Revising OpenWrt Rules

[Sam Kuper]

On Wed, Oct 28, 2020 at 07:57:51PM +0000, Imre Kaloz wrote:
> On October 28, 2020 at 19:52, Fernando Frediani wrote:
>> On 28/10/2020 13:09, Sam Kuper wrote:
>>> I also propose a different criterion for forcing retirement of
>>> Decisionmakers.  I think it should be different because:
>>>
>>> - Decisionmakers with long-term dedication to OpenWrt might
>>>   occasionally experience life events that legitimately keep them
>>>   away from OpenWrt for >3 months (unexpected bereavements, serious
>>>   illness, etc.).  It seems healthiest (for OpenWrt *and* for the
>>>   Decisionmaker) if Decisionmakers are not under strong pressure to
>>>   vote during such periods, but can instead resume voting when
>>>   recovered & clear-headed.
>> 
>> Why not if it happens, they don't show up for 3 months and losses its
>> Decisionmaker status, whenever they resume their activities they may
>> request again and it will be natural thing they be welcomed back
>> again.

Hi Fernando,

Compared to my proposal, yours would seem to create more work for:

- the project, which would have to (1) de-list Decisionmakers and (2)
  (if they re-apply) hold a vote on whether or not to restore them;

- de-listed Decisionmakers, who would have to (1) submit a new
  application and (2) address any scrutiny directed at them (e.g. from
  newer Decisionmakers who might not be familiar with their record of
  contributions) while their application is pending.

My proposal should avoid that needless effort, letting people spend time
better  :)


>> The important thing is to not leave it opened for too long waiting
>> for some Decisionmakers to vote and that may not happen and things
>> need to move on. Even more importantly it is necessary to have the
>> total number of active Decisionmakers to be able to know what is the
>> minimum number of votes necessary for something to pass. If there are
>> too many sleepy Decisionmakers the minimum number of necessary votes
>> will be higher and that may compromise something that need to move
>> on.

I mostly agree.  So far, I think the rest of Rich's RC1 meets those
needs well.


> This is a funky thread. I don't want to get into details on how Rich's
> original mail or some of the current ideas go against the project
> remerge agreement nor how trying to fix governance issues with more
> regulation shows how many history teachers failed people here.

Hi Imre,

I used OpenWrt before the fork, and used both LEDE and OpenWrt during
it.  I was happy when the remerge happened, because I prefer healthy
collaboration over competition or duplication of effort.  But I
certainly don't know all the details.  If I inadvertently upset you or
anyone else with any of my remarks in this thread, I apologise.  I
intended only to make constructive comments to help OpenWrt go from
strength to strength.


> People don't vote for numerous reasons. For example I didn't vote
> about dwmw's access given it has been agreed years ago once already.
> You can also decide to not vote if the decision is either going in the
> way you want or goes so much into the other way that it doesn't
> matter.
> 
> Most of these problems raised here are nonexistent - if you want to
> have a decision in 2 weeks you make the voting process last for two
> weeks, that's all.
> 
> Imre
> 
> P.S: A lot of people might not understand the reason why former core
> developers are not that active. Your favorite search engine might give
> you some ideas if you look for "state capture".

I'm grateful for these points, which helped me see this differently.
I'm aware you have deeper and longer knowledge of the project than most!

It sounds like you aren't in favour of removing a person's Decisionmaker
status for some non-participation in votes.  (I'm not saying that
judgementally, just to check my understanding.  If I misinterpreted you,
please correct me.)

I also see that the current rules[1], as linked from the remerge
announcement[2] do *not* remove Decisionmaker status for
non-participation in votes, only for being "unreachable for three months
in a row".  So, the current rules seem consistent with your preference
on this point (AIUI, see para above) whereas Rich's RC1 is not.

In other words, I see why you might be aggrieved.

(The current rules say "committers" rather than "Decisionmakers", but
that's a separate matter and maybe not contentious.  I don't have an
opinion on this.)

The current rules and RC1 both have good intentions, but also flaws.

The current rules let committers take breaks without consequence, as
long as they are still alive and contactable.  This helps foster a
low-stress environment in important ways.  But it risks an
ever-increasing list of inactive committers, which could make it harder
for ballots to be quorate (as Fernando pointed out above).  The current
rules also don't specify when a ballot should be considered closed,
which might create stressful situations by leaving committers unclear
whether/when to proceed with a proposal even if some committers have
voted.

RC1 has the benefit of addressing both those risks.  But, by requiring
frequent ongoing participation, it might cause undue stress for
Decisionmakers and induce them to either quit or to adopt an
anti-pattern: "busywork".

The ideal would be to avoid the risks the current rules leave
unaddressed, while still capturing those rules' minimisation of
unnecessary stress.  Win-win  :)

In that light, does my proposal (see my previous email in this thread)
seem to you a reasonable compromise *between* these relative extremes:

- revoking Decisionmaker rights after 3 months' non-voting (per RC1); vs

- allowing committers to retain rights indefinitely, even if they no
  longer ever vote or otherwise contribute, as long as they are
  "reachable" (per current rules)?

If not, would my proposal seem reasonable to you if it instead had a
more lenient threshold, e.g. 10% or 25% rather than 50%?


Thanks again,

Sam


[1]: https://openwrt.org/rules

[2]: https://forum.openwrt.org/t/announcing-the-openwrt-lede-merge/10217

-- 
A: When it messes up the order in which people normally read text.
Q: When is top-posting a bad thing?

()  ASCII ribbon campaign. Please avoid HTML emails & proprietary
/\  file formats. (Why? See e.g. https://v.gd/jrmGbS ). Thank you.