Re: Question for DNS propronents?

Ronald E. Daniel (rdaniel@acl.lanl.gov)
Wed, 14 Jun 1995 15:35:28 -0600


From: "Ronald E. Daniel" <rdaniel@acl.lanl.gov>
Message-Id: <9506141535.ZM25173@idaknow.acl.lanl.gov>
Date: Wed, 14 Jun 1995 15:35:28 -0600
In-Reply-To: Michael.Mealling@oit.gatech.edu (Michael Mealling)
To: Michael.Mealling@oit.gatech.edu (Michael Mealling), uri@bunyip.com
Subject: Re: Question for DNS propronents?

On Jun 14,  1:34pm, Michael Mealling wrote:
> Subject: Question for DNS propronents?
> This question is for those who propose to use the existing DNS namespace
> for URNs:
>
> How do you deal with information spaces and server hierarchies that
> don't correspond to actual DNS name/server hierarchies?
>
> For example, on our campus our Grants and Contracts department handles
> documents that are 'owned' by other departments. This information hierarchy
> is completely unrelated and almost completely opposite of how our
> DNS server/name tree is organized.
>
> How do you deal with it?

Hi Michael,

Good question. To some extent the solution for handling this will depend
on the policies at individual sites, but here are some suggestions.

1) Push the problem up a level. Make gatech.edu the naming authority
for such documents, not individual departments like G&C or an academic
department such as CompSci.

2) Just what is it that G&C is doing with the documents anyway? They
may help prepare them, check them, ... but they still belong to the
department that originated them. Publishe the thing under cs.gatech.edu,
but allow G&C to have write permission to the URC so that they can
play around with it.

3) Give G&C a domain - the lawyers will rule the whole pace soon enough
anyway?


This sort of thing is very closely related to establishing a workflow
procedure where a faculty member drafts a proposal and submits it to
the system in the "pray for NSF funding" category. Some luck administrator
at gatech has developed a workflow script that says such submissions are
routed to a choice of 2 peers for technical comments, the dept. head for
authorization, G&C for whatever, a vice-provost for final approval,
then back to the hapless faculty member to be Fed-Ex'ed to the NSF.
The script modifies the URC to give read permission to the approapriate
parties, and sends email notifications. The document's URN always
remains the same - whatever is assigned by this horrible workflow
monstrosity - but gatech.edu is the most likely publisher.

Later,