On 4/27/2021 4:17 PM, Brian E Carpenter wrote:
On 28-Apr-21 01:26, Vittorio Bertola wrote:
Il 27/04/2021 10:41 Lars Eggert <lars@xxxxxxxxxx> ha scritto:
There was a suggestion recently to not serve I-Ds from ietf.org domains until they were adopted by the IETF. Do you think serving individual drafts from another domain would help make that distinction clearer?
URIs can help, because they are posted around to refer to documents, so they can contain prominent semantic "messages", either in the hostname or in the top path element. However, I think it would be even better to do this in the filename, as the filename persists even when the file is downloaded or attached. For example, you could reverse the order of the initial elements and things would already be much clearer:
ietf-draft-<wg>-<subject>
irtf-draft-<wg>-<subject>
independent-draft-<author>-<subject>
I think this, like the suggestion to post drafts to differently named servers, would make matters worse for several reasons:
1) If people already have intellectual difficulty understanding that a filename starting with "draft-" is only a draft, this difficulty will be even greater with remixed names or multiple servers.
2) I hesitate to write this, but "draft-" has beome a brand, although less so than "RFC".
3) We have a lot of tooling that (for better or worse) assumes the draft- naming conventions, and a single server for all drafts. That isn't only IETF official tooling; I'm sure many of us have our own scripts and working habits. That's a real economic cost.
4) We have a de facto agreement that all RFC streams use common practices as far as possible, including IPR regimes. Remixing names and splitting servers wouldn't change that, but would make it less evident.
5) It is not uncommon for drafts to jump ship (from a WG to AD-sponsored; from IRTF to IETF; from IRTF or IETF to Independent, etc.). Forcing name changes for this reason is just makework.
I definitely agree with what Brian said. In addition, as a rule, it is a
rather poor idea to twist names to incorporate properties. The chances
of ending up with an unstable mix of combinations are just way too high,
and the effects on document management chains are not great.
Yes, there is a risk of confusion because a draft that I just submitted
will look and feel exactly like one that a working group has been
working on for three years. But that risk can be addressed in more
productive ways than twisting names. If we believe that documents
acquire different cachet and statuses by going through different hoops,
then we should say so in the status of the document. Saying things like
"this is an individual contribution to the IETF", or "this is a working
document of the FOO working group", or "this is a research contribution
to the IETF", or "this work was not updated since April 4, 2019, and is
considered abandoned in the IETF" are all statements that would fit well
in an updated status.
-- Christian Huitema
By the way, "independent" is not immediately clear. "Personal" or "unofficial" would IMHO be better.
"Independent" refers to the Independent Submission RFC stream.
Brian
