On 8. Dec 2024, at 02:15, Stephen Farrell <stephen.farrell@xxxxxxxxx> wrote: > > > Hiya, > > On 07/12/2024 23:51, Carsten Bormann wrote: >> any I-D) being inappropriate to cite > > Well, I-Ds truly are not great things to cite because: > > - if you only cite the file name(e.g. [1]) then the > content may have changed when the reader gets to it […] > [1] https://datatracker.ietf.org/doc/draft-thomson-gendispatch-no-expiry/ Don’t do that then :-) Adding back rfc-interest, because the way we cite I-Ds in RFCs is quite considerate for some of this. References to I-Ds in RFCs include the revision number in the bibliography entry. E.g.: https://www.rfc-editor.org/rfc/rfc9595#CORE-COMI ...references revision -18 in the citation. The reference here is an example where it actually makes sense to (and the intent is to) reference the entire progression of revisions — this is an informative reference that just says “this is one way the present document is intended be used”. > - if you cite a specific draft number and a newer draft > is ever created the reader won't know which was meant unless > the author called that out, which is extremely rare - much > more common would be that neither author nor reader really > know any of these IETF/I-D minutiae. As I said, references to I-Ds in RFCs always point to a specific revision. We indeed often keep it implicit whether we mean that specific revision. Sometimes we actually explicitly reference a specific historic revision, e.g. in: Reference to progression of I-Ds: https://www.ietf.org/archive/id/draft-bormann-nemops-coreconf-00.html#I-D.ietf-core-comi References to specific revisions: https://www.ietf.org/archive/id/draft-bormann-nemops-coreconf-00.html#I-D.draft-vanderstok-core-comi-00 https://www.ietf.org/archive/id/draft-bormann-nemops-coreconf-00.html#I-D.draft-vanderstok-core-comi-05 Here, the information that a specific revision is intended is expressed by the revision number in the anchor (which is sometimes lost during RFC editing, unfortunately), but also should be clear from the (historic) context of the citation. (I’ve long argued that there should be a third subsection of the References section besides normative and informative: historic.) > I don't think the above is at all affected by supposed > expiry. I agree. If I follow an I-D reference and find an expired draft then this is a hint that the draft may no longer be pursued. (I could look at the date of the draft and find the same information, but agreeing on a cadence for how often a draft should be updated is a useful way to have datatracker indicate that the draft may no longer be active.) > I-Ds can be very useful things to reference and some such > references are done well, but most in the academic literature > are done seemingly carelessly or without really understanding > what can change. Indeed, there are additional considerations for references in academic literature, and here the “work in progress” designation may be useful (or may be wrong and should be “work no longer being pursued” for some historic references). Grüße, Carsten
