looking at xev's output, it doesn't seem to be receiving any XF86MonBrightness{Up,Down} keypresses, it only sees XF86WakeUp upon pressing the fn key

I have a super weird error on pfctl inside the jail. apparently, my pf is still working and I got this error everytime I run: # pfctl -sr

Assertion failed: (elems <= maxelems), function pf_nvuint_32_array, file /usr/src/lib/libpfctl/libpfctl.c, line 147.

rwp, guess what... VT-d was disabled in the BIOS... it was hidden under "System Agent Configuration". The "Enable Virtualization Technology" switch only enabled VT-x

now passthru is working :>

is there any way to stop files from being saved into /tmp from inheriting "wheel" as the group

tm512: who is writing the files?

tm512: you have to map the keys if you want brightness and shit to work

wheel gets inherited regardless of which user writes to /tmp

then it sounds like /tmp has the wrong chmod

drwxrwxrwt 9 root wheel 10 Feb 16 00:46 /tmp

also I *can't* map the keys for brightness control. as I said, only XF86WakeUp gets sent to the X server when the fn key is pressed. no distinct events are sent for the fn+brightness combos

supposedly the firmware is supposed to handle this

the mode and ownership for /tmp is the exact same on my system

`file /tmp` says ?

is the person writing files a member of wheel?

"sticky, directory", and as for being a member of wheel, my main user is, the user I'm running firefox under is not

files saved in /tmp from firefox get browser:wheel and then my main user basically loses write privileges to those files

jbo, Very good! I don't know why those are always vendor disabled by default.

I'm 100% okay with it being disabled by default (and that is what I would expect)

I am not okay with the convoluted graphical menu crap we have nowdays

tm512, In *BSD systems files always inherit the group of the parent directory. In Linux/SysV that is what the sgid bit on directories does. But it's the default for BSD.

tm512, Why doesn't the file ownership stay the owner of the file?

Isn't your browser running as your uid? That's where things are getting disconnected.

And alternately you could change the tmp to be $HOME/tmp so that it is written into a tmp directory in your $HOME and then the group would be your group.

well as I said I'm running firefox as a separate user, as an alternative to FF's lack of sandboxing on FreeBSD

Well, if it is running as a separate user then that separation of user thing is going to be working to keep you separated. :-/

But anyway the group setting to the group of the parent directory is core BSD behavior since forever. It's part of the landscape.

I added myself to the browser user's group, I guess I need these files to be set g+w by default though

seems like a lot of extra effort

maybe capsicumizer would be a better solution though it seems like development on that has been abandoned, the github repo was archived iirc

bit distracted from this now since I'm back to getting "fault errors on pipe A" when I run picom. I'm starting to get worried that this laptop has faulty hardware, because the laptop is also back to making weird noises at bootup. maybe these things aren't related though and I'm making spurious correlations

The laptop fan might be failing, a typical problem, and at boot the bios usually turns the fan on at high speed before reducing the fan speed. That's the only thing that would make noise on newer machines.

Older machines with a cdrom/dvd would seek the mechanism. If you have one of those. Those are becoming more rare as time goes by.

rwp: it seems to be playing through the speakers

drm-515-kmod pretty reliably gets picom to throw fault errors. I thought things were working fine here on drm-510-kmod, but now it's throwing them

Weird noises through the speaker at boot time? I guess I have heard clicks and such from sound cards as drivers are loaded.

rwp: looks like a safe and effective way to get rid of open source

it's like buzzing/whining accompanied by popping at the end of the tone. one time, it was an incredibly loud tone blaring from the speakers that tbh scared the shit out of me

hernan, It seems like the Law of Unintended Consequences being known ahead of time to everyone else making these Known Consequences but apparently Ignored Consequences by the politicians.

rwp: in another safe and effective measure, they want to get rid of farmers (and the food that peolpe eat, which comes from farmers).. certainly the politicians are not coming up with that nonsense by themselves.. they cant think 2+2. They are receiving orders from above.. the groups that already control everything, every company, every pharma, every main stream media, every politician. Do you know the 3

company names ? that control everything?

so these sounds play when the machine first starts, before the BIOS splash shows, and then again while FreeBSD is enumerating disks during boot

I'm wondering if one of these drives is faulty and messing with the rest of the hardware

the sounds don't show up reliably though, they'll go away for a while before coming back

hernan, I don't believe in conspiracies when incompetence is sufficiently explanatory.

tm512, I can't imagine how a storage device (SSD? Surely not a spinner...) would be able to affect other hardware. Other than by shorting out the power supply which is just about it. It's a power connection plus a SATA data connection. It's not like that can reach out to other parts of the machine.

"Sufficiently advanced cluelessness is indistinguishable from malice."

tm512: picom backend=glx, vsync=true ?

rwp: this laptop came with both a SSD and HDD

might also want to toggle the use-damage flag

I haven't actually used the HDD for anything yet

rwp: its no conspiracy, see for yourself in under 1 minute (the top 3, top 4 holders that own most of these companies) finance.yahoo.com/quote/BKNG/holders finance.yahoo.com/quote/META/holders finance.yahoo.com/quote/MSFT/holders finance.yahoo.com/quote/AAPL/holders finance.yahoo.com/quote/AMZN/holders finance.yahoo.com/quote/TSLA/holders

Title: Booking Holdings Inc. (BKNG) Stock Major Holders - Yahoo Finance

Title: Pfizer Inc. (PFE) Stock Major Holders - Yahoo Finance

topcat001: it's basically the default configuration with shadows and fading disabled, but yes using GLX and with vsync enabled

I might also have glx-no-stencil enabled

tm512, I am a craphound and routinely buy older used equipment. And routinely remove spinning hard drives and replace them with solid state drives. It's a cheap upgrade. Spinning drives on mobile equipment was always problematic but it was just the best available option at the time. Time has passed and now better options exist cheaply.

rwp: I was planning on leaving the HDD in and using it for storing media but if I want to test the hypothesis that something is faulty regarding the storage, I guess my one option right now is to pull the HDD

if the noises persist, I suppose I could load a spare SATA SSD into the space where the HDD was, and clone my FreeBSD install over to it, before removing this M.2 SSD

If you are not yet using it then that would be an easy test. It's going to be making noise at power on reset time as it seeks the head. That might be part of the noise you are hearing.

I need to relocating. BBIAB.

this thing is making noises that do not sound anything like a hard drive

the sound is punctuated by popping noises that sound a lot like the end of an audio buffer not aligning with the phase of a waveform

It's haunted! It's either poltergeists or gremlins.

or like the kind of popping that I think would happen if the waveform had a DC offset? basically what happens when the DAC hits a brick wall where the signal centers back to 0

also if it were the HDD I'd expect to hear it every time the machine starts up or every time I access the HDD while it's idling, but I just don't

it's all in your head

got a video of this though man, my tablet's microphone is terrible. a lot of this sound is just drowned out by noise

computer is wideband emi generator

anyway

platter hdds are still useful though right

maybe not as OS drives. but like for general storage

cheap, don't lose data over time?

actually that sucks in flash

youtube.com/watch?v=QVndKgz87gg here's a video with the noises, though yeah this mic did not pick it up with the clarity I would've liked. the "fault errors" spam makes a cameo at the start when I reboot the machine

Title: ThinkPad noises - YouTube

tm512: those pops?

i hear something in bg too

ketas: the pops are coming from the laptop, but there's a high-pitched whine between the two pops

hmm mic sucks then

or my phone speaker

the whine isn't unusual sadly :/

i guess only fix is usb sound card

i heard pssssh pop?

the noise isn't persistent, it only happens right before the BIOS splash, and right as FreeBSD starts detecting disks. and it's intermittent. it doesn't happen on every boot

only

if it happens before bios

that's shitty

but otherwise no noise?

i guess they optimized protections out

not just before BIOS, but when FreeBSD detects disks as well. this doesn't seem like something normal or expected, it seems like something is malfunctioning

it's not disks

it also inits sound

i guess that's that card doing init

not unusual to heard something but still

that's too much

yet, just hw is like that :/

I know the video is pretty blurry, but the sound shows up precisely when FreeBSD spits out info about the connected disks. it does not seem to coincide with sound hw initialization

hmm

but only on init?

so perhaps on that platform, doing init for storage causes sound to make weird output

again, it's only occasional. it seems like maybe it only starts happening when I haven't had the laptop powered off for a while, but that's speculation from an admittedly small dataset

weird

but only "cosmetic" issue?

this isn't the only noise I've heard it making, either, though the other noises it makes, I think it only happens right before the BIOS splash, and not when FreeBSD is booting. those other noises are louder, sometimes significantly so

one time it blared a sawtooth-wave like tone

I don't know if it's only "cosmetic", I'm worried it's a sign of some serious issues

kinda worried that the GPU fault error messages I'm getting have some relation to this, because those are also intermittent at least on drm-510-kmod

eh

dropped it?

it could be anything

I haven't dropped this laptop at all, but it was a refurb, so I dunno its history

anyway, huh, it did seem like there was a loose screw inside the laptop. when I heard that noise inside of it I thought it was a broken clip. though if that were short-circuiting anything I'm guessing the whole machine would be powering off at random

wtf

you never know actually

had one nos t60, did drop it a bit, nothing immediately, but i guess i just used it up too, i replaced inverter, then it lost bt, losed wifi if moved, lost fp, eventually stopped booting without any reason

i should take it apart and look

I guess to see if the drm fault errors are hardware related, *maybe* I could spin this thing up with Linux 5.10 or 5.15 and see if the same version of picom triggers the same messages? though maybe that's not definitive since the GPU drivers aren't identical

I can't even tell if this loose screw even belongs to this laptop, or if the refurbisher was just sloppy. the bottom cover screws are all accounted for

tm512: See if you can find a teardown video. It'll show you all the screws.

tm512: so this laptop just has some screws loose

:p

tm512: regarding your noise issue, I also use FreeBSD on some older machines, but use a Ubuntu live USB to check the hardware in such cases.

mason: it seems like the screw was supposed to be holding down something that covers the M.2 drive. a heat sink? it's entirely missing in my laptop

topcat001: the noises occur before the BIOS splash even shows, so I already know it's not OS-dependent

I was talking about trying Linux specifically in relation to this issue where picom causes my dmesg to be spammed with fault errors

still getting noises even with the HDD removed, though now they're different

ketas: so closely examining the video, it does seem like the noise is when the sound hardware gets initialized. or at least it coincides with the messages about hdacc/hdaa/pcm

hrm, can I blacklist the driver from even loading in order to test?

i think it's in

also, here's a video that much better showcases the noise though it's a bit more buzzy this time, compared to the first vid: youtube.com/watch?v=0PB8YesP36Y

Title: more ThinkPad noises - YouTube

iirc there is no way to disable built in module

this noise also plays when I wake the machine up from suspend, though nothing gets printed regarding the sound hardware

what the

!

I'll see if there are settings in the BIOS to disable the sound hardware entirely. might still play before the splash screen, but if it no longer plays while booting into FreeBSD, that narrows things down

but why use fbsd to fix it

I dunno what you mean

it does same in bios

what I'm talking about is preventing FreeBSD from initializing the sound hardware, see if that shuts it up. if it doesn't, then it probably isn't related to the sound hardware

tm512: ah ok so the noise is an electrical issue it seems.

tmiw: which DDX driver are you using in X?

sorry tm512 ^^

BTW even in Linux I sometimes get the fault errors, but they are very rare. Maybe one every couple of months.

pretty sure I'm using modesetting, though I did try intel without success

which intel igpu is this again?

I dunno that I tried every combination of drm version and modesetting/intel though

it's a comet lake mobile iGPU, the CPU is an i3-10110U

rule of thumb is old = intel, new = modesetting

10th gen, ok

most likely the bsd driver needs an update (not a useful comment sorry)

both Linux 5.10 and 5.15 came out a while after 10th-gen intel

and in this case, the 5.15 drivers seemed to be worse regarding the fault errors, at least on 5.10, for a while I wasn't dealing with any fault errors

but yeah, this is more than just a few, it completely fills up dmesg's buffer within like 10-15 seconds

Ubuntu 20.04 LTS has kernel 5.15, for comparison

It would be relatively quick to install on an external USB drive to check with some simple WM+picom

I was thinking Void Linux which has a much more vanilla kernel. I'm worried that there are a lot of confounding variables though, even though it sounds like FreeBSD tries to make the fewest number of changes possible to the Linux drivers in order to get them working

cool

maybe I'll just have to try 15-CURRENT in order to get drm-61-kmod, though I don't really want to

so picom's xr_glx_hybrid backend doesn't produce the fault errors, though this is not optimal as far as performance goes

oh nevermind picom just didn't launch

have to specify --legacy-backends to get picom to use that backend, and when I get it to launch the fault errors show up

btw, drm-61-kmod won't work with an Intel Arc GPU (or at least an Arc A770) since loading the i915kms module kernel panics for me, even when I already provided the correct firmware bins by hand. Don't get your hopes up there

this is a bit older than the Arc GPUs, I think

alright, well, this seems rather odd. I uninstalled 510, reinstalled 515, went to check it with both the modesetting and intel DDX drivers, both gave fault errors, so I moved back to 510, and the errors are no longer occurring even though nothing has changed

I am pretty worried that this laptop isn't doing too good

are you trying to revive an old laptop?

not really old. it's a ThinkPad E14 from 2019, I bought it as an upgrade from my previous laptop which is actually quite old now

wasn't prepared so I only caught the last bit of it, but here's that loud tone, one of the other sounds this laptop is making. though this one is less common. it plays from before the BIOS splash and continues playing until partway through the fbsd boot process as shown in the video: youtube.com/watch?v=sK_A9MKgNOg

Title: worse ThinkPad noises - YouTube

tm512: I believe the most vanilla Linux kernel is the one you can get from kernel.org.

remiliascarlet: but then you're stuck compiling from source

And?

and there's no need for me to do that when Void provides essentially vanilla kernel builds

especially pointless for some throwaway install just to test the i915 driver

The Linux distro I'm using is CRUX, so I already got used to having to compile my own kernel on every release, not even a big deal once you know your hardware configuration.

faced with the choice between compiling a huge piece of software from source and grabbing a binary that someone has already compiled, I'll generally take the latter

Void's Linux 5.15 package makes no modifications to the i915 driver, it's a good fit for this test

with the drm-510-kmod drivers, I would really like to know what makes picom cause the fault errors on some boots, but not on others. I'm also quite confused about how replacing 510 with 515, booting into 515, then downgrading back to 510 and rebooting seems to bring the system into a state where the fault errors don't happen

how set are you in picom ?

as long as I'm using X11, quite set

I *suppose* I could try the TearFree option on the intel DDX driver, but in my experience I'm not sure that's ever done what it's supposed to do

at least in recent years, I get screen tearing even with TearFree turned on

i will say, with configuration i've had to do a lot less horsing around with wayland

i'm partial to sway, myself

I would prefer to actually figure out what the issue is instead of attempting to side-step it especially since I'm concerned that the fault errors are a manifestation of a hardware issue (along with the noises this laptop makes). I dunno if those errors are something that would be seen if the hardware was malfunctioning, though

ah that noise is weird

I've got like 9 days left in the 30 day return period from the ebay seller, so I'm gonna have to do what I can to troubleshoot, and if there's nothing I can do I'm going to have to make an annoying decision

I don't want to have to return this. rather annoying process especially since this isn't just an envelope I can carry down to the neighborhood drop-box, and I'm pretty sure the seller doesn't have a replacement ThinkPad E14 they can send me, and I've got FreeBSD running fairly well on this thing, outside of needing wifibox and having these fault errors

but like suspend/resume? basically flawless outside of that seemingly common issue with GLX dropping hardware acceleration whenever there's a VT switch. just had to toggle the suspendswitch sysctl to sort that out, though

maybe I could keep the machine and get a partial refund for being sent malfunctioning hardware that might give out on me sooner than it otherwise would

I think I'm gonna try out 15-CURRENT w/ drm-61-kmod in a separate boot environment to see if that has any effect on the fault errors (I still don't even know what those are, tbh, or what effect they're having). I might also start pulling more hardware out of this thing to see if I can pin the noises to any particular component. clearly isn't the HDD

I'll take out the M.2 SSD, though first I'll need to clone my install over to a separate SSD which I'll probably just put in the now vacant 2.5" bay. I'm guessing I just clone the ZFS datasets over, but what about the bootloader. what's analogous to grub-install for FreeBSD's bootloader?

I could pull the wireless card and the battery as well

You'll want to use gpart(8) to partition your disk before doing zpool create and subsequently zfs sending your datasets.

Your partitioning will depend a bit on how your firmware is configured to boot, but it either involves GPT with UEFI ESP or GPT with UEFI-CSM.

zfs-send(8) will make the transfer itself quite simple, at least.

kinda regretting that it took me this long to get aboard the ZFS train tbh

tm512: do all OSes beep like that on boot?

rtprio: I've only had FreeBSD on this (well, Windows 11 was on it briefly, before I installed over it) but since the noises also occur during what seems to be POST and occasionally during the UEFI splash screen, gives me the impression that it's not related to the OS

I only booted into Win11 just to make sure the machine wasn't DOA, and I didn't get further than the part where it was demanding I bring the machine online and log in to a MS account

if I'm putting in a 2.5" SSD maybe I'll just install Linux first. hell, I could just pull the drive from my previous laptop, wouldn't even have to go through install/setup, I'd just have to configure picom to use glx instead of xrender (previous laptop is so old that it doesn't meet picom's GL requirements)

I have a super weird error on pfctl inside the jail. apparently, my pf is still working and I got this error everytime I run: # pfctl -sr

Assertion failed: (elems <= maxelems), function pf_nvuint_32_array, file /usr/src/lib/libpfctl/libpfctl.c, line 147.

spmzt: that's a bug, you should file it at bugzilla.freebsd.org

Title: FreeBSD Bugzilla Main Page

I will, thanks

lw

having a bit of a problem with IPv6. i set up addresses on freebsd and on my machine, but i can ping my machine from freebsd but not freebsd from my machine

i don't have any firewall configured in the freebsd host... wth?

hjf: if you are sure that you don't have any firewall enabled, check if you recv the icmpv6 packet from your machine on your freebsd? (and if does it reply to your machine and the machine somehow ignore it?)

dump the traffic on both hosts, check the routes

jbo

lw!

jbo: not to be rude, because i appreciate that you commit all my ports, but you could say your actual quesion rather than just saying my nick when you ping me :-P

lw, well I have no reason :p

#feelingLonely

ok tcpdump shows it receives the ARP request but doesn't respond

jbo: do you want some ports to commit so you feel less lonely?

ff02::1:ff00:2: ICMP6, neighbor solicitation, who has fd10:c467:a267:1000::2, length 32

jbo: bugs.freebsd.org/bugzilla/show_bug.cgi?id=277020 bugs.freebsd.org/bugzilla/show_bug.cgi?id=277017 bugs.freebsd.org/bugzilla/show_bug.cgi?id=276347 - these are all trivial

Title: 277017 – multimedia/libmpeg2: set LICENSE to GPLv2+

hjf: but not the icmpv6 echo?

do you have a ll address ?

ifconfig $IFACE inet6 auto_linklocal

sysctl net.inet6.icmp6.nd6_onlink_ns_rfc4861=1

Title: 263288 – IPv6 system not responding to Neighbor Solicitation

lw: done. bugs.freebsd.org/bugzilla/show_bug.cgi?id=277093

Title: 277093 – pf: Assertion failed: (elems <= maxelems), function pf_nvuint_32_array on stable/14 with RACK

Question, i have installed VirtualBox after using it, some time gui's (from firefox to thunar) start to freeze (also on the screen) for more then 2minutes

then everthing is laggy as hell, then normal again. After some time it starts to freeze again. I could trigger that espacially with qt-gui's

it's FBSD14 with the amdgpu driver.

has anyone else the same problem? When not loading the vbdrv-module everything is snappy again

lw, you're able to run factorio in wine?! :o :o :o

jbo: can't everyone do that?

lw, factorio is the only game I care about these days and the only reason why I still boot windows sometimes. but has gotten rare because booting windows.

jbo: it's working fine for me in wine-devel 9,2 (not in ports, patch in bugzilla)

i had one complete system hang but since then it seems to be ok

lw, let me get back to you on that over the weekend or something

jbo: just make sure you install d3dcompiler_47 from winetricks, wine's shader compiler can't compile Factorio's shaders

lw, testport just failed on 277020

jbo: log?

lw, local build, have to copy pasta

lw, pastebin.com/VXHaixUL

Title: tools/install-sh -m 644 include/windows.graphics.directx.direct3d11.idl /wrkdir - Pastebin.com

that's odd. i'll try testport again later

that was on a 13.2-RELEASE jail

i only support -current, but this error doesn't look release-specific

shouldn't we have a warning/error in the makefile if OS versions are restricted?

i don't think this error is related to OS release, it's just weird

that I get

lw, same error happened on 14.0-RELEASE (just as info)

i'm trying to run Flatout, it's free on gog right now. their nix version is basically just windows build packaged with old version of wine. when i try to run it, i get this error:

wine: failed to initialize: libm.so.6: failed to map segment from shared object: Cannot allocate memory

any ideas if there is a fix/workarond for this? or is it trying to do something freebsd just won't allow?

jbo: i'll have a look later, feel more in a play games mood than debug ports mood right now :-)

lw, that is legit. I'll leave a note in the PR. no stress from my side.

btw, Factorio did with with wine-devel 9.1 (which is in ports) so don't feel like you have to commit this just to play it

s/did with/did work with/

lw: what tools do you use to debug something like that where it's a program on top of wine

johnjaye: i've never had the need to debug something running under wine

ok. i don't know much about debugging on freebsd for that kind of thing.

if i did need to do that, perhaps i'd start by seeing if WinDbg works under wine

jbo: testporting wine-devel on -current now

i don't have a testport script for 14 or 13, but if it doesn't fail maybe i'll make one

lw, in general, please also test for all currently supported -release versions when submitting port updates

(in general - if it's something that only works on -current then... you know the drill).

makes committer's life easier and your submissions will land earlier :)

jbo: that's a lot more effort than you might think. like, to test mediaelch on 14 i'd need to create a whole X/Wayland environment

lw, not for testport

testport != 'test a port'

all it does it test that it builds, not that it works

hence I say testport and not test port ;)

note: if you are the _MAINTAINER_ of a port you are expected to also do runtime tests.

well, you said 'test'...

true

s/test/testport

committers don't care about runtime issues (technically). that is the maintainer's job.

i didn't see anything in the porter's handbook about this, it seems more reasonable to me to wait for users to report problems on odler releases, if there are any

which aspect are we talking about right now?

runtime testing

i'd like to help out if possible. I've used Windbg but just for really basic things.

like, i did not test mediaelch or sublime-music on 13 because i don't have a complete install of 14/13 sitting around just to do that

if that's requirement to maintian a port... i will probably just not maintain ports :-/

obviously if i was maintainer of something like Qt or glib then yes, i would do more testing and probably ask for exp-run before submitting patches

but the amount of testing required can vary based on port, right?

lw, docs.freebsd.org/en/articles/contributing/#adopt-port

Title: Contributing to FreeBSD | FreeBSD Documentation Portal

> Build, install and test your port on as many platforms and architectures as you can.

"as you can" - I know what you're going to say ;)

so you have to have virtual machines or something right to do that?

everybody has a different solution/setup for that I think

also depends the ports

jbo: i find this quite hilarious because the only port i took over maintainership of (rather than submitting as a new port) has been broken for months because no one can be bothered to commit a fix for devel/py-urwid

if it s non-GUI upstream with a built-in testing framework you can do a lot with testport

if no one else cares about this, why should i...

lw, that is not the spirit we're looking for :)

jbo: it's not the spirit i'm looking for either because i'd like my port to work

but apparently we don't care if everything that depends on urwid is just broken?

i mailed ports@ (twice) and python@ about this, no response

I cannot speak for python(@). I actually ran into an issue that would need some python patching which I wanted to ask you whether you'd be able to handle that :D

(platform detection crappery)

that aside (but let me be clear that this whole thing has really soured me on maintaining anything, tbh), the way i see it is, it's not feasible for me to test my port on every architecture (i386, amd64, arm64, armv7, riscv64) x every release (13, 14, current), but if someone reports a problem on some release, i will look ingo that and fix it is possible

i actually submitted this a month ago bugs.freebsd.org/bugzilla/show_bug.cgi?id=276347 (trivial patch to mark the port as BROKEN since it is) and not even that is committed

Title: 276347 – net-im/toot: mark as BROKEN since devel/py-urwid is broken

so, idk, what is the right spirit here

wait, you're the maintainer of it?

of net-im/toot, yes

can you update that BROKEN= message to specify the py-urwid version rather than "current"?

"Does not work with devel/py-urwid-X.Y currently in ports"

or something like that

the version of py-urwid doesn't actually matter, the issue is that setuptools_scm is broken

this is the fix: reviews.freebsd.org/D39288/new

Title: ⚙ D39288 devel/py-setuptools_scm: update to 8.0.4

but someone could fix this for py-urwid specifically by committing the fix in bug 274411

274411 – devel/py-urwid Update to 2.2.x breaks several python packages bugs.freebsd.org/bugzilla/show_bug.cgi?id=274411

I don't know the standard/requirements for BROKEN= values but personally I'd suggest putting a message in there referencing that diff or whatever. but I'll ask somebody who knows.

tbh i would rather someone would fix urwid, i don't understand how there's been a patch for this serious bug since November and no one cares

I currently don't have enough experience to mingle with python@. If it would be a "person" maintainer I wouldn't mind looking into maintainer-timeout but python@ might have different ideas for reasons I am not aware of.

i thought adding a BROKEN= tag would be a trivial change in the mean time, not something that requires a review by itself

BROKEN= does not require a review on your end but I am still under mentorship. My belt is being loosened lately but I have not been given permission to touch something like that without mentor approval yet :)

for example, the wine-devel PR I could commit without mentor review by now

lw, so you e-mailed python@ about this?

uuuf, 274411 is not even that big of a "patch"

unfortunately, I don't know the python ecosystem well enough to call a shot here. I'd try to get in touch with python@

jbo: i mailed ports@ in December lists.freebsd.org/archives/freebsd-ports/2023-December/005096.html and in January i mailed python@ cc: ports@ lists.freebsd.org/archives/freebsd-python/2024-January/005480.html

Title: devel/py-urwid and devel/py-setuptools_scm (PR# 274411)

"stymied" holy f*

new word day

the TL;DR is that I'll not interfere with python@'s business.

I can ask around tho

that's fair, i don't expect you to fix this. but you can see why this makes me laugh a bit when someone asks me to test my patches on 3 different freebsd releases, right? :-)

jbo: didn't you do stumpwm?

Demosthenex, nope. I never wrote a single line of lisp. I'm a perl guy :p

PERL!

Demosthenex, and I'm a user of bspwm

must have been another j name

i was chatting with the maintainer of stumpwm port ;]

lw, well the testing stuff should be unrelated to this, tbh. but I get your point of course. I just think it's irrelevant :p

Demosthenex: that's jrm@

Demosthenex, that would be jrm

lw, do you feel like using your anger for something helpful? :D I'd have a little something that needs fixing and I'm not a python guy.

jbo: sure, but i need more vodka + a cigarette first, feel free to type your question in the mean time :-)

lw, devel/emscripten.... upstream ships with emrun.py and that does platform detection which currently only handles Windows, Linux and MacOS. I did quick-n-dirty just default to LINUX=True yesterday but that is not proper and they apparently do gpu checks with glxinfo and all that crap.

yay!

lw, github.com/emscripten-core/emscripten/blob/main/emrun.py

Title: emscripten/emrun.py at main · emscripten-core/emscripten · GitHub

i appreciate your time packaging something so exotic ;]

the secret is that usually (!) the exotic stuff is much easier to maintain because it has less bloat :p

cgit.freebsd.org/ports/tree/www/chromium/files ....

Title: files « chromium « www - ports - FreeBSD ports tree

jbo: yeah, stumpwm should be one whole single binary file ;]

given common lisp programs are typically distributed as a image file

I did not mean to diminish jrm 's efforts of course!

The StumpWM port/package is back to a single bundled binary. For a few years I switched to a tiny package that had a few deps (SBCL and a few CL libraries), but dealing with the FASL stuff was a pain.

jbo: ah i do not know anything about emscripten

isn't that llvm thing?

lw, yeah but that's irrelevant. emrun just spawns a webserver to serve an HTML file generated by em++

lw, emscripten generates webassembly from anything that has an LLVM frontend

jbo: well, you know the Internet is only supported on three platforms nowadays (Windows, macOS, Linux), so it sounds like they did it right

(do i seem bitter about this? i am)

alright, I'm gonna stop bothering you for now :)

i would help but i know absolutely nothing about emscripten, sorry

if you have more obscure C++ questions i can probably help with that

you don't have to - but it's totally fine :)

lw, yeah, I am trying to use coroutines for the first time :D

co_<3

jbo: testport showed the same issue with emulators/wine-devel 9.2, odd, i will investigate this later

Hello , I have crash when I want to use Pluma with MATE desktop , on AMD64

what kind of crash?

lw, ack

Demosthenex: I don't know but I have a core dump

i mean did the WM crash, did X crash, did the whole kernel crash and reboot?

But I don't know to use this core file.

you'd typically open it in the corresponding debugger (lldb,gdb, ...) along with the binary (ideally with debug symbols) and start investigating :D

Demosthenex: No , when I clic on , it's await and avort and I have a coredump, Pluma.core

jbo: Sorry , I don't know this, I'm beginner on freebsd. But I want to learn to make this.

dwho: ok, so youhave X11 running, and MATE desktop up. you try to launch and application (pluma) and it coredumps

generally you need to be a C developer to read a core dump, but you should certainly try to open it and see if there's an obvious cause

Demosthenex: Yes

I read the Pluma.core with ghex but I don't to extract info.

you need a debugger, not hex dump

I haven't the knoledge to exploit core.

Title: Chapter 2. Programming Tools | FreeBSD Documentation Portal

Title: Chapter 2. Programming Tools | FreeBSD Documentation Portal

dwho: that's a place to start. also check your logs in /var/log, the messages file may have something useful

Thanks, I read this and I learn.

having a single failed app is an easy place to start

Demosthenex: Yes I read /var/log/message and I have "exited on signal 11 (core dumped)"

progress!

I learn to will have the knowledge to exploit coredump, thanks fpr help.

lw, you're running jellyfin, right?

Does Plex have hardware acceleration with FreeBSD and a Nvidia 3060 GPU?

Hello I want to use wireshark without root but I use a user with wheel group and

I modyfing /etc/devfs.conf with append own bpf* root:network and perm bpf* 0660

but my user don't capture with wireshark ?

@SponiX I believe my Plex box has acceleration with an Nvidia card. Not the 3060, though. I'll double check.

dwho, is your user a member of the "network" group?

Yes, my plex box has it enabled.

beowuff: thanks, I appreciate you checking for me. I am still on the fence about migrating from Linux to FreeBSD

The FreeBSD plex server lags behind when Plex is updated a few days, but the devs maintaining the port seem to be on top of things pretty well.

I haven't had issues.

hello

has anybody managed to compile LocalAI in freebsd ?

wow, csh might get functions, finally: tcsh-org/tcsh #77

Title: Introduce 'function' built-in. by Krush206 · Pull Request #77 · tcsh-org/tcsh · GitHub