IRC Logs for #crux-devel Monday, 2019-06-17

*** xor29ah has quit IRC04:22
*** xor29ah has joined #crux-devel04:23
*** stenur has joined #crux-devel11:40
*** stenur has quit IRC15:12
TimB_jaeger: so, my latest commit fixed most of it, but ssd won't read the pid file because it's not owned by root. cardinal suggests changing the owner:group attribute after starting up avahi - can you say if this is a valid way to go?18:26
jaegerMight just have to skip opt and xorg ports? I think that's what they usually are18:41
jaegerer... what the fuck, irssi18:41
jaegerthat's 2 completely different lines stuck together18:41
TimB_he :D18:41
jaegerMaybe just make the pidfile owned by root:root? I think that's what they usually are18:41
jaeger^ that's what I type. the fuck18:41
jaegers/type/typed/18:41
TimB_okay, will push an update to fix that in a bit18:42
TimB_tlovoo19:43
TimB_foo19:43
TimB_i obviously need eye focused window controls19:44
jaegerheh19:50
TimB_either that, or more sleep. I'll try to get the eye focused gimmick :P19:51
*** stenur has joined #crux-devel21:10
*** frinnst has quit IRC22:04
*** frinnst has joined #crux-devel22:06
*** frinnst has quit IRC22:06
*** frinnst has joined #crux-devel22:06
stenurHm, indeed, core/s-s-d has not been updated to the version from s-s-d.git :(22:13
TimB_stenur: can you shed some light for me? does s-s-d need the pidfile to be owned by root?22:20
stenurJürgen said something else, i just cannot find it in my archives...22:20
stenurI think all you need is another selector, for example..22:21
stenurFor example --name.22:21
stenurSomething to make s-s-d n22:23
stenur... uniquely identify the program <-> pid relationship.22:23
stenur(So understandable my patch from April 30th did not make it, even though now rc files need to become adjusted.)22:24
TimB_--name works fine with avahi22:25
stenurGreat!22:25
TimB_is it worth to be changed for another revision?22:26
stenurMe? s-s-d??22:26
TimB_oh, no. sorry22:26
TimB_https://crux.nu/bugs/index.php?do=details&task_id=1741 <- talking with this in mind22:26
stenurs-s-d as in core refuses to work otherwise...  I had pushed a commit to start-stop-daemon.git which would have delayed the problem a bit..22:28
stenurWith a big fat warning in the manual ("hihi")22:28
stenurI have lost the conversation with Jürgen, dammit.22:29
TimB_:P22:29
TimB_so.. would you say I should change the rc in contrib/avahi to use --name instead of chaning the pid files owner?22:30
stenurHe said he already turned some ports over, it -- was it on IRC?? -- .. Sorry anyway.22:30
stenurWell yes, i would.  The lesser root the better, if possible, no?22:30
TimB_I agree22:31
TimB_on my way to push -3 of the port 8) thanks for the hint22:32
stenurI don't understand the s-s-d policy, matching non-root ok, but if writable only by the daemon user, not group or other.. why??22:32
stenursure.22:32
TimB_I was iritated by that too22:34
stenurHmm. I think --name is ok, but i thought more in direction "--pidfile $PID --name NAME", i.e., both. I have no idea of ahavi though.22:37
stenurIt is only that s-s-d can verify that the pid in $PIDFILE _really_ refers to the desired program, which is why they change the policy.22:38
stenurTo avoid that higher privileged daemons etc. can be fooled to act for unprivileged users.22:38
TimB_oh, mh.. darn it. it does work without troubles like that though.. I'll try to have a more detailed look into it tomorrow22:40
TimB_I'll go to bed now... alarm goes in 6 hours. x_x22:41
stenurYep, more than six i won't get, too.  Good night!22:41
TimB_Thanks, you too!22:42
*** stenur has quit IRC23:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!