Mick wrote:

But how should Freevo be able to control a child of a child?



Freevo just has to worry on killing all the spwaned childs and it's sub-childs. Since most of the helper processes do not notify anything about spwaning sub-childs the only way i see to make this work is to assign a process group to all your spawned helper-childs.
When you send a SIG_WHATEVER to the process group, all the processes in that group(including (sub)+-childs) receive it.
It's full-proof and it eliminates unexpected behaviours from the helpers.


I wish I knew.. This could more easily be broken into how can any python
know the pid of a child of a child.

I've been searching through python docs and mailing lists and I can't find a
solution to this. Any ideas are welcome.. The fact is any child of a child
which does not die when the child is sent SIGTERM is a badley progmramed
child, and should be punished ;-).~


No it's not :) There are plenty of justifiable reasons for this to happen. And of course, this should not be punished, the child should be spwaned having in mind that it MIGHT spwan a subchild.

But, this is transcodes current state.. as a child of a child it does not
listen to SIG_ANYTHING. It needs them sent directly... Maybe on the fly
mpeg2 was justa pipe dream after all ;-)


No it doens't, if this child of child is in a process group it will listen to SIG_ANYTHING sent to it or to the process group. And i think this is the way to go since is the standard way to do this in *nix

Bruno



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Freevo-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/freevo-devel

Reply via email to