Re: [Emailcore] Navigating 5321bis (and predecessors)

[John C Klensin <[e-mail]>]

--On Sunday, January 10, 2021 07:26 +0100 Julian Reschke
<[e-mail]> wrote:

>> I hope it is sorted out quickly --certainly before 5321bis is
>> ready-- but the xml2rfc vocabulary defines index ("<iref>")
>> elements which presumably means it is someone's problem (not
>> ours) to implement them in a way that is sensible for all
>> relevant output formats.  My personal opinion is that will
>> mean index entries have to end up targeting section numbers,
>> not page numbers, but I wait to see what others come up with.
>> However, unless there is a proposal and community consensus
>> to declare that markup element useless and/or obsolete, I
>> think we get to assume that they will work in some meaningful
>> way.
>> ...
> 
> +1, but note
> <https://trac.tools.ietf.org/tools/xml2rfc/trac/ticket/418>
> -- index generation currently is broken in xml2rfc (in v3
> mode).

That had been called to my attention after I noticed that
converting the (v2) xml for draft-ietf-emailcore-rfc5321bis-01
(which carries forward the index from RFC 5321) to v3 produced
perfectly good <iref> entries but that processing it into either
text or PDF produced some rather badly damaged output without
any index targets.

My perspective as editor is that the index was added to RFC 5321
after IETF Last Call and at the insistence of the IESG.  Unless
the IESG wants to undo that decision (and undo it RSN rather
than during or after IETF LC when the document gets there), I
hope the problem is fixed before the WG finishes its work.  The
fix should ideally reference section numbers, not page numbers,
for obvious reasons.  Otherwise either the RPC is going to have
a really interesting problem on its hands or this WG should stop
its work until the tools are ready.

So I agree with your end of December priority change from
"minor" to "major".

    john