00:04:28 j/buffer 15 00:55:50 I'm trying to get OpenZFS 2.2 to work on FreeBSD 13.2. I've installed from source per the directions on the OpenZFS wiki. However, whenever I load openzfs.ko, I get dropped to a mountroot shell with complaints that zfs is an unknown filesystem 01:00:46 Oh, are we showing puppy pictures? 01:00:49 https://trioptimum.com/~crtxreavr/roxie/ 01:00:50 Title: Single File PHP Gallery 01:05:43 for folks who didn't disable sendmail, do you read messages in the console, or do you forward to your cozy webmail? if in console, what program? 01:13:08 If I were going to read it in console, I'd use pine or mutt. 01:13:26 Though, mailx is in base. . . it's only semi-horrible. 01:13:38 scoobybejesus: I've used mutt in the past. I'm not sure if there are others theses days. 01:14:22 Oh mail/alpine 01:14:34 They added features and tweaked the name. 01:15:00 excellent, many thanks. i will look into those 01:15:43 mutt was easy to get working with GnuPG, which is why I used it. 01:16:49 i recently became curious about the mails processed by sendmail 01:17:39 if you don't need the mails, you can turn off sendmail 01:19:39 i recently saw in zfs diff'ing snapshots that /var/spool/clientmqueue will still fill up as a result of being unable to deliver mail. any idea how to deal with that? 01:20:50 scoobybejesus: pkg search -c mail | grep client | sort :-) 01:21:09 scoobybejesus: are you sendint out email from that host where you see that ? 01:21:20 i don't know yet if i don't need the mails. i want to try to get comfortable with what it sends. seeing all the immediate ssh login attempts in there make me put setting up sshguard on my todo list 01:22:34 in particular, this is a jail snapshot with the four sendmail items set to "NO". it's a laravel app running. i'm not sure if php is sending mails or this is from cron jobs or what. with sendmail off, all i know is these files tell me that connection to 127.0.0.1 was refused 01:24:10 since the emails are not deliverable, something's not configured properly so that that's why its filling up. Probably a better approach might be to lok at hte config to see why its not getting delivered ? 01:25:28 i figured they were not getting delivered because sendmail was disabled so there's no daemon listening 01:26:33 cron, for example, may be sending mail, periodic or whatever 01:26:40 scoobybejesus: I just checked my jail. I have sendmail_* set to NO as well, and no sendmail processes running. 01:27:26 i have syslogd_flags="-ss" and cron_flags="-J 60" as well, so maybe related to one of them 01:29:12 most likely syslogd. I have syslogd_flags="""". cron_flags is the same as yours. 01:29:34 hmmmm 01:29:55 ok I just checked my jail and I have the same problem as well. clientmqueue is filled up. 01:31:31 Interesting. For syslogd, -s says "Operate in secure mode. Do not log messages from remote machines. If specified twice, no network socket will be opened at all, which also disables logging to remote machines." 01:32:41 I was super curious why i had notable snapshot diffs to send over the wire, so i ran zfs diff on the snapshots, and there were lots of clientmqueue items, so that's why i'm going down this rabbithole 01:38:29 I've been learning a lot today about device.hibts nids and everything. I'm exciting freebsd makes me want to learn 01:38:46 Device.hints* 01:48:25 makes me want to learn too :) 01:55:54 It's like freebsd wants you to enjoy your pc and hry have so much documentation and hints. I bet the experts can run freebsd on a toaster with everything working 03:21:15 nobody is not allowed to log in on /dev/pts/5 03:21:18 nobody is not allowed to log in on /dev/pts/5 03:21:33 nobody is not allowed to log in on /dev/pts/5 03:21:50 ok once more with feeling... 03:22:21 nobody is not allowed to log in on /dev/pts/xxx WTF? 03:23:58 I am thinking its a login.conf issue but it was SO long ago that none of my team remembers what was done to fix it... 03:33:46 feurig: why are you trying to login with 'nobody' somewhere? 03:41:04 I am not trying to login as anything 03:41:23 I am trying to start the node_exporter service. 03:41:23 `nobody` is trying to authenticate somewhere, the questions are where and why 03:41:54 ah node_exporter, let's take a look here 03:42:18 ``` 03:42:23 service node_exporter start 03:42:24 Starting node_exporter. 03:42:24 nobody is not allowed to log in on /dev/pts/5 03:42:24 su: Sorry 03:42:24 /usr/local/etc/rc.d/node_exporter: WARNING: failed to start node_exporter 03:42:27 ``` 03:43:01 so i pkg install node exporter and then try to start it. 03:43:49 My peers have seen this before but dont remember what the solution was. 03:44:23 It works on most of our systems (12.1..13.2) 03:44:27 do you have any configuration for the node_exporter service in rc.conf, or just enable set? 03:44:58 nothing other than enabling it. 03:46:35 `su -m nobody -c 'echo hi'` same result? 03:48:28 yep. 03:49:14 anything interesting in /etc/pam.d/su ? (please pastebin or cat ... | nc termbin.com 9999 if you're going to paste it) 03:53:25 everything points to system defaults 03:53:41 from earlier, mns, `mailq -Ac | grep "Nov 4" | wc -l` in the jail returns 1000! 03:55:15 and they seem to all be at the same time 03:55:16 feurig: /etc/login.access all comments? 03:56:26 * kevans is learning a lot about pam tonight 04:00:52 haha.. the the mailer.conf(5) bugs, it says: The entire reason this program exists is a crock. Instead, a command for how to submit mail should be standardized, and all the "behave differently if invoked with a different name" behavior of things like mailq(8) should go away. 04:01:40 i mean, it's got a point 04:05:55 the same BUGS section can be found in MAILWRAPPER(8).. fascinating 04:06:05 interesting description of things in there 05:20:45 Can somebody explain to me device.hints and exactly how they work and how do I know what to change I'm looking at https://man.freebsd.org/cgi/man.cgi?device.hints 05:20:46 Title: device.hints 07:01:07 Good morning 11:45:45 jb1277976: at the most basic, it let's you disable troublesome devices. but it also changing certain configurations. 11:46:40 many drivers have configurations of their own, see for example, https://man-dev.freebsd.org/em.4#LOADER_TUNABLES 11:46:41 Title: em(4) - FreeBSD Manual Pages (Dev) 11:49:17 but this hooks in earlier into how the kernel discovers the device and what state it finds it in, before loading the appropriate driver (often handled by https://man-dev.freebsd.org/devmatch.8 ) and handing control to it 11:49:19 Title: devmatch(8) - FreeBSD Manual Pages (Dev) 12:26:47 hello freebsd people! 12:29:06 hi 12:37:01 o/ 12:37:20 I cant wait to have a backend freebsd server. I want one soo bad 13:10:31 Hi 13:10:51 I have screen tearing 13:10:54 amdgpu 13:11:51 nevermind I fixed it. But how can I make touchpad click on tap? 13:26:39 see xorg.conf(5) 13:26:39 https://bsd.to/wobt/raw 13:26:40 Title: wobt 13:37:19 ok 13:45:43 babz: Thank you so much! 13:48:04 At first I thought FreeBSD was a hassle. But because it is a bit more hassle than OpenBSD I was motivated to put in a bit more effort to trim it down, so now I feel really happy with the results. E.g. instead of installing drm-kmod I picked the specific packages needed, so drivers and all for graffix is less than 15mb now. Also the bluetooth firmware thing is broken which is a plus since bluetooth 13:48:06 is a security issue and I never use it anyway so BT can just, dissappear from my systems 13:52:11 > bluetooth is a security issue 13:52:14 how ? 13:52:47 a) It is unencrypted 13:52:56 b) It is bloat, extra surface area 13:54:01 it is encrypted 13:54:03 c) the protocol has numerous security issues 13:55:00 d) its usually networked hardware, running opaque firmware blob, that has DMA access to the rest of your computer 13:55:24 given (c) then (d) is considerably less secure, and there's a LOT you are forced to trust 13:56:43 Ok so it seems I was wrong about the encryption. You're right, it is encrypted 13:56:47 see biab, braktooth, blurtooth, bleedingtooth, bluebugging 13:57:21 well, we could say the same things for wifi or gpus (or the whole PC platform, really) 13:57:23 Yea it is proprietary protokoll as well, running on network card 13:57:24 https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=bluetooth and $$PROFIT$$ 13:57:25 Title: CVE - Search Results 13:57:36 bu yeah, if you dont use it, dont load a driver... 13:57:38 babz true, but you *can* get opensource wifi cards 13:57:47 Yes, I got wifichip that allows open source firmware 13:57:52 or use a wired nic even (maybe with a very long cable :D) 13:58:06 I don't know about fully open gpu tho 13:58:10 I guess no good ones are 13:58:25 computers are bad, and widely regarded as a mistake 13:58:41 babz: That was the pluss, the bluetooth firmware for my specific chip seems broken on FreeBSD (or I skipped a step to set it up properly because I really don't want or care about bluetooth) 13:59:16 Also I remove wifi chips from all my devices that don't really need wifi 14:00:05 I accidentally broke a raspberry pi 4 trying to figure it out, but the schematics are either a lie or the wifi circuit is so integrated that if you disconnect it it bricks the pie. 14:00:37 Ok the pie isn't really broken, it just refuses to boot or do anything, I still get the red light indicating it works 14:02:07 I am considering yeeting the wifi out of my home. Just, throw away the router (main router has no wifi chip, secondary I just use for wifi, but mainly as a switch) 14:34:37 zyxer: where did you find the schematics ? 15:07:35 babz: Ok so it was more like a diagram and not actual schematics 15:08:36 Like, image of the RPi and then arrows with "This is wifi chip, this is CPU, this is RAM etc. etc." 16:14:36 anyone have experience using tmux and "handling" the configuration files.. i have my tmux.conf in ~/.config/tmux/.tmux.conf (as per the man page) but it seems to NEVER source it 16:19:17 looks like you have an extra dot 16:21:24 bahhhhh 16:21:38 it is ~/.tmux.conf OR ~/.config/tmux/tmux.conf 16:21:46 ill go back to my hole, thank you nimaje 16:22:46 (and I think it was a bad decision to promte that bug in ls to be a feature and hide files when they start with a dot, it was lazy coding to hide . and .. without checking for NUL) 16:25:07 I would like to create jails with deterministic nic names, that just go, vnet0, vnet1, etc… how do I do that with epair(4), when I have more than one jail? 16:25:43 nimaje: agree and i am here.. trying to "reshift" linux tutorials on tools such as tmux and neovim and i am trying to get my brain to translate on the fly... 16:25:51 Thanks meena 16:28:00 meena: can't you re-route stuff with device.hints like say there is sound in the hdmi port but you want to be in the speaker intead. device.hints do that right ? 16:29:21 jb1277976: most likely, but I don't touch real hardware, only the virtualized. I have a cool talent for manifesting bugs, and hardware can explode. It's for safety. You'll have to ask other people who enjoy making their computers make noises what to do about sound, or you could try punch that into a search engine and see what others have already 16:29:21 found 16:30:51 Thanks 16:34:45 ifconfig epair create name vnet0 # gives me vnet0, and epair0b, and I basically need this to go the other way around 16:35:05 I might have to go read code I've helped write… 16:37:05 with "deterministic nic names" you mean computed from the jail name? 16:39:38 nimaje: nah, i just want every jail's nic's *inside* the jail to be vnet0, vnet1, etc… 17:20:01 meena, I create epairs one pair at a time with a name I know and then rename them immediately, two renames. Here is a paste demo'ing what I do. https://bsd.to/VbXH/raw 17:20:02 Title: VbXH 17:21:09 rwp: thank you 17:24:32 I have my jails all numbered with a unique number for my convenience. I use that number to make the epair and then rename each side (two renames) to other names using the same number. By using the same number I never have a name collision no matter what order I use to start/stop jails. 17:25:30 I did it the other way originally and did actually run into collisions in my hacking around on things and so switched to always assigning a unique number globally to avoid that problem. 17:41:22 rwp: aye, that's what I'm doing right now, but I'm not renaming them yet. 18:00:09 I've seen someone using /etc/jail.conf and had epairs and vnets setup in a uniform, deterministic way. 18:00:30 I thought it was your setup meena, but might have been someone else. 18:16:17 Hmm... "freebsd-update fetch" said that it downloaded files for 13.2-RELEASE-p5. I then ran "freebsd-update install" and rebooted. The system now says "freebsd-version -kr" both 13.2-RELEASE-p4 which I expected to say -p5. 18:16:44 Is it me? Or did a patch release sneak out without updating the number to -p5? 18:35:08 "No updates needed to update system to 13.2-RELEASE-p5." is reported by freebsd-update fetch. 18:37:23 rwp: Try freebsd-version -kru 18:38:27 To the best of my knowledge the p5 update only updated the userland. the -u option will show you what patch level your userland has. 18:38:38 thorre, That's what I already reported above. The -u is the userland and it also reports -p4. 18:39:27 It's an odd thing when updates come through but the version number is not changed. 18:39:31 Sorry, I missread your post. 18:40:12 No worries. I noticed in passing that, for example, /rescue was in the upgrade list that it showed me. I forget what other things. I could diff I suppose. 18:40:32 This may be a long-shot but did you have the possibility to re-start your system after the update? 18:41:11 thorre, Oh! My apologies! On my other machine freebsd-version -u does report -p5 there. My bad! Sorry. 18:41:28 No problem at all :-) 18:42:10 That's the problem with me doing ad-hoc upgrades this morning without a plan and without a systematic flow to things. 18:42:35 rwp: Do you know what CDO is? 18:43:03 I'll bite. No. What is CDO? 18:43:30 It is OCD in the correct alphabetical order. Sometimes OCD is good, for example upgrades are best performed with OCD. 18:43:49 LOL! Yes. 18:45:20 I got confused between the differences of the base kernel, the base userland, the pkg userland. I confused the base userland files with the pkg files. Since pkg feels like it should be userland, and it is, but it's a different userland. That was my confusion this morning. 18:45:55 Well, I am glad that the problem now is solved. 18:46:12 Thanks for helping me get my mental confusion straightened out this morning! :-) 18:47:35 Okay. With that understood now I have one more machine to upgrade and I am typing on it right now so I will BBIAB. And maybe get a bite of lunch too. 20:25:27 i wonder if my jail vnets work now 20:42:17 thanks rwp after a while i could get to ttyv6 and yea alt+f9 worked 20:42:35 Question is there a reason why freebsd dosen't use sof-firmware ? 20:42:49 i think it would help a lot of pc's get sound working 20:42:58 especially on freebsd 20:43:56 isn't there a port? 20:44:45 nothing on freshports 20:44:51 i could be typing it wrong also 20:45:33 I can't find it either 20:47:16 what is sof-firmware; i've never had much trouble getting sound to work 20:47:57 Yea sound on here worked out of the box 20:48:33 https://www.sofproject.org/ when i was on arch and debian it was the only way to get sound working 20:48:34 Title: Home - Sound Open Firmware 20:48:53 im trying to setup sound on this chromebook 20:50:11 and the chromebook is running freebsd? 20:50:44 yep its on it right now. 13.2-RELEASE 20:51:06 wifi works using iwlwifi 20:51:13 i even have X working 20:51:46 I have also not heard of Sound Open Firmware either. Sound has always Just Worked for me. Glad to hear you have things working and X up and running. 20:52:12 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=267817 20:52:14 Title: 267817 – No sound with Intel Comet Lake PCH-LP cAVS sound card 20:52:26 nobody has responded yet. i found that the other day 20:52:26 I want to hint that on the vt consoles Alt-Fn gets to that numbered vt console. But under X the Alt key is usually used by the window manager. So under X it is Control-Alt-Fn to switch to the vt console. 20:52:50 Got it 20:54:00 unless you plan on writing a driver, it's probably not worth it. get a pair of usb headphones and call it good 20:54:27 Another thing is that FLOSS software often has to react to new hardware being available and that can take some time for people to reverse engineer new hardware. Most of us run on hardware a few years old and by then well known and well debugged. 20:55:27 +1 for one of those USB sound adaptors. Several are fully FLOSS and those work very well. 20:57:09 rwp: my microsoft usb nic worked in freebsd without drivers, yet somehow needed a driver in windows 20:57:14 i always thought that was amusing 20:58:19 on that note, iphone tethering works OOTB on Linux and FreeBSD, but on Windows you have to install iTunes first... 21:03:42 rtprio, That does not surprise me actually. FLOSS is possible to add drivers more quickly than large corporations. 21:03:56 Add existing drivers I mean as a clarification. 21:04:48 Got it. im giving up and putting linux back on here. no big deal. i just like challenges. i just talked to the chromebook people on discord and they said they just haven't ported it to freebsd yet 21:05:01 least my main pc has freebsd 21:10:25 cool 21:14:19 If it is a new hardware problem lacking driver support then that's going to be annoying. Best to go with what works in that case. 21:15:24 I am extremely happy with FreeBSD on servers. It's the best! I am also running FreeBSD on my main desktop. It's my daily driver. On laptops I hit rough spots like graphics drivers suspend resume issues that still need polish and maturity. 21:21:12 Is there good documentation on how to format a usb to fat32 ? i know i need to destroy i first or something 21:23:12 you want to lose any existing data / partitioning? 21:23:17 found it 21:23:28 https://forums.freebsd.org/threads/how-to-repartiton-format-a-usb-flash-drive.11260/#post-67523 21:23:30 Title: How-To repartiton (format) a USB flash drive. | The FreeBSD Forums 21:24:04 that's really out of date 21:24:30 if you want it to have MBR partitioning and one fat32 partition, then: 21:25:58 gpart destroy -F daN && gpart create -s MBR daN && gpart add -t fat32 daN && newfs_msdos -F32 /dev/daN 21:27:39 for gpt, use -s GPT and -t ms-basic-data 21:39:18 Thanks RhodiumToad now i know 21:40:07 I have a question. somebody said they don't use X like how do you use your laptop/pc then ? im assuming they only use it for a server right ? how do you right documents and stuff. if i could stay in the terminal i would but i have to get stuff done 21:40:30 s/right/write/ 21:41:05 there are perfectly good editors for text mode 21:41:55 also, maybe they were using wayland instead of X? (I don't, so I haven't checked how usable it is these days) 21:43:15 you can do almost everything in the terminal/tty session if you wanna learn. biggest exception is anything involving video. 21:44:06 then again, even video editing can be done w/ ffmpeg. but you'd have to start X to watch and confirm the video came out as you intended. 21:45:16 neal stephenson famously uses emacs to write books 21:45:16 pretty sure mpv can just run from a tty 21:47:09 but RhodiumToad is proably right. "somebody" is probably using wayland if the device they are referring to is their daily driver. 21:49:56 nimaje: sure it'll run and you can even get audio in the tty, but how is it going to play a video stream? 21:51:05 ZedHedTed, Framebuffer? 21:52:18 what the...i just tried it (on MX Linux from a tty) and it worked 21:55:48 parv: thanks. biggest TIL in awhile! 21:56:09 I think the biggest problem would be a webbroweser, but I think netsurf has a framebuffer port 21:57:29 ZedHedTed , Thanks for trying 21:59:08 just tested on freebsd, mpv works fine from a tty 21:59:10 nimaje, w3m-img does graphics in the terminal (but not Javascript AFAIK) 22:00:02 nimaje, Will try netsurf; thanks 22:00:21 hm, but we don't have nsfb in ports, only netsurf-gtk :( 22:00:37 You can get video in the terminal ? 22:29:06 I use XTerm text terminals for everything except Firefox (24x7), mpv video player (often), LibreOffice (rarely). I don't miss X for anything other than web browsers due to the way the Interwebs require Javascript and graphics for everything. 22:30:40 I am always logged in with ssh to other systems and so of course those are pure text console interfaces. I always have a tmux running remotely to be able to resume that session after network disconnects. I use text terminal console editors when editing files and don't miss clicking with the mouse. 22:31:31 But I always have my fingers on the keyboard and hate reaching for the mouse. If you are the opposite and always have your hand on the mouse and hate reaching for the keyboard then this type of work flow of text with fingers on the keyboard you will hate. Opposite mental models. 22:42:05 webapps are tricky. good luck checking your bank account or payroll site w/ a TUI browser. 22:51:40 death to webapps 23:03:55 Banking is one example where I must have X and Firefox with Javascript available. That's the world we live in now. 23:04:56 you could have a small machine somewhere purely for banking 23:05:41 and you can complain 23:06:06 you'd be of service to a lot of people with dissabilities 23:08:45 I would not want to be in the middle of other people's banking transactions. Banks would not allow this because they require the client to be a current web browser with javascript. 23:08:45 rwp: znc with irssi or what I'm guessing ? 23:10:06 rwp: I mean by complaining to your bank to implement non-webapp banking options, not by having a banking machine :P 23:10:27 jb1277976, Yes. irssi is my main IRC client on my server. irssi has excellent irc bouncer proxy capability to mirror IRC to *multiple* connections. That's a feature I don't find anywhere else but have not learned weechat so don't know about it. Then connect a znc to irssi so that I can get a buffer scrollback on my laptop when I connect mobile from places. 23:10:46 that's just a way to keep your workhorse clean of x/firefox 23:11:49 jb1277976, Then I mostly use my Emacs erc irc client to connect to znc which is connected to irssi and here we are with me with my full Emacs capability which makes me happy. So on my end it is a long chain of goodness. 23:12:19 Do you irc from your phone? 23:12:24 entikan, My bank has non-webapp options, by me going into one of the branch offices. So I use the web with it. 23:12:41 a physical office? your bank has those? :-) 23:13:23 (the moderate-size town where I live used to have branches of two different banks, but they both closed) 23:14:07 And being a FreeBSD desktop user let me hint this important note, chase.com is one that blocks UserAgent strings that are not Linus,Windows,Apple. Blocks FreeBSD. Therefore I must use a UserAgent switcher with Firefox to spoof one of those so I can connect to chase.com with a browser. I use User-Agent Switcher plugin and it works excellently. 23:14:26 yeesh. 23:15:11 I know of one site (not a bank) that blocks user-agent strings containing "amd64" (rather than "x86-64" or whatever linux calls it) 23:15:15 rwp: do you only IRC from the desktop/laptop ? 23:15:31 jb1277976, You mean like I am doing now? 23:16:05 Yes. I'm at McDonald's with I've of my kids ircing from Mobile 😁 quassel ftw 23:16:14 One* 23:16:27 Oh, I missed the question about IRC from my phone above. I never use my phone for IRC. I have used my tablet on very rare occasions for special reasons. 23:16:59 I'm always in the loop 23:17:30 I need to test drive quassel because znc has a well known connect rate limit that can create a delay of *several minutes* between connecting and being in sync ready for use. 23:18:16 never used irc on my phone either. just doesn't seem right. matrix otoh, i use the shit out of that on mobile. 23:18:49 The problem is that znc tries to get a /names sync across all channels upon connection (I think that's the problem) and it would get kicked for ircd server abuse if it fired that all off all at once for many channels. So it rate limits the interaction to be nice to the server. Which causes this long connect delay. 23:19:03 rwp: you on Android or iOS ? Quasssel has to setup on a server then you have to connect with a quassel client they are limited they have a desktop client and iOS and Android client the iOS one sucks. I know cause I have an iPhone 12 but Pixel 7 for my daily driver 23:19:04 IF you are connected to many large channels. If you are in only one channel or three then the delay is short and not a problem. But I am in many channels. 23:19:54 jb1277976: are you using text-to-speech too? 23:20:01 Nope 23:20:15 I use Android. But the phone and tablet are not good interfaces for me for IRC which is mostly text. Remember I like keeping my fingers on the keyboard and hate reaching for the mouse. 23:20:29 Aw 23:20:48 oh. usually when someone uses run-on sentences it indicates TTS usage. 23:21:07 maybe that's just how my aunt uses it though. 23:21:19 Brb gotta drive home 23:22:01 I need to run to something IRL. (Hey my sentences in IRC are almost always long run-on sentences and I am lovingly typing each character in one at a time here, know they are run-on, but it's conversation "in the wind" and think that is the best way to do it, but does a much more prepared job on email.) BBIAB