Re: [URN] Protocol Action: URN Syntax to Proposed Standard

["Martin J. Duerst" <[e-mail]>]

On Thu, 1 May 1997, The IESG wrote:

[This mail is currently only addressed to the URN group.
I will write to the IESG if the need to do so arises from
our discussion.]


>   The IESG has approved the Internet-Draft "URN Syntax"
>   <draft-ietf-urn-syntax-04.txt> as a Proposed Standard.

> Working Group Summary
> 
>    There has been considerable discussion in the working group,
>    regarding the compatibility of URNs with URLs (including the
>    notion of relative URNs) and the possible internationalization
>    of URNs.  This document leaves relative URNs (with syntax
>    compatible with that of URLs) for future study, by reserving
>    the '/' character.  Internationalization is also left undefined,
>    but this document specifies that URNs must be transmitted in
>    a pure-ASCII representation, and that all URN aware applications
>    must make URNs available for display in the pure-ASCII form
>    to enable reliable transcription by humans.
> 
>    With these compromises, there is strong working group consensus
>    on the document.

While I agree that there was indeed a good working group consensus,
the above explanation, in particular the phrase "Internationalization
was also left undefined" is not very fortunate. It had me fear that
the consensus about UTF-8 was actually thrown out at the last minute,
or whatever. It also gives a bad impression to other protocol writers
in that it seems to say that it is okay to not address internatio-
nalization issues. This is not what we have done. We indeed have
addressed internationalization issues in a way suitable for URNs.

I would therefore suggest that we ask the above text, from the
word "Internationalization", to be rewritten as follows:

Internationalization was addressed by specifying that URNs must
be transmitted in a pure-ASCII representation and by defining an
encoding based on UTF-8 for characters outside ASCII for legacy
namespaces. User interface issues were not addressed but for the
requirement that URN aware applications must make URNs available
for display in the pure-ASCII form to enable reliable transcription
by humans.


Regards,	Martin.