To: timbl@www0.cern.ch
In-Reply-To: timbl@www3.cern.ch's message of Sat, 9 Jul 1994 14:49:07 -0700 <94Jul9.144915pdt.168963@nebula.parc.xerox.com>
Subject: Re: Another snapshot of the URL document.
From: Larry Masinter <masinter@parc.xerox.com>
Message-Id: <94Jul20.161116pdt.2761@golden.parc.xerox.com>
Date: Wed, 20 Jul 1994 16:11:05 PDT
I'm making another pass through the URL document (belatedly, as
usual), and trying to deal with the various comments.
Guido van Rossum said:
>> Unfortunately it seems that there are now (for some protocols) four
>> different categories: characters that are safe and reserved
>> (e.g. user@host is different from user%40host),
and Tim Berners-Lee replied:
>No -- @ is the same as %40 in all cases. The user and host strings
>may neither contain either.
but this seems unfortunate:
Do you really intend to disallow %40 in user names? If there's a
system which requires you to login with user= your email address, you
won't be able to supply a URL for it? This seems unfortunate and
unnecessary. I think making "@" reserved will solve a lot of problems
and not introduce new ones.
In addition, adding "&" to the list of characters that may be reserved
in specific schemes doesn't add more problems and corresponds well to
current practice.
I hope to have a new draft out tonight...