[Tools-discuss] Re: Finding authors [was: [e-mail] is not long for the world]

[Brian E Carpenter <[e-mail]>]

On 16-Aug-24 18:22, Toerless Eckert wrote:
> I would love a rendering option of RFC which has updated metadata.

I agree, but that's the main reason for always citing an RFC via its info page, e.g.
https://www.rfc-editor.org/info/rfc6826
There you learn that the RFC has been updated by another, and that it has no confirmed errata.

As far as I can see, showing current email addresses (when known) is "a small matter of programming." But of course they are not in the metadata today.

However, this line is presumably generated from the metadata:

>> Discuss this RFC: Send questions or comments to the mailing list [e-mail]

    Brian


> Such as current standard track status - and maybe current email
> addresses of authors/contributors - although i am aware of an important
> use-case for this. Maybe it is not a bad thing if you need to apply some
> IETF knowledge to do the mapping. Old authors may like their peace an quiet ;-)
> 
> Doing the mapping today is not really be that big a problem
> for somebody who loves to do web programming to do this. But agreeing
> on some standard add-on header with such "current" metadata would be nice
> anyhow.
> 
> The mapping from old-email to new email is already publically possible
> with some heuristics:
> 
> https://datatracker.ietf.org/person/<old-email>
> 
> Gets you the persons datatracker page
> if <old-email> was once captured by datatracker due to role or draft/rfc.
> 
> Then you need to apply the heuristic to find one or more "current" email
> from the role section or latest draft/rfc section on that page. Aka: what
> i guess we're all doing manually anyhow when we need to find such an email.
> 
> Of course that's just a heuristic.
> 
> Most simple way to allow datatracker persons to post an "authoritative"
> personal email on datatracker would be to add a tag "personal_email"
> to the list of tags for "Additional person resources".
> 
> Cheers
>      Toerless
> 
> On Thu, Aug 15, 2024 at 05:12:52PM +1200, Brian E Carpenter wrote:
>> [Attempting to switch lists...]
>>
>> The topic is finding current addresses for authors of older RFCs.
>>
>> George,
>>
>> We cannot change the RFC text; that's a rule. So the solution has to be external to the RFC itself. (See my comments in line below.)
>>
>> The datatracker already has the information needed. Taking myself as an example, it knows that [e-mail], [e-mail], [e-mail], [e-mail], and [e-mail] are all the same person. Also, my datatracker profile knows which address is currently primary.
>>
>> Therefore, writing a function that delivers current addresses for the authors of RFC N, in cases where the tracker has this information, is entirely possible, for people who know how to access the datatracker's database. authors_of_rfc(2119) would return ("[e-mail]") authors_of_rfc(1671) would return ("[e-mail]")
>>
>> So we could have an API and a web tool for this.
>>
>> Whether the effort for this is justified is for discussion.
>>
>> On 15-Aug-24 15:19, George Michaelson wrote:
>>> I think I must misunderstand what you're trying to say, because this
>>> reads as the other half, and does not address what  I think Terry
>>> proposed: The embedded values in the RFCs are held to be immutable,
>>> but these contact strings are anything but immutable, as we all know.
>>> user@host is not a constant which alters the normative force or
>>> semantics of a document, its a reference to the authors.
>>
>> It doesn't matter. Published RFCs are immutable. This can only be handled
>> as metadata.
>>
>>>
>>> If the author reference in an RFC was abstracted to a DOI or an ORCID
>>> or similar, noting Phils comments to the need to make it trustable
>>> through cryptography, then the document suffers no loss of
>>> information, when the user@host has to change. That crytographic
>>> management is btw completely outside the RFC process. It manages a
>>> value expressed into an RFC.
>>>
>>> An abstracted contact ID could be external, or could be internal. I
>>> prefer internal, managed inside IETF process, and amenable to an XML
>>> definition so it can be tokenised properly in the web display and
>>> datatracker.
>>
>> s/IETF process/RFC Editor process/
>>
>>>
>>> Ie its not "listing the emails" its tying the identity information of
>>> the author to something outside of the RFC which can itself remain
>>> substantively immutable, when emails change.
>>
>> Certainly one could imagine this added to the metadata for an RFC,
>> but that would be a complicated discussion over in RFC Editor land.
>>
>> Regards
>>     Brian
>>
>>>
>>> G
>>>
>>> On Thu, Aug 15, 2024 at 11:47 AM Brian E Carpenter
>>> <[e-mail]> wrote:
>>>>
>>>> Terry,
>>>>
>>>> We effectively have that already. Try these:
>>>>
>>>> https://datatracker.ietf.org/person/[e-mail]
>>>> https://datatracker.ietf.org/person/[e-mail]
>>>> https://datatracker.ietf.org/person/[e-mail]
>>>>
>>>> The only issue I see is that if you have no formal role (lucky me!), no current email address is listed. That could be an option in the user's profile, or "author" could be added as a new role. (If you like that, we could discuss it at [e-mail])
>>>>
>>>> Regards
>>>>       Brian Carpenter
>>>>
>>>> On 15-Aug-24 11:46, Terry Manderson wrote:
>>>>>
>>>>>
>>>>>> On 15 Aug 2024, at 7:54 AM, [e-mail] wrote:
>>>>>>
>>>>>> Although I appreciate the impact this has to our RFCs, we all experience this ([e-mail] is no more as well), though perhaps not to the same degree.
>>>>>>
>>>>>> I’ll step in here to defend Harvard’s decision; having an email available to someone who no longer holds an official position is a significant legal risk.
>>>>>>
>>>>>> Emails, URLs, and even RFC numbers change (remember back when TCP was “always” RFC793?). Search engines mitigate this problem, as would (preferably) a bounce message from Harvard providing the next known email, at least for a while.
>>>>>>
>>>>>> Joe
>>>>>>
>>>>>
>>>>> I'm looking at this from the impact to the RFCs and the link between RFC authors and other inquisitive minds. Especially while the author is still interested in responding to email questions.
>>>>>
>>>>> I wonder if a level of abstraction can be created through an "author profile" that ties together all past author's address blocks and can provide the "latest known" address.
>>>>>
>>>>> Just a thought.
>>>>>
>>>>> Cheers,
>>>>> Terry
>