[rfc-i] xml.resource.org

Nico Williams nico at cryptonector.com
Thu May 22 15:09:10 PDT 2014

On Thu, May 22, 2014 at 4:32 PM, Julian Reschke <julian.reschke at gmx.de> wrote:
> On 2014-05-22 22:45, Nico Williams wrote:
>> But I'll repeat myself: IETF URNs for RFCs (and I-Ds, and ...) refer
>> to RFCs (and I-Ds, and ...), yes, of course, but there's no reason
>> that they can't also refer to their corresponding bibxml entries when
>> used in a context that needs bibxml.
> In which case you're not using a URN, but just a string and overload the
> meaning. It's not much different from the xml2rfc seriesInfo element.

Not really.  If I say urn:ietf:rfc:2119 I'm clearly referring to
RFC2119.  I may or may not be referring to the text rendering of it,
or the HTML rendering, or...  It's indeterminate unless the context
makes it clear, and the default should be "the canonical published

And clearly we've been using "[RFCxxxx]" as a reference.

It'd be silly to paste in an RFC's contents in the middle of a
document just because the author of the latter referenced if as
urn:ietf:rfc:xxxx.  Context matters.

>> Nor is there any reason that we couldn't extend the existing URN
>> schemes to support "fragments" to identify specific aspects of RFCs
>> (and I-Ds, and ...).  "Aspects" like: specific output formats, bibxml
>> references, I-D tracker URL, ...
>> ...
> Yes, we could do all of that, but now it seems that you're re-inventing the
> Web :-)

Not really.  I explicitly don't want URLs.  I want something where
location is NOT part of the URI.  That makes it a URN.

> If we want reliable resolution of RFC metadata, then we should instead just
> make sure that this is an IETF service and that it gets the proper support.

That can be mirrored locally.  Such that I can point xml2rfc to any
number of mirrors without having to alter the XML inputs to xml2rfc
(which I just had to do to get past xml.resource.org being down, and
which I want never to have to do again).

I don't care what you choose to call this, or even if you choose not
to use URNs for it.  What I really want is to divorce location of
bibxml entries from my XML sources.  That seems like a perfect fit for
URNs; please explain how that is not so.


More information about the rfc-interest mailing list