[crossfire] crossfire source code control systems
Tchize
tchize at myrealbox.com
Wed Aug 23 04:27:34 CDT 2006
A quite common way of voting is to let people mark all thing they want.
For example if we have item a,b,c,d and i can
vote a; that mean a has my preference
vote a, b; that mean a and b have my preference over c and d
vote a, b, c; that mean i don't like d
For the delay, i think we can afford a 1 month time voting (to give
opportunity to vote for commiters on holidays). We are not in a hurry to
switch :)
For the notions of weighting vote, that's quite stupid, every time you
try to weight something, you always introduce arbitrary that can
frustate people it's definitly more unfair than fair.
I proposed the idea of having people select different items because, if
i remember well my student time, it's the system that vote for a
candidate that is the most widly accepted, but not necessary the most
widely preffered, and it take into account the case were similar
candidates have to share the voters.
Also, the idea of a 'first choice' and 'second choice' is bad, because,
again, it introduce a weighting system and people have several choice
don't necessary have a first and second choice and you fall back in case
where similar cnadidate have to share the first choice, leading the path
to another less accepted candidate.
However, because we can't argue on the voting system nor vote for a
voting system ;) i suggest the project manager take a wise decision on
how we will vote on the source repository change and by the same time,
how we will vote next time we have something important to vote on
(its probably worth noting that different people work on different things,
which probably also weighs on what they want to use. Rename feature is going to
be much more important for maps and archetypes than for the code area, where
files are seldom renamed)
I don't agree, i think all commiters are aware that we have different
part in crossfire and different need, and commiter can vote based on
such requirements.
Mark Wedel a écrit :
> Any selection method is flawed at some level.
>
> Voting probably does make the most sense - but even then, you get some
> complications - should the amount of commits a person do also have weight on the
> number of votes?
>
> A person who does 5 commits a week is using the system a lot more than the
> person that does 1 commit a month. Do we attempt to weigh that at all? And if
> so, how?
>
> (its probably worth noting that different people work on different things,
> which probably also weighs on what they want to use. Rename feature is going to
> be much more important for maps and archetypes than for the code area, where
> files are seldom renamed)
>
> Is the voting method just vote for what you want to use? Or does each person
> provide a list, and you weigh the bottom vote getters with some negative points?
> (for example, if lots of people put the same thing as their bottom choice, that
> suggest some strong dislike).
>
> And how long should the voting process be for? Need some finite time -
> probably a week or so.
>
> Should we vote on the voting method? :)
>
>
> _______________________________________________
> crossfire mailing list
> crossfire at metalforge.org
> http://mailman.metalforge.org/mailman/listinfo/crossfire
>
>
More information about the crossfire
mailing list