Re: [xsl] Understanding why <tag></tag> is the way it is (was Re: [xsl] IE Client side transformation issue)

Subject: Re: [xsl] Understanding why <tag></tag> is the way it is (was Re: [xsl] IE Client side transformation issue)
From: Norman Gray <norman@xxxxxxxxxxxxxxx>
Date: Fri, 3 Aug 2007 17:33:53 +0100
On 3 Aug 2007, at 16:45, David Carlisle wrote:

<b/>abc

If you send it with an html mime type, then what you _should_
get is the same as if the input had been >abc with a visible > sign, but

This reminds me that I got my previous account of the sequence of SGML syntax curiousities slightly wrong.


Just for the sake of insane completism: the 'NET' hack was the enabling of the feature where, if the start tag ended with '/' rather than '>', then the next '>' closed the element, without requiring a complete close tag. Thus "<b/bold text>" was valid, and equivalent to <b>bold text</b> (!). Thus, immediately, <b/> generated an element with no content. I think. I'm not sure I could face ISO-8879 again[1].

David is right: I don't believe there was _ever_ a browser which contained an SGML parser.

Norman

[1] It disturbs me that I didn't have to look this number up.


-- ------------------------------------------------------------ Norman Gray : http://nxg.me.uk eurovotech.org : University of Leicester, UK

Current Thread