[RFC] Introduce inactive member status

[Daniel Golle]

On Mon, Oct 25, 2021 at 07:42:31PM +0100, David Woodhouse wrote:
> On Thu, 2021-10-21 at 12:35 +0100, Daniel Golle wrote:
> > Hi everyone,
> > 
> > 
> > 
> > as discussed in the previous developer meeting on 19th of October
> > we need to find ways to improve our voting mechanics. It has became
> > increasingly difficult to on-board new members as this currently
> > requires a majority of all current members to be in favor of accepting
> > the new member[1].
> 
> What is the problem we're trying to solve with the rules as they stand?
> 
> I understand that where meetings could be held in person at a specific
> time, it makes sense to require a majority of the *electorate* instead
> of just a majority of those present. Especially for 'constitutional'
> changes, and I suppose that accepting a new voting member is one of
> those.
> 
> That prevents a hostile takeover by declaring a meeting when *only*
> those who would vote for the takeover are present, then declaring that
> it was "by the rules".
> 
> But where votes are conducted by email with a two-week voting period,
> that's a bit less of a concern (even if it was really a practical
> concern for OpenWrt in the first place).

We would need to change the rules to introduce a sensible timeout for
voting and that should not be too short because then the before
mentioned hostile decision-making would happen around X-mas, during the
summer holidays or Chinese New Year. 6 weeks timeout would be sensible
imho and that would also solve the problem, we'd just have to be more
patient when it comes to on-board people (as usually, 2-3 weeks are
enough to conclude under the current terms and pushing people to
participate a bit).

> 
> Why not just switch to allowing votes based on the majority of the
> ballots in that particular vote? If we declare members "inactive"
> because they aren't voting (or don't want to be held to doing so), and
> call them "active" again when they do... just so that we don't have
> to... erm... be sad when they don't vote... isn't that overkill? Why
> don't we just *not* be sad if/when they don't vote, on a per-vote
> basis?
> 
> 
> Put another way — a variant on the existing proposals:
> 
> Why not call a member "inactive" automatically for a *specific* vote,
> if they don't actually cast their ballot in *that* vote.

Absolutely fine if the timeout is long enough and everything happens
on openwrt-adm via email.