Re: Conformance ratings

Eric Gauthier (eric@gauthier.centre.edu)
Mon, 12 Feb 1996 23:17:31 -0500 (EST)


From: Eric Gauthier <eric@gauthier.centre.edu>
Message-Id: <199602130417.XAA08381@roxanne.centre.edu>
Subject: RE: Conformance ratings
To: www-html@w3.org
Date: Mon, 12 Feb 1996 23:17:31 -0500 (EST)
Cc: eric@gauthier.centre.edu (Eric Gauthier)

Well,

First I'd like to clear up something I said in 
my original posting about bandwidth.  I messed 
up.  I meant that it might put an unnecessary
drain on their system/processor load.  Not that
parsing a short HTML file is difficult, but 
I'm sure they are constantly trying to speed
up their database searches.

Second of all, there has been a lot of discussion
about what "errors" should pass and what "extra"
stuff should pass (like netscape-acisms).  I believe
the whole point of this was to give people incentives
to conform to the HTML standards.  All we would have to
do is say "Yes, this page has no errors according to 
RFCxx".  Now, if the page also contains java links,
extensions, or non-standard tags -- Great!  But,
I was under the impression that we just wanted a way
to cut down on pages with blatent errors in them and
pages which, although a particular browser can handle, 
do not conform and can not be viewed.  

Now, you might add a second category like:
Netscape Enhanced <yes/no>
Jave/VRML   <yes/no>
Or something like that, but it should be EXTRA.  The rating
should merely reflect the standard.  Netscape, although adding
a bunch of extra stuff, should still be able to view the
original.  Oh well, this seems to have sparked MUCH debate.
Maybe this is a problem which needs to be tested in practice 
instead of just debated in theory.  I don't have an HTML parser,
but it shouldn't be to difficult to write.  Does anyone just
want to do this and see where it takes us?

Eric :)

PS: sorry about talking about netscape so much, its the browser
I'm most familiar with.  No bias intended... :)