W3C home > Mailing lists > Public > www-svg@w3.org > January 2010

(additions and revisions) Proposed Agenda Items for Today's SVG Conference Call

From: Patrick Dengler <patd@microsoft.com>
Date: Mon, 25 Jan 2010 16:53:34 +0000
To: "'www-svg@w3.org'" <www-svg@w3.org>
Message-ID: <4A2DB3AE4504E944AF122BBFBA7FBA1F3E7A31E7@TK5EX14MBXC114.redmond.corp.microsoft.com>
I talked to Doug and Jay on Thursday and we identified (as Doug sent out) some proposals for how to handle hit testing so this is less of an issue now (though clear behavior on pass through in XHTML and transparent PNG's might be needed if we missed it in the spec)

-W3C Test Suite not respecting HTML4/5 parsing rules
It seems that the SVG Test Suites around <object> have self-closing <object> tags.

<object data="../svggen/color-prof-01-f.svg" width="480" height="360" type="image/svg+xml"/>

>From what we understand, these do not conform to HTML4 or the HTML5 spec around <object>, though major browsers tolerate it.  Having said that, we are going to run a web crawl on how often this is seen in practice.  We will report that back to the SVG WG as well as the HTML5 WG.

-HTML5 SVG parsing Rules
I was talking to Doug and Jay about "web developer" scenarios and SVG integrated into HTML.  He had anecdotally noted that a developer he was working with was stumped on a link, as he had used an href/src combination instead of xlink.  We had a discussion around this for SVG 2.0 linking improvements.  However it brought to mind "general SVG web development scenarios".  When I look at the HTML5 parsing and SVG case fix up I think this similar issue might occur more broadly. Since HTML markup is case insensitive and a number of SVG elements are camelCased, HTML 5 specifies a conversion table to go from the lower-case version of an element name in the SVG namespace, to the properly-cased version (http://dev.w3.org/html5/spec/Overview.html#parsing-main-inforeign).  Are "web developers" going to be tolerant of this constraint? Will they get lost as the developer did in the case of href?

Thanks in advance for your consideration.

Patrick Dengler
Senior Program Manager
Internet Explorer Team

-----Original Message-----
From: Patrick Dengler 
Sent: Thursday, January 21, 2010 10:35 AM
To: www-svg@w3.org
Subject: Proposed Agenda Items for Today's SVG Conference Call

Hosting SVG Face-to-face
We are excited to be able to host the face-to-face for SVG; I just need more details on exact dates (i.e. whether or not we want days beyond May 26th)

Closing down SVG 1.1 2nd Edition
We've discussed in previous conference calls the desire to finalize the 2nd edition.  I wanted to note that it might be worth keeping this window open a little longer as we dig deeper into the SVG specifications and browser behaviors in order to get crisp interoperability.

Potential Spec Improvement Topics
CSS and <use>:
We are seeing some inconsistent behaviors with CSS sibling, child and descendant selectors and <use>. If there are CSS selectors that apply to an element that is referenced via 'use' but the selectors don't specifically apply to the 'use' tag itself, they are not applied to the 'use' instance in Firefox and Webkit. Opera applies all selectors that apply to the referenced element to the 'use' instance, even if those selectors don't match the 'use' tag.

How should we interpret the last two sentences as "CSS selectors that apply to the referenced element do not necessarily apply to a 'use' instance of that element".  We have the impression that a 'use' instance of an element should contain all of the styles that apply to the element that it references.  And does this include events?  Examples are at the end of this mail.

"stroked-dasharray" inconsistent across browsers and tools:
In the attached screen shot we notice different behaviors with stroked-dasharray

		<rect x='30' y='30' rx='5' ry='5' width='250' height='250' stroke-width='20px' stroke='red' stroke-dasharray='4px, 3px' />

Since Opera and Chrome are very close, did we miss some details? We actually think Inkscape looks the best.  Do we care to clarify this in the spec?

Hit Testing:
We are not sure what the expected behavior is for hit testing.  For example, the spec states for onclick: "occurs when the pointing device button is clicked over an element."  Yet we are not finding clarity for the following scenarios and are wondering if the spec should be updated according to consensus.
-Is it expected when SVG overlays XHTML, the transparent portions of the SVG should pass events through to underlying XHTML?
-Related to the above issue, if a transparent PNG SVGImage overlays SVG or XHTML, should events pass through to the underlying XHTML?
-Another item is that the text of the spec states, for example, "the button is clicked over an element", but we identify that for stroke-dashed array (in the example above), or whitespace surrounding text (i.e not the bounding box), events do not fire in some browsers unless it is actually on rendered pixels.  We think this is probably due to graphics library implementations and aren't sure if this behavior should change, but we thought we would note it for the WG.

Thanks in advance as I'll talk to folks on the WG today about these if there is room for discussion.

Patrick Dengler
Senior Program Manager
Microsoft

Examples for CSS <use>
Sibling Example:

<?xml version="1.0" encoding="UTF-8" standalone="no"?> <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN" "http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd">
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"  width="100%" height="100%" viewBox="0 -10 480 360" contentScriptType="text/javascript">


    <style><![CDATA[
        g + rect {fill: green}
    ]]></style>

 
    <g></g>
    <rect x="0" y="0" width="96px" height="96px" id="rectElem" fill="red" />

    <use xlink:href="#rectElem" x="96px" y="96px" />

</svg>

Child Example:

<?xml version="1.0" encoding="UTF-8" standalone="no"?> <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN" "http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd">
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"  width="100%" height="100%" viewBox="0 -10 480 360" contentScriptType="text/javascript">


    <style><![CDATA[
        g > rect {fill: green}
    ]]></style>

 
    <g>
        <rect x="0" y="0" width="96px" height="96px" id="rectElem" fill="red" />
    </g>
    

    <use xlink:href="#rectElem" x="96px" y="96px" />

</svg>

Descendant Example:

<?xml version="1.0" encoding="UTF-8" standalone="no"?> <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN" "http://www.w3.org/TR/2001/REC-SVG-20010904/DTD/svg10.dtd">
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"  width="100%" height="100%" viewBox="0 -10 480 360" contentScriptType="text/javascript">


    <style><![CDATA[
        g rect {fill: green}
    ]]></style>

 
    <g>
        <rect x="0" y="0" width="96px" height="96px" id="rectElem" fill="red" />
    </g>
    

    <use xlink:href="#rectElem" x="96px" y="96px" />

</svg>
Received on Monday, 25 January 2010 16:55:22 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:44 GMT