[aur-general] TU Bylaws amendment proposal

[Xyne]

Loui Chang wrote:
> > > > Another wording which I find easier to understand would be a
> > > > modification of the new sentence introduced by Xyne from:
> > > > "The quorum is counted among TUs who are active throughout both the
> > > > discussion period and the voting period."
> > > >
> > > > to:
> > > > "The quorum is counted among TUs who are active at the point of time
> > > > the voting is started at the AUR but every TU is allowed to vote."
> > 
> > If a TU then becomes inactive during the vote and prior to casting his or her
> > own vote, he or she would still count towards the quorum, which doesn't make
> > sense.
> > 
> > Second proposal:
> > 
> > > The quorum is normally counted among TUs who are active from the
> > > beginning of the discussion period until the end of the voting
> > > period. Other TUs may participate in the vote if they wish, in which
> > > case they shall be among those counted to establish quorum.
> > 
> > 
> > Example:
> > * 20 active TUs during the whole procedure (start of discussion to end
> > of vote):
> >     quorum is 14 (0.66 * 20, rounding up)
> > * 2 new and one inactive TU decide to participate:
> >     quorum becomes 16 (0.66 * 23, rounding up)
> > 
> > The second sentence thus prevents skewed quorums.
> 
> Yeah I agree. A TU should be active throught the discussion period and
> the first day of voting for them to be counted in the quorum.
> 
> I don't think we should necessarily assume that a new TU has been active
> and following the other issues though.
> 
> It may also become an issue if an inactive TU votes on a heated or
> controversial issue, so I think we shouldn't count any inactive TUs in
> the quorum, or votes. They could badly skew the vote if they aren't
> counted in quorum, but their vote is counted.

I interpret the last paragraph to mean that you would block inactive TUs from
voting. Is that because it would be difficult to count the quorum?

The voting interface is aware of who has voted, but not who is active, right?
If we had an interface for updating status (active/inactive) then it could
determine quorum automatically:

A = { TUs who voted }
B = { active TUs }
( |A| / |(A V B)| ) >= quorum

i.e. the number of TUs who voted divided by the number of TUs who voted and any
active TUs who didn't vote.

As for implementation, it should only require a table of TUs with a boolean
field, and a simple interface to check active or inactive. Is that something we
could do?