Re: UA: indicate missing content

Hi Jonathan,

> on what basis were height and width excluded for <use> in svg1.2 spec?

It would appear to be on a technical basis.

In 1.1 the width and height attributes of <use> are only meaningful when 
referencing <svg> or <symbol> elements.  The attributes are used to form 
a new viewport for the referenced content.  The width and height 
attributes are not used for references to any other elements.

In 1.2T the <use> element cannot reference an <svg> element, and there
is no <symbol> element.  There can be no new viewport, so the width and
height attributes do nothing.

> there is a clear use case that the author wishes to specify a
> different area to that in the remote file.

Yes, there are use cases for specifying a new viewport.  That is an 
argument for allowing child <svg> elements in 1.2T.

> it would also provide one method to specify the size of a checkbox in
> the case the resource is not loaded.

In practice, the width and height attributes
  a) may control the size of the referenced content (1.1 only)
  b) may have no influence on the referenced content
  c) may be absent
  d) only one is present (subset of (a))

In case (a), there is a meaningful default region for fallback content. 
  Otherwise, only (b) is applicable for your method and in this case the 
author is defining a known or nominal fallback region of their choosing.

Not so different to Doug's proposal (in part below) which goes a long 
way to consistently addressing the issue for all elements which can 
reference content

> the user agent must display instead any fallback content supplied as
> a child of the element

If the author wants meaningful fallback then the author can define their 
own meaningful fallback.

eg
<use xlink:href="another.svg#unresolved" x="50" y="50">
  <text x="10" y="20" font-size="10">:(</text>
</use>

or

<audio xlink:href="welcome.unsupportedaudioformat" begin="0s">
  <use xlink:href="#myMissingAudioIndicator" />
</audio>


Ken Stacey


On 29/07/2008 8:15 PM, Jonathan Chetwynd wrote:
> My email was to erik,
> 
> however...
> 
> on what basis were height and width excluded for <use> in svg1.2 spec?
> 
> there is a clear use case that the author wishes to specify a different 
> area to that in the remote file.
> 
> it would also provide one method to specify the size of a checkbox in 
> the case the resource is not loaded.
> 
> regards
> 
> 
> my logo <http://www.openicon.org>
> 
> Jonathan Chetwynd
> 
> j.chetwynd@btinternet.com <mailto:j.chetwynd@btinternet.com>
> http://www.openicon.org/
> 
> +44 (0) 20 7978 1764
> 
> 
> On 29 Jul 2008, at 10:13, Doug Schepers wrote:
> 
>>
>>
>>> the case against re-introduction of height and width needs to be 
>>> stated clearly, hence my request to erik
>>
>> Your meaning here is unclear.
>

Received on Wednesday, 30 July 2008 04:57:21 UTC