einhyrningsins - graceful restarts for socket-based daemons

	Commit message (Collapse)	Author	Age	Files	Lines
*	work in progress on control socket	bnewbold	2016-10-11	4	-10/+197
\|
*	more work in progress	bnewbold	2016-10-11	3	-22/+68
\|
*	more signal handlers	bnewbold	2016-10-11	1	-6/+36
\|
*	refactor of Offspring+brood code	bnewbold	2016-10-11	1	-47/+122
\|
*	rip out RPC stuff for now	bnewbold	2016-10-10	1	-5/+1
\|
*	more signal handling	bnewbold	2016-10-10	1	-11/+11
\|
*	add ipv4/ipv6 forcing; nonblock and reuse socket opts	bnewbold	2016-10-10	1	-10/+52
\|
*	use logging in python example	bnewbold	2016-10-09	1	-5/+9
\|
*	upate python example with KeyboardInterrupt catching	bnewbold	2016-10-09	3	-14/+59
\|
*	more refactoring progress	bnewbold	2016-10-09	3	-28/+87
\|
*	split refs from README	bnewbold	2016-10-09	2	-9/+11
\|
*	start refactoring to chan-signals	bnewbold	2016-10-09	3	-63/+204
\|
*	basic/naive signal handling	bnewbold	2016-10-03	1	-0/+45
\| \| \| \| \| \| \| \| \| \|	The goal here was to hook Ctrl-C and send USR2 signal to childred instead. Turns out this is doesn't work easily because Ctrl-C is sent to the entire process group, not just the top running process. Making this work would require either masking SIGINT in the childred (children might change this) or mucking with process groups, both of which are "meh". But, unsafe code, woooo!
*	start addings refs to README	bnewbold	2016-10-03	1	-0/+9
\|
*	simple python3 http example script	bnewbold	2016-10-03	1	-0/+40
\|
*	basic file descriptor passing	bnewbold	2016-10-03	3	-10/+85
\|
*	more skeleton code	bnewbold	2016-09-27	3	-48/+36
\|
*	gh-pages helper	bnewbold	2016-09-27	1	-3/+11
\|
*	init repo	bnewbold	2016-09-27	8	-0/+957