Message-Id: <199410251949.PAA15534@postman.osf.org>
To: brennan@hal.com (Dave Brennan)
Subject: Re: Current URN syntax is unacceptable
In-Reply-To: Your message of "Tue, 25 Oct 1994 01:29:09 CDT."
<9410250629.AA17292@hysteria.hal.com>
Date: Tue, 25 Oct 1994 15:49:11 -0400
From: "Norbert Leser - OSF DCE: (617)621-8715" <nl@osf.org>
>> For the purpose of simple name comparison only, instances (such as
>> client interpreters) that don't have the naming authority's knowledge,
>> case matching is insensitive and white spaces are not significant.
>
> Why does it need to be case insensitive? If the string is opaque,
> then so be it. Why can't a client just do a straight comparison?
I cannot answer your first question why it "needs" to be case insensitive.
I've posed the same question before and couldn't get a satisfying answer.
One could certainly list pros and cons for both (case exact and case
insensitive matching).
Anyhow, I think that for this "simple name comparison" purpose, this case
insensitiveness appears to be an acceptable compromise. This would at least
accommodate naming systems such as DNS (and actually also X.500 that
specifies case insensitive for RDNs) whose names are case insensitive.
Clients that don't know the service's matching rules would then be able
to process names more efficiently in terms of network traffic
(i.e., use caches and other means). This doesn't mean that there are no
other ways in dealing with this issue (as we did in XFN, for instance),
nor does it mean that it doesn't cause other nasty problems (i.e.,
ambiguities and potentially wrong mappings) in case-exact type of naming
systems. I'd therefore be more than happy, if we could come up with
another solution.
It was implicit in my note, but I want to point it out here specifically:
I'm against the proposal in the URN draft that any client or proxy performs
a conversion of names to lower case (i.e., it's suggested in the BNF there).
Names should be passed on "as is". Otherwise, the problem with case-exact
type naming systems would become a very serious one. The case
insensitiveness is only for comparison, in order to increase hit rates of
caches and such.
Norbert