OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   Re: [xml-dev] Re: Can XLink be fixed?

[ Lists Home | Date Index | Thread Index ]

Simon St.Laurent wrote:

>>Having said that, I think that certain HTML idioms, such as
>>
>><img src="URL of picture" alt="local text" longdesc="URL of text">
>>
>>are really bad markup design and aren't worth grandfathering.  Why
>>bad? Suppose you want to have descriptions (long or short) in more
>>than one language, and suppose you want to provide multiple sizes of
>>the picture, etc; 
> 
> The URI people keep telling me that's a job for content-negotiation,

Perhaps, but content-negotiation is kind of a blunt instrument and isn't 
going to help you with for example different picture resolutions or 
expert-vs-introductory level material or a lot of other things.

I think that if for any X, you want to use markup to represent it, then 
if you're really sure there's only going to be one of them, attributes 
are OK, but if there's a chance there's going to be more than one, then 
elements are the way to go.  This is one of the motivators for the 
design of the complex XLink.

> I guess
> there's a good question about whether the longdesc resource is just a
> representation of the same resource identified by the src attribute...
> How deep do you really want to go with this?

Without going very deep, the current way IMG is done just feels like bad 
markup design for a bunch of fairly obvious reasons.

>>in fact, while you can argue about the appropriate
>>uses of elements and attributes almost forever, this is a good
>>example of how not to use attributes.  
> 
> > I dunno about that, unless maybe we want to get into requiring all
> URI-containing values to be child elements rather than elements.  I
> don't see any reason that src and longdesc don't both qualify as
> metadata about the img element, though I'm sure I could come up with
> something vaguely plausible.

I've never found it very helpful to agonize over what's metadata and 
what isn't.  I think there are simpler decision procedures: want 
white-space normalization?  Use an attribute.  Possibly multiple 
occurrences?  Use an element.  I don't think whether or not it might be 
a URI is part of the argument. -Tim






 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS