W3C home > Mailing lists > Public > uri@w3.org > January 2005

Functional requirements for URI scheme processes.

From: Weibel,Stu <weibel@oclc.org>
Date: Thu, 20 Jan 2005 10:12:30 -0500
Message-ID: <8CC50D49B6828C4FBAB7DA1FCAB0526A2713DB@OAEXCH1SERVER.oa.oclc.org>
To: <uri@w3.org>

Larry Masinter writes:

> In any case, if you don't like the proposed process, 
> what would you propose in its place? 
> This seems to me like the best we can do, given all the requirements.

I'd like to have a go.  Please (anyone) help me understand the
requirements and design constraints.  Here is my current understanding:

1. Allowing unrestricted reservation of unique provisional URI scheme
tokens permits persons or organizations to speculatively reserve such
tokens. Many tokens are unlikely to have much branding value (eg. mms:).
Others will be desirable either because they already have brand
recognition, or because they have innate semantics that are attractive
and concise: (ID: DOI: DATA: are possible examples).  The social and
business value of reserving and protecting these tokens is potentially

2. Reserving a token that then lapses for whatever reason removes a
potentially desirable token from consideration by others; a good policy
should provide a recycling mechanism for abandoned tokens.

3. Registration and review processes are costly in proportion to the
human attention they require for effectiveness.  Policies, registration
procedures, and review requirements must be designed to incur the
minimum of human intervention consistent with effectiveness.

4. Policies should provide incentives (or at least avoid disincentives)
for good net citizenship (avoiding name collisions, averting name
speculation, etc.).

5.) Compliance with policies should be subject to quantitative
assessment, not philosophic or ideological assessment.  This constraint
is in a sense a derivative of #4:  negotiating ideological positions is
very costly in terms of time, and occurs again and again (the ground-hog
day effect).  This is not to say that there are no architectural
principles or technological philosophies embedded in the policies, but
rather that once agreed, compliance with those principles should be
determined largely algorithmically.

What have I got wrong? Left out?


Stuart Weibel
Senior Research Scientist
OCLC Research
Received on Thursday, 20 January 2005 15:12:59 UTC

This archive was generated by hypermail 2.4.0 : Sunday, 10 October 2021 22:17:46 UTC