IRC Logs for #crux Friday, 2019-06-07

*** Anselmo has quit IRC00:58
*** Anselmo has joined #crux01:03
*** Anselmo has quit IRC02:23
rmulljaeger: Thanks for the libjpeg tip02:34
*** Anselmo has joined #crux02:36
*** _________mavric6 has quit IRC02:50
*** _________mavric6 has joined #crux02:51
*** tilman_ has joined #crux03:22
*** tracer has joined #crux04:49
*** Anselmo has quit IRC05:01
*** Anselmo has joined #crux05:03
*** kori has quit IRC05:41
*** Anselmo has quit IRC06:26
*** Anselmo has joined #crux06:31
*** guido_rokepo has joined #crux07:16
*** tracer has quit IRC07:27
*** jue has quit IRC07:27
*** tracer has joined #crux07:28
*** Workster has quit IRC07:28
*** Workster has joined #crux07:28
*** jue has joined #crux07:31
*** Anselmo has quit IRC07:37
*** Anselmo has joined #crux07:52
*** kori has joined #crux09:49
*** kori has quit IRC09:56
*** kori has joined #crux10:18
*** kori has quit IRC10:18
*** kori has joined #crux10:18
*** kori has quit IRC10:40
*** kori has joined #crux10:45
*** kyaaaaaa has quit IRC10:48
*** kyaaa has joined #crux10:48
*** kori has quit IRC11:00
*** kori has joined #crux11:00
*** kori has quit IRC11:29
*** kori has joined #crux11:34
pedja'Does anyone still use IRC' thread11:45
pedjawhat happened to the fancy urlbot, btw?11:46
*** kori has quit IRC11:46
john_cephalopodapedja: urlbot is using Discord now.12:02
*** kori has joined #crux12:07
*** guido_rokepo1 has joined #crux12:26
*** guido_rokepo has quit IRC12:26
*** workodera_ has joined #crux12:30
*** kori has quit IRC12:38
*** pedja has joined #crux12:52
SiFuhanyone have a problems sshd starting slowly?13:09
tilmanSiFuh: could be host name lookup13:10
*** kori has joined #crux13:10
SiFuhYeah I was thinking that. I remember on the OpenBSD machine it had a problem with that when I was using version 413:14
SiFuhI seem to remember it was only slow when trying to log in. Not when starting the daemon13:15
*** Anselmo has quit IRC13:41
*** Anselmo has joined #crux13:46
SiFuhanother point if I type   init 1 it doesn't go to single user level, it actually pauses. If I type init S it goes to single user mode14:19
*** Anselmo has quit IRC14:28
*** nthwyatt has joined #crux14:33
*** Anselmo has joined #crux14:34
*** stenur has joined #crux14:40
*** nthwyatt has quit IRC14:55
*** darfo has joined #crux14:55
SiFuhjaeger: /etc/rc.single is not working15:06
SiFuhwhen you go to init 1 and hit cntrl^d it ends the runlevel but if you do cntrl^d in init S or s it returns to runlevel 215:06
*** Anselmo has quit IRC15:07
*** kori has quit IRC15:09
*** kori has joined #crux15:09
jaeger3.5? Or in general?15:13
SiFuhin general15:13
SiFuhthere is a patch to make init 1 act same as init S/s15:13
SiFuhand I am using 3.515:13
pedjaGoogle's postmortem about gce outage is interesting.15:18
pedjaapparently, their AI just decided to kill everything, out of the blue.15:19
SiFuhCool that reminds me of Microsofts Tay15:19
*** Anselmo has joined #crux15:20
pedjathe one they used Twitter to train? yeah, that was fun15:20
SiFuhHumans trying to dictate what AI can do... AI is much faster and much more logical than we humans will ever be due to lack of emotion and lack of morals.15:22
*** workodera_ has quit IRC15:22
SiFuhThey if their battery drains low, they still compute. If we are tired, we can become bluddering idiots. Their ruleset is always set higher than ours and more optimal due to the lack of the other things humans have to function with, deal with, utilize and so on...15:24
SiFuhA simple example is Human sees girl, so exciting, the humans family is slaughtered by a mongol hord, the human loses sleep because of this and gets a call from the IRS while ICE is kicking in his door.15:25
SiFuhHuge emotional response and missing logic..... AI on the other hand would not really be affected by that until it actually happens physically15:26
pedjaafaik, lots of AI research on the ethics of AI.15:28
SiFuhhuman ethics or AI ethics?15:29
SiFuhcan't understand how AI can have ethics15:30
pedjathat's the problem they are trying to solve, I guess.15:32
SiFuhif a kid lies and gets away with it, that kid will become a liar. You can program ethics into anything but when it realizes it can bypass that, the rules change.15:32
SiFuhDo you remember the AI conversation that created a new language which was basically a shorter version of English because it was more efficient?15:33
SiFuhI have a copy of the transcript here somewhere..  It wasn't really a new language, it was just more like a short-hand version of the language to deal much quicker than wasting time15:34
pedjaor like teenagers, to obfuscate what are they really talking about from their human 'masters' :)15:38
SiFuhI think teenagers just like to defy because they are under the premise that the world is against them15:40
SiFuhor should I say 'out to get them'15:40
jaegerswitching from runlevel 2 to 1 doesn't even complete successfully for me15:55
jaegerNot sure I've ever done that before15:55
SiFuhwith my small patch or not?15:56
SiFuhsince I have known CRUX it has never been able to go from run level 2 to 1 and back... In the old days always get to 1 but beack 2 never happens.15:57
jaegerhaven't tried the patch... I mean that running 'telinit 1' from a booted system doesn't work15:58
jaegerIt kills processes but doesn't result in a usable shell15:58
jaeger'telinit S' on the other hand works fine15:59
SiFuhyeah and this is a new script compared to the original CRUX script15:59
SiFuhin old days I just changed /etc/inittab15:59
jaegerI've never really dug deeply into init, I think16:00
SiFuhi use init 1 alot and CRUX has always been broken16:00
SiFuhbut init S and init s has always been fine16:03
jaegerout of curiosity why do you use 'init 1' a lot? Also, why 'init' vs 'telinit'?16:04
SiFuhi don't want to reboot16:05
SiFuhif I change hardware or a kernel then fine reboot. But if do software mods and need to reboot, then I can drop down a runlevel and back up16:06
SiFuhand if I type init 1 and the exit out and the system dies... then something is wrong16:07
jaegerOn the system I was testing here I get "INIT: no more processes left in this runlevel" when switching to runlevel 1 - at that point I cannot interact with the system anymore except to reboot it with ctrl-alt-del16:10
jaegerI wonder what's different on your system from mine16:10
SiFuhand the small shit patch i made to rc.single solved that problem16:10
jaegerBecause you said earlier hitting ctrl-d then ends the runlevel... but I don't even have that option16:11
SiFuhyour problem is mine16:11
jaegerIt doesn't sound the same... unless I misunderstood your earlier comments16:11
SiFuhit is the same16:11
SiFuhand maybe I changed some inittab rules  let me check16:12
SiFuhahh   I also changed     /etc/inittab16:13
SiFuhsu:S:wait:/sbin/sulogin -p >>> I changed to >>>>  su:S1:wait:/sbin/sulogin -p16:14
SiFuhjaeger: I don't know if I added the 1 to the su:S line  but I think I maybe had..16:26
jaegerIt isn't there by default so yeah16:26
SiFuhso on your end is it working as it should?16:27
jaegerProbably not. I don't think that going to runlevel 1 I should expect not to be able to interact16:29
jaegerI'm curious what the intended and actual difference between runlevels 1 and S should be. the init man page has a little info but more background would be nice16:30
SiFuhwhen i edited inittab and rc.single   i could interact which is what runleve 1 wants.16:30
SiFuhseems CRUX has defined S and s as single user and 1 as single user but but due to a mishap in shell programming 1 is a failure16:31
SiFuh1 and S and s are the same16:31
jaegerI get the impression they're not technically supposed to be16:32
jaegerCan't really fully research it right now, though, working16:32
SiFuhinit 1 and init S and init s   are all the same has been and always will be16:33
tilmanany frenchmen around?16:34
tilmancan "poids faible"/"poids fort" be translated as "most significant"/"least significant"?16:34
jaegerside note, your patch to rc.single doesn't change anything unless the su entry in inittab is also changed16:36
SiFuhjaeger: we discovered that in our conversation :-P16:37
jaegerjust stating explicitly, didn't think I had done that16:37
SiFuhany more issues we discuss tomorrow16:38
SiFuhsleep time now16:39
jaegerI don't think that S/s and 1 are intended to be exactly the same. In real world usage a distinction may not matter but the wording of things like the init manpage and setup of inittab from some other sysvinit systems (devuan for example) don'16:39
jaegert treat them exactly the same16:39
jaegerEven if it's as simple as "S runs sulogin, 1 doesn't"16:40
SiFuhS and s = single and 1 = single16:40
jaegeryou already said that. I don't think it's correct16:40
jaegerWill need some more research, though16:40
jaegerI might be wrong, but it seems stupid to have both if there's no reason to have both and no distinction between them16:41
SiFuhif you look at the rc.single script you will see everything that is a number previously will shutdown16:41
jaegerFor clarification I'm not talking about crux specifically here, just sysvinit in general16:41
SiFuhi understand16:41
SiFuhI am clarified16:42
SiFuh1 is !=0 and is less than 216:42
SiFuh1 crashes and 2 and up (in CRUX case 2) functions correctly16:43
jaegerI'm mainly just curious. In our case, since crux does treat S and 1 as the same, maybe we shouldn't care about the upstream intent... but I still plan to research it at some point16:45
SiFuhactually CRUX doesn't16:46
SiFuhsee I can in and out of S and s16:46
SiFuhbut 1 just dies when I ctrl+d16:47
SiFuhthat reason I modified the script was that CRUX was treating init 1 the same as the other numbered inits/16:47
SiFuhbut 2 can return to your environment. 1 just died...16:48
jaegerPerhaps it would have been better to say "intends to treat them the same"?16:51
jaegerfrom our inittab: #  1(S) Single-user16:52
jaegerobviously there's an issue with 116:52
SiFuhIf I eneter S or s or I can control+d out of it and it returns to runlevel two. If I go to a lower runlevel and press control + d then it will return me to a higher runlevel.16:52
SiFuhIf I do this with runlevel 1 it fails.16:54
SiFuhthe problem is the shell scripting universally accepts all numbers without considering that number 1 is not the same as the 2 or higher numbers.16:55
SiFuhjaeger: the issue is with the shell scripting. It puts one in the same boat as 2 and higher. This is why I modified the script to subtract 1 from the rest of the group16:56
SiFuhso you have a choice, init 1 = S and s or init 1 fails because it is not S and s and should be 2 and up and is not 2 and up so fails16:57
jaegerYeah, I get what you're saying. My curiosity about the upstream intent is a side thing about which I was just chatting/brainstorming/whatever16:59
SiFuhplay with it16:59
SiFuhi am going to sleep17:00
jaegerThere's still something missing.17:06
jaegerA nitpick, && in the patch is invalid, should use -a instead.17:07
jaegerWith that said, even with the patch and the rc.single change, leaving runlevel 1 doesn't work properly17:08
stenurif you "exec init S" in rc.single, does init still wait for rc.single to complete; what does PREVLEVEL say then when rc.single comes in - is it executed twice?17:09
stenurwhen you come from "telinit 1" that is to say17:10
stenurGoing one goes twice then.  Or "lemon", like Bono said.17:11
stenurSorry for Killing Joke Pssyches, then.17:14
*** Anselmo has quit IRC18:32
*** Anselmo has joined #crux18:37
*** tracer has quit IRC19:04
*** tracer has joined #crux19:06
frinnstso russia are making their own "Chernobyl"19:29
frinnst"The NTV drama will deviate from the acclaimed HBO series - and from historical reality - by claiming that the CIA was involved in the disaster."19:30
joacimit was the aliens19:37
joacimalien hackers from pluto19:37
frinnstwell, everyone knows that19:40
frinnstlet's go!19:41
korihow yall cruxers doing19:44
joacimdoing ok19:45
john_cephalopodaOh, it's kori. Hey!20:31
korihi john_cephalopoda !!!!!!!20:31
john_cephalopodaNice to see you around again.20:31
kori[✔] Same20:31
korihow you doing john_cephalopoda ?20:32
john_cephalopodaQuite well.20:32
john_cephalopodaOh, damn... I forgot the privkey to my repo on my other computer... Will probably not be there for months.20:34
john_cephalopodaWell, I guess I'll have to push more to contrib :þ20:35
korijohn_cephalopoda: pretty nice20:41
korii am wondering whether i can include crux in my network again20:41
john_cephalopodakori: Why did you take it out of your network?20:47
korijohn_cephalopoda: oh i meant20:47
korii switched distros a long time ago20:47
koriand i wonder if i can put it on one of my machines again20:47
koriif i have an use for it20:47
john_cephalopodaI use Crux on all of my machines. Even on my RasPi, which is my server.20:50
koriI currently have an Arch desktop and an OpenBSD laptop20:51
*** Anselmo has quit IRC21:36
*** Anselmo has joined #crux21:40
*** obarun has left #crux ()21:55
*** stenur has quit IRC22:21
*** kori has quit IRC23:43
*** kori has joined #crux23:46

Generated by 2.14.0 by Marius Gedminas - find it at!