We are using win32 process waitpid2 in our project, and it behaves slightly different then Process.waitpid2.
Could you tell is it a bug?
Scenario: at service_stop of win32daemon we are killing another process (with '9' signal) and waiting while it actually exits.
inspect of the object returned by Process.waitpaid2:
and a process is actually stopped (I check it by sending '0' signal)
inspect of the object returned by win32process.waitpaid2:
and a process still alive after the wait
Which version of Ruby, btw? We may need to revisit win32-process for 1.9.x.
ruby 1.9.2p290 (2011-07-09) [i386-mingw32]
I think I'm going to release 0.7.0 soon, and remove the custom definitions of waitpid, etc. I just don't think they're necessary for 1.9 any more, and may even be problematic now that they've added Process.spawn, etc.
Thank you for response