Boost logo

Boost :

Subject: Re: [boost] encouraging review managers -- was Re: Review Request: Variadic Macro Data library
From: Joachim Faulhaber (afojgo_at_[hidden])
Date: 2011-02-21 05:28:10


2011/2/21 Gordon Woodhull <gordon_at_[hidden]>:
>
> On Feb 20, 2011, at 8:54 AM, Joachim Faulhaber wrote:
>> Gordon Woodhull wrote:
>>> Second, Joachim proposed the role of Review Manager
>>> Assistant as a way for new authors like Ed (and myself
>>> and many others) to manage reviews in conjunction
>>> with more seasoned boosters.  The Assistant would
>>> do most of the work of summarizing the debate,
>>> and then the senior Review Manager would make
>>> decisions and produce the final report.
>>>
>>> I would like to suggest a generalization which is
>>> more fluid (and which would allow RMAs to put
>>> the nicer title "Review Manager" on their CVs):
>>> simply allow multiple Review Managers for a review.
>>> The managers can decide how to split up the work:
>>> in sequence like Joachim's RMA/RM idea, or in
>>> parallel by divvying up topics for the report.  Or
>>> some mix of the two.
>>
>> Personally I am more interested in clear commitments for a particular
>> role and taking responsibility for that role within the Boost
>> community. A "Review Manager Crowd" I dislike ...
>
> Okay, I will admit it, it's only the name I don't like about
> Review Manager Assistant!  Unfortunately in the USA,
> with the bizarre way that Political Correct language evolves
> and devolves, Assistant is synonymous with Secretary,
> and has the same negative connotations.  Let's be realistic,
> we want something that looks nice on a CV - but I do hope
> that if anyone shoddily tries to manage a review just to
> pad their resume, they'll be forced to actually do the work.
>
> Really, the RMA is doing all of the work and making
> all of the decisions, and the Review Manager should
> approve, or disagree and redo part of the work.
>
> That is why I like the idea of just calling the RMA a
> Review Manager, and it may happen that their work is
> being checked by another Review Manager.  But I'd
> also be open to other names, such as
>
> Review Compiler (a pun)
> Review Summarizer (awkward)
> Review Reader (too mild)
> Review Investigator
> ...
from the Boost Bureau of Investigation (BBI ;-)

what about Review Comanager.

>> More fundamental to my thoughts
>> http://lists.boost.org/Archives/boost/2010/05/166423.php
>> is the idea to make the RMA job
>> (1) A precondition to an own first library submission
>> because obviously the most energy, excitement and motivation is in the
>> endeavour of library contribution.
>
> I don't think someone who has some insanely brilliant library
> should be deterred from submitting it because they haven't
> "compiled" a review.

Depends on the rules. Generally my proposals are trying to strengthen
the effects of action and weakening the effects of inaction. If there
is a great interest in a library, such that seasoned boosters or
moderators have an interest of getting a library into boost such a
library could be *invited* for a review. Like on a conference: Most of
the submitted talks are reviewed and then accepted or rejected in a
standard way but the key note speaker is invited.

> But I think this should be one of the
> ways to show that they have the cooperative instincts and
> stick-to-it-iveness to be a good maintainer and member of the
> community.
>
>> First time contributors should
>> (2) learn thoroughly all aspects and standards around boost libraries being RMA
>> (3) help to enhance the quality of submissions of libraries of others
>> and the quality of the review queue as a whole.
>> (4) have an opportunity to establish themselves in a role of
>> contributing for others
>
> Above all, to understand the review process and learn
> how debates are resolved and consensus is reached
> before they head into their own review.

+1

>> (5) empower the group of contributors and make them more independent
>> of the boost functionaries
>> (6) unburden the boost functionaries
>
> I think it might be possible for for the Wizards to encourage
> these connections to happen,

currently the Wizards seem not to be very much in favor of my
proposal. I asked them off list. John Phillips dislikes the idea and
thinks the RMA role made everything only more complicated. But they
didn't take much time to discuss things in detail. As I said, they are
only humans no real Wizards. I can understand they won't be able to
comment on each and every one of the numerous proposals on the review
process that come from the mailing list.

> but I have to think about it.
>
>> (7) Finally, functionaries can not discourage contributions by mere
>> inaction anymore. They would have to veto. Which I think is
>> appropriate because contributors deserve a response. Not necessarily a
>> yes, but a response.
>
> I think you are suggesting that the Wizards get to veto a
> review report if they don't think it was done properly or fairly.

No, the crucial point in my proposal is the strengthening of the
active contributors.

(7.1) RMAs are not necessary for a review. So it can be done as before
with an RM only
(7.2) yet in addition, the review process can be scheduled after the
RMA has thoroughly checked library for all preconditions.
(7.3) The submitter and RMA are supposed to find an RM but if no one
steps up, the formal review can be conducted by the RMA. I think this
is acceptable, because the "sovereign" is the community itself, not
the RM nor the RWs. If a library gets a lot of yes votes, it hard for
an RM to reject it anyway.
(7.4) But a library thats review is conducted without RM should have a
weaker position. Because it lacks the person that is designed for the
usual way of quality control. Therefore there should be a mechanism of
control that depends on action: A veto. If e.g. the review went pretty
controversial and the RMA seems to vote in favor of the submitter
without convincing subject-specific arguments, RWs or Boost moderators
or a booster that would otherwise quality as RM could issue a veto
giving a substantial justification, of course.
(7.5) This is a mechanism of control or blockade, if you will, that
demands action. Mere inaction should not discourage contributions.

> I want to democratize that suggestion, and say
> that it should be possible for a review to be challenged
> by anyone on the list.  If the Wizards think a challenge
> has merit -- using the same anti-troll (and anti-"I just won't
> use this ridiculous stuff") criteria that Review Managers
> use to judge No votes -- then it can go for a second review
> with a different review manager.

I'm not sure if I understand you completely here. But challenging a
review result should be a serious action and should not happen too
often. So I would like to couple it with competence.

My concern is to maximize contribution, throughput, quality and fun
and minimize blockade, waiting, false expectations and frustration.

Cheers,
Joachim

-- 
Interval Container Library [Boost.Icl]
http://www.joachim-faulhaber.de

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk