it's tagged at 201 (the vid)

ok. got it.

i saw what i did wrong there.

sweet! it's doing exactly what i need it to do. now for the real test lol

great

zi doesn't want more connections

whoa

is dho!

tis!

how've you been

good man, how bout you?

been many moons

it has -- sorry, running kiddo bedtime, back in a bit!

good morning! when I running this script both as root and regular user dpaste.org/7u7vo it´s working and detect external-monitor, but when I running this script (via devd) it doesn't work the it always suspending, what Im missing? my lid.conf (devd) looks like this dpaste.org/oZFw4

or is it like this that I cant run devd and xrandr?

It’s working but rclone is super slow. I guess the trials and tribulations of using fuse?

Maybe I can tweak rclone a bit to speed it up.

rclone can "patch" large files instead of transfering them again. this may or may not be a good idea, depending on the data, the fs, the transport. make sure, you are using the mode best for your use-case

just like freebsd-update?

hmm

btw

how to get missing feature out of tcsh and into sh

it's not in bash either

i have no idea which other shells have this partial history search

at quick glance as if nobody ever used heard wanted what i used form 4.6

from

:p

exec echo\ndate\nw\n, then type e and up, in csh you get echo, in sh and bash you get w

i can't find what it even is called

when i used linux, i don't think i even knew what shell i used, let alone change it, so i didn't change it in fbsd either :p

`!e\n` would work at least with bash afaik

ridcully: that executes it directly?

yes it does

but what just gets it on prompt?

it's like, oh i had this oneliner, i think it all started with fgrep so you hit up on f

likely find it so you can reuse parts of it

! would just russian roulette run something from past

zsh can tab-complete ! - everything beyond that i use fzf for. the feature you want is named `history-beginning-search-backward` in zsh. maybe this gives you a hint for the other shells or something to search with

hmm

yeah it can do a little bit like that

cursor is wrong tho

for some funny reasons bash, at least here, never has it on

sh also has no {} while bash bash that

has

bash bash wtf

some people never know {} either :p

one was like wait what did type {,.bak}

you

or ^d on prompt :p

oh look

how many additional bugs it created

the issue

s/bugs/bug reports/, someone just can't use search or better yet post to current/stable mailing list

ketas: I think bash enables incremental history search by default with CTRL-R?

I'm trying to run WINE in a jail, and when I try to run /usr/local/share/wine/pkg32.sh install... I get: "/usr/lib32/libc.so.7" not found; exiting. I'm running bastille and using a thin jail. Is there something I need to run to fill in the lib32 dependencies on the jail?

cyric: yeah i did botg

h

ahh I think I found it.. bastille_bootstrap_archives=

Koston: i tried but i can't use it somehow, i have no idea how it works, i type chars and something just inline appears

pertho, find /usr -type f -name 'libc.so.*'

CrtxReavr: yeah I just had to add 'lib32' to the bastille_bootstrap_archives list and re-bootstrap and it was happy

BarnabasDK: you know about sesutil?

as linux_base-c7 is deprecated, what do we use as linux base now? I didn't find another linux base in ports

nimaje, not really

well, sesutil seems to be what you asked about some time ago, it lets you control the state leds of disk

yeah, if it is scsi

hm, I should test on my desktop what the sesutil commands do, as it somewhy has ses, but only SATA connectors and I put normal SATA disks in there, I don't think they have leds anywhere; I already tested sesutil show, it can show me which disk is connected to which slot

ketas: and it keeps piling up

cyric: where?

yeah people build their worlds i guess

:p

yeah 5 total reports

funny, it's hard to test thing too

kevans: where those changes should be put to for feedback i wonder, i've tried to put them into some ml before, later filed pr, there none were rejected, were useful ideas, but it's not like i can do anything else

ketas: throw them in phabricator and we can tag a few people that might be willing to review

i guess i could make account there

some them i couldn't figure out fully either

as if more than one brain is needed :p

i watched bunch of phabs, seemed like changes came after reviews, etc

was curious of process

but yeah i'll see

while looking around i found wontfix from old times that proposed 24h clock into many places

but ps did get 24h now eh

maybe that was the right place

and uptime everyone just configures locally

i find it fun how i created special script to search for 24h locales so i can use them solely for that purpose

because apparently it's easier for coders to do it that way

even made my own special locale because nothing else did fit

at least it would be easier to get diffs there :p

directly from wd

hmm

Hi, I never worked with different routing tables ( fibs ) so let me ask some questions

1 have configured 3 nets igb1, igb2 - two ISPs ... igb0 - lan

igb1 is in fib 0, while igb2 in fib 2

lets imagine situation when igb1 is down - isp1 - down

the routing still holds but the local dns resolver - bind which sitting in FIB0 cannot do queries

so If i do `setfib 1 service named restart` the bind starts working, locally.... but I cannot query DNS from igb0 ( lan ), all requests is timed out ...

despite the bind still opens sockets on a local interface igb0

are you the same one

or maybe not

damn routings suck

a shorter version of a question

it's a problem i can't imagine how to get right

why I can query bind running in fib0 and cannot while it runs in fib1

it sends reply into wrong place

would be it

maybe

now how to make bind route both right

no idea

use route get 1.3.4.5 to unfuck this

somehow

maybe pf needs to be involved

ketas, are you some kind of AI ?

lol

no

consider using pf instead of interfaces

pass in on em0 to !<localnets> rtable 1

etc

via interfaces only the incoming packets are tagged on the interface fib

so I need to somehow route local - this machine traffic ...

and if you want some bind responding on two fib , must be two instances of bind ...

ok, need to dive deeper into IPFW ...

Teraii, I cannot connect bind running in different fib !

that is the issue at the moment

and you can't ever

but why ?

only one fib at a time

for one process

and natting between them ?

(with setfib X process)

natting if not routing

natting is not routing

something like ipfw add xxx setfib 1 ip all from any to any rcvd in . ,, ??

sorry

rdr it?

i do not use ipfw for long time ago :)

nat has other uses too i found

btw

but i think it's the same way as pf

I stick to it .. it faster... and my brain is not big enough for changin things to PF ..

i had to connect into device which has no routing in it

mine too :)

no defgw

so you can't access it from far away

so i had to nat it

yeah ,... i'm getting old

fun

nat made all packets come as if from same ip and it worked :p

what would happen if you rdr the bind traffic?

tho outgoing is still sucky

oh i don't like to think of it

feels like my brain wants out

yea it's not easy

meanwhile, how to make openvpn listen on all interfaces and both families?

kind of same

ensure you have default route on each fib

me?

it's just 1 fib here

yes

i gave up on v6 and just ran ovpn on v4 only

fib0 or fib1 ? :)

fib0 is the default

managed to make it work

i have other fibs tho

Teraii, thanks ...

but i run ovpn in 0

;)

the machines from the local network has no issues reaching outside world with both providers

i have wan #1 that's inet & wan #2 that's my iptv

I'm having issues with bind only ..

should have wan #3 that's lte so there it's real fun

well you can't run bind with two tables

so put bind into fib0

could maybe even nat it?

that's a stupid idea maybe

in any way packets need to route right

so

bind works where?

if bind only works towards network it has default route for

then you know :p

should probably make some virtual test to try here

but why rdr won't work here?

like imagine if your dns is separate machine igb3

how would you get traffic there :)

s/igb3/on igb3/

fun, i didn't know, igb was merged into em but it represents itself as igb

i looked what i have created in my pf

it has extensive rdr's

and 2 nats which aren unconventional

eg they aren't on wan

ketas: please don't scare people seeking help here

what

mzar, i fail to see how ketas is scaring folk, although i have him dimmed in my irc client so it's harder for me to read what he says

why the dims

:p

funnily, certain types of translations have like certain use cases

but they have others too

Reinhilde: no worries, it was a joke, I tried to cheer him up only; but why you dimmed him... it's scary

hah

:=D

it's so dark now

He scares me, especially when he's in the dark

where did daddy touch you

Right in the PS/2 port.

Is there a way, with zdb or something, to find the recordsize of a specific file on ZFS?

zfs get recordsize path/to/dataset

*file*, not dataset

I have a dataset that was created with recordsize=1M, and some very large files. I have since changed the recordsize to 128K (it had side effects I did not consider), and now need to rewrite those files that still have recordsize 1M.

vkarlsen, show me on this doll where he hurt you.

hmm

Ltning: I think `stat -f %k` would be sufficient if you're fine with the fact that files < recordsize will have a smaller blksize

basically, you can't assume == 1M, you have to assume that >128K means they were written with 1M recorsize

Yeah, that works. Brilliant, thanks!

Too bad zfs rewrite does not handle record size changes :(

I feel that bears mentioning in the man page

hmm, yeah, that'd be a good note

So, bummer!

That would be even better, obviously.

CrtxReavr: It was in the ps/2 port, all the way

And he used a DIN plug

lmfao

Ltning: word on the street is that making zfs-rewrite do it would be a major rework, unfortunately, so I suppose documenting it will have to suffice

kevans: thanks for checking. It's strange, but .. i won't pretend to have the brain power to understand why. ;)

the short version is that the bits behind it don't actually rewrite the layout, just dirties up blocks