Re: async-mqtt5 is CONDITIONALLY ACCEPTED

From: Klemens Morgenstern (klemensdavidmorgenstern_at_[hidden])
Date: 2024-11-05 14:35:15

On Tue, Nov 5, 2024 at 8:31 PM Mateusz Loskot via Boost
<boost_at_[hidden]> wrote:
>
> On Tue, 5 Nov 2024 at 11:31, Andrey Semashev via Boost <
> boost_at_[hidden]> wrote:
>
> > However, as I've already said, I do think that the discussion should
> > have happened on the ML.
>
>
> Legitimate or not, personally, that is also my expectation.
>

I think we need to distinguish three things here:

1. What a review manager takes into account.
2. What reviewer bases his review on.
3. How an author acts.

This discussion started on point 3, as Ivica was being accused of
dodging questions. In response it was brought up that he was very
responsive on Slack.
Generally, I think the author's behaviour can be taken into account
for a review result, but can't invalidate the review process itself.
The solution is writing a review, expressing concerns with the
author's lack of responsiveness.

The policy for (1) is that an RM may take anything into account, but
is only required to take into account reviews that are posted to the
ML or sent to his EMail.

This is stated as follows (https://www.boost.org/community/reviews.html)

> Proposed libraries are accepted into Boost only after undergoing a formal review, where Boost mailing list members comment on their evaluation of the library.
>
> The final "accept" or "reject" decision is made by the Review Manager, based on the review comments received from boost mailing list members.
>
> Boost mailing list members are encouraged to submit Formal Review comments:
>
> - Publicly on the mailing list.
> - Privately to the Review Manager.

"review comments" refers to what we call review these days, thus the
policy states that the RM bases his decision on these. This does not
include all comments on the ML.
The process explicitly states that this is for members of the ML and I
think that's correct and should not include comments on other media.
Considering other posts on the ML at his own discretion seems appropriate.

Regarding (2): I think it is just a fact of life that reviewers will
have discussions with the authors on other platforms.
Even if we make it a rule that all discussion about the library has to
take place on the ML during the review period,
a reviewer may have made up his mind based on slack discussions that
happened earlier.
I don't think there's a good solution here other than to ask reviewers
to include what communication he had with the library author
and encourage discussion on the ML.

We should define a policy soon, as the next review starts on the 13th
with a RM (Richard Hodges) and an author (yours truly) that are both
active on slack.