Re: SVGt 1.2: processing invalid external URI references

David,

whether "much more catastrophic" may be dependent, however it does  
seem evident that this instance is something that it may be helpful to  
deliberate publicly, seek opinions and incorporate into SVG standards.

it might for instance be considered appropriate to include a  
description of the need for appropriate title content when linking to  
external references.

equally it may be useful to discuss with, provide advice to and learn  
from UA developers at this stage.
ie afaik no UA provides either title content, or a 'missing image'  
graphic.
It's also not clear to me at least how this would work.

as you say, small details can dramatically change the meaning of an  
image.
whereas in the past we know that copyists frequently amended text,  
whether deliberately or not.
We have little current knowledge as to how to usefully phrase advice  
on this matter, regarding graphics.

regards

~:"

On 2 Apr 2008, at 11:28, David Woolley wrote:

>
> Jonathan Chetwynd wrote:
>
>> would it not be necessary or at least helpful, if as expressed and  
>> provided for in html, that a marker should be displayed, indicating  
>> that an image, or some graphical content is missing?
>
> I don't believe that HTML mandates the behaviour for broken images  
> and there are at least two de facto behaviours:  to display a broken  
> image icon; and to fall back to the alternative text as though the  
> browser couldn't handle images at all for that image.  HTML also  
> provides object, which if implemented properly, would give finer  
> control, but, in practice doesn't do images well.
>
> SVG is a presentational language, so a failure to display an image  
> is much more catastrophic.  A viewer may well want to limp along in  
> that case, but the missing image is likely to have seriously  
> compromised the author's intent, whatever the the viewer does as a  
> workaround. (I'm sssuming HTML used as intended, rather than the  
> create a specific visual design, even though the latter is rather  
> common.)
>
> I'm assuming HTML 4.01.
>
> -- 
> David Woolley
> Emails are not formal business letters, whatever businesses may want.
> RFC1855 says there should be an address here, but, in a world of spam,
> that is no longer good advice, as archive address hiding may not work.
>

Received on Wednesday, 2 April 2008 12:10:24 UTC