W3C home > Mailing lists > Public > www-html@w3.org > January 2001

RE: 2 HTML tags in one document

From: Dave J Woolley <david.woolley@bts.co.uk>
Date: Fri, 19 Jan 2001 11:43:04 -0000
Message-ID: <81E4A2BC03CED111845100104B62AFB50102A3E6@stagecoach.bts.co.uk>
To: "'www-html@w3.org'" <www-html@w3.org>
> From:	wong yenchin [SMTP:galilee5@hotmail.com]
> 
> Neither Javascript : URLs or the meta-finish seems favourable to u.  
> Javascript : URLS should be banned and meta-finish is non-standard. That's
> a 
> bit tough.
> 
	[DJW:]  Firstly, javascript URLs are URLs, not HTML, so 
	off topic, and Refresh purports to be an HTTP header (and is
	actually recognised as such by the big 2, even if it appears in
	no HTTP standard - you can have self refreshing JPEGs with no
	HTML wrapper) so is also off topic.

	javascript: URLs can, I believe, always be replaced by onclick 
	(but note my recent note about really needing an onactivate) and 
	doing so allows more accessible HTML in that it frees href for
	a non-scripted link (I get thoroughly frustrated by the number of
	dead links resulting from javascript:), although it doesn't prevent
the
	increasing problem of href="#" with an onclick popup.  Good
accessible
	design would, in any case, start with a valid href and then augment
it with
	the onclick.

	The on topic issue with meta/refresh is whether meta http-equiv
serves
	any valid purpose.  The only reliable way of getting HTTP headers
actioned
	is to put them as real HTTP headers.  Proxies do not look at meta,
so
	you cannot do proxy cache control with metas, and not all browsers
look
	at it, even though the HTML spec makes a special exception for
content-type,
	but only when there is no character set in the HTTP headers.  The
reason
	that the use of meta http-equiv with CGI was questioned is that,
	if you have CGI permission on a server, you can always generate
proper
	HTTP headers (there are basically two reasons for the prevalence of 
	meta/http-equiv:  
	- hosting sites that charge extra to properly configure the server;
	- ignorance and the single tool psychology that wants to do
everything in
	  one place.)

	In theory, the only reason for meta/http-equiv, is to instruct the
	server to add the relevant real headers, but none that I have used
	do that - it involves them in treating HTML specially.

	Refresh with a timeout of zero is bad, because the same effect can
be had
	with real HTTP headers and works for all HTTP 1.0 onwards browsers. 
	Also, with a fully compliant browser, you can even redirect in a way
that
	causes bookmarks to be automatically corrected.  Also, on some, if
not
	all browsers, it makes it difficult to get back past the page using
the
	back button. 

[DJW:]  

-- 
--------------------------- DISCLAIMER ---------------------------------
Any views expressed in this message are those of the individual sender,
except where the sender specifically states them to be the views of BTS.
>  
Received on Friday, 19 January 2001 06:43:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:45 GMT