] Jan 8, 2009 telcon

From: Doug Schepers <schepers@w3.org>
Date: Fri, 09 Jan 2009 12:06:54 -0500
Message-ID: <4967842E.80306@w3.org>
To: "public-svg-wg@w3.org" <public-svg-wg@w3.org>

Hi, SVG Fans-

The minutes for the SVG WG telcon of 2009-Jan-8 can be found at:


Or as plaintext below:


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

                               - DRAFT -

                   SVG Working Group Teleconference

08 Jan 2009



   See also: [3]IRC log

      [3] http://www.w3.org/2009/01/08-svg-irc


          Shepazu, ChrisL, ed, [IPcaller], heycam, Shepazu.a,
          +39.524.9.aaaa, ed__, +95249aabb




     * [4]Topics
         1. [5]Agenda for Sydney F2F
         2. [6]ISSUE-2181
         3. [7]SVG 1.1 Errata
         4. [8]SVG Tiny 1.2
         5. [9]SVG Tiny 1.2 Errata
     * [10]Summary of Action Items

   <trackbot> Date: 08 January 2009

   <ed> Scribe: shepazu

   <ed> Scribenick: shepazu

Agenda for Sydney F2F

   ed: several possible topics...
   ... errata, filters, layout, compositing, modules, next version of
   the spec...
   ... what's happening with Print spec?

   ChrisL: Print spec is suffering from lack of implementations...
   ... open source folks are interested in the color profiles

   ed: how can we speed up the process?

   ChrisL: I've talked to folks from Scribus and Inkscape, but they are
   worried about the spec changing under them
   ... they got burned before with the textflow stuff we changed to
   please to the browser vendors

   ed: what phase is it in?

   ChrisL: LC

   shepazu: let's move it to CR

   <ChrisL> ended Feb 2008. Next step is CR

   ed: yes, we'll add that to the agenda
   ... any other topics?

   heycam: I'm going to try to write up some layout stuff to discuss

   ed: do we need to publish another snapshot of (more tests) of the
   SVG Tiny 1.2 testsuite anytime soon?

   heycam: good question, I've been looking at bugs and we do need to
   make some fixes to the test suite

   shepazu: we do need more tests, but maybe new implementers can help
   contribute tests
   ... I think we need to put a priority on 2.5D and other things that
   we are coordinating with the CSS WG on

   ChrisL: some of the CSS proposals have been checked into CVS
   ... Bjoern has remarked that they are underspecified

   ed__: I will put those agenda items on the wiki page


   <ChrisL> issue-2181?

   <trackbot> ISSUE-2181 -- Look at Solutions for 'filter' Property
   Conflict with IE Filters -- OPEN

   <trackbot> [11]http://www.w3.org/Graphics/SVG/WG/track/issues/2181

     [11] http://www.w3.org/Graphics/SVG/WG/track/issues/2181

   ed__: I couldn't find WG discussion about this

   ChrisL: there was discussion on the public lists about this

   [chris summarizes public discussions]

   <heycam> [12]http://markmail.org/message/p4qkphrqkhilszxq

     [12] http://markmail.org/message/p4qkphrqkhilszxq


     [13] http://lists.w3.org/Archives/Public/www-svg/2004Nov/0257.html


     [14] http://lists.w3.org/Archives/Public/www-svg/2004Nov/0259.html

   ed_work: we should coordinate with CSS and HCG on this

   shepazu: and the HTML WG



   heycam: IE is making changes in IE8, we should look at that

   ed_work: they are now supporting -ms-filter, too, so that's a step
   in the right direction

   me @ChrisL: can you rejoin?

   <ed__> element.style.filter = "foobar"

   <ed__> some people "detect IE" by checking if element.style.filter
   is defined

   shepazu: let's talk about this at the f2f

   ChrisL: my suggestion is that we find out what hackers are doing
   when using that hack, and give them a better way to do it
   ... this is not something that happens in a stylesheet, it's a test
   in script

   <ChrisL> its used to do opacity, and transparency in PNG images. Its
   only for old versions of IE, too - IE7 and up do PNG correctly

   ChrisL: and otherwise just hide the filters test when it's only a

   ed_work: I think your proposal is the best way forward
   ... and we could push IE to drop the proprietary filter thing

SVG 1.1 Errata





   I need to finish all those actions

   ed__: we should set aside time to finish this up at the F2F
   ... and publish the errata and SVG1.1 2nd edition
   ... it would be nice to have all the errata done before then
   ... it would require new tests
   ... is that too optimistic


   ChrisL: we've been going through them quickly

   ed__: we could triage them and do some of them in the next version
   of the spec

   ChrisL: it would be useful to have a 2nd ed, especially where it
   brings it into line with SVG 1.2

SVG Tiny 1.2

SVG Tiny 1.2 Errata

   ChrisL: going forward, we should just be use the same syntax for all
   lists: optional comma, optional space

   shepazu: heycam and I had come to a similar conclusion

   ChrisL: I noted that you wanted to make exception for URIs
   ... since they can have commas in them
   ... but IRIs can also have spaces
   ... so a better solution is simply to require that spaces and commas
   be escaped

   ed__: should we make this an errata item for SVG 1.1?

   shepazu: we would be changing conformance criteria for SVG 1.1 UAs?

   ChrisL: yes, but older content would still be conforming

   shepazu: and ASV is less relevant now...

   ChrisL: and has been for some time

   heycam: so this affects the 1.2 errata... it should now say any
   combination of comma or whitespace

   shepazu: what about viewBox?

   heycam: in the DOM, it's exposed as an animated rect, but in the
   spec it's a list of numbers, and that it has to have 4 members
   ... so it's not an arbitrary list of values

   shepazu: that's also the case for transforms

   <ed__> Proposed resolution: for 1.2T the 'g1', 'g2' and 'glyphname'
   attributes can be comma-whitespace separated

   heycam: how about using "list of xml names"?

   ChrisL: that wouldn't allow a digit at the beginning

   general agreement

   <ed__> for the 'u1' attribute : Comma is the separator character;
   thus, to kern a comma, specify the comma as part of a range of
   Unicode characters or as a glyph name using the 'g1' attribute.

   shepazu: but uDOM is typed, which list of string doesn't give us

   <heycam> heycam: we state explicitly in the udom appendix what types
   different xml attributes are exposed as, so changing the attribute
   list syntax in use by a particular attribute shouldn't affect that

   <heycam> [18]http://mcc.id.au/temp/2009/font-name-list-syntax.svg

     [18] http://mcc.id.au/temp/2009/font-name-list-syntax.svg

   shepazu: I'd like to check that this is okay with SVG 1.1

   heycam: we could start by compiling a list of affected attributes

   ChrisL: agreed

   shepazu: smil semicolon-separated values are essentially lists of
   lists; that could be a datatype

   we will propose changing values in SVG 1.1 to use a uniform list
   style, research what changes this would entail, and get feedback
   from SVG 1.1 implementers

   Resolution: we will propose changing values in SVG 1.1 to use a
   uniform list style, research what changes this would entail, and get
   feedback from SVG 1.1 implementers

   trackbot, end telcon

Summary of Action Items

   [End of minutes]
Received on Friday, 9 January 2009 17:07:13 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 9 January 2009 17:07:13 GMT