[aur-general] TUs and their following of the Bylaws

[Lukas Fleischer]

On Sun, Aug 04, 2013 at 05:45:04PM +0200, Sébastien Luttringer wrote:
> On Sun, Aug 4, 2013 at 3:35 PM, Lukas Fleischer
> <archlinux at cryptocrack.de> wrote:
> > On Fri, Aug 02, 2013 at 03:40:53PM -0300, Angel Velásquez wrote:
> >> Hi people,
> >>
> >> [...]
> >>
> >> Having that set, I am shocked about how the bylaws are being just used
> >> just for addition process, but for somehow are being ignored for stuff
> >> like quorum and removal procedures, some TUs look to otherside when we
> >> mention this subject.
> >>
> >> [...]
> >>
> >> Let this discussion begin.
> >
> > Trying to automate this, I just submitted a couple of AUR patches to
> > aur-dev [1].
> >
> > Based on this, we could:
> >
> > * Add another simple patch that simply displays whether a vote is
> >   accepted or not. This leaves no room for discussion on the results of
> >   future votes.
> I think we should not compute the active TU at the beginning of the vote.
> Some TU may have miss to set the inactive state, or be back from
> holidays during a vote.

I don't think it matters a lot whether we compute the amount of active
TUs at the beginning or at the end. All arguments against computing it
at the beginning have a negative counterpart (and vice versa): If you
argue that a TU might be back from holidays during a vote, you might as
well argue that a TU might go on vacation during a vote. If you argue
that a new TU might join during a vote, you might as well argue that a
TU might be removed during a vote. If you argue that a TU might have
missed to set the inactive state, he might have missed it at the end of
a vote.

The only difference is that a new active TU appearing during the vote
results in an overvalued quorum if the number is computed at the
beginning (and if the TU votes), whereas an active TU disappearing
during the vote results in an overvalued quorum if the number is
computed at the end.

> 
> Should we not remove the "active" users in our bylaws?
> We consider that ALL TU should vote.
> Saying we remove inactive TU in vote to reach the quorum, is an hijack.

Does that mean that you expect TUs on vacation to vote? A single TU vote
accounts for ~3%. If there are three TUs that are away during a vote,
the calculated quorum might be off by 10%. Just saying.

> 
> > * Implement auto-initiation of removal procedures for repeated quorum
> >   offenses.
> We could also add auto removal of MIA? There is a field in archweb
> with the "last action" of a dev/tu.

Not yet. Also note that our Bylaws only mention automatic triggers on
"repeated quorum offenses" in the "removal due to unwarranted and
undeclared inactivity" section. Not sure if it even applies in the MIA
case.

> 
> >
> > That do you think about that?
> 
> This is an excellent initiative!
> 
> I heard some complain about people missing the vote beginning.
> Maybe adding some kind of
> - direct TU email (like for notification) when proposal beging,

That should be easy to implement.

> - direct TU email 1 day before end of the vote to non votant

That is a bit harder to implement since it requires some kind of
scheduler (e.g. cron job). This also is the reason why I decided to
compute the number of active TUs at the beginning, basically...

> - public email (aur-gen) when to announce the status of a proposal.
> 
> Cheers,
> 
> -- 
> Sébastien "Seblu" Luttringer
> https://www.seblu.net
> GPG: 0x2072D77A