Comment 2 for bug 172286

More discussion:

<jml> mwhudson: I think that what ddaa says sounds right (although kind of crappy)
<thumper> mwhudson: perhaps we up the timeout to 30 minutes? :-|
<mwhudson> thumper: worth a try, i guess
<mwhudson> jml: ddaa said several things, which did you mean?
<mwhudson> SIGINT, wait, then KILL?
<jml> yeah.
<jml> (sorry, am doing a couple of other things atm)
<thumper> followed by explicit unlock
<thumper> does suck somewhat though
<mwhudson> if the process is siginted it should unlock the branch
<mwhudson> (i think?)
<thumper> that's my understanding too
<thumper> however I think it'd still be worth testing in the code
<thumper> basicly check if locked, and if so, unlock
<mwhudson> makes sense i guess
<mwhudson> do either of you desperately want to fix it or shall i assign it to me?
<thumper> I didn't even get to look at code today :(
<jml> mwhudson: I'm full.
<mwhudson> i'll take it then