SVG Working Group Teleconference -- 06 May 2010

<trackbot> Date: 06 May 2010

Updates on F11 2nd Edition

<ChrisL> http://www.w3.org/Graphics/SVG/WG/track/products/1

<anthony> ED: Went over entire spec

<anthony> ... to look for default values

<ed> http://dev.w3.org/SVG/profiles/1.1F2/publish/struct.html#SVGElement

<anthony> ... that were not mentioned in the DTD

<anthony> ... if you look at the SVG element for example

<anthony> ... and scroll down to the bottom of the element defs

<anthony> ... I've added a few things Zoom, Pan, PreserveAspectRatio

<anthony> ... Should the default value be defined, is it the same on all elements?

<anthony> ... Not sure if it's true

<anthony> ... I could remove what I've added and add default values else where

<anthony> ... should I remove the things that have links

<anthony> ... or list them explicitly

<anthony> CL: Might be better to list them explicitly

<anthony> ... but I don't mind either way

<anthony> DS: Same

<anthony> ED: A good part of it is that I found no places that were missing things

<anthony> ... except for the filters chapter

<ChrisL> The above discussion relates to ACTION-2676

<anthony> ... the only place is PreserveAspectRatio

<anthony> CL: Just add the definition once

<anthony> ED: Ok, I'll go back and do some minor changes

<anthony> ... I'll reopen the ACTION and make the changes

<anthony> ... Then I did a bunch of updates on the testsuite

<anthony> ... some of that was based on the QA staff from Opera

<anthony> ... they would like to see a test harness that doesn't have all the stuff

<anthony> ... that a person would look at

<anthony> ... like a test description etc

<anthony> ... they just want a basic test suite

<anthony> ... with test, ref image, and link

<anthony> CL: How do the automated test suite move from one link to the next?

<anthony> ... list of links?

<anthony> ... if you have a HTML page that links the SVG and PNG image then that's all you need

<anthony> ... what does the HTML do for you?

<anthony> ED: Have the reference image and SVG side-by-side for checking

<anthony> ... I haven't checked that in anyway

<anthony> ... and the QA staff seem to think the amount of text on the test is too much

<anthony> ... in some cases the description is too long

<anthony> CL: We've talked before about having the information there

<anthony> ... it's valuable if you want to know more about the test

<anthony> ED: One suggestion was to collapse the description field

<anthony> CL: That would be an easy thing

<anthony> ... what to do to pass the test

<anthony> ... and what to do to run the test

<anthony> ... we starting to get some standard text

<anthony> ... I think it's good to start putting that everywhere when it's valid

<anthony> DS: I think we're getting two different cases here for the text

<anthony> ... one running the test

<anthony> ... and two using the tests as examples

<anthony> ... having an explanation is useful for the later case

<anthony> ... but not so useful for the initial

<anthony> ED: Just wondering if I should make another harness

<anthony> ... that collapses the extra wording

<anthony> CL: The harness is not normative

<anthony> ... and so we can make as many versions as we like

<anthony> ED: Ok, I'll have another chat with them

<anthony> ... I moved the description last on the page

<anthony> ... because it was pushing the images out

<anthony> ... status update on actions

<anthony> CL: Sent an email about spec history checking because we made some edits we agreed to then undid them and change them to something else

<anthony> ... JWatt had proposed some wording at an F2F that we all agreed on

<anthony> ... and put that back in

<anthony> ... all the stuff EBNF we thought was wrong is not

<anthony> ... so I closed that ACTION

<ChrisL> ACTION-2765

<ChrisL> ACTION-2730

<anthony> CL: Above are the actions I closed

<anthony> ... and then there's the long running action about the IRI stuff

<anthony> ... I did that

<ChrisL> ACTION-2697?

<trackbot> ACTION-2697 -- Chris Lilley to make funcURI consistent, and update tests -- due 2009-11-30 -- CLOSED

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/actions/2697

<anthony> ... I also checked for all the properties in the appendix that has the funcURI and checked to see if we have tests for them

<anthony> ... and we do

<ChrisL> ACTION-2753?

<trackbot> ACTION-2753 -- Chris Lilley to investigate BNF fix for path and pologyon commonality and see if it is easier to fix 2nd edition spec or move to 2.0 spec -- due 2010-04-15 -- OPEN

<trackbot> http://www.w3.org/Graphics/SVG/WG/track/actions/2753

<anthony> ... the above is what I'm currently working on

<anthony> ... the BNF is going to a big change

<anthony> ... while I was doing the action there was something in the spec that we don't test

<anthony> ... polyline and polygon they are the same as the path with an absolute move to

<ChrisL> http://dev.w3.org/SVG/profiles/1.1F2/test/svg/shapes-polyline-02-t.svg

<anthony> ... and I wrote a test there

<anthony> ... I'm about to make a test for polygon

<anthony> ... just need someone to review the tests

<ChrisL> ACTION: anthony to review shapes-polyline-02-t.svg and /shapes-polylgon-02-t.svg [recorded in http://www.w3.org/2010/05/06-svg-minutes.html#action01]

<trackbot> Created ACTION-2779 - Review shapes-polyline-02-t.svg and /shapes-polylgon-02-t.svg [on Anthony Grasso - due 2010-05-13].

<anthony> CL: Another thing I noticed

<anthony> ... what I was working the types chapter

<anthony> ... the ones that have the type IRI or type funcIRI don't have it

<anthony> ... is it represented as a DOM string?

<anthony> ED: Do you have the wording?

<ChrisL> http://dev.w3.org/SVG/profiles/1.1F2/publish/types.html#BasicDataTypes

<ChrisL> In the SVG DOM, <angle> values are represented using SVGAngle or SVGAnimatedAngle objects.

<anthony> CL: So for example it says the above

<anthony> ... it says that for most of the types except for things that are a sequence of 0 or more characters

<anthony> ED: I think there is an interface

<ed> http://dev.w3.org/SVG/profiles/1.1F2/publish/types.html#InterfaceSVGURIReference

<anthony> ED: I don't think it is used for all of them

<anthony> ... only used for non-properties

<anthony> ... so it wont use that interface

<anthony> CL: So IRI uses that, but funcIRI doesn't

<anthony> ED: Right

<anthony> CL: Ok, I'll add it for IRI

Mercurial is available

<anthony> ED: Did you guys see the announcement?

<anthony> ... we can put it as a topic for the F2F

<ChrisL> http://mercurial.selenic.com/wiki/TortoiseHg

<anthony> ... was thinking that we might be able to move to it

<anthony> CL: There is a TortoiseHg which gives you a GUI to interface to it

<anthony> ED: I will add that to the agenda page for the F2F

http://mercurial.selenic.com/wiki/TortoiseHg

TPAC Schedule

<ChrisL> http://www.w3.org/2002/09/wbs/19480/SVGLyon/

<anthony> ED: I haven't had a time to look at the schedule

<anthony> CL: CSS and Webfonts are on Mon/Tue

<anthony> ... and SVG is on Thur/Fri

<anthony> ... XSL-FO might be a clash

<anthony> ... I don't see any hard overlaps

<anthony> DS: Is HTML only meeting 2 days?

<anthony> ... so we can't meet with the HTML group at all?

<anthony> CL: We'd have to do a joint session

href

<ChrisL> batteries in my headset are dead. will rejoin in a moment

http://www.w3.org/Graphics/SVG/WG/wiki/Href

<anthony> DS: I made the page on the wiki

<anthony> ... I tried to summaries all the issues in the resolution

<anthony> ... and put the resolution up the top

<anthony> ... I front loaded what we actually plan on doing

<anthony> ... then underneath is all the discussions

<anthony> ... and decisions we made

<anthony> ... before we release it to the public

<anthony> ... if you guys can look over it

<anthony> ... that would be good

<anthony> ED: I think in the resolution it shouldn't say something will be parsed

<anthony> ... both attributes will be parsed

<anthony> ... so that's something you don't have put there

<anthony> ... but it's useful to say which namespace the href and xlink:href attributes end up in

<anthony> DS: What about in the 3rd one?

<anthony> ED: That I'm not so sure about

<anthony> ... that's not consistent

<anthony> DS: So say "The resulting DOM attribute will have the attribute name with a NULL namespace?"

<anthony> ... added the wording in

<anthony> ... look at it again

<anthony> ED: I'm not really sure I understand the 3rd one

<ed> "the resulting DOM attribute will have the href value of the href attribute and the namespace value null."

<anthony> ED: I think you need to be a bit more specific then

<anthony> ... because it sounds like both will have one attribute in the DOM

<anthony> ... so we agree that both attributes will be on the element and you did a getAttribute, which one would you get?

<anthony> DS: you'd get the one you specify

<anthony> ... get href gets href. get xlink:href gets xlink:href

<anthony> ED: I think it's good to say explicitly what it does

<anthony> DS: I'll add a section for getters and setters there

<anthony> CL: This is for SVG 2.0 right?

<anthony> DS: Yes

<anthony> ... that is something I should say right upfront

<ChrisL> "If this is going to be done, the time to do it is now, while SVG is not yet ubiquitous. " could be interpreted as, well, now as in SVG1.1SE

<anthony> DS: Chris I've now indicated that in two different places that this is SVG 2.0

<anthony> CL: This looks like a good summary

<anthony> ... It covers the issues I'd expect to see there

<anthony> DS: There was one particular issue

<anthony> ... if you guys reload the page

<ed> http://dev.w3.org/SVG/profiles/1.1F2/publish/types.html#InterfaceSVGURIReference

<anthony> ... that interface doesn't have a namespace it's just a string

<anthony> DS: That's a mistake

<anthony> ED: Maybe I'm miss reading things

<anthony> ... the interface doesn't really care where it comes from

<anthony> DS: There wont be a namespace value

<anthony> ...I should look at the equivalent HTML interface

<anthony> ED: I guess that makes sense

<anthony> ... there's the location interface

<anthony> ... and window

<anthony> DS: Window is part of HTML

<anthony> ... there's three different ways in which this is exposed

<ed> http://dahlström.net/svg/dom/changing_href_on_use.svg

<anthony> ... one as the attribute value itself

<anthony> ... the second is the href interface or the URI interface

<anthony> ED: The read only part is only for that particular interface

<anthony> ... it points to an "SVGAnimatedString href"

<anthony> DS: We should at least note in SVG 2.0

<anthony> ... how that works

<anthony> ... we should just say

<anthony> ... for any attribute is read only, because it writable because it inherits from a different interface

<anthony> ... it is writable

<ed> ED: you can write to the href.baseVal, but you can't assign something to SVGURIReference.href

css style comments

<ChrisL> http://dev.w3.org/csswg/css-style-attr/issues-lc-2009

<ChrisL> our comments http://lists.w3.org/Archives/Public/www-style/2010Feb/0041.html

<scribe> scribeNick: shepazu

<ChrisL> their responses http://lists.w3.org/Archives/Public/www-style/2010Feb/0191.html

ChrisL: they have accepted 1 comment, rejected 1, ruled 2 out of scope

shepazu: please CC the SVG WG list, so we're aware of it

ChrisL: will do next time

<ChrisL> Peter Linss agreed with outr comment #3

<ChrisL> http://lists.w3.org/Archives/Public/www-style/2010Feb/0192.html

ChrisL: comment 3 was not listed in their DoC, though they agreed to it
... comment 2, they clarified the spec, we should accept that one

<ed> http://dev.w3.org/csswg/css-style-attr/

ChrisL: it doesn't seem clear to me whether curly braces are allowed within a declaration block
... but they said it is clear from the grammar

<ed> http://www.w3.org/TR/CSS2/syndata.html#rule-sets

<ChrisL> their grammar had declaration but not declaration block

<ChrisL> http://www.w3.org/TR/CSS21/grammar.html

DS: for example <circle style="{fill:red;}"/> ?

ChrisL: their grammar doesn't include the term "declaration block"
... they do have "declaration list"

ed: its probably clear anyway

<ed> http://dev.w3.org/csswg/css-style-attr/#syntax

<ed> http://dev.w3.org/cvsweb/csswg/css-style-attr/Overview.html

ChrisL: I think it can be read either way, but on rereading it, I guess it is clear enough
... the next issue is that CSS 2.1 could be read that only CSS2.1 properties are valid, therefore CSS3 and SVG properties should be ignored
... they seem to be saying in their reply that that's not the intent, but I'm not sure what the resolution is

ed: we could ask them to include the wording change more explicitly

ChrisL: this could be solved by changing "CSS 2.1 properties" to "CSS properties"

ed: I think we should push for this change

ChrisL: in other words, we shouldn't accept it until we see the change in CSS 2.1
... their change would be fine if they actually commit it
... I asked if unitless values and scientific notation were allowed
... they responded no
... that are using FLEX grammar, but with their own definition of NUMBER
... SVG is consistent currently with this, because it allows scinot for attribute values but not in CSS declarations
... Mozilla seems to support this, but Bert rejected any change to the grammar, which he insists must be frozen

shepazu: if an implementation (at least one, maybe more?) supports it, CSS shoudl be changed to reflect reality

ed: it seems it would be better to be consistent

<scribe> ACTION: ChrisL to ask CSS WG to include clarification for "CSS 2.1 properties" as noted in response [recorded in http://www.w3.org/2010/05/06-svg-minutes.html#action02]

<trackbot> Created ACTION-2780 - Ask CSS WG to include clarification for "CSS 2.1 properties" as noted in response [on Chris Lilley - due 2010-05-13].

<scribe> ACTION: ChrisL to ask CSS WG to revisit scientific notation issue for grammar [recorded in http://www.w3.org/2010/05/06-svg-minutes.html#action03]

<trackbot> Created ACTION-2781 - Ask CSS WG to revisit scientific notation issue for grammar [on Chris Lilley - due 2010-05-13].

trackbot, end telcon

- DRAFT -

SVG Working Group Teleconference

06 May 2010

Attendees

Contents

Updates on F11 2nd Edition

Mercurial is available

TPAC Schedule

href

css style comments

Summary of Action Items

Scribe.perl diagnostic output