[rfc-i] Proposal for v3 to simplify most references

Paul Hoffman paul.hoffman at vpnc.org
Sun Feb 9 15:13:01 PST 2014

On Feb 9, 2014, at 2:39 PM, Nico Williams <nico at cryptonector.com> wrote:

> On Sun, Feb 9, 2014 at 10:14 AM, Paul Hoffman <paul.hoffman at vpnc.org> wrote:
>> On Feb 8, 2014, at 9:20 PM, John Levine <johnl at taugh.com> wrote:
>>> If that's in some RFC published today, and in a couple of years, we
>>> publish RFC 10979 that updates the IPR rules and becomes the new BCP79,
>>> what will people get when they render the XML?
>>> This isn't an insoluble problem, but if the XML is definititive, even
>>> the current situation implicitly depends on the entries in the
>>> citation libraries never changing.
>> Hrm, good call. It works for BCPs that have changed, but not ones that will change.
>> So, in the proposal I sent, make the "expanded" attribute mandatory for "bcp" and "std".
> When formatting an RFC.  For an I-D it makes no real difference.  If
> you want ease of use, why make the I-D author look up the RFC number
> (and then, when the BCP is changed to refer to a newer RFC, the I-D
> author will get nits about it).

That seems like a *feature*. That is, an author who is referring to a BCP kinda needs to prove she or he knows what is being referred to. The RFC Editor could trivially host a BCP->reference tool. In addition, the nits checker could in fact say "you didn't specify the 'expanded', and you probably meant '3456,4567', but you should look that up <here>".

> On the whole, I think a notion of a) reference libraries, b) standard
> reference libraries with standard names in it, c) the ability to
> assign vanity names to anchor for use intra-document, and d) the
> ability to assign vanity names for what should appear in [brackets] in
> the rendered text... will be very welcome.
> Something like:
>    <library name="RFC-Editor"/>
>    <library name="my-lib" ref="URI to my reference library">

I hope you don't mean that for published RFCs.

>    <reference library="RFC-Editor" ref="RFCxxxx" anchor="foo"
> display_name="FOO"/>

In such a case, what would be the purpose of both "anchor" and "disply_name"? Is the latter only for people who want to start reference names with numerals?

--Paul Hoffman

More information about the rfc-interest mailing list