08:45:07 <polarian> wavefunction: nope because I only have a 802.11n card
08:45:19 <polarian> for a reason, I could have gotten an ac card but why would I when it wasn't supported at the time :P
08:45:25 <polarian> and also I do not plan to use iwlwifi
08:45:41 <polarian> 802.11n is fast enough for basic things, and if I wanted better internet I would plug in an ethernet cable
08:46:05 <[tj]> you know with ac you can do like a gigabit over wifi?
08:46:17 <polarian> I dont even have gigabit internet haha
08:46:32 <polarian> the only benefit would be if I was downloading/uploading a lot to my local servers
08:46:33 <[tj]> I do (normally)
08:46:44 <polarian> well then, have you tried an ethernet cable?
08:46:46 <polarian> much more reliable
08:46:48 <[tj]> yes
08:47:08 <[tj]> its a moot point at home, my office is the far end of a mesh bridge
08:47:08 <polarian> I have also concluded that wifi repeating/relays and meshes are ewaste
08:47:20 <[tj]> nope, mine is great. Spend more money
08:47:25 <polarian> only way to have decent wifi is to have an AP directly attached via ethernet proferably in every room :)
08:47:39 <[tj]> even the the fritzbox one I'm using now is pretty good
08:47:51 <polarian> [tj]: not when you live in London and there is about 20 different ssids, all with their own meshes (multiple APs) in your area
08:48:03 <polarian> and also UK houses tend to have thickkkk brick
08:48:20 <polarian> I swear that the kitchen is lead lined, no wifi singles could get in and out
08:48:30 <[tj]> probably just a lathe and plaster wall
08:48:46 <polarian> besides its fun drilling holes in the walls...
08:48:46 <[tj]> I can still do 500mbit on wifi in my home, in the uk
08:48:51 <polarian> the landlord doesn't agree
08:48:56 <polarian> but they dont know or ask :)
08:49:21 <polarian> [tj]: lucky you, I have vDSL so yeah
08:49:45 <[tj]> gotta pick where you live
08:49:46 <polarian> I pick that over the faster gigabit capable virgin media service because virgin media goes down as often as shit breaks on freebsd
08:49:55 <[tj]> and not in the big red bit on the internet coverage map
08:50:05 <[tj]> you are in london just use gigabit 5G
08:50:21 <polarian> I want stable, low latency internet
08:50:30 <polarian> vDSL, despite all the hate it gets, provides that
08:50:32 <[tj]> 5G is for controlling robots
08:50:44 <polarian> 5G is less reliable than virgin media
08:51:12 <polarian> and the only decent 5G provider is EE
08:51:33 <polarian> also more speed is not always required, my line sits idle most of the day
08:51:37 <[tj]> move house :D
08:51:41 <polarian> orrr
08:51:45 <polarian> maybe I am fine with what I have>?
08:52:26 <polarian> Openreach replaced the cable about 8 months ago because the old one from the pre-openreach days was so old it was powder...
08:52:56 <polarian> Openreach came out the next day too...
08:53:07 <polarian> When Kelly turns up, you're fucked
08:53:12 <polarian> when openreach turns up, you are in safe hands
08:53:35 <polarian> also Openreach will have fibre to my area by 2026
08:53:41 <polarian> by the end of 2026 :)
08:54:02 <polarian> it may be asymmetric and more expensive than the competition, but it is something :P
08:54:20 <polarian> although I dont know if I want to get rid of my vDSL now, it has grown on me a lot :c
08:54:35 <polarian> I might actually miss it when I get fibre coverage
08:56:02 <BinGOs> OR are pushing getting rid of their copper infrastructure, they want to be out of the telephony business
09:25:01 <polarian> BinGOs: well they want to digitalise it
09:25:44 <polarian> but the whole system has been made obsolete by voip pretty much
09:25:56 <polarian> and integrating calls in with other platforms
09:26:06 <polarian> people rarely whip out their phone and dial a number anymore
09:45:10 <polarian> Ugh...
09:46:17 <polarian> wpa_supplicant: ioctl[SOICS80211, op=20, val=3, arg_len=7]: Can't assign requested address
09:48:06 <polarian> weird left it about 2-3 mins and wifi works now...
09:48:20 <polarian> it keeps erroring out on boot though :/
09:48:30 <polarian> drm borked let me fix that first
09:49:39 <jgh> and turn the tx power down, so they don't interfere with each other.  And go for the highest possible RF band, so the walls really do block.   Remind me... why not just used wired?
09:50:20 <polarian> mason: there is no FreeBSD-kmods in /etc/pkg/FreeBSD.conf
09:50:57 <polarian> jgh: I want to but in lazy... when I get a chance I will setup lagg to use wifi as a failover iface
09:51:11 <polarian> then I will run ethernet from the switch to my desk
09:53:58 <paulf> ll -rt
09:55:07 <paulf> wrong window, as always
10:05:06 <beastie> ls
10:05:10 <beastie> oops
10:05:12 <beastie> sorry
10:06:03 <polarian> xD
10:06:12 <polarian> how come everyone has done that...
10:06:40 <polarian> I guess its what happens when a sysadmin uses too much IRC :P
10:48:56 <polarian> hmmm
10:49:13 <polarian> FreeBSD 14.3-RELEASE doesnt seem to work
10:49:21 <polarian> drm-kmod is broken
10:49:38 <polarian> iwn/wpa_supplicant is buggy
10:49:55 <polarian> ugh
10:51:26 <polarian> holddd up I forgot to finish the install, its only partial
10:51:40 <polarian> (this is why you dont sleep mid-update folks!)
11:13:36 <duncan> isn't it just that the packages haven't been built for 14.3 yet?
11:14:44 <ivy> duncan: they have been, they're in the new FreeBSD-kmods repo which should be added to /etc/pkg/FreeBSD.conf automatically when you do the /etc merge
11:59:01 <polarian> duncan: I compiled from src anyways
11:59:07 <polarian> ivy: helps if I finished the install lol
11:59:20 <polarian> although its been installing updates for 1 hr now... I know I use a HDD but geez
11:59:29 <polarian> how big is this version
12:18:22 <polarian> ok I see the repo has been merged in which is cool
12:18:36 <polarian> wifi still fucked though
12:19:24 <polarian> right where are the freebsd wifi guys again?
12:19:31 <polarian> they have an IRC channel somewhere...
13:00:33 <ketas> polarian: you are in it still, in efnet/#freebsd-wifi
13:00:36 <ketas> :p
13:00:51 <polarian> ketas: ahhh efnet
13:00:52 <polarian> ok thx
13:01:10 <polarian> I cant keep track of all the networks my bouncer is in
13:01:15 <ketas> it's esp. fun since you are still there
13:01:17 <ketas> hah
13:18:10 <mns> What is the FBSD replacement for sar?
13:19:30 <ivy> mns: what is sar?
13:20:14 <ivy> based on the Linux manual page, you may be looking for vmstat, iostat or systat
13:20:49 <polarian> I like how ketas has to tell me where my bouncer it xD
13:20:51 <mns> sar == system activity reporter,  used to be available on most Unix systems
13:20:57 <polarian> you would have thought I would keep track of this stuff
13:23:14 <polarian> mns: as ivy pointed out there is numerous tools
13:23:20 <polarian> is there something specific you are trying to monitor
13:24:08 <polarian> top is best for cpu, iostat I would say is the best for io activity (breaks down each dev) and you can use systat with the -ifstat flag to monitor network
13:24:25 <polarian> cpu/memory/swap/ARC (if you are using zfs) for top sorry
13:25:33 <polarian> ooo sar is a lot more advanced than that...
13:26:02 <mns> polarian: not something specific yet.  Just trying to setup a reporting system.  I'm noticing that recently my SSH connectivity is slow.  Takes about 24-27 seconds from outside my network to get a login prompt when it used to be about 2-3 seconds.
13:26:23 <polarian> according to a quick google, there is sysutils/bsdsar
13:26:50 <polarian> oh wait... that is gone :P
13:26:53 <polarian> ignore that comment
13:27:55 <polarian> https://man.freebsd.org/cgi/man.cgi?query=sar&apropos=0&sektion=0&manpath=SunOS+5.8&format=html there is a man page entry for sar...
13:28:06 <mns> that's for SunOS 5.8
13:29:08 <polarian> ahhh I am blind haha
13:29:12 <polarian> thanks for pointing that one out :P
13:29:23 <ketas> i bet you could find bsdsar src but
13:29:35 <polarian> did it go EOL or something?
13:29:43 <mns> I was thinking there might be something in base.  I'll check out the single purpose tools and see if I can cobble something together
13:30:12 <ketas> depends what one needs
13:30:29 <ketas> there's top, gstat, systat
13:31:09 <polarian> geez bsdsar went EOL in 2013
13:31:58 <ketas> systat -v 1, systat -if -s mbit 1 -m 'em*,vlan[3-79]*,tun*'
13:32:00 <ketas> eg
13:32:04 <ketas> but
13:32:10 <polarian> man systat :P
13:32:40 <mns> I have htop already.  systat might be close to what I want.  Seems like FBSD has never had sar as part of base.  The 1.0-RELEASE man pages don't have sar in them.
13:32:47 <ketas> top -qaHSPs1
13:32:52 <ketas> hmm ok
13:33:26 <polarian> whats with all the flags lol
13:33:28 <mns> My unix history needs updating lol
13:33:35 <polarian> is regular top without flags not good enough for you ketas?
13:33:52 <ketas> zpool iostat -v 1, sockstat -46
13:33:53 <ketas> no
13:34:05 <ketas> i mean ton of stat utils
13:34:10 <ketas> around
13:34:43 <polarian> atop as well
13:34:59 <polarian> although its GPL
13:35:46 <mns> there is btop as well
13:36:00 <ketas> netstat -n -b -f link -i
13:36:11 <ketas> so you wanted tops?
13:36:22 <mns> btop does some fancy screen ui
13:36:23 <ketas> :p
13:37:03 <ketas> for network traffic counts there's vnstat
13:37:43 <polarian> cant systat do like... everything though (according to the man page)
13:37:49 <mns> ketas: me?  nope.  I have htop already, always running.  I wanted something that would report on system activity like sar does and then I can put it in a cronjob and get regular activity reports.
13:38:17 <mns> polarian: systat, based on the man page, seems to be the closest to what I want at the moment
13:38:24 <ketas> but what did sar even report?
13:38:37 <ketas> esp. regularly
13:38:53 <polarian> ketas: sar was less reporting but monitoring/archiving of system statistics according to wikipedia :P
13:39:40 <ketas> there's plenty of that too i think
13:39:52 <ketas> graphs maybe? :)
13:39:53 <polarian> but most of the time I see people do that with promethius and grafana
13:40:05 <polarian> fancy web interface with fancy graphs
13:40:47 <mns> and I dont' need the fancy stuff
13:41:03 <polarian> it would help to know what you want to do :)
13:41:25 <ketas> i'm unsure if those are too big hammers
13:41:40 <ketas> munin exists too
13:41:41 <ketas> and
13:41:52 <ketas> lessfancy
13:42:08 <ivy> systat really should be improved to work on the larger, colour terminals we have today...
13:42:32 <polarian> you dont like old school? D:
13:42:38 <rtprio> ivy: yes, make it look like btop
13:42:55 <ketas> who's going to colorize it
13:43:33 <ketas> who wrote gstat?
13:43:47 <polarian> who wants to colourise it... it looks nice as it is :c
13:43:51 <ketas> "appeared in FreeBSD 5"
13:45:52 <mns> I just want to monitor the system I have, a general monitoring of activity. Was looking at basic tools, and remembered sar from way back when in the early 90s. Don't really want to setup things like nagios or such as I don't need stuff like that.
13:59:43 <Demosthenex> munin is pretty fast to setup
13:59:55 <Demosthenex> sar is ok, but you need a sar viewier like ksar
14:09:18 <polarian> hmm weird, using freebsd-update with -b flag with the path to the root of a jail to update to 14.3-RELEASE won't let me, as it is detecting the host is on 14.3-RELEASE
14:09:30 <polarian> this used to work, or well I thought it did...
14:09:37 <polarian> -j requires the jail to be running
14:09:45 <polarian> oh well I will just do it with -j
14:22:29 <mason> polarian: Ah, was using freebsd-update here so I ended up with the updated /etc/pkg/FreeBSD.conf automatically here. Didn't know to explicitly refresh it, but when I did things worked.
14:27:47 <polarian> mason: I didnt need to explicitly refresh it, I did the kernel update last night and forgot to the userspace update when I got up
14:27:56 <polarian> so hence I only had half of 14.3-RELEASE installed
14:36:24 <mason> That sounds like an exciting state, certainly!
17:24:50 <shbrngdo> ivy, polarian: colors in POSIX tools need a UNIVERSAL OFF SWITCH.  Half the time I install a Linux or create a login [let's say shared hosting even] the 'ls' is by default aliased to show colorization (and is nearly unusable), and nano uses colors unsuited for ANY background.... (must add -Ynone to make it go away, not always automatic like when 'EDITOR is implied and sudo is used) and it's **NUTS**.  "Old School" is
17:24:51 <shbrngdo> BETTER, especially for old eyes.
17:25:06 <ivy> shbrngdo: systat is not a POSIX tool
17:25:33 <shbrngdo> yeah ok.  prob'ly who no insane colrizing
17:26:00 <shbrngdo> 'why no insane colorizing'
17:26:09 <ivy> i don't really know why just mentioning using colours causes everyone to react like i killed their cat, i'm not even planning to do this, i just suggested someone could
17:26:56 <rtprio> systat | lolcat
17:27:20 <shbrngdo> true.  Well let's just say that add 20 years to your eye age and you realize that colors are too faded to read things and no background color makes it easier.
17:28:53 <rtprio> make the fonts bigger, old man
17:29:56 <shbrngdo> already do.  It's not the size, ity'sthe contrast.  color:color contrast does not work over 50.
17:31:19 <wavefunction> I make it a point to have my zsh / whatever aliased appropriately and, when working on remote machines, do my best to not deal with color craziness
17:32:29 <ivy> shbrngdo: since you mentioned GNU ls, i do agree that the default colours for ls in most Linux systems are extremely bad
17:32:29 <shbrngdo> ack - same here with new logins and new setups.  'unalias ls' is one of the first edits to ,bashrc etc.
17:32:36 <ivy> that doesn't mean colours are bad, it just means GNU coreutils is bad
17:32:48 <rtprio> we knew coreutils is bad
17:33:59 <shbrngdo> it's hard to find a color set that is as readable sd pore black on white or white on bacl [though I dim the blue on white colors to avoid eye strain]
17:34:27 <shbrngdo> pure not pore - stupid keyuboard
17:35:52 <ivy> colours are good when used to enhance readability, e.g. vim's default colour schemes are generally very good
17:36:17 <ivy> they are not garish and you hardly even notice them, they just differentiate elements which would other look similar
17:36:58 <shbrngdo> I guess the summary is that SOME people's readability is enhanced, but for many others it's completely backwards.
17:37:17 <shbrngdo> 'enhanced' = 'unreadable'
17:38:55 <wavefunction> https://uutils.github.io/ ?
17:42:23 <oxbar> nxjoseph: you ever get battery working ? like a status on the dwm bar ? i got dwmblocks working but thats it
17:51:47 <nxjoseph> oxbar, i don't have a laptop so i don't know, never experienced
18:16:15 <oxbar> Dam.. do you have anything that says charged or charging in the bar or plugged in ?
18:16:28 <oxbar> Lots of ors lol
18:16:48 <rtprio> can you not just look on the side of the laptop and see if it's plugged in or not
18:17:03 <nxjoseph> oxbar, no, i don't have anything like that, never needed them cuz it's desktop
18:17:24 <oxbar> Thanks
18:17:26 <nxjoseph> rtprio, that may be make sense but battery level would be important
18:17:51 <nxjoseph> oxbar, you are welcome
18:40:40 <rtprio> interesting, a long captive smart test doesn't pick up errors that ZFS does
18:40:44 <rtprio> i was not expecting that
18:41:37 <ivy> rtprio: it probably would do if you had checksums enables on the storage device, but that's never enabled by default and i'm not sure freebsd even supports it
18:42:13 <rtprio> checksums when i created the pool?
18:42:36 <ivy> no, checksums when you formatted the storage device, you need a SCSI command to enable it which switches the device to use 520-byte sectors
18:42:45 <ivy> (it's useless with ZFS since ZFS does its own checksums)
18:42:58 <rtprio> oh yeah, i sure did not do that
18:43:41 <ivy> i just like to point this out since some people believe only zfs can detect on-device data corruption :-)
18:44:02 <rtprio> so do i toss it
18:44:09 <rtprio> yolo it
18:44:23 <rtprio> i don't think it's in warranty
18:44:28 <ivy> i would not throw out a disk for a couple of checksum errors, those can occur for lots of reasons... cosmic rays, loose cable?
18:44:39 <rtprio> usb chassis :(
18:44:42 <ivy> if you clear the errors and it happens again i would replace it, though
18:45:18 <ivy> oh, if USB add controller firmware bugs :-)
18:45:38 <rtprio> it happened the last time i had it plugged in too; managed to copy things off of it so now it's vacant
18:47:54 <ivy> i wish we supported UASP
18:48:11 <ivy> that would make modern USB (10Gbps or faster) an awesome way to connect storage devices
18:50:19 <ivy> on the subject of storage, does anyone know what the smartctl ports expects daily_status_smart_devices to be set to in /etc/periodic.conf?
18:50:28 <nedko> IIRC UASP mattered at 480Mbps too, on linux it helped i mean
18:50:29 <ivy> e.g. daily_status_smart_devices="/dev/ada0 /dev/ada1" or daily_status_smart_devices="ada0 ada1" or something else?
22:00:52 <demido> if my daemon binds to 2000 on localhost (127.0.0.1) can it also separately bind to 2000 on the public ip of the box?
22:05:57 <wavefunction> There's nothing to stop that, no
22:10:51 <demido> you mean yes?
22:11:04 <wavefunction> demido: Sorry, yes. It absolutely can
22:11:43 <demido> ok ty that's really powerful
22:12:02 <demido> every address, even local private ip, can have 65k ports
22:15:38 <rwp> demido, Yes.  You can either bind to the wildcard address * or you can bind to a specific address such as 127.0.0.1 or to a global scope IP address either IPv4 or IPv6.  Binding to the loopback address is often done in order to guarantee that only local processes can access it as a security measure.  And I recommend a firewall to prevent accidentally exposing something under development to the hostile Internet.
22:16:34 <demido> if my server binds to a localhost port, no firewall is needed to protect that since no external ip can communicate with localhost ip, right?
22:17:16 <rwp> Right.  No remote network can connect to the loopback network.  Not directly anyway.  Tunnels and VPNs connect around it explicitly.
22:17:49 <demido> tyvm guys
22:18:14 <rwp> People with non-vnet jails often need to bind daemons to specific addresses when by default they would use * they must pick an alias address to keep separate from other services in other jails and on the host.
22:19:09 <rwp> Generally though it is a PITA to have to map everything explicitly like that.  It's one of the reasons for vnet jails which have a full network stack and can have daemons binding to * as they would normally but only effectively have a single IP that is assigned to that vnet device.
22:39:34 <demido> ya i just wish vnet jails had the mac port acl support
22:39:39 <demido> that'll be killer feature
22:54:48 <ketas> demido: like per jail?
22:56:42 <rwp> I don't know what mac port acl support means.  You can have a separate independent set of PF rules for each VNET device.  If that isn't powerful enough then I don't know what would be.
22:56:58 <demido> ketas ya
22:57:30 <demido> i needed to let each vnet jail bind to a <1025 port and the easy mac portacl method didn't work iirc
22:58:03 <demido> was a real drag. one of the only holes i found in vnet jails
22:59:26 <ketas> as a workaround one can use jail allocated special uid/gid
22:59:33 <ketas> it's still same kernel anywaya
23:02:08 <ketas> but that like doesn't have any limits iirc if jail root is untrusted?
23:02:16 <ketas> if that's issue
23:02:26 <ketas> hell knows
23:02:44 <ketas> who's going to implement it properly :/
23:14:07 <demido> i ended up using devd or some other hack. should just have mac portacl support added to vnet jails to make it nice for everyone
23:20:51 <wavefunction> "mac portacl" where you restrict what mac addresses can access what ports?
23:21:43 <wavefunction> https://wiki.freebsd.org/ThomasHurst/Caddy demido
23:21:57 <demido> i just needed to easily let a vnet jail bind to port under 1025
23:23:12 <wavefunction> I'm advocating from a point if ignorance here, but it looks like there are settings that allow exactly that, per the caddy instructions?
23:24:10 <rwp> I was thinking the exact same thing here.
23:26:38 <rtprio> hrm, i should move my caddy to a jail
23:27:34 <demido> ya it's been a while i gotta dig back in sometime
23:27:48 <ketas> wavefunction: no it's the other mac
23:27:58 <ketas> sadly
23:28:06 <ketas> so many macs
23:29:01 <ketas> Mandatory Access Control
23:29:10 <ketas> as opposed to voluntary?
23:29:23 <ketas> it doesn't mean anything
23:29:33 <ketas> maybe if you live in enterprise or govt
23:29:45 <ketas> didn't it came from there eh...
23:33:08 <wavefunction> ketas: :(
23:33:16 <wavefunction> Acronyms are awful.
23:33:24 <rtprio> AAA, great one
23:34:03 <ketas> and it's noy triple alcoholics club...
23:34:08 <ketas> not
23:35:30 <ketas> while PCMCIA is always People Can't Memorize Computer Industry Acronyms
23:35:34 <ketas> :p
23:36:10 <wavefunction> Thing is, I hate acronyms *unless* you're doing a shorthand with someone who you've collaborated with long enough to know / share the references.
23:36:43 <ketas> since security mac is often used around networking mac, it's hell
23:36:46 <wavefunction> Every one of those acronyms is an abstraction, and the problem is computer folk keep poking leaks in those abstractions
23:37:10 <ketas> it's a shortcut
23:39:30 <ketas> The longest acronym is NIIOMTPLABOPARMBETZHELBETRABSBOMONIMONKONOTDTEKHSTROMONT with 56 letters (54 in Cyrillic
23:39:34 <ketas> :)
23:39:52 <ketas> at this point, it's secure password instead
23:40:10 <dkeav> isn't that a volcano in iceland
23:40:35 <ketas> that's eyajallayökull or something
23:41:00 <ketas> actually: Eyjafjallajökull
23:41:11 <ketas> :p
23:41:46 <ketas> funnily it's possible to learn them but
23:45:08 <ketas> saying it properly is another thing
23:45:15 <ketas> (the volcano)
23:50:10 <ketas> full bg in man -P 'less -p "contri"' 9 mac
23:51:55 <ketas> spawar sounds like splash battle in water park
23:55:39 <ketas> for added confusion, there are ugidfw and mac_portacl manpages
23:57:18 <ketas> who uses fbsd as wifi ap btw?
23:58:02 <ketas> apparently noone, judging from support :p