00:03:32 1- What do you have in your rc.conf for WiFi configuration? 2- (about not seeing debug output) In /etc/syslog.cionf, check whether "*.=debug /var/log/debug.log" is commented out. If it is, uncomment it, save, then run: service syslogd reload 00:03:47 s/cionf/conf/ 00:04:08 (For 1, use a pastebin.) 00:04:25 wlans_iwn0="wlan0" 00:04:25 00:04:36 ifconfig_wlan0="WPA DHCP" 00:05:11 The debug sysctl does nothing since debug is not compiled in, I just found that and I'm recompiling the kernel. 00:05:52 options IWM_DEBUG # is required for the sysctl to do anything. 00:07:25 you have 'wlans_iwn0' in there 00:07:31 syslog debug file is enables. 00:07:33 not wlans_iwm0 00:08:00 Oh I have that, too. Sorry, mental mixup. 00:08:23 adjust it, then run service netif restart, and wireless should come up 00:08:55 keys should go in /etc/wpa_supplicant.conf which you probably have filled already 00:09:01 No, I already also have: wlans_iwm8="wlan8" ; ifconfig_wlan8="WPA DHCP" 00:09:22 I also have a USB wireless in there that works fine. 00:09:24 8 ? 00:09:25 what's in dmesg? 00:09:41 for.. iwm, wlan 00:09:55 iwm0: mem 0xf7c00000-0xf7c01fff irq 17 at device 0.0 on pci3 00:09:56 00:10:06 iwm0: hw rev 0x140, fw ver 17.352738.0, address 7c:5c:f8:dc:f9:3d 00:10:06 00:10:22 The only wlan entries are for the usb wifi dongle. 00:12:05 Hmm, that's iwm0 (zero), not 8 (eight). 00:12:26 As I said, the same FreeBSD install works for other iwm cards, just not this one. I am compiling with IWM_DEBUG now. 00:12:39 why? 00:13:28 kldstat would show if_iwm being loaded, right, here 00:13:29 What does "grep iwm8 /var/run/dmesg.boot" show? 00:13:32 font too small, i guess 00:14:26 er, yeah, if dmesg doesn't align with rc.conf it's also not going to work 00:14:37 wait. 8? 00:14:47 Holy... 00:14:59 Thanks for spotting that. 00:15:05 if your keyboard has a numpad on the right, use '0' on the numpad since it's far away from '8' 00:15:19 ya 00:15:49 I use an old VGA font that puts a dot in the 0. Easy to mistake. 00:16:07 they still should look nothing alike 00:18:45 ... and it works fine. Just the 8 instead of the 0 that was wrong. Thank you so much. 00:20:20 You're welcome. 00:37:46 heh 01:00:00 ... and now my root filesystem is full because iwm debug goes to /var/log/messages :-D 01:01:32 Logs are working! 01:01:33 :) 01:02:13 mv /boot/kernel /boot/kernel.debug; mv /boot/kernel.old /boot/kernel && reboot 01:02:23 I would have wished that it goes to /var/log/debug, but I think the kernel can't do that. 01:20:15 Is there a discussion of what "jv" is rc.d/jail? 01:22:14 s/is/is in/ 01:31:05 looks like it's the name of the jail 01:31:51 er, short identifier 01:32:15 rtprio: There's functionality that seems to care about it. 01:32:31 eval _jconf=\"\${jail_${_jv}_conf:-/etc/jail.${_j}.conf}\" 01:32:34 yeah, i'd say so 01:32:38 Yeah, but there's more than that. 01:32:41 Follow it down. 01:32:48 I'm going to start digging through commit logs. 01:33:42 bbiab - I'll summarize what I find 01:34:57 i should really start using jails 02:18:09 does anyone here use freebsd? 02:18:31 sorry 😂 i meant: does anyone here use xmpp? 02:18:48 i created a muc, for freebsd 02:19:18 happy to hand it over to freebsd devs, but otherwise feel free to join: xmpp:freebsd⊙xf?join 02:19:51 ... i meant, xmpp:freebsd⊙mxf?join 04:19:12 Hi. We have an old 7.4-STABLE system (Pentium 4) with a Cyclades PCI card connected to some external devices. We are using the cy(4) driver to control the devices. I checked https://cgit.freebsd.org/src/tree/sys/dev and don't see a cy driver. Anyone know the last release of FreeBSD to support this driver? 04:19:13 Title: dev « sys - src - FreeBSD source tree 04:45:17 https://www.freebsd.org/cgi/man.cgi?query=cy&sektion=4&manpath=FreeBSD+7.4-RELEASE 04:45:19 Title: cy(4) 04:45:31 use the drop list to find it 04:46:15 you can't move too much forward on that class of machine (about 2004-2005) 04:48:13 12.4 has it 04:48:42 read carefully release notes on the major versions 04:50:42 lystra, you could probably upgrade 'carefully' to about ~10 and prpobably later reaseases with some caveats (attention to changes) 04:56:42 epony: 11.4 is the latest release that will install on our 32-bit Pentium 4 system. Trying to rebuild a custom kernel now with the rp(4) and cy(4) drivers. 04:57:58 If the driver is in the tree, does that mean it compiles? cy(4) isn't in the GENERIC kernel. 04:59:58 can't say without trying 05:00:56 if I was doing this, I would not skip major branches without their release notes an understanding what is happening in the kernel and sub-systems (and file systems) 05:02:11 if you're reinstalling from scratch, you can just prepare various last minor releases from the major branches 05:03:18 so 8.4 9.3 10.4 11.4 12.4 05:03:25 and compare between these 05:03:52 in your care about up to 11.4 05:03:55 "case" 05:04:08 Ok, thanks. 05:51:07 epony: Looks like cy(4) won't build on anything past 7.4 as the tty layer was replaced in 8.0. 05:52:00 interesting 05:52:17 I tracked well FreeBSD up to 12 but not with that use case 05:52:31 (back in those years) 05:53:15 does it say it in specifics on the 8.x release notes? 05:54:54 https://www.freebsd.org/releases/8.0R/relnotes/ third point 05:54:55 Title: FreeBSD 8.0-RELEASE Release Notes | The FreeBSD Project 05:55:42 Yep. 05:55:52 that should not breat it though 05:56:15 but the multi-port devices are more picky 05:56:51 I forgot I already tried 8.0 12 years ago: https://lists.freebsd.org/pipermail/freebsd-stable/2010-April/056403.html. My memory is going... 05:57:20 I was about to say someone had reached that conclusion before you.. 05:57:34 to place the system on hold at the 7.4 release 05:57:57 and I recall well the SMP unlocking improvements in the 7-8 releases 05:59:13 https://www.freebsd.org/releases/12.4R/hardware/ still lists cy(4) 05:59:14 Title: FreeBSD 12.4-RELEASE Hardware Notes | The FreeBSD Project 05:59:44 Yeah, probably a mistake. rp(4) builds on 11.4 though. 05:59:51 and a number of other multiport boards around it 06:03:07 The nice thing about the rp is that there is a PCI card that interfaces with https://comtrol.com/rocketport-multi-port-serial-cards/rocketport-upci-interfaces/rocketport-32port-rs232-rackmount-interface. Much better than all those cables. 06:03:09 Title: Comtrol | RocketPort 32-Port RS-232 Rack Mount Interface 06:03:51 breakout cables are not a new invention ;-) 06:04:07 there are also breakout boxes in the more durable variant 06:04:35 multi-point breakout cable are expensive and failure prone (cheaper version) 06:05:24 network appliance vendors make as much money on such cables than their computer-appliances 06:06:31 We started with breakout cables and then upgraded to breakout boxes. The 1U rackmount RJ45 interface is much cleaner. 06:44:40 mmm, i have some 8 port rocketport cards 07:30:46 lystra: wow 07:34:08 rtprio: thank you. 07:34:08 rwp: thank you 07:34:08 futune: thank you. 07:34:08 http://paste.scsys.co.uk/2018 07:34:09 Title: magnet_web paste from Someone at 217.168.150.38... 07:35:40 huh 07:43:31 Which is the best tutorial or man for bectl (ex beadm) to handle upgrades? 08:05:33 ferz: the numbers after the man pages aren't a ranking, unfortunately, they are sections… so "the best" is "the only one we have" 10:09:18 xd 12:57:16 Ummm hi 12:57:24 I am wondering - how come I cannot redirect an output? 12:57:50 Software in question: Suckless' Quark: http://tools.suckless.org/quark/ 12:57:51 Title: tools | suckless.org software that sucks less 12:58:20 I get a clear output, but cannot redirect it into a file... 12:58:32 For example messages that I get when someone visits my site(s) 12:58:54 If I am delegating a DNS zone, but it will be on the same name server, do I need the glue records? 13:08:28 I can redirect the output, but it is empty 13:08:46 the file gets created if doesn't exist, so it means that the permissions are sufficient? 14:19:31 PredatorONormies: how do you redirect? 14:20:40 exec.start+="/bin/sh -c '/usr/local/bin/quark -p PORTLY -d /var/www >> /var/log/quark/quark.log &'"; 14:20:52 Inside Jail config 14:34:48 Is the output you want maybe on stderr? Add `2>&1` at the end. 14:40:15 i'd rather go with rc.d script 14:54:07 > Is the output you want maybe on stderr? Add `2>&1` at the end. < I tried that 14:54:36 > i'd rather go with rc.d script < What do you mean? 14:54:54 And it still prints to the terminal? 14:55:24 PredatorONormies: i would recommend to use a script in LOCALBASE/etc/rc.d to start the applications 14:55:42 cracauer, it prints out nothing, not even in the file (wanted) 14:56:06 otis, hmmm this way isn't recommended? 14:56:15 Why do you think it should? 14:56:16 Like I said - it's a Jail 14:56:17 inside a Jail 14:56:27 cracauer, should what? 14:56:38 What is the expected output? 14:59:39 A line that says someone connected 15:00:09 I think the printout is found in connection.c 15:00:29 Like I said - without any redirection - it prints normally 15:00:47 but I want the output to go inside an log file as I exit the jail and well - I don't have a terminal 24/7 open with it... 15:01:11 I could in theory use TMUX and then detach the terminal with Quark open, but that wouldn't log, just output to the terminal 15:09:50 So it refuses to print anywhere except the controlling tty? That's odd. 15:11:26 WhAt? 15:11:34 It refuses to log to the log file 15:11:41 I want it to log 15:11:46 I tried >> 15:11:49 I tried 2>&1 15:11:53 I tried >& or &> 15:12:04 I tried also >/dev/null, and I believe all combinations with each other 15:12:11 -probably al 15:12:12 all 15:12:19 But it prints to the terminal? 15:13:07 NoO 15:13:09 are you trolling me? 15:13:36 > cracauer, it prints out nothing, not even in the file (wanted) 15:14:27 But you said that without indirection it prints what you want, no? Not trolling, just confused. 15:16:09 indirection? 15:16:31 Without any redirection - it prints out to the terminal, yes 15:16:46 Where does that text disappear when I do redirection to a log file? No idea xD 15:16:51 would like to know 15:17:46 I took a look at Quark a little bit, and it seems like there's a printf in connection.c that does that printing out, but probably not as it has too much values given for it to be equal to the output(unless it's like not shown) 15:22:34 PredatorONormies: quark doesn't have built in logging? 15:24:58 Not that I know 15:25:06 you might consider starting it with daemon(8), it will capture stderr and log it to a file for you 15:25:08 Although I also think it wouldn't be that difficult, either 15:25:11 to implement I mean 15:25:20 daemon? 15:25:24 log it to a file? o.o 15:25:37 I don't want to capture stderr, I want all Quark output to be captured (normal text, too) 15:25:43 daemon will do that 15:25:46 Unless I don't know what I want 15:26:12 zfs-send'ing between two different pools doesn't have any rsync like features? Before my home dir was on a UFS that was rsync'ed to a mirrored zpool where snapshots were created. That home dir is now on a single disk zpool. There's no zfs'ish like features to get a snapshot of it to the existing mirrored zpool that would use less CPU/bandwidth/resources? 15:26:16 OWO 15:26:30 It seems rsyncing it to the mirrored zpool is still the best option? 15:26:52 Otherwise zfs-send is sending all the data? 15:27:10 daemon -S quark and shit will go to syslog. which you can configure to spit into your own quark log 15:27:51 or daemon -o /var/log/quark quark 15:28:28 skered: no, unless you do an incremental zfs-send 15:30:05 o.o 15:31:23 is that doing what you want? 15:32:21 1min 15:33:00 is it okay to launch quark/daemon as I did until now? 15:33:07 from /etc/jails.conf ? 15:33:26 > otis> PredatorONormies: i would recommend to use a script in LOCALBASE/etc/rc.d to start the applications < This 15:34:08 with exec.start? i don't care, i don't think it matters 15:34:57 if quark doesn't have a rc.d file and you can't be bothered to write one just keep doing what you're doing 15:37:36 :) 15:40:40 > daemon: quark: No such file or directory 15:40:48 full path to Quark? 15:41:18 yep 15:41:47 does not work 15:41:48 nothing 15:41:54 did /usr/local/bin/quark 15:41:58 and now nothing prited out 15:42:57 let me try removing the file completely 15:43:08 well it should go to the log file you specified 15:44:33 hrm, it's not seeming to work 15:44:37 xD not loading wait 15:44:45 huh you tried? 15:45:00 Yeah nothing 15:45:10 the file didn't even get created this time 15:45:16 why 15:45:22 hmmm DIR permission problems? 15:45:43 drwxr-xr-x 15:47:10 you start it as root, yeah? so daemon should be able to write where ever 15:47:57 It's in /etc/jail.conf 15:47:58 wait 15:47:59 umm 15:47:59 XD 15:48:24 brb 15:48:36 i might mention if you need logging, you might consider a less sucky web server 15:48:52 which is ironic considering the domain name this one came from 16:34:30 you mean a more sucky web browser? 16:34:32 xD 16:34:53 Logging apparently works for them 16:34:57 I forgot to mention that, rtprio 16:35:05 That's why I came to #FreeBSD 16:35:14 Talked to Quark's creator, I believe. 16:36:02 By the way this is the command I did now "exec.start+="/bin/sh -c 'daemon -o /var/log/quark/quark.log /usr/local/bin/quark -p 28000 -l -d /var/www'";" 17:50:33 no, i mean a less sucky webserver than can actually write real logs somewhere other than stdout 17:50:50 PredatorONormies: and is it working? 17:52:10 Yeah.. it's a bit stupid to write logs to stdout and not to a log lol 17:52:17 not working lol 17:53:31 I like daemon's -S option 17:54:36 let me try that 17:54:41 in conjuction with -o 17:54:52 i thought it was one or the other 17:55:02 :/ 17:55:02 we'll see 17:55:51 even lighttpd logs to a file 🤷 17:56:28 yeah, but how big is it? 17:56:34 nope tried both options 17:56:34 nada 17:57:07 > wc -l *c *h 17:57:09 3240 17:57:10 line 17:57:16 with comments and everything 17:57:22 Quark 17:59:19 why do you even need to log static files 17:59:43 In case of an DDoS, I guess? XD 18:00:03 And also my ego might like it 18:00:08 gotta stroke da ego 18:00:10 3kilolines is nothing 18:00:28 exactly 18:00:35 Ultra minimalist 18:00:51 ok. well good luck getting your logs 18:01:22 but instead of a printing-out message they could have written to a file 18:01:24 I forgot how to do it 18:01:27 I used to write C 18:01:36 fprintf? 18:02:05 I might try to modify Quark, but thanks for your help attempt :) 18:05:15 Isn't this funny, thought? 18:05:20 > connection_log(const struct connection *c) 18:05:23 it's called a log 18:05:46 but prints-out to a non-redirectable unuseful output 18:10:34 daemon should really be able to grab it, yes 18:10:51 i don't know if it needs a fflush or what 18:11:43 stderr is normally hot stdout is normally buffered in most applications 18:12:30 So it IS Quark's fault? 18:13:14 If that's so - I should copy-paste the last 3 lines you just said 18:13:17 messages 18:13:23 They must go quite out of their way to make a stream output that appears on a terminal but not when redirected. 18:14:19 yes, it is quarks fault. what did the quark author say about not being able to redirect logs 18:14:26 since you talked to him 18:14:57 rtprio: it's the new standard 18:19:51 when python -m http.server works better than the webserver you picked out, perhaps you should reconsider some of your life choices 18:20:05 daemon -o /tmp/access python3.9 -m http.server 18:20:11 ~$ less /tmp/access 18:20:12 ::ffff:127.0.0.1 - - [05/Jan/2023 18:19:09] "GET / HTTP/1.1" 200 - 18:27:50 PredatorONormies, instead of calling quark directly after sh -c; instead call a init.sh and put this in it: stdbuf -i0 -o0 -e0 daemon -o /var/log/quark/quark.log /usr/local/bin/quark -p 28000 -l -d /var/www 18:28:21 its just a guess but I do not have quark to test with 18:28:38 so 18:29:01 "exec.start+=/some/path/init.sh" 18:29:03 oh that's a utility i haven't come across 18:29:21 rtprio, me neither was just idling googling; looks handy to remember though 18:30:13 might be better to even call the init.sh from daemon 18:30:19 instead of the other way around 18:30:25 that means stdbus is called in the child 18:31:15 > yes, it is quarks fault. what did the quark author say about not being able to redirect logs < The 1st time I e-mailed him on the mailing list he said that he can do it I believe.. the 2nd time - he didn't respond 18:31:26 possibly because he thought I was joking/trolling him 18:31:51 lol rtprio 18:32:17 a init.sh??? 18:32:33 sure why not 18:32:48 start.sh or begin.sh 18:32:54 like write a script and call it with exec.start from within /etc/jail.conf ? 18:32:56 you could even chain the commands with && really 18:33:01 but a startup shell is tidier 18:33:06 yes 18:33:18 can I just have it the way I had it now? 18:33:28 that did it, daemon 18:33:33 \o/ 18:33:35 REALLY? 18:33:36 which side 18:33:38 before daemon or after 18:33:40 rtprio, you spoiled the fucking surprize nooo 18:33:51 just have to type faster 18:33:57 daemon -o /tmp/quark stdbuf -i0 -o0 -e0 quark -p 8000 -d /tmp -l 18:33:58 oh 18:34:02 so yes 18:34:04 faster? 18:34:05 it was quarks faul 18:34:06 oh 18:34:06 t 18:34:12 if it was before daemon it was daemon's fault 18:34:19 if it was within daemons exec child it was quarks 18:34:27 daemon detaches so that doesn't make any sense 18:34:33 of course it wouldn't have stderr open 18:35:04 if its on the inside of what daemon starts, the only thing buffering writes can be the child 18:35:15 because that is all you changed 18:35:36 by having stdbug inside, the child daemon created got told it has no buffers 18:35:45 IT WOOOOOOOOOOOOOOOOOORKSFUCK 18:35:47 FUCKY 18:35:47 YEAH 18:35:51 Y OU ARE A FUCKING GENIUS 18:35:57 just lucky with google ;p 18:36:05 you're joking, right? 18:36:09 What did you fucking type? lmao 18:36:16 what magic spell 18:36:16 "exec.start+="/bin/sh -c 'daemon -o /var/log/quark/quark.log /usr/local/bin/quark -p 28000 -l -d /var/www'";" 18:36:17 err 18:36:23 bourne shell no stdout buffer 18:36:28 in duckduckgo 18:36:35 0.o 18:36:53 its the second response 18:36:58 https://stackoverflow.com/questions/3465619/how-to-make-output-of-any-shell-command-unbuffered 18:36:59 Title: stdout - How to make output of any shell command unbuffered? - Stack Overflow 18:37:09 because it was clearly a buffer write/flush problem 18:37:22 I see. 18:37:37 but anyhow learned a cool new command 18:37:57 How can I repay you? How about a blowjob 18:38:06 (gets banned) 18:38:12 nah im good, the dog / mrs would not approve lol 18:38:19 donate to something that deserves it :) 18:38:26 :OOOO 18:38:40 I like you 18:38:47 also nice name lol the irony 18:39:00 I know that is why I seen the problem irc client highlighted me to heck lol 18:39:40 daemon: lol, best be intimately familar with /usr/sbin/daemon 18:39:47 I use it regularly 18:39:51 hahaha 18:39:54 PredatorONormies: $20 to the freebsd foundation 18:40:01 damn 18:40:07 rtprio, oof 18:40:14 I might one day 18:40:22 But I only use Monero 18:40:28 Does FreeBSD accept Monero(XMR)? 18:40:42 i once made $50 with a one liner shell command to save someones ass in production, $20 is a steal 18:40:56 lol 18:42:04 i was really proud of it too, it used xargs twice! how often do you get to do that 18:42:05 there are plenty of worthwhile projects out there; that accept monero though; not sure! but you do not need to donate money anyhow, got an old computer or or something you do not need; perhaps some blankets for a local animal shelter? do whatever you can :) 18:42:11 Damn I feel so wholesome right now 18:42:18 haha rtprio 18:42:22 xargs is weird 18:42:33 I always forget to use !! and sed with it 18:42:41 xargs is a command I have never managed to use without both a search engine (examples) and the man page 18:42:48 its worth that writing regexs with perl -e 18:43:00 s/worth/worse/ 18:43:28 daemon, I hope I can help the world with software.. alike Suckless'.. and try to help people like you do (well.. not nearly good like you, but yknow) 18:43:49 local animal shelter? Not sure if there are any (nor how that would help animals 0.o) 18:44:01 lol daemon 18:44:04 i use it all th time; ls downloads/*/*.rar |xargs -n1 unrar e 18:44:07 for example 18:44:23 daemon, can I ask a question? Do you eat meat? 18:44:24 I always seem to lose track when you use place holders 18:44:33 PredatorONormies, I do 18:44:44 I don't even have to say anything .. 18:45:00 ANyways.. writing an e-mail right now 18:45:31 :) have a good one! time for a little gaming for me 18:45:43 use the placeholder when the argument isn't at the end 18:47:49 cd /usr/local/etc/rc.d; ls s* |xargs -n1 -J% service % status 18:48:22 > daemon... instead of calling quark directly after sh -c; instead call a init.sh < why? 18:54:06 soooo 18:54:10 that was an idea to make it easier to write that stuff, but in the current form you shouldn't need a shell anymore exec.start+="daemon -o /var/log/quark/quark.log stdbuf -i0 -o0 -e0 quark -p 28000 -l -d /var/www"; probably works fine 18:54:20 anything I should tell him in e-mail other than copy-pasting most of what you two said? xD 18:54:50 1 (on FreeBSD) 18:54:50 2 18:54:50 3 ... 18:54:50 4 daemon should really be able to grab it, yes 18:54:50 5 i don't know if it needs a fflush or what 18:54:50 6 stderr is normally hot stdout is normally buffered in most 18:54:52 7 applications 18:54:54 8 ... 18:54:56 9