Woof!

Nice!  I like it.  Some minor quibbles that really don't need to be changed, 
but I've got to give "value add" or else they might not keep me around.

sipXsupervisor.cpp: Line 114:
   The comment says "Kill of the supervisor-in-waiting" but the code actually 
just waits for it to die.
the killing is done in a different place.

sipXsupervisor.cpp: Line 306:
   The child is reading stuff from the parent, and happily ignoring it.  I'd 
like to see the the child wait for a specific command (like "die\n" or even 
just a single byte, like "!" or something) and then cause itself to exit.  This 
does several things that will be helpful later.  There's no need to send a 
signal to it, just write a command down the pipe.  Thus finding it's PID isn't 
needed, and the supervisor-in-waiting could maintain it's root privledges, 
which it cannot at the moment because later the parent wouldn't have the rights 
to send it a signal after it dropped it's privledges.

sipXsupervisor.cpp: Line 441:
   The supervisor-in-waiting is created AFTER sipXsupervisor drops root 
privledges.  At the moment this is okay, because there is nothing to be done as 
root at the moment.  In the future, there may need to be some root activity, 
which the supervisor-in-waiting couldn't do when it reigns supreme, because it 
was created after the privledge drop.  By moving the creation BEFORE the drop, 
this situation could be avoided.

sipXsupervisor.cpp: Line 205:
   The supverisor-in-waiting is killed via a signal.  I'd prefer to see this as 
a command written down the pipe, as described previously.  Signals can come 
from outside, commands over private pipes cannot.

--Woof!
_______________________________________________
sipx-dev mailing list
[email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev

Reply via email to