W3C home > Mailing lists > Public > public-svg-wg@w3.org > October to December 2008

Minutes, 11 December 2008 SVG WG telcon

From: Chris Lilley <chris@w3.org>
Date: Thu, 11 Dec 2008 22:10:19 +0100
Message-ID: <1016704257.20081211221019@w3.org>
To: "public-svg-wg@w3.org" <public-svg-wg@w3.org>

Hello public-svg-wg,

 Minutes of the 11 December call are here

and below as text for trackbot

                   SVG Working Group Teleconference

11 Dec 2008

   See also: [2]IRC log

      [2] http://www.w3.org/2008/12/11-svg-irc


          ed, heycam, anthony, Shepazu, ChrisL




     * [3]Topics
         1. [4]HTML5
         2. [5]Vacation break
         3. [6]Selectors review
         4. [7]clipping
         5. [8]new template for 1.1 tests
     * [9]Summary of Action Items

   <trackbot> Date: 11 December 2008

   <scribe> scribe: ChrisL

   <shepazu> scribeNick: ChrisL



     [10] http://lists.w3.org/Archives/Public/public-svg-wg/2008OctDec/0420.html

   (taking comments one by one)

   ED: is it good to collect high level goals?

   CMcC: Yes, it halps as long as they are not too abstract

   ED: Suggested rewording avout remaining in XML

   CL: Main aim was that folks using existing xml would not need to
   change it to copy and paste

   DS: Henri said that nonwelformed XML should be exposed as WF by the
   browser after correction

   CMcC: the later ones are higher level goals

   DS: So rearrange, and put this one as a consequence. Say why, and
   then it naturally falls out

   CMcC: that more convincing, yes

   ED: neither mail captures what we want

   CL: Not sure we should push on validity. Well formed is more

   ED: Want the fixups to be arked as errors

   CMcC: No implied elements, would restrict ability to change the
   ... Parser will need to fix up idf the end tag is wrong

   CL: At TPAC, agreement that the doctype etc would not be pasted in

   DS: Until there are tools, they will not understand why its not
   working. Paste in direct from illustrator and inkscape
   ... Need to ediucate people on what is allowed, some are allowed
   within XHTL, others within HTML

   CL: No doctype etc in the middle, obviously

   CMcC: Some examples in my mail

   CL: Entities are kind of a transitional phase, just use Unicode

   AG: Omitted quotes is abad idea because with complex syntax its not
   always clear where the attribute value ends

   DS: Lots of attrs take space-less strings and would be fine here,
   and would be error corrected

   <g title=a space string /> with orr characters/>

   CMcC: This is error correction?

   ED: Not in HTML5, its not an error

   DS: Effect is the same.
   ... depends whether it reports an error or not

   AG: I think its inconsistent

   DS: Should be flagged as an error to help people who are trying to
   do standalone svg
   ... need to be aware of quoting html5 quirks. makes it harder to
   adapt content

   CMcC: Fixed up export helps there

   DS: yes
   ... But mashups, not using a browser, would be effected
   ... concerned that people have svg-like files, they download the
   file and try to open in an authoring tool and it does not work. They
   don't know why

   AG: An artist will understand right click, save as, and then open
   the file

   CL: If browsers do silent fixup then there will be svg files that
   are not well formed and cannot be edited

   DS: This is why I said it should be seen as error correction
   ... no 'visible indication of error' but a log would show a warning
   ... eg in a validator

   CL: Warning is very reasonable.

   DS: Would not be surprised to see xml allow this also in the next 5
   ... SVG requires quotes
   ... Its ok as long as its acknowledged as error correction

   ED: Perhaps we should discuss these in email

   CMcC: Henri made a fork of Mozilla that uses the old HTML handling
   of SVG

   <heycam> [11]http://about.validator.nu/htmlparser/

     [11] http://about.validator.nu/htmlparser/

   <heycam> sorry not that url

   <heycam> [12]http://hsivonen.iki.fi/html5-gecko-build/

     [12] http://hsivonen.iki.fi/html5-gecko-build/

Vacation break

   CL: here next week then off two weeks

   <shepazu> [13]http://www.intertwingly.net/blog/

     [13] http://www.intertwingly.net/blog/

   AG: Working to Xmas, then off to 13 jan

   DS: Working holiday

   CMcC: next week ok?

   (agreement seems to be to cancel from22 and restart from 5 Jan

   ACTIN: Erik announce the xmas break

   <scribe> ACTION: Erik to announce the xmas break [recorded in

   <trackbot> Created ACTION-2377 - Announce the xmas break [on Erik
   Dahlström - due 2008-12-18].

Selectors review


     [15] http://lists.w3.org/Archives/Public/public-webapps/2008OctDec/0465.html

   ED: Some feedback from lachlan

   CL: Important thing is that the spec now clearly states that ns
   functionality is not there. So thats good

   ED: Node selector text is also improved
   ... I mentioned some potential non-interop, and he pushed back. i'm
   happy to drop those. Selectors on subtrees outside the document will
   not be interoperable however
   ... on list creation, its not very precise but its ok i guess
   ... mention of downlevel clients sparked some discussion
   ... refuses to add a DOM3 xpath informative reference

   DS: Sicking makes good points

   ED: On last issue he says he is reconsidering adding an example

   <scribe> ACTION: Erik to respond to Lachlan on selectors api review
   [recorded in

   <trackbot> Created ACTION-2378 - Respond to Lachlan on selectors api
   review [on Erik Dahlström - due 2008-12-18].


   DS: Not fully convinced by Thomas' argument. Seems to conflate
   clipping paths with viewports?

   CMcC: No, his example is a viewport one but he understands the
   ... he brought in viewports and an example of something that clipped
   out pointer events

   CL: Consider a symbol, which has a clippath. The pointer events are
   clipped in the other three quadrants unless overflow is enabled

   DS: And with visiblePainted, thats what you get

   <heycam> [17]http://mcc.id.au/temp/2008/clip-window.svg

     [17] http://mcc.id.au/temp/2008/clip-window.svg

   DS: Rules fr clip paths apply. Clip paths do not clip events al the
   time, depends on value of pointer events. Whie whena vireewport is
   established, events are always clipped itrtespective of the value
   ... asv clips the events no matter what. safari clips no events at

   CcC: Proposal changes =it to allow either clipping or not

   DS: Under author control

   CMcC: Thomas' concern is that people have existing content that
   depends on this. But behaviour is not fully defined in spec.

   DS: Is this a real use case?

   CMcC: yes, its how people create invisible event targets now
   ... but there are workarounds

   DS: So I don't think this combination is very common as there are
   better and more direct ways to do this

   ED: Interesting to hear Mozilla view on this, would liked to have
   asked JWatt or RoC

   <scribe> ACTION: Doug ask Mozilla about this clipping issue
   [recorded in

   <trackbot> Created ACTION-2379 - Ask Mozilla about this clipping
   issue [on Doug Schepers - due 2008-12-18].

   ED: Wonder how much content uses clip path plus pointer-events
   nondefault values

new template for 1.1 tests


     [19] http://lists.w3.org/Archives/Public/public-svg-wg/2008OctDec/0423.html

   DS: If we make a schema we could have the validator at W3C be more

   CL: oXygen ships with a 1.1F XSD, we should get them to add an
   SVGT1.2 RBG as well

   DS: No need for an xml prolog

   CL: Apart from non UTF-8, UTF-16, no need for an xml encoding

   Resolved: New 1.1F template has no doctype and no xml declaration

   (we agree that having one template is a good goal derived versions
   can be made with xslt)

   (discussion on version and baseProfile desirability in tests)

   CL: version and baseProfile should only be aded where actually
   needed. most tests do not require it
   ... and i agree about the 'approved' value, never understood why
   that had a different value

   ED; There was a distinction between approved an=d accepted, which
   was that it was published for the public. Used to be a list. Now its

   CL: Only downside is some (small) changes to the scripts

   <scribe> ACTION: Anthony to look through scripts for any breakage
   that would happen from conflating approved and accepted [recorded in

   <trackbot> Created ACTION-2380 - Look through scripts for any
   breakage that would happen from conflating approved and accepted [on
   Anthony Grasso - due 2008-12-18].

   ED: Need to clarify what to do when dropping a test back to reviewed
   or created

   Resolved: the approved and accepted should be one value

   ED: Wonder if there are tests which are accepted but not approved

   meeting adjourned

Summary of Action Items

   [NEW] ACTION: Anthony to look through scripts for any breakage that
   would happen from conflating approved and accepted [recorded in
   [NEW] ACTION: Doug ask Mozilla about this clipping issue [recorded
   in [22]http://www.w3.org/2008/12/11-svg-minutes.html#action03]
   [NEW] ACTION: Erik to announce the xmas break [recorded in
   [NEW] ACTION: Erik to respond to Lachlan on selectors api review
   [recorded in

   [End of minutes]

 Chris Lilley                    mailto:chris@w3.org
 Technical Director, Interaction Domain
 W3C Graphics Activity Lead
 Co-Chair, W3C Hypertext CG
Received on Thursday, 11 December 2008 21:10:55 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:20:10 UTC