RE: [xsl] The "%" in DTD

Subject: RE: [xsl] The "%" in DTD
From: "Julian Reschke" <julian.reschke@xxxxxx>
Date: Sun, 17 Jun 2001 14:44:54 +0200
> From: owner-xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> [mailto:owner-xsl-list@xxxxxxxxxxxxxxxxxxxxxx]On Behalf Of Tobias Reif
> Sent: Sunday, June 17, 2001 2:10 PM
> To: XSL-List@xxxxxxxxxxxxxxxxxxxxxx
> Subject: Re: [xsl] The "%" in DTD
> Mike,
> > >
> > This is also wrong in that it calls character references "entities",
> calls them "Character entity references"
> some abbreviate that to "entities".
> > and,
> > as noted, fails to be a reliable reference when used with Netscape 4.x.
> NN4 has problems with rendering large tables; IE5 does that better;
> NN6/Moz can render progressively: the table rows get displayed as the
> are generated and downloaded (in the other two browsers, one has to wait
> for the complete table to be downloaded until the first line gets
> displayed)
> > All this reference does is demonstrates how your browser renders those
> > character references when it encounters them.
> Exactly.
> That's all it's intended for: it says:
> "Get numeric character entities for special characters in HTML, XHTML,
> XML, SVG, etc. ;
> see how special characters look like in different fonts;
> and check which are available in different fonts."
> I hope that it fulfills this.

No, it doesn't.

After looking at your page, people might think that &#153; is the correct
way to encode a trademark symbol in XML. It isn't, and next thing they'll do
is complain on this mailing list that their XML parser doesn't accept it.

> > It is heavily dependent on
> > the browser, the fonts you have installed, and the mechanisms
> your browser
> > uses for selecting fonts.
> It should; it's to be used to find out which character entity references
> work in which browser, and how the look in different fonts.

But then it shouldn't claim to give you any meaningful information for
anything XML related (because it doesn't).

> > Please, folks, if you are trying to find out what a character reference
> > *should* and *does* mean, just refer to the specs that define this stuff
> > without ambiguity -- XML, HTML and Unicode. Testing "&#12345;" in your
> > browser is not the way to go about it.
> Absolutely: then, after the correct numbers are found in the resources
> you mention, everyone working on real-world applications needs to test
> which work in which browsers with which fonts installed.

That's fine, maybe the title paragraph should state that.

 XSL-List info and archive:

Current Thread