W3C home > Mailing lists > Public > www-amaya@w3.org > July to September 1999

RE: error tolerance within amaya

From: John Russell <VE3LL@RAC.CA>
Date: Wed, 8 Sep 1999 08:17:00 -0400
Message-Id: <199909081210.IAA07676@admin.cgocable.net>
To: www-amaya@w3.org
verifiers and editors should of course be error intolerant so that
well formed documents are prepared......
however BROWSERS should be error tolerant to allow the viewer
to get the message even if there is a bit of noise in it.....

the web is about COMMUNICATIONS, not about correctness.
the more correct we are, the better our image but if the 
content isnt there or meaning passed to receiver then the most
correct document is useless.......

having done my editorial for the day and if Amaya Is a BROWSER
that MAY receive a bad colour situation   then what should it do ------
[1] neglect the command and use no color --- that is go to default setting
[2] set the color to black !!!!
i think #1 is correct option as this leaves things readable (not correct
but readable)  
if it were an editor/verifier  it should raise an error message saying gris
is not an acceptable colour .....

Date forwarded: 	Wed, 8 Sep 1999 07:11:17 -0400 (EDT)
From:           	Dave  J Woolley <DJW@bts.co.uk>
Send reply to:  	www-amaya@w3.org
To:             	www-amaya@w3.org
Date sent:      	Wed, 8 Sep 1999 12:09:04 +0100 
Subject:        	RE: error tolerance within amaya
Forwarded by:   	www-amaya@w3.org

> > From:	John Russell [SMTP:VE3LL@RAC.CA]
> > 
> > if a style sheet  happens to miss on the background
> > setting  for example    background : gris
> > in amaya i get a solid black background.........
> > in explorer it just says lets go with the default setting of
> > the browser
> > 
> 	HTML editors, and browsers used to verify pages,
> 	should never apply error recovery, except that an
> 	editor may apply error recovery by correcting the 
> 	source.
> 
> 	One of the reasons for the fact that most web pages
> 	are invalid HTML is that people validate them with 
> 	browsers which attempt error recovery, so they end up
> 	designing for the error recovery behaviour, not to the
> 	specification.
no the reason is that many do not validate at all ......
how many people even use their spellcheckers when 
virtually every wordprocessor and mail utility has them built in!



john russell  VE3LL@RAC.CA

homepage: http://web.cgocable.net/~jrussel
Received on Wednesday, 8 September 1999 08:09:44 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 April 2014 11:01:33 UTC