steam... stuff? like... what? games?

or that sourcemaker thing

mainly just steam and i guess jetbrains ides if they dont work on freebsd

im just trying to decide weather i should only use freebsd and use a linux vm with gpu passthrough for steam or dual boot freebsd with a linux distro using zfs (if i can actually get it working for once)

since i have about 800 gigabytes of files i need to preserve and only 512 gigabytes of external storage so i cant format all the ssds at once

storage i cant format all the ssds at once*

there are ports for most of the jetbrains stuff, although i couldn't get CLion to work here

honestly steam is the only thing that has me worried about switching over to freebsd completely

I'd love to see Steam work smoothly.

i have dual booted linux and freebsd before but i had to manually add freebsd to the grub file

grub can be a real downer sometimes

do people only run steam on freebsd with some kind of linuxy compatibility layer?

i guess wine doesn't exist on freebsd for too much linuxisms

wine works fine on freebsd, i'm using it right now

(to play factorio, although not the steam version)

thegman: It's a bit of hassle to dual boot. You have to choose one or the other. And FreeBSD might become an after thought if you're playing games all the time. I'd suggest FreeBSD plus a VM or other alternatives, but keep one main driver.

is gpu passthrough solely a thing so people can game in a vm?

i think it serves other purposes

how well does bhyve gpu passthrough work with an amd cpu and an amd gpu

i heard that it doesnt work with amd gpus

No clue. I don't play video games.

tuaris, Hmm... Interesting. I do not know. I looked through various docs but could not find the "pause" state documented. So... I don't know.

lw, you up?

jbo: ish

lw, so I am looking at ztd::text

lw, I have an existing application that uses std::string and STL algorithms such as "replace", "starts_with", "ends_with" and so on. That application currently just "assumes" that all strings are single-byte strings (I don't have to explain you more, I guess).

lw, I want to improve that situation heavily by using ztd::text. however, it is unclear to me how that library would tie into algorithms. there is ztd::text::text (which seems WIP, but whatever).

lw, do you have any insights into how that would look like practically?

jbo: i really don't know anything about how it works, sorry

alrighty

wait why does freebsd have fetch instead of wget

or should i say why does linux have wget instead of fetch

thegman, wget is available in ports -> ftp/wget and www/wget2

im just wondering what happened along the line to give it a different name

because it's an entirely different tool?

s/tool/program

fair enough

thegman: fetch was written specifically to be an HTTP download tool in the base system. it is simply/has fewer dependencies than cURL, and is not GPL licensed like wget

so its basically just superior wget

i don't think it's better than wget, it simply has different design goals

why does it provide such confusing instructions: bsdio.com/uefi/building-under-freebsd ?

Title: | Bex' Site

what does it even mean "Create the following symlinks in BaseTools/Bin/FreeBSD-amd64"?

Is FreeBSD-amd64 a text file?

is it a directory?

can someone please explain to me what it means?

presumably a directory base on the verbiage

'symlinks in ...' doesn't make any sense unless that's a directory

kevans: but symliinks are created as "ln -s /usr/local/bin/gcc13 /usr/local/bin/gcc", so what does FreeBSD-amd64 have to do with it?

kevans: so I don't get it. what does it mean?

yes, they're saying you need to reate BaseTools/Bin/FreeBSD-amd64/python as a symlink to /usr/local/bin/python2.7, etc

okay

does anyone have any idea how to set up ssh tunneling with non-root users? ssh -w any fails with:

Tunnel device open failed.

Could not request tunnel forwarding.

ideally I would like to create the tun0 interface when I ssh

check the sshd config for PermitTunnel

I have it enabled,

PermitTunnel yes

hello fam

trying to upgrade from 12.1 to 14

any advice will help please.

Title: Chapter 26. Updating and Upgrading FreeBSD | FreeBSD Documentation Portal

VimDiesel thanks!

src component not installed, skipped

No updates are available to install.

im getting this error

Hi there

Does anyone uses here i3wm desktop ?

any guidelines how to install it on BSD ?

warsoul, BACKUP first :D

nerozero how i do it?

do you have zfs filesystem ?

if so - you can make a recursive snapshot

im new in BSD

installed couple of years ago have 12.1 version

any important files on the system ?

want to upgrade to latest version

no important files

then the fastest way to reinstall the latest bsd

otherwise - you have to go step-by-step upgrades

you mean like do a netinstall

from scratch

12.1 is pretty old version

so I guess there is a lot of steps to upgrade that to 13.x or to 14.0

upgrade to latest 12.x then upgrade to 14.0 should work fine

lw how can i upgrade having the errors i have?

what command are you running and what is the output?

If you have no important files - just do a fresh install

copy the configs you like to save first

like /etc/rc.conf, and entire /usr/local/etc

# freebsd-update fetch

# freebsd-update install

freebsd-update fetch install

freebsd-update upupgrade -r 13.2-RELEASE

or - freebsd-update upupgrade -r 14.0-RELEASE

read carefully the output of each command after execution, it will guide you to the next step you should take ]

updating to 14.0 without updating to latest 12.x will trigger the /usr/include udpate bug that breaks the C++ compiler

I'm still trying to avoid 14.x till 14.1 release, so :D

freebsd-update install

src component not installed, skipped

No updates are available to install.

Run '/usr/sbin/freebsd-update fetch' first.

im getting this error

because you have not downloaded this

freebsd-update fetch install <<< do that

No updates needed to update system to 12.1-RELEASE-p13.

get this after running that command

freebsd-update -r 12.4 upgrade

Inspecting system...

it stays here

it can take a while

ok

will let you know when its done

ls it started to update :)

To install the downloaded upgrades, run "/usr/sbin/freebsd-update install".

lw got this

run that command?

yes

have to reboot will be back

back

Installing updates... done.

lw done

lw whats next?

can i upgrade from 12.4 to 14?

warsoul: did you follow the complete procedure from the handbook for upgrading to a new release? usually you need to run freebsd-update several times

i just finished upgrading to 12.4

so you rebooted and ran freebsd-update install twice?

lw yes

Installing updates... done.

ok, then you can follow the same procedure to upgrade to 14.0, except you have to run install three times for a major release upgrade - be very careful about the order you do install and pkg upgrades, the handbook does explain the correct order

ok where can i see the handbook

you have the link?

i just asked if you followed the handbook instructions to update to 12.4 and you said yes

Title: Chapter 26. Updating and Upgrading FreeBSD | FreeBSD Documentation Portal

lw no, you helped me lol

that guide is talking about 13.X versions

im in 12.4

it's the same procedure for any version

so i have to upgrade from 12.4 to 13.X

?

no, upgrade to 12.4, then upgrade to 14.0, you can skip 13.x

ok

The following file could not be merged automatically: /etc/master.passwd

Press Enter to edit this file in /usr/bin/vi and resolve the conflicts

manually...

lw im getting this error now

that's not an error, it just wants you to merge changes in your local /etc/master.passwd with changes in the upstream version

lw how do i do that

when i press enter it takes me to a edit file

i'm not sure if there's any good documentation on this, but essentially the file you're editing has three types of marker: "<<<", "===" and ">>>". the text between "<<<" and "===" is what's in your current copy of the file, the text between "===" and ">>>" is what's in the new version. you need to merge those two versions together into one block, then remove the markers

ok

after doing that how i exit that config

you mean save the changes in the text editor? it's vi, so type :wq

ok

I'm building my own gcc-13.2, not from ports, on a quad core mini-pc, using gmake -j 6, and the load acoording to htop is just around 6.25, even with all four cores at 100%. FreeBSD rocks

that should not be a surprise? 6.25 is pretty much exactly what you'd expect running a -j6 job

(on a quad core machine)

jbo: ssshh, let people enjoy things :-P

oki :<

jbo: do you like spaghetti? le-fay.org/tmp/30d/Screenshot.from.…024-02-25.at.16_30_59.469447404.png

lw, I envy you. let me know if you ever wanna make pasta together.

lw, "starter base" - right? >:D

jbo: yeah, except i normally use a main bus even for starting so this mess is a bit new for me :-)

trying bob's mods which adds a ridiculous amount of new materials (tin, lead, ...)

lw, the only mod I ever bothered dicking around with was LTN. and that became somewhat obsolete in later versions with improved circuits

lw, I am really, really looking forward to the expansion

lw, I need to invest some time figuring out how this wine non-sense works. never used it

can you do LTN with circuits now?

jbo: wine is easy: install it, run "winetricks", install d3dcompiler_47 (which is required for most games), then just "wine64 factorio.exe"

well, not all of it. stuff like depots etc. don't work in vanilla. but "back in the day", you had very limited (to no) circuits information from a train at a train station. also stuff like setting train limits via circuits was not possible and that is the feature I usually use to manage my trains: circuits to figure out how many trains can be sourced or sinked at any given station.

that way you can have a mostly working many-to-many solution in vanilla

lw i mess the upgrade lol

will have to install 14.0 from scratch

lw, and you're getting 60fps out of that? I did use wine once back in 2009 or something and it was one of the slowest experiences I can recall

jbo: 60 fps/ups and my cpu isn't even at max frequency (currently 2.2-2.8GHz)

lw, to be fair, the factorio C++ dev team did an absolutely formidable job.

lw, I take it that you're familiar with dosh's work?

wine 3D is *very* good these days because, aiui, it just runs everything through dxvk so it's basically native vulkan... although in the case of factorio i think the game is opengl anyway

no, what's a dosh

lw, youtube.com/playlist?list=PLtyo3aqsNv_OL_ZAwr9NkjBUuejZM68gZ

Title: Factorio - YouTube

lw, software engineering playing factorio the way I would want to if only I had the time :D

i don't need any more crazy ideas right now :-P will have a look once i get to the next tier in this save maybe

lw, oh don't worry. the stuff he shows is usually not something you end up doing yourself :D

jbo: remind me what you want tested for the wine-devel update, is it that it builds on 13+14 or do i need to check it runs there too? (it's got no maintainer so there's no one else to do this except me and you)

lw, I mostly care about the testport failures on 13R and 14R. If you can runtime test it that would be great

ok testport is easy, i'll see if i can run some command line app in a 13/14 vm

powershell? :D

(the testport failure is on 15 too, i don't know how i missed that when submitting the patch)

:)

lw, so d3dcompiler_47 is not something that comes from ports - is that correct?

jbo: correct, winetricks gets it from weird place like... the Firefox installer? it's a part of the Microsoft Direct3D redistributable

wtf

(specifically it's the shader compiler; wine d3d has its own shader compiler but it's very incomplete so you normally want the official D3D one instead)

ack

fancy nick btw

what, lw? it's my name

no, the one you disclosed earlier :>

oh

also, triple letters or it's not :p

#oldpeople

i have two middle names, i'm llfw on github (because lw was taken)

ls whitch is best release or stable version

?

warsoul, if you have to ask then the answer is most likely: release

warsoul: right now? i'd go with 14.0 but if you're very conservative, 13.2 might be the more tested release

a lot of people stay one release behind at least until the next .1 is out (so 14.1)

lw, jup

i run 15.0 everywhere so i'm probably not the best person to ask about most stable release :-P

I found a 13.1 jail that I upgraded to 14.0 and now mysql is not starting because of some PAM stuff. as far as I can tell, because opie was removed and for whatever reason it used it. I am still trying to figure out how to fix this one (other than installing security/opie)

jbo: it's extremely unlikely your mysql actually uses OPIE, just remove it from the pam config

opie was in the default pam config in <= 13

and with "found a 13.1 jail" I of course mean that I knew it was there but just continued to neglect it because non-production :p

# service mysql-server start

Starting mysql.

su: pam_start: System error

jbo: grep opie /etc/pam.d/*

lw, I did comment out pam_opie.so and pam_opieaccess.so

oh hm

ah, should probably restart the service? :D

is mysqld linked to opie?

(if there is a pam service to restart?)

there is no pam service

what does mysqld.log say?

lw, ldd `which mysql` does not indicate that it's linked to opie.

yeah i didn't think it would be

whats in the ~~box~~ log

I'm still looking for the mysql logs. I migrated "everything" to psql so memory is not fresh

wtf is ~~box~~ log?

jbo: oh i forgot you're old. what's in the box^H^H^Hlog

also mysqld.log is usually in the data directory

unless you configure it to use syslog but i don't think freebsd does that by default... although i haven't used mysql for ages

lw, I think I'm in need for some explanation of what box log is.

("requires mysql" is basically an automatic nope for me when evaluating new software)

jbo: sorry, i was trying to make a joke. obviously badly :-P i want to know what's in the log. "what's in the box" is a JJ Abrams-themed pop culture joke

oh shit sorry :D

still looking for dem logs

I assume those are not the *.bin-xxx under /var/db

no, but if the log isn't in /var/log it's probably in the same directory as those files (those are database binary logs)

(binary logs meaning updates to the database data itself, that's the wrong sort of log)

jup, thought so

still looking

did you check the rc.d script? maybe it has an option to set the logfile

nope. I remember this to be a major pain back in the day

config file has no log= entry...

omg why is download.freebsd.org so slow

/aux0/vm/.iso/FreeBSD-13.2-RELEASE-amd64-booto 77% of 372 MB 9258 kBps 05s

at least you have log files

ok... freebsd 13.2 installer hangs at "Launching APs: 1 3 2"... can tell this is going to be a fun day

anyone have an idea what's up here le-fay.org/tmp/30d/VlQYBx.txt

lw, so... I was launching a previously working windows 11 vm yesterday and it hung every time in the bootloader

that was on stable/14

lw, so here's my modified /etc/pam.d/sshd: termbin.com/56pl

/etc/pam.d/sshd definitely has nothing to do with mysql

ok freebsd 14.0 hangs in the same place, wtf. did someone break vmm?

lw, regarding vmm: I'd really like to know too.

and whether it affects current stable/14

it works fine if i only give it 1 cpu

that's obviously no need for building ports though

that is _EXACTLY_ the symptoms I encountered too

jbo: i'll mail current later about this if no one has an idea (i asked #bsdcode just in case)

lw, there's the freebsd-virtualization@ ML

oh god not another mailing list

Title: FreeBSD Mailing lists

:>

the archive for freebsd-women@ is extra compact

I was delighted to see that we have a freebsd-tcltk@ ml

lw, does your vm use disk_type="nvme" by any chance?

jbo: it does

lw, can you try not using that?

(as do all my VMs on a 14.0 host - and they work fine)

hmpf

i'll try virtio block for the 13.x one

just waiting for 14 to install

jbo: yes the 13.2 installer boots fine with virtio-blk

lw, and it didn't with nvme?

correct

so my hunch was correct :(

lw, time to e-mail virtualization@

yeah, let me get wine-devel done then i'll do that

lw, I can commit that one without asking for explicit mentor approval (unless your changes go nuclear now)

jbo: how come

jbo: ok this is interesting, i booted my installed 14.0 system with virtio-blk and it hang for about 15 seconds at "Launching APs", but then it did boot

i wonder if this is a timing issue since i'm running testport at the same time

5:30PM up 1 day, 3:36, 4 users, load averages: 8.95, 7.26, 6.90

it's also *really* slow to boot, much slower than i'd expect with a build running

jbo: ok i switched back to nvme and it still works fine

just slowly

I have an issue with indel HD Graphics 620 on a dell notebook witn intel Xeon E3-1200 7th gen

right after I enable drm i915kms in rc.conf - notebooks constantly goes to auto reboot

havent found anything in logs

jbo: should i take maintainership of wine-devel? i am not in wine upstream development and may not be the best person to maintain it, but i do use it

s/not in/ not involved in/

lw, you definitely should :>

i noticed Gerald Pfeifer <gerald⊙Fo> commits to it a lot but hasn't taken maintainer for some reason

oh now wtf is wrong

% ldapvi cn=thyme.eden.le-fay.org

ldap_start_tls_s: Connect error (-11)

additional info: error:0A000086:SSL routines::certificate verify failed (certificate has expired)

ok, forgot to include slapd's certificates in dehydrated's hook.sh... god how i wish we had something like Windows' certificate storage on freebsd

lw, might be worth shouting gerald@ an e-mail first as he seems to maintain the other ports too. but technically you can just take it.

lw, regarding the current PR: just update the patch when you're ready and I'll handle it

jbo: sent mail to gerald@ cc: you

jbo: i am working on the patch, just need to set up nfs/kerberos/ldap/etc on my new 13+14 VMs

ok, gerald@'s mta rejected my mail...

|<gerald⊙pc>: host mailwash60.pair.com[66.39.2.60] said: 550 5.7.1

| le-fay.org: Sender is not authorized by default to use 'lexi⊙lo' in

| 'mfrom' identity (mechanism '-all' matched) (in reply to end of DATA

| command)

what the heck does that mean

are my SPF records wrong

kerberos & ldap? wut?

SPF would be my guess

jbo: kerberos and ldap so i can log into them?

easier to test ports if i can log in :-P

why not just ssh?

and why kerberos AND ldap?

jbo: i do use ssh... logging in as my LDAP user, authenticated by Kerberos

fancy

I guess I must be forgiven for only using plain SSH and RDP

you always need Kerberos AND LDAP because LDAP only provides naming, Kerberos provides authentication

jgh__: my SPF records look fine though

le-fay.org.3492INTXT"v=spf1 ip4:81.187.47.194 ip6:2001:8b0:aab5:107::10 -all"

thyme.eden.le-fay.org has address 81.187.47.194

thyme.eden.le-fay.org has IPv6 address 2001:8b0:aab5:107::10

lw, e-mail arrived here w/o issues

jbo: could you please reply and copy gerald and include the error i pasted above?

lw, also, not really nice disclosing the dev's personal e-mail addr in a publicly logged channel

there's a reason why people hide behind @freebsd.org :D

true i suppose although i don't really consider email addresses to be secret...

hence just "not-nice" :)

lw, can you paste that MTA message somewhere for me to copy? IRC formatting...

I'm sorry to have to inform you that your message could not

be delivered to one or more recipients. It's attached below.

For further assistance, please send mail to postmaster.

If you do so, please include this problem report. You can

delete your own text from the attached returned message.

The mail system

<gerald⊙pc>: host mailwash60.pair.com[66.39.2.60] said: 550 5.7.1

le-fay.org: Sender is not authorized by default to use 'lexi⊙lo' in

'mfrom' identity (mechanism '-all' matched) (in reply to end of DATA

command)

fuck

i meant to send that via PM, sorry

jesus christ :D

pasting is hard :-(

Well, don't. :P

Don't paste, problem solved?

debdrup: don't PM, don't paste, or don't jesus christ?

not really my fault X/Wayland has like 37 clip buffers

ok maybe a little bit my fault

lw, done.

lw, gotta run, ttyl o/

jbo: no come back you have to keep me company while i testport this crap

who else will understand my random complaining

lw, sorry, can't. will be back later (2hrs?)

lw, I am sure I'll hear your complaining anyway :p

"the wind is blowing"

really gotta run, ttyl o/

wow rude

huh? that was not supposed to be rude, apologies.

it was meant as a funny thingy.

i was joking :-d

-__-

ttyl

cu

One feature of irssi is that if you try pasting something that's more than 3 lines, I think, it'll attempt to catch it and prevent it from being sent before you explitly tell it to.

Which probably says more about how easy it is to mispaste, than anything else - but it's still a very nice little feature.

yeah, in this case i actually meant to copy the URL containing that text and paste it to jbo in /msg though

so wrong either way

if you use irssi you can also have it catch your paste and turn it into an url

Nei: huh, cool!

How's that done?

you just load a small script like anti.teamidiot.de/static/nei/*/Code/Irssi/pastebin.pl

and then when you paste something, irssi will show you a prompt where you have multiple choices like confirm, or turn into an url (ctrl+u)

Oh, it supports the null pointer, neat. Any chance to get paste.sh added, though?

shouldn't be too hard, either call out to the paste.sh provided script or port it to perl

Yes, please. :P

teamidiot.de ... :D

reminds me this : img.izismile.com/img/img6/20130320/…that_lowers_your_iq_permanently.mp4

;)

I still haven't figured out anything useful to do with the domain I grabbed..

debdrup, build your own mailserver :)

nerozero: I already run one on another domain.

then I guess this is enough :D

its ajust a placeholder for your name...

need some sleep, if anyone can suggest me what tp do with "drm_atomic_helper_wait_for_dependencies" please leave a comment at: forums.freebsd.org/threads/intel-62…encies-hangs-reboot-very-slow.92487

Title: intel 620 - drm_atomic_helper_wait_for_dependencies, hangs, reboot, very slow | The FreeBSD Forums

postfix with blacklistd and milters to filer spam via rspamd, dovecot, solr, and sieve

add sieve on dovecot-pigeonhole

that's what i'm doing

ah ok

ok, this is weird: i can log in via ssh via Kerberos if /home is not mounted, but if /home is mounted, then authentication fails

lw: via pam_krb5?

debdrup: no, GSS

maybe try pam_krb5 instead?

i really don't want to type my password every time i log in

ssh-agent?

sure, i could stop using kerberos entirely and use ssh keys instead, but i'd rather find out why kerberos doesn't work

that's understandable - well, ssh with enough verbosity might reveal something

sadly it did not

how many levels of verbosity are you on?

-dddd

did you put any .k5* files in your home?

i was thinking -vvvvv

it did log 'NFSv4 error WrongSec: You probably need a Kerberos TGT' which suggests it's trying to access something in my home directory, but i don't understand why that causes authentication to fail when it works fine if /home is not mounted at all

might also help to increase the log level of the server

debdrup: sshd doesn't have a -v option

lw: i said ssh, not sshd

ssh as in both ssh(1) and sshd(8)

sorry, i should've made myself more clear

oh, well the problem is on the server... ssh -vvv doesn't return anything interesting, it tried GSS auth and fails

freebsd's NFS client seems to have no mount option to use the host credential for root access which i find odd

there's syskrb5, but that disables user tickets entirely

ah wait got it

or, no. using gssname=host,allgssname lets me log in, but i can't access my home directory

maybe i need autofs, but last i heard freebsd's autofs is broken with nfs

ok, i'm pretty sure this is a bug in sshd. it works fine via console login, with automount of my home directory

hello

lw i just done installing 14.0 from scratch

but when i try to run startx

it says cannot run in framebuffer mode. please specify busIDs

how can i fix this

ok, yes, this is a bug in Heimdal: heimdal/heimdal #368

Title: Inaccessble ~/.k5login breaks sshd gssapi-with-mic · Issue #368 · heimdal/heimdal · GitHub

368 – Lpd doesn't log errors after failed exec bugs.freebsd.org/bugzilla/show_bug.cgi?id=368

NFSv4 error WrongSec: You probably need a Kerberos TGT

fstatat(AT_FDCWD,"/home/lexi/.k5login.d",0x2792356252b0,AT_SYMLINK_NOFOLLOW) ERR#5 'Input/output error'

Failed gssapi-with-mic for lexi from 2001:8b0:aab5:106:3::10 port 63175 ssh2

i guess this means i need to build sshd with MIT Kerberos

how odd, good find

Nei: maybe not that odd since our Heimdal is ~12 years old and probably full of bugs...

Cy is working on replacing it with MIT

I'm wondering how the k5login scenario is supposed to be handled though

you need a ticket to read it but you want to read it first to check if your user is even allowed to log in

also why doesnt it use the host principal to try and read the file

right, i don't see how this could ever work with nfs+krb5 home directories

Nei: freebsd's mount_nfs doesn't have an option for "use the host principal only if user ticket is not available"

the only options are always use the host ticket or never use the host ticket

this is probably something that, unrelatedly, should also be fixed

might mail rick about that as he's helped me with kerberos nfs bugs in the past

does it try to read the k5login as root? maybe running k5start would be enough

that's not a bad idea, i can't remember what package k5start is it though

ah kstart

alas I have no idea what I'm doing so please be leery

ok, it does work if i "kinit --keytab=/etc/krb5.keytab host/freebsd14.eden.le-fay.org" then run sshd... so it should work under k5start... let's see

yes it works! root@freebsd14:~ # k5start -u host/freebsd14.eden.le-fay.org⊙LO -a -f /etc/krb5.keytab /usr/sbin/sshd -p 2222 -dddd

Nei: thanks... i'll hack up /etc/rc.d/sshd for now, i feel like this should be fixed in a better way though

cool

add this to item #57 on my things-to-fix-in-freebsd todo list...

oh no, i rebooted and now it doesn't work

ok, putting kinit in /etc/rc.local fixes it... that'll do for now

I believ if you only use kinit the ticket will expire and you're in for a random surprise

even for a host ticket? i thought those didn't expire

or does the TGT still expire

check klist

the keytab ofc is valid until you rotate host keys but the tgt should expire just the same

you are right

Principal: host/freebsd14.eden.le-fay.org⊙LO

Issued Expires Principal

Feb 25 19:21:56 2024 Feb 26 05:21:55 2024 krbtgt/LE-FAY.ORG⊙LO

you can also run kinit in a loop; poor mans kstart

well, 24 hours is enough to do what i need to do, then i'll nag -current@ :-P

freebsd has a install including the desktop enviroment?

not that i know of, you just install whatever desktop you want from ports/packages

follow all the steps to install xfce

but when system boots

i hit enter

and screen stays black

nothing pops up

mm

its like will come up something but it stays black

any advice

jbo: icmp echo req

lw, pong

lw, I'm back

hello the jbo

jbo: random question, why do you keep calling me "he", do you know many men named Lexi?

lw, I assume everybody's gender to trigger deep emotional trauma

lw, (obviously just kidding, I try to use "they/them" universally but when I am stressed I might screw up sometimes)

also, where I come from "Lexi" is used as short-hand for "Alexander"

also, I honestly don't care about your biological sex just tell me if I offended you because that was almost certainly definitely not intentional :)

does that answer the question sufficiently? :D

ah, over here alex is coded masculine, alexa/lexi are coded feminine (alexandra). interesting that that's not universally true

yeah I see I screwed up in that e-mail I sent just before I had to run - apologies. shall I rectify retroactively?

lw, testports running on my end now

jbo: i'm waiting for 14.0 to finish then i'll start 13.2

wait... you submitted the patch before the testport completed? :D

ah, you disclosed that in the PR :)

yes, but i literally said that in comment: "testport on 15.0 is fine, testport on 14.0 and 13.2 are running now."

just wanted to provide the working patch as the previous patch was definitely broken

jup

*twiddle* [14-wine_devel_92] [2024-02-25_19h47m35s] [parallel_build] Queued: 160 Built: 23 Failed: 0 Skipped: 0 Ignored: 0 Fetched: 129 Tobuild: 8 Time: 00:24:36

at least 14.0 packages seem up to date today

[00:00:17] Building 25 packages using up to 12 builders

meh

I hope I don't have to throw this on the big box

curious that wine-devel on freebsd14 seems to compile with gcc instead of clang

need help

with xfce not starting up

jbo: 14.0 testport passed, 14.0 bulk running now (so i can runtime test)

cant even get to the terminal

warsoul: if no one here can help, you may ask freebsd-questions mailing list: lists.freebsd.org/subscription/freebsd-questions

Title: FreeBSD Mailing lists: subscription for freebsd-questions

lw, sounds good. keep me posted

jbo: as if i wouldn't

if a mouse moves in my port i @ you

warsoul: screen stays black, like do you have a login: screen or what

yeah

stays black

so you need to log in and start xfce

lw, real men use trackballs (I hope the joke about the men part lands well right now)

how

i cant see nothing

its all black

is the monitor turned on?

jbo: i own a trackball but i haven't used it since i was a man (this is a trans code)

yes

s/code/joke

i see when freebsd is loading and everything

lw, the code part just made it that more enjoyable lol

what did you install to make it black?

i install nvidia driver

xorg

xfce

lightdm lightdm-gtk-greeter

enable it to run auto after booting

that's probably it; i'd disable that until you get the nvidia driver figured out

ok

when i boot how do i enter terminal

does ctrl-alt-f keys do anything?

let me check

not

nothing

can join terminal in the freebsd menu when it boots?

you could boot into single user mode and disable lightdm

How do I tell which C compiler is being used by ports/lang/gcc13 ?

mns: what do you mean

you want the exact version of gcc for lang/gcc13 ?

it logged in as single user

shows root user

didnt ask for passwd

when i try to disable lightdm says permission denied

even with sudo

warsoul: / is mounted read only; `mount -u -o rw /; mount -a`

worked

let me see

rebooting

lw, llvm17/18 does not yet implement std::jthread and std::stop_token -_-

jbo: huh are you sure? you did -std=c++23 right?

lw, those are c++20 features

i'm sure i used std::stop_token in like clang 15... unless that was with MSVC..

Title: Compiler support for C++20 - cppreference.com

17 (partial)*; 18 (partial)*

.__.

I have code using std::stop_token which does the thing under gcc13 but fails to compile with llvm17 so pretty sure, yep

rtprio: what I mean is that which is compiler is used to compilet ports/lang/gcc13. Is there a particular CC that the build uses or just whatever happens to be first in the PATH ?

the systme default, likely clang, unless the port specifies otherwise

lw, looks like emulators/wine is behind a full major release too

really, i thought it was on 9.1

8.0.2

latest stable release is 9.0

ah true, well, that sounds like Alexander's problem (do you see what i did there)

oh and i see my wine-devel is out of date because 9.3 is out now

I agree, definitely something we should tell lexi about

fuck sake...

lw, I didn't know how to tell you so I thought I'll ease it in by taking about stable release first :D

jbo: do you want me to finished testporting 9.2 or should i start over with 9.3? i'd rather do tha ltter

s/tha ltter/the latter/

lw, I finished testporting 9.2 so we can commit that if you like and then you get started on 9.3

ok sure

lw, but that is mainly because 99% of the work is done already

at least i've externsively tested 9.2 so we know it works

yep

extensively

always good to test stuff externsively

i rather test it externsively

not sure that's a word though

he who shall test externsively

extern "sively" { int wine(void); }; "

template<char* C> [[nodiscard]] consteval bool gen() { if constexpr (std::is_same_v<lexi, T>) return m; else return f; }

jbo: this is an honest question, don't get mad: for something like wine-devel which release frequently and is an unstable release, do you think it's more important to test on 13.x and 14.x or is it better to keep up with current release and fix old releases as reported?

ah I screwed it up... of course I did.

rust would be so much safer **cough**

lw, a -devel port is still expected to work under 13-RELEASE and 14-RELEASE. if upstream has a bug that was fixed in a later version that would be covered by -devel that just results in a new port PR

from what I can tell, the wine-devel situation is a bit unfortunate.

but seems like the right thing to do so I'm not complaining

lw, so you good with the 9.2 commit?

\

who can help me to install the right nvidia driver

?

warsoul, unless you have a very old GPU, just install nvidia-driver-535

Title: FreeBSD Display Driver – x64 | 535.43.02 | FreeBSD x64 | NVIDIA

see "Supported Products" tab

warsoul, as in: pkg install x11/nvidia-driver and you should be good.

lw, did you timeout?

jbo: went for a fag, testport 14.0 still running

jbo: i'm fine with 9.2 fcommit on 15.0 but i haven't tested it on 13/4

lw, is that major effort or... a matter of minutes?

matter of many minutes

actually testport on 14.0 passed, i'm waiting for bulk so i can install and test it

cpu use seems weirdly low actually, i wonder what it's doing

PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND

11 root 8 187 ki31 0B 128K CPU0 0 234.5H 686.68% [idle]

59509 root 15 20 0 8252M 7046M kqread 1 160:50 104.01% bhyve: freebsd14 (bhyve)

lw: wait so you're literally just updating the wine port?

johnjaye: i'm updating emulators/wine-devel

i didn't realize the distinction

but yeah that's cool.

i can't even work out what it's doing

PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND

11 root 4 187 ki31 0B 64K CPU0 0 340:31 297.21% [idle]

19260 root 1 68 0 15M 3520K nanslp 1 0:02 0.19% sh: poudriere[14-wine_devel_92]: html_json_main (sh)

29862 nobody 1 68 0 88M 68M wait 2 0:01 0.11% gmake -f Makefile all

9732 root 1 20 0 14M 3472K CPU3 3 0:00 0.03% top -Sa

it's just kind of doing nothing

i hate when that happens.

you're testing it in pourdriere?

yes

lw, have you tried poking it with a stick?

i have a custom script that does testport (to make sure it builds) then does bulk and sets up a repository to install the package

this meets jbo's testing requirements for a port :-d

it only counts if you do it by hand :p

oh now it's installing

oh now it's creating the package

well this looks to be fun.. clonding from ssd -> nvme to replace a boot drive is not as easy as it sounds

see - sometimes things just need to be poked with a stick

i guess i could just put freebsd on my spare drive. it's gpt partitioned already so

#farmerKnowledge

jbo: i didn't poke it, i let it be itself and do its own thing

lw, did it change gender half way in?

jbo you use xfce

?

see you southrners don't know it all

warsoul, no

jbo: the place i used to live had a pub called "the farmer's" because it was were all the farmers went

like, the pub wasn't actually called that, everyone just called it that

that applies to the whole area I live in

jbo i did before pkg install nvidia-driver

i need a Windows command-line .exe i can test wine with

any suggestions

will x11 work better for me?

lw, wget?

oh, cURL is a good idea

that ^

notepad.exe

warsoul, I don't understand the question. whether you use xfce shouldn't matter to your nvidia driver installations

notepad++.exe for the more sophisticated

warsoul, _if_ you dare to use wayland, keep in mind that nvidia supports a different API than some compositors use.

jbo: freebsd 14.0 tested good [37!] freebsd14 ~/curl-8.6.0_2-win64-mingw# wine64 bin/curl.exe -i le-fay.org

0024:fixme:ntdll:create_logical_proc_info stub

0024:fixme:ntdll:init_cpu_info Failed to get logical processor information, status 0xc0000002.

002c:fixme:ntdll:create_logical_proc_info stub

002c:fixme:ntdll:init_cpu_info Failed to get logical processor information, status 0xc0000002.

fuck me i did not mean to paste all that

why is pasting so hard

let me try this again

lw, you really need to go through pasting bootcamp :D

jbo: freebsd 14.0 wine-devel testported + runtime tested: le-fay.org/tmp/30d/8EBKJR.txt

lw, so any chance taht this is because vmm stuff?

Title: 301 Moved Permanently

pastebin is bsd.to. that's almost enough to type on one hand

lw, update the PR for historic completeness and I'll get things going. I completed my testports.

jbo: i still need to test on 13.2

ah

is that the stable of 13?

sucks to be you :p

(kidding?)

johnjaye, 13.2-RELEASE is the current 13 release

johnjaye: no, the stable of 13 is 13-STABLE. 13.2 is the latest release of 13

i see

13.3 builds have started today I think

RC1 that is

and 12 is eol so you don't need it?

yes, 12 is EOL but jbo will probably force me to test on 12 anyway becuase he hates ports maintainers

(this is a joke don't @ me)

all the way down to 9 on PowerPC and MIPS

jbo: testporting on Windows NT 3.51 for PowerPC now

ack, let me know when 3.11 completed

did you know notepad and write were separate things even on windows 1.0?

true story bro

lw, don't forget to test on TempleOS please

jbo: i bet pkgsrc doesn't make you do all this crap

this is probably why pkgsrc doesn't work on freebsd though

lw, what you failed to realize is that you're not the maintainer of emulators/wine-devel, so this is more of a common courtesy than anything else tbh :p

_if_ you are the maintainer you're expected to do this as per the porters handbook / general rules.

jbo: what you're forgetting is you demanded i do this to commit the patch :-P

lw, what you're forgetting is that I'm a mentee and technically have no authority :p

aha

hmm, but i can't assign my bug to someone else because i don't have special bugzilla privileges

so you're holding my bug hostage

let me know when 13.2-RELEASE runtime tests completed and we're good to go :-*

alright, Jack Bauer, let's play

oh, I can gladly re-assign it

you can always e-mail portmgr@ with personal complaints (but honestly - although we're joking around - please don't)

root@freebsd13:/usr # git clone -b wine-devel-92 git.eden.le-fay.org/ports.git ports

git: Command not found.

everything hates me today

*padpad*

oh right i pushed it to github

> git: Command not found.

I hardly think it matters

ok clever guy, what do you think about this

root@freebsd13:/usr # git clone -b wine-devel-92 git.eden.le-fay.org/ports.git ports

Cloning into 'ports'...

fatal: repository 'git.eden.le-fay.org/ports.git' not found

oh it's because the repository is on github!

oh, so you failed at pasting again is what you're saying? >:D

no, you just failed at reading my mind

smh

stronger smh

S my H SO HARD

committing my Sing H right now

sorry, I meant: more externsive smh

> externsive

oh, you picked up on it?

lol, I myself have had a battle with ports for the last day or so. I don't know why, but for some reason textproc/xmlcatmgr fails at the package stage on newer patched versions of 13.2. I've already tried fixing up the make file, but then that just ends up breaking dependent ports in some other way.

jbo: go paste yourself

lw, I'm gonna revert wine-devel back to 7.1 if you don't stop now

lw, at least I have more PCIe lanes!

jbo: i will write the 9.2 package to DVD and submit to archive.org

okay _mister_ starter base

wow reporting you to core@ for transphobia

<lw> jbo: went for a fa

<lw> jbo: went for a fag

damit, paste fail

[#freebsd] <jbo> lw, you really need to go through pasting bootcamp :D

yeah I guess you win this one :p

gerald will be proud

his email server won't

still don't know what's up with that

should have pretended to be a dude, lexi.

(oh god I hope this doesn't end up somewhere out-of-context)

tuaris, do you have a build log to look at?

ok, seriously pasting question, i'm trying to paste my testport script and it only pastes like the first 20 lines? wtf

* lw has quit (Excess Flood)

installed xfce and xorg again but without auto run on boot

on terminel when i type startx says

jbo: pkg.ent.morante.net/poudriere/data/…58s/logs/errors/xmlcatmgr-2.2_4.log

cannot run in frambuffer mode. please specify busids for all framebuffer devices

warsoul, before you mess around with x11/xorg: did you successfully install & load a driver that is compatible with your GPU?

If I try to fix the paths in the makefile, other stuff just breaks as a result; pkg.ent.morante.net/poudriere/build…rnlegacy&build=2024-02-25_02h00m10s

tuaris, I just built textproc/xmlmgr and it builds out of the box without issues (2.2_4)

official builders seem fine too

i386 oh

jbo i installed pkg install x11/nvidia-driver

but didnt know have to load it

how can i load it

warsoul, docs.freebsd.org/en/books/handbook/x11/#x-configuration-nvidia

Title: Chapter 5. The X Window System | FreeBSD Documentation Portal

It failed for me on 13.2-RELEASE-p9

64-bit

but it worked fine on 13.2-RELEASE-p3: pkg.ent.morante.net/poudriere/data/…_04h39m24s/logs/xmlcatmgr-2.2_4.log

tuaris, I cannot reproduce this. 13.2-RELEASE-p9 amd64 builds fine out of the box

ok i think this is an alacritty bug, or maybe a wayland bug, it's randomly truncating my pastes

hmm, very strange

le-fay.org/tmp/30d/Screenshot.from.…024-02-25.at.21_57_16.612848831.png i pasted the left file into the right file and it's clearly just truncated halfway through, wtf

lw, I was using alacritty for some time but ended up having pain everywhere, switched to kitty eventually. no more pain. don't ask for details please.

jbo: i know someone else who uses kitty... please provide details

also isn't there one that begins with 'z'? not zenity....

lw, is that someone better at pasting than you?

jbo: did you look at my screenshot

urxvt

literally my terminal emulator can't paste

no no urxvt

something that works under wayland

maybe one day I'll try wayland again

and maybe one day you'll complete your runtime tests :>

jbo: maybe you ask the alacritty maintainers why she didn't test pasting

I'm too busy:

ID TOTAL ORIGIN PKGNAME PHASE PHASE TMPFS CPU% MEM%

[18] 09:45:50 misc/pytorch | pytorch-2.2.1 build 09:37:00 1.98 GiB 0% 0.1%

nothing like having 28 cores and waiting on that one port to complete

do you even PARALLEL_JOBS

only on Tuesdays

wtf shell

root@freebsd13:/usr/ports/emulators/wine-devel # mv ~lexi/testport /usr/local/bin/

Unknown user: lexi.

root@freebsd13:/usr/ports/emulators/wine-devel # id lexi

uid=10006(lexi) gid=10006(users) groups=10006(users),10007(admins),10008(filer)

root@freebsd13:/usr/ports/emulators/wine-devel # mv ~lexi/testport /usr/local/bin/

Unknown user: lexi.

file ~lexi

root@freebsd13:/usr/ports/emulators/wine-devel # file ~lexi

Unknown user: lexi.

root@freebsd13:/usr/ports/emulators/wine-devel # sh

# file ~lexi

/home/lexi: directory

exit; echo $SHELL

I smell csh

not sure what you're after there but

root@freebsd13:/usr/ports/emulators/wine-devel # exit

logout

FreeBSD/amd64 (freebsd13.eden.le-fay.org) (ttyu0)

login:

it's 13 so yeah probably csh

oh I thought you spawned another sh there hence the exit. wanted to know $SHELL on the "failed" 'file ~lexi' attempt

login: root

Password:

Feb 25 22:06:40 freebsd13 login[1560]: ROOT LOGIN (root) ON ttyu0

Last login: Sun Feb 25 22:06:28 on ttyu0

FreeBSD 13.2-RELEASE releng/13.2-n254617-525ecfdad597 GENERIC

root@freebsd13:~ # echo $SHELL

/bin/csh

fucking csh

pkg install -y zsh

:D

i already installed zsh so i could log in

now imaging having to do all of this again when you updated emulators/wine-devel to 9.3 :>

9.3 will be easy as i've already done this shit

except reporting the sshd+kerbeos+nfs bug which i'll rely on cy@ to fix

or rick

rick and cy are my bestet buds

ey!

now it's getting personal :<

rick and cy and jbo are my bestet buds

:>

also imp is nice since he commits my src patches

cy and rick and jbo and imp can come to my birthday party

I'll bring some PCIe lanes

if they come with a motherboard you're welcome to

if the motherboard has proper SFP+ cages so i don't have to put a NIC in it, even better

that reminds me that I need to order a spare heating element...

root@freebsd13:/usr/ports/emulators/wine-devel # testport

Branch: wine-devel-92

Port: emulators/wine-devel

*** Creating ports tree

[00:00:00] Creating wine_devel_92 fs at /poudriere/ports/wine_devel_92... done

jbo: now you have to wait because unlike you i don't have a 256-core CPU just sitting around to run testport on

must be nice to be rich

my gtx 1080 ti paints a different picture

hahahahahah pwned

vgapci0@pci0:41:0:0: class=0x030000 rev=0xc1 hdr=0x00 vendor=0x1002 device=0x73bf subvendor=0x1eae subdevice=0x6701

vendor = 'Advanced Micro Devices, Inc. [AMD/ATI]'

device = 'Navi 21 [Radeon RX 6800/6800 XT / 6900 XT]'

except as i mentioned earlier i'm wondering if i should have gone with nvidia for freebsd

except i guess nvidia don't provide binary drivers for main? so

vgapci0@pci0:1:0:0: class=0x030000 rev=0xa1 hdr=0x00 vendor=0x10de device=0x1bb0 subvendor=0x10de subdevice=0x11b2

vendor = 'NVIDIA Corporation'

device = 'GP104GL [Quadro P5000]'

class = display

subclass = VGA

your 1080 is a quadro?

i actually have a 1080 Ti here and to be honest the main reason i upgraded was because it was the Founder's Edition and the fan was fricking ridiculous

you'd start X and it would spin up to ten gazillion rpm

nah I'm just sitting at the office where I got a P5000

1080ti is at home

is there a guide for that device subclass subvendor terminology?

% sysctl hw.model

hw.model: Intel(R) Xeon(R) CPU E5-2690 v4 @ 2.60GHz

also dated af :(

johnjaye: the USB Forum probably publishes one but in short: 'vendor' and 'device' are 16-bit numbers which are set by the PCI card. 'class' described the generation function of the device

s/generation/general/ (damnit jbo)

tahahahaha <3

ok

i don't know why i just said 'USB Forum' when we're talking about PCI cards

but USB kind of does the same thing

lw, to get real: any ETA on this 9.2 effort? I would not mind going home but I got everything setup here rn

jbo: 13.2 testport is running now

[13-wine_devel_92] [2024-02-25_22h11m47s] [parallel_build] Queued: 160 Built: 5 Failed: 0 Skipped: 0 Ignored: 0 Fetched: 132 Tobuild: 23 Time: 00:08:57

didn't we already conclude 13.2 testport? -__-

no, we did 14.0 before

then you demanded i do 13.2

so here we are doing 13.2

"demanded"

and now you're complaining about it

fricking jbo

man, I really hope people don't get the wrong idea if reading this log :D

literally the worst

lw, we'd be making better progress if you'd know how to paste

jbo: buy me some vodka and i'll delete the log

i know how the paste, my god man, you would not believe the shit i've pasted

lw, the time it takes you to do things I can probably make some from scratch

lw, did you just assume my gender?!?!?!?!?!

i pasted under artillery fire in vietman

my friends call me "Hunter Paste Thompson"

clearly they must have very low standards

jbo: instead of sending PCIe lanes why don't you send me a new CPU then this would finished quicker

that's probably why 'nam took you 20 years

lw, I have no comeback to that

i just want to finish this shit and go back to my started base :-(

starter

s/started/starter :>

typing hard

pls Elon send me a neutalink keyboard[1;5

neuralink

pls tell me ur kiddn'

i am so much kidding

not sure if sarcasm

i would never put Elon device in my brain, only Apple devices

lw, earlier when I had to leave for a couple of hours, I shared the anecdote about the MTA with a friend. had a good laugh.

I'm also still enjoying the box^H^H^H

did you ever find the mysql log file or are you just terrible at logs

ugh krb5-ldap still isn't built for arm64

deos the arm64 builder just run once a month or what

lw, didn't find it. just fixed the problem.

what was the problem

still had pam_opie*.so elsewhere in /etc/pam.d/

...

:D

i'm sure i said grep opie /etc/pam.d/*

took me longer than you'd think even after you pointed it out

that must have gotten lost in-between your terrible pastes

but yes, you did.

which reminds me - instead of bs'ing here with you I could finish fixing the jail because the nullfs mounts are currently still broken

heh jails

I found some left over Git merge conflict lines in the Makefile under textproc directory. Not sure how those got in there. Maybe that's the problem? (seems unlikely). Other than that the ports tree looks legit. Going to give it another try

As many of you know, bhyve for aarch64 has been added. I'm trying to build the bhyve utility on aarch64, but after adding the needed things into Makefile, it still doesn't work.

did I miss anything?

tuaris, from what I can tell, you should be able to build textproc/xmlcatmgr as found in the ports tree without any modifications.

antranigv: i think only the kernel bits were merged?

lw true, but how to boot a VM?

antranigv: i suggest send mail to whoever merged it (Andrew Turner i think?) and ask :-)

yeah, I should :), but it's not working for me on a few diffrent systems. I might resort to spinning up a grgesh VM and giving it a try if this stil lfails

*fresh

antranigv: may well depend on uncommitted / private changes

tuaris, since you're already using poudriere, can you poudriere-testport on the default (unmodified) tree?

jbo: i realised i testported wine-devel against the wrong tree so i have to rerun both 13.2 and 14.0

lw, I completed testports for both 13.2 and 14.0 from your patch against main so we should be good there.

yep, getting that going

but we haven't tested them at runtime

lw, yes, but no need to testport again (if it's the same patch applied to main)

jbo: ok, do you want me to make a new patch to take maintainership then?

lw, right now I'd recommend to leave the patch of your existing PR as is - if possible, still do runtime tests. regarding maintainer ship: I assume you'd want to coordinate with gerald@ first?

but again: technically you're free to just take over.

jbo: i'm confused, you just said we don't have to do runtime tests

i can do another build with the correct version, no problem, but i've spent like 6 hours on this so far, would like to know exactly what is required

lw, maybe I misunderstood: when you said that you testported against the wrong tree - does/did that also apply to the runtimetest(s) you performed so far?

jbo: at least for 13.2, yes, once the test finished i noticed it was using the wrong version of the port

i'm rerunning the 13.2 test now, i'll check 14.0 after that

lw, when you say "wrong version" does that mean you runtime-tested emulators/wine-devel 9.1 instead of 9.2? >__<

jbo: yes, but don't worry: it was just a problem with my auto-generated pkg config. i re-ran the test on 14.0 with wine-devel 9.2 and it's fine

lw, awesome - what's the time investment cost for you to runtime-test wine-devel 9.2 on 13.2-RELEASE right now?

jbo: who the hell knows but i'm doing it anyway

spoken like a true maintainer

probably an hour or two

reminds me of: forums.freebsd.org/threads/how-to-i…-server-v-8-0-29.84950/#post-565320

Title: How to install mysql80-server v 8.0.29 ? | The FreeBSD Forums

your're doing all this testing in jails?

johnjaye: no, a VM

oh

> I spent 6 hours today to test [...]

> How has your weekend been so far?

:D

testing in jails is useless because it still uses the host kernel

oh ok

to actually test an application you need a VM

is the url you clone from the WWW in pkg-descr?

jbo: 13.2 testport finished already, hopefully bulk won't take long

johnjaye: no, cloning from WWW would be unusual, that's usually the homepage of the software

although if WWW is a github url, you *can* clone from that... but that's a special (if common) case

ok. and testport is poudriere testport ?

no, testport is a custom script i wrote

oh

and pr is project report not pull request

johnjaye, when most people talk about testport it is usually poudriere-testport

lw is just lazy :p

just making sure i know el vocabulario

call it lexitest already

this is my script that runs poudriere testport *and* runs poudriere bulk *and* configures a local pkg repository so you can install and test the port

lazy? i guess, if you think doing things correctly is lazy

(i'm joking, don't @ me unless you're jbo)

non-lazy people don't use scripts, they type everything by hand each and every time (error free, 100%)

echo '*** TESTPORT FAILED'.

echo y | poudriere ports -d -p "$portsname"

?

?

!

?!

jbo: there's no point keeping the ports tree around if it fails

i'll fix the tree then next time i run testport it'll create a new ports tree

> Poudriere doesn't allow '-' in ports tree names.... why?

why knows

who*

i think it's because '<portstree>-<setname>' is special to poudriere

so you can't use - in the ports tree name

wow, it was the corrupted Makefile. After fixing that, it's working now, lol

i will copy your script. then i too can build and test ports

jbo: bugs.freebsd.org/bugzilla/show_bug.cgi?id=277020#c4

Title: 277020 – emulators/wine-devel: update to 9.2

ack

thanks for testing properly

jbo: i'll see if gerald@ replies but considering i can't mail him... ?

so wait how was it fixed? poudriere-testport is failing on 13.2-RELEASE and 14.0-RELEASE as discussed on Irc

johnjaye: the original patch was wrong, i fixed the problem in the patch i provided in comment #2

the problem was pkg-plist was missing some new files

lw, if he's reply-all'ing you should get the mail. I'd let you know if I get one. resolving that would be part of the discussion I guess.

you added include/wine/windows/inputpaneinterop.h ?

i added several files to plist that you can see in the patch: bugs.freebsd.org/bugzilla/attachment.cgi?id=248744&action=diff

Title: Attachment #248744 for bug #277020

277020 – emulators/wine-devel: update to 9.2 bugs.freebsd.org/bugzilla/show_bug.cgi?id=277020

i see, i see. so that's how it's done

i didn't notice this originally because 'make install' of a port, even with DEVELOPER=yes, doesn't flag missing files in plist

[00:00:20] Building 8 packages using up to 8 builders ...

because jbo is a good committer, he picked up the issues so i could fix it

i'm going to finish the porter's handbook and then maybe i can become a ports maintainer like lw!

i don't actually maintain this port

i just submitted an update since i use it and it has no maintainer

i see...

the porter's handbook also tells you what the expectations towards maintainers are

which includes being responsive to e-mails, PRs, resolving issues, tracking down problems, testing and stuff like that.

and ideally collaborating with upstream

in fact, since i amn't the maintainer, i could have just left the broken patch for someone else to fix

i don't know why i didn't just do that, maybe i'm too nice of a person

only low-pcie-lane-count people do that

don't become like them

jbo: i'm still waiting for your feedback on my youtube videos

the spark has been extinguished

the spock? are you a trkkie?

jbo: vote on the next nonsense i should look at le-fay.org/tmp/30d/Screenshot.from.…024-02-25.at.23_50_19.030973775.png

where are we on postgres-exporter, that's on my right? test it and rename the port

lw, fix your other PR(s)

yes

and sublime-music I think we concluded to wait until 3.11 lands

ok, we can make a deal, commit wine-devel and i'll give you a new patch for postgres-exporter

but I can push that one - would just ask for approval

I am working on it.

sublime-music should just be committed, that you who's nervous about it

it works fine on python 3.11, i'm literally using it right now

then why commit it, if you're using it already? :>

tbh i often wonder about this

it works fine for me, why submit a port...?

why is your git user.email set to .ORG

jbo: sublime-music is a new port, it works, committing it won't break anything, just commit it

it can only be a net benefit

why is my what?

lw, it was a bad joke (yet again - apologies; withdrawn)

the TLD on your author e-mail is uppercase - you want that?

oh

or is that a paste issue?

yeah my domain is le-Fay.ORG

like FreeBSD.ORG?

so I leave it?

yeah

ack

i'm considering changing it to LE-FAY.ORG

on the basis that mixed case domains are confusing

and domains are caonically upper case

landed

which one

sublime?

oh wine-devel

y

neat, thanks

thank you too!

can I go home now?

no

jbo: commit sublime ffs, who cares if it breaks, it'll be exciting

jbo: why is your mentor approval 'implicit'?

because I gave it to myself :p

wat

isn't the point of having a mentor that she has to approve all your commits

towards the middle/end of the mentorship program you start to get some freedom if your mentor(s) think that you're not a complete idiot (at which point you'd never have become a committer in the first place)

ah ok

so you're not a complete idiot?

opinions differ on that subject

can i mail tcberner@ with my opinion?

sure