Re: [xml2rfc] xml2rfc online: references to I-Ds and associated dates

[Julian Reschke <[e-mail]>]

On 14.07.2019 00:17, John C Klensin wrote:
> ...
> I fear we are not communicating.
> ...

Are we not?

> Let me give you two reasons.  The second one is the important
> one; I wonder whether the first one is enough of a distraction
> that I should omit it.

Examples are good.

> (1) Suppose that there was some external event or announcement
> that interacted with the subject matter of an I-D (a product
> being shipped or the announcement of a really nasty vunerability
> come to mind as possible examples).  It would then be helpful
> for the reader to be able to tell, quickly, whether there is any
> chance that the I-D might have addressed those issues.
> Certainly that date information could be determined by going to
> the tracker with the version number and looking up the date
> there, but I don't think our goal should be to see how many
> hoops (or page clicks) we can get the reader to jump through.  I
> hope you agree.

I agree with the goal, but not with the conclusion.

First of all, with a sane format, you need exactly one click to get to
the referenced Internet Draft, and that will have the date on the front
page. (This is true today for the HTML version on tools.ietf.org, and
will also hopefully be the case for xml2rfc's v3 HTML output).

That said, it is already possible to annotate a reference with any prose
you like. So if you want to say something like "this draft addresses the
security issue raised by xxx on yyy", you can already do that.
(<https://greenbytes.de/tech/webdav/rfc7749.html#element.annotation>,
<https://mailarchive.ietf.org/arch/msg/xml2rfc/JGsffnj-JKZRF80qvh6dLB_4r0w>).

> (2) As long as the purpose of posting documents as I-Ds includes
> allowing authors or groups to expose ideas to the community for
> discussion, editorial decisions about what belongs in those
> documents should be in the hands of those authors.  Neither the
> community, nor the authors of a production tool that is
> recommended by the community for use with I-Ds, should make
> stylistic decisions that affect what information is presented
> without a compelling reason.   So the question isn't why I might
> want to display the day of the month, it is why you (or some
> other tool author or specifier) should tell me I can't have it
> displayed.
> ...

xml2rfc tries to formalize/automate things that historically happened at
RFC production time. One of these things is consistent style of
bibliographic references. The goal is to produce the format that the RFC
Style Guide asks for.

Now that is for RFCs. Using xml2rfc to produce *Internet-Drafts* is a
slightly different use case. My understanding is that people use xml2rfc
for this because it will reduce the amount of formatting changes when
going to RFC. So here the question would be: do we need more freedom in
references styles in Internet Drafts, as opposed to RFCs? That's
definitively an interesting discussion, and maybe the outcome is that we
need to make changes on the RFC Style Guide side.

And yes, I understand that what you say applies to all of the text, not
only reference styles. That's a broad topic, but then, I'd prefer to
discuss concrete use cases, not the whole philosophy.

Best regards, Julian