RE: idea for HTML addition

> From:	Theatronix [SMTP:theatronix@navicom.com]
> 
> 
> One aggravating feature of the WWW is the automatic-forwarding action of
> some web pages that pass you on to some other page without any actions on
	[DJW:]  
	This is not a feature of HTML; it is achieved by one of
	three mechanisms:

	1) a proper Status 301, 302, etc. redirect from the web
	   server - I suspect this is not what you are describing,
	   and probably doesn't cause the problem.  It is the right
	   way of doing it.

	2) Scripting which sets document.location.

	3) Abuse of META to invoke the, non-standard and illegally
	   named++, Refresh HTTP header, and abuse of the Refresh header
	   to simulate case (1) by using a delay of zero.

> your part.  This is a big benefit UNTIL you try to go "BACK".  Then, if
> the
> time constant is set too short, you wind up where you started, rather than
> going to the previous webpage.  
> 
> A possible solution would be to add something that would alert the
> browser's "BACK" button to NOT just go back one step, but to go back 2
	[DJW:]  
	I doubt there is a problem in case 1.

	For case 2, the page has already used scripting, so they
	could use scripting to provide their own back button on the
	page.

	For case 3, you should take this up with the designers of
	the browsers that introduced this non-standard feature in the
	first place.  As Refresh is an HTTP, not an HTML feature, any
	standardisation should be addressed from the HTTP, not the
	HTML point of view.

	NB, in some cases, the breaking of the back button is probably
	considered a desirable side effect.

	++ Non-standard names should have an X prefix.  Refresh really does
	work as an HTTP header.

	[DJW:]  PS I'm pretty sure that this auto-redirect is on Jakob
	(Designing Web Usability) Nielsen's black list.
<http://www.useit.org/>

Received on Friday, 7 April 2000 07:36:44 UTC