W3C home > Mailing lists > Public > www-style@w3.org > October 2000

Re: Entities in XML (Re: New version of the Selectors module of CSS3)

From: Simon St.Laurent <simonstl@simonstl.com>
Date: Mon, 09 Oct 2000 08:40:41 -0400
Message-Id: <200010091237.IAA04270@hesketh.net>
To: bert@w3.org, <www-style@w3.org>
At 11:52 AM 10/9/00 +0200, Bert Bos wrote:
>Of course, XML is just a syntax, and you could use it in any way you
>want. But you may encounter difficulties, since most XML tools
>(parsers, transformation tools, renderers, etc.) implicitly assume a
>certain way of using XML. E.g., as you discovered, CSS has no way of
>styling an entity, since it assumes it only ever has to render the
>expansion of the entity.

I'd agree entirely, if it weren't for the fact that XML allows
non-validating parsers to skip the resolution of entities entirely if
external resources (DTD or entity content) are involved.

Because XML allows this, I'd strongly suggest that CSS have some way of
presenting that information, if only to warn readers that they may not have
a complete copy of an XML document. 

(I've had enough trouble explaining this behavior to programmers that I
doubt those just trying to read documents are going to be able to figure it
out on their own.  Style sheets can provide an unfortunate layer of masking
incomplete document structures, since XML 1.0 doesn't specify how
unresolved entities should be treated.)

If it were possible, it'd be great to support some mechanism for letting
users specify that entities should be resolved, but that would likely
require starting the behaviors food fight all over again.

It's not the fault of CSS, but it'd be great to see something done here.

Simon St.Laurent
XML Elements of Style / XML: A Primer, 2nd Ed.
XHTML: Migrating Toward XML
http://www.simonstl.com - XML essays and books
Received on Monday, 9 October 2000 08:37:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:06 GMT