Re: draft-finseth-eid-url-scheme-00.txt

Craig A. Finseth (fin@finseth.com)
Thu, 5 Nov 1998 09:32:39 -0600 (CST)


Date: Thu, 5 Nov 1998 09:32:39 -0600 (CST)
Message-Id: <199811051532.JAA27230@isis.visi.com>
From: "Craig A. Finseth" <fin@finseth.com>
To: masinter@parc.xerox.com
Cc: uri@Bunyip.Com
In-reply-to: <003401be0849$af0aada0$aa66010d@copper.parc.xerox.com>
Subject: Re: draft-finseth-eid-url-scheme-00.txt

   Your proposal for an "eid" URL scheme fails the first
   fundamental criterion for a "Uniform Resource Locator",
   namely that it be "Uniform", and have the same meaning no
   matter what context.

Three points,

Frist, not all existing URLs have this property (consider "file:").

Second, Many -- if not most -- data transfer and specification
protocols have provisions for what is essentially "vendor private"
data.  It is my opinion that the URL scheme _should_ have made
provisions for this type of escape mechanism at the outset.

Third, you are making an incorrect assumption.  _Within_ any given
implementation, I expect that the "same meaning" concept would still
apply.  It's just that in this case it is up to the implementation to
ensure that its usages are kept within the system.

So, why go to all this bother?  Why not just keep the private data
private?

The reason is that people who make these implementations need to be
able to carve off a space that they know won't be used by anyone else.
Again, a standard concept.

   Please reconsider,

You do point out a significant point of confusion and I am certainly
willing to rewrite the document to clarify the justfication.

It may also make sense to publish it as an Informational rather than
standards track RFC.

However, I do feel that the concept is an important one and it should
be part of at least the RFC (as different from Internet Standards)
process.

Craig