From: Steve Putz <putz@parc.xerox.com>
To: uri@bunyip.com
Subject: Re: Summary of Re: URN wrappers and URL:
Message-Id: <93Oct18.143221pdt.2445@spoggles.parc.xerox.com>
Date: Mon, 18 Oct 1993 14:32:19 PDT
Perhaps the the URL: prefix debate is less important than we are making it.
Do we agree that most of the static URL references that now occur in
documents should really be URNs? We are now using URLs in documents
because the necessary light-weight URN creation and lookup mechanisms
are not defined and implemented yet.
URLs have their place and importance, but it is not properly as links
or citations in documents (including email). URLs are like numeric IP
addresses. They should be looked up, cached, and used on a temporary
basis only.
I sincerely hope the URN creation and lookup mechanisms we invent will
be easy enough to use that there will be no desire to use URLs
explicitly. URLs will fade into the background as part of the
retrieval mechanism for certain kinds of URNs.
* * *
So back to the question of whether there should be a URL: prefix to
distinguish URLs from URNs? Maybe it doesn't really matter.
Since there are strong arguments for continuing to accept the existing
URL prefixes (http:, gopher:, ftp:, etc.) in places like HTML
documents, the near-term benefit of introducing a URL: prefix is
minimal since no one will be using it for a while (for fear of breaking
client software). The long-term benefit is also minimal since by then
URL references will hopefully be replaced by URNs.