[rfc-i] For v3: Better identification for multi-document sets

Jim Schaad ietf at augustcellars.com
Sun Dec 29 10:36:00 PST 2013



> -----Original Message-----
> From: Julian Reschke [mailto:julian.reschke at gmx.de]
> Sent: Saturday, December 28, 2013 1:59 PM
> To: Jim Schaad; 'Paul Hoffman'; 'RFC Interest'
> Subject: Re: [rfc-i] For v3: Better identification for multi-document sets
> 
> On 2013-12-28 22:28, Jim Schaad wrote:
> >
> >
> >> -----Original Message-----
> >> From: rfc-interest-bounces at rfc-editor.org
> > [mailto:rfc-interest-bounces at rfc-
> >> editor.org] On Behalf Of Julian Reschke
> >> Sent: Friday, December 27, 2013 10:56 AM
> >> To: Paul Hoffman; RFC Interest
> >> Subject: Re: [rfc-i] For v3: Better identification for multi-document
> >> sets
> >>
> >> On 2013-12-24 04:42, Paul Hoffman wrote:
> >>> There are many multi-RFC BCPs and multi-document standards from
> >> another SDOs. Maybe we should add and element or attribute to make
> >> reference to these normalized.
> >>>
> >>> --Paul Hoffman
> >>
> >> +1
> >>
> >> See proposal from John Klensin in
> >> <http://trac.tools.ietf.org/tools/xml2rfc/trac/ticket/106> from which
> >> we should start:
> >>
> >>> It would allow the right thing to happen if, e.g.,
> >>>   (i) <reference> were modified to allow an "aggregate='yes'"
> >>> attribute so that, with that attribute turned on,
> >>>    <reference anchor="BCP0101">
> >>>       &rfc3777;
> >>>       &rfc5633;
> >>>       &rfc5680;
> >>>       &rfc6859;
> >>>    </reference>
> >>>
> >>> would suppress the individual anchors on output but otherwise do the
> >>> right thing.  And
> >>
> >> The tricky part is whether the container itself should allow certain
> >> child elements such as <title>, and also what to do with the
> >> seriesInfo elements
> > in
> >> nested <references> that repeat the container information.
> >
> > Another issue that needs to be considered is the question of recursion.
> > Should I be able to do
> >
> > <reference anchor="Keywords">
> >    <reference anchor="BCPXXXX">
> >      <reference anchor="RFC2119">
> >
> > Jim
> 
> It's an interesting thought; do we have a use case?

Part of this is a question of the best way of addressing things like the STD
and BCP references.  If there is an easy inclusion mechanism, it might be
better to include the RFC indirectly rather than directly.  This makes it
easier to include only the leaf reference.   This also makes it easier to do
references as documents change over time in some respects.  If I do the xref
to RFC2119 rather than to BCPXXXX then I would potentially get a different
tag and reference if the BCP is updated to point to a new document.  I would
then be able to make a decision later about what I am referencing.

This may be an interesting question if what we are going to keep is XML as
the base document.  It is not an issue if the XML has no includes, but if it
does then it may make a difference.


Jim

> 
> DTD-wise the right answer might be to pick a different name for the
> reference container.
> 
> Best regards, Julian



More information about the rfc-interest mailing list