We continued this with Callan aka wizzo on IRC. I'll summarize, without the gnashing of teeth and pieces that incriminate me^W^W^W^W^W other fluff.
<vegai> I was looking for that Roman's mail you mentioned.
<wizzo> http://archlinux.org/pipermail/aur-general/2007-December/006387.html
<vegai> thanks.
<vegai> Ok, that certainly is a problem.
<vegai> although it doesn't remove from the point that voting for new TUs is largely pointless
<wizzo> why?
<vegai> you didn't seem to reply my mail fully
<vegai> especially the part,
<vegai> We simply cannot protect ourselves fully
<vegai> from cases like that, since the one and only way to know if a person is
<vegai> worth anything is to give that person a chance to work and see what happens
<vegai> in half a year.
<vegai>*
<vegai> by this I don't mean letting xterminus hang on for that long
<vegai> but that when we get new people in, there's no way to know ahead of time if they're good or not
<vegai> i.e. whether they'll take 1000 packages and go or not
<wizzo> ok I interpreted your email wrong, I apologize
<vegai> could we have seen that when he applied?
<wizzo> of course not
<vegai> so what is the point of voting?
<wizzo> I understand we always run that risk, we can't stop it
<wizzo> because it gives us all a say in what happens
<wizzo> getting a sponsor is normally a good indicator that you can make packages
<wizzo> the vote is so everyone can give their say on it
<wizzo> I think people who just vote yes without looking at the applicant's packages is just as bad as not voting at all anyway
<wizzo> but that's not easy to stop
<vegai> also, it seems that a mailing list is not a very reliable medium for this
<vegai> since 3 of my previous votes were lost
<vegai> dunno what would be.
<vegai> perhaps there should be a TU application somewhere for all these things
<vegai> not try to tweak something existing to poorly support what we need, but something we create specifically
<vegai> can I/should I post this to the ML?
<wizzo> I agree with what you're saying but I think this need some further discussion with more people than just us
I suppose the two new things that came from there are
* 1) TU voting. Is it pointless in its current form? Are we pretty much reduced to rubber stamps in those elections?
* 2) Are our tools lacking? Are we willing to create and maintain additional infrastructure specifically designed for TU activities?
( 3) wizzo agrees with what I'm saying**
Also, I think we agreed on that no active TU should feel they should leave because of this thing.
--vk vegai