Review action holder manipulation on draft-revision submission.

[rjsparks]

From @cabo at https://mailarchive.ietf.org/arch/msg/tools-discuss/kRL2lI4KASkF040nPkpdP6I_Gu8:

https://datatracker.ietf.org/doc/draft-ietf-lake-edhoc/ is in state Publication Requested::AD Followup

You would expect that this makes the AD action holder?

No. When a new version was posted, the state changed, and the action holder was removed:

"Removed all action holders (IESG state changed)”

So the AD needs to do the followup now but isn’t an action holder?

(I had the same problem earlier with one of my drafts, so this is not a fluke.)

[jennifer-richards]

The action holder updates are controlled by update_action_holders(). When the state changes, it clears the action holders when the new state is in DocumentActionHolder.CLEAR_ACTION_HOLDERS_STATES, which is currently

    CLEAR_ACTION_HOLDERS_STATES = ['approved', 'ann', 'rfcqueue', 'pub', 'dead']  # draft-iesg state slugs

For this draft, the holders were cleared because it entered the "approved" state.

[cabo]

It now is approved, but surely wasn't on July 07 when I made this observation.

https://datatracker.ietf.org/doc/draft-ietf-lake-edhoc/history/

2023-07-07	20	(System)	Removed all action holders (IESG state changed)
2023-07-07	20	(System)	Sub state has been changed to AD Followup from Revised I-D Needed
2023-07-07	20	Göran Selander	New version available: draft-ietf-lake-edhoc-20.txt

2023-07-07
20 (System) Removed all action holders (IESG state changed)
2023-07-07
20 (System) Sub state has been changed to AD Followup from Revised I-D Needed
2023-07-07
20 Göran Selander New version available: draft-ietf-lake-edhoc-20.txt

[jennifer-richards]

Ah, had not noticed the date on your email and had not looked back far enough through the history to see the previous clearing of action holders.

The issue is that the code only adds the AD as an action holder if the state changes. The "sub state" is handled separately - removing the "need-id" causes the authors to be dropped but there's no corresponding rule to add the AD for this transition.