To: miked@cerf.net
In-Reply-To: miked@cerf.net's message of Thu, 25 Aug 1994 15:12:54 -0700 <94Aug25.151310pdt.2766@golden.parc.xerox.com>
Subject: Re: 'relative URLs', status
From: Larry Masinter <masinter@parc.xerox.com>
Message-Id: <94Aug26.023042pdt.2760@golden.parc.xerox.com>
Date: Fri, 26 Aug 1994 02:30:37 PDT
> I would personally agree, but then the issue becomes: Is the spec "what
> we would all like someday", or "what is currently in practice" ?
The practice is to reach consensus on what the spec should be.
Certainly current practice influences the standard, but does not
control it.
In this particular case, however, there isn't a serious issue: current
browsers handle many cases that are 'illegal' but currently are widely
used. We don't need to add things to the specification if browsers
expect them.
If the use of 'relative URLs that use a scheme' was widespread, or had
some demonstratable utility, I'd push for retaining them, but I can't
actually see any interpretation for which they are useful.
Larry