I still feel the general approach is insufficient and that more needs to
be done to make it easier for people outside of given clique-headspaces
to see what is actually worth spending time on.
To Stephen's wonder - yes, there will be a few places where updating the
tooling to deal with the lack of an Expires header won't be trivial at
the moment - idnits, in particular, may take more than a second to
change, but the cost of that won't be huge. Breaking external things is
an interesting question to ask. I'm not aware of any external tools that
will care, but we really don't know what we might break.
If this draft goes forward in its current form, I would like to see this
changed:
When a new
version of a draft is published, it is immediately marked as
"active", and all earlier versions of that draft are marked as
"inactive".
Please remove ", and all earlier versions of that draft are marked as
"inactive". "
Make this change throughout the document - speak of marking things as
"active" and have inactive just be the consequence of not being active.
This is metadata that is associated with a thing that has versions. the
thing is active or inactive, the versions are not. There is a most
recent version.
If you need to say anything at all about earlier versions (and I don't
think you need to), say something like
"To the extent possible, tooling should make it obvious that newer
versions exist when looking at any previous version".
If you are trying to accommodate the current notions of the "draft
repository" vs "draft archive" and feel like the draft needs to say
something, please only say
"The tooling will make it straightforward to find the set of active drafts".
RjS
On 1/24/24 4:49 AM, Stephen Farrell wrote:
Hiya,
On 24/01/2024 09:21, Lars Eggert wrote:
Before I initiate a formal last-call, I'd like to do a substitute of a
WGLC and ask interested IETF participants to give this a final read and
indicate whether they are OK with seeing this go forward.
The draft seems fine. I do wonder how much work it'll
create for the tools folks and whether just omitting
the expires header might break external things, so it
could be worth adding "Expires: 1 April 2124" or some
such, instead. (If that was discussed already, that's
ok, forget I suggested it.)
Cheers,
S.
