IRC channel logs


back to list of logs

<mubarak>I searched for it and I found a answer on, I added that code to /etc/config.scm
<mubarak>and it give me this error "/etc/config.scm:55:2: error: (udev-configuration (inherit config) (rules (cons brightnessctl (udev-configuration-rules config)))): invalid field specifier"
<mubarak>I see that David made a patch for Xfce desktop
<nckx>mubarak: That error sounds like you're adding (udev-configuration ...) directly under (operating-system … <here> …).
<nckx>It should look (very) roughly like this untested fragment:
<nckx>You probably got a few typos for free. Sorry.
<mubarak>yes I added it just below services
<mubarak>nckx, after adding it it give me this error "/etc/config.scm:9:0: error: %desktop-services: invalid field specifier"
<nanounanue>Hi everyone
<mubarak_>Hi nanounanue
<nanounanue>How can I check which user defined services are running? When I run `herd status` I am getting `error: connect: /run/user/1000/shepherd/socket: No such file or directory`
<nanounanue>But if I run it with `sudo` I only see the system wide services
***malaclyps is now known as mala
***mala is now known as ma
***ma is now known as mala
***mala is now known as dob
***dob is now known as dannyob
***dannyob is now known as dannyobrien
***dannyobrien is now known as malaway
***malaway is now known as mala_
***mala_ is now known as mala
<nckx>mubarak_: The modify-services syntax isn't second-nature to me as I've never used it at home, but that random ‘%desktop-services’ popping up near the end *does* look out of place. See my example. There's nothing after it.
<nckx>nanounanue: Then you have no user-defined services running 😉
<nanounanue>nckx: (o_o)
<nckx>Guix doesn't manage them, at all, nor does it start a user Shepherd, so unless you've created your own ~/.config/shepherd and launch ‘shepherd’ somewhere in your .xsession or whatever, it's expected that there is none running.
*nckx → 😴💤
***chipb is now known as chipb_
***chipb_ is now known as chipb__
***chipb__ is now known as chipb
<mubarak_>anybody here use Mate Desktop. did you managed to increase/decrease backlight without the need to enter your password?
<vagrantc>i would guess you just need to set the appropriate permissions on files in /sys
<nanounanue>nckx: I am very tired (after fixing the inodes, and preparing reinstall, etc). You are right. I made a mistake in my `init.scm` that's why I didn't have any user defined services
<vagrantc>i use a udev rule from brightnessctl to do that
<vagrantc>but i use brightnessctl and not a GUI to adjust brightness
<mubarak_>vagrantc, I used to use Fn+F7 keys to increase brightness. After reading your post in that mailing list and using brightnessctl to adjust the light.
<mubarak_>It would be great to use laptop function keys instead
<vagrantc>hear you
<drakonis>how's xen support in guix right now?
<drakonis> oh this is good
<leoprikler>btw. since wip-emacs has been merged, can we gc it? :P
***lukedashjr is now known as luke-jr
<drgibbon>greetings all, general question here about packaging for Guix (I know next to nothing about the Nix/Guix style of doing things): How often (generally) does source need to be patched with snippets to make it work with the Guix store approach?
<drgibbon>Many things about the Guix approach sound appealing to me, but having to patch to make it build does sound a bit painful.
<drgibbon>Especially if you're just a pleb :D
<robin>drgibbon, it really depends on the program. a lot of c/c++ programs using gnu-build-system will "just work" given appropriate inputs, for example
<robin>ime the most common problems are hardcoded paths in source code (which you can sometimes fix with substitute* or a patch) and test suites that fail under guix (and for initial packaging you can simply disable the "check" phase)
<tissevert>hi guix
<drgibbon>robin: ok thanks. I'm thinking of getting into it, but there's a few packages I'd need to package (signal-desktop is one, amongst others). I'll probably check it out, thanks.
<efraim>if signal-desktop uses electron then I'd suggest using the flatpak version for now
<MysteriousSilver>Hi! - Assuming i install a package foo from guix and it's currently the latest version, is it normal that foo is re-downloaded while running `guix package --upgrade foo`?
<cbaines>MysteriousSilver, yep, that can happen when a dependency of foo has changed for example
<solene>crap, most of my patches were mangled by my mail client, I forgot to configure it after I saw it was wrong for the first mail -_-
<solene>sorry to people who had to read my mails, I've sent updates to the tickets numbers addresses ^^
<MysteriousSilver>cbaines: thanks
<mange>I'm trying to do some analysis on a system of mine, and work out the size cost of various things. I've run "guix size" on the built system, but now I'd like to run something like "guix graph" to find out what's bringing in each store item. I can't see an easy way to do that. Is it easy/possible to get a references graph for a system?
<cbaines>I'm back trying to login through the GDM to Gnome, so if anyone has any suggestions for things to delete, or places to look for information about why logging in doesn't work, I'm all ears
<efraim>does /var/lib/gdm have the right permissions? and can you log in from a TTY?
<efraim>cbaines: ^
<cbaines>so, I think the state stuff might have been a distraction
<cbaines>as well as updating Guix, I removed some deprecated bits from my config, including (gnome-desktop-service)
<cbaines>adding that back has got me back the warning, but I can also login
<cbaines>my assumption at the time was that I didn't need that bit, since I was using %desktop-services
<cbaines>but looking at the actual code, I realise that the procedure is deprecated in favour of using the service type directly
<cbaines>desktop services will only get you as far as the login manager
<efraim>interesting, quassel on guix shows overlapping "face + palm + female-sign" black and white symbols and quassel on android shows "female face palm emoji"
<paul_j>/msg NickServ VERIFY REGISTER paul_j D8qOtJE6TR9YTUHv
<cbaines>hexchat for me just doesn't show those characters very well
<cbaines>I have to copy things in to emacs and ask it what the characters are sometimes
<paul_j>good morning guix!
<cbaines>morning o/
<cbaines>you should change your NickServ password paul_j
<paul_j>Thanks wingo for the pointer here. cbaines - I have made a new account on with a brand new secure password. I now need to see which other channels have jumped here from freenode...
<paul_j>I was paulj there, but that nick is already taken on
<stikonas>you can connect to all the same channels... Some channels that have not officially jumped also has some activity
<paul_j>Thats good to know - thank you!
<paul_j>I notice the message at the header is still "1.3.0 is out!..." - who can update it?
<nckx>paul_j: Update to what?
<paul_j>2.0 released...?
<paul_j>Sorry - ignore me!!!!! I am mixing nyxt up with guix... I'll just crawl back into my corner and shut up!!!
<paul_j>Is it normal for stuff in the store to be dated 1/1/1970? Or is this an indication I may have an issue?
<solene>it's normal, it's to have reproducible builds, so they are not dependent on time if they all have the same time
<paul_j>Thanks - I did have an idea that might be the case.
<dstolfa>hmm, guix strongswan is misconfigured
<tversteeg>Hi all, I have a question regarding the packaging of fonts, is this the right place to ask?
<catonano>efraim: This is Polari on the Ubuntu desktop
<catonano>tversteeg: yes, this is the right place to ask. There's also the mailing list
<tversteeg>Great, thanks. My question is how font packages should be/are installed? I've created a package for a nerd font (, and it installs and I can access it in `~/.guix-profile/share/fonts`, but none of my applications detect the font. What could cause this?
<tversteeg>I've based the package on, the only thing I've added is the removal of some redundant files
<dstolfa>does guix aim for its packages to be minimal or cater to the most usecases as possible by default?
<dstolfa>asking because the strongswan build template is as minimal it gets, to the point where it's unusable
<dstolfa>i have to change it to get my VPN working, but if the preferred is "enable as much as possible so that most configurations work", then i can do that
<dstolfa>what is the preferred way to install packages that are compiled locally with a modified .scm file (this one is a part of networking.scm)?
<dstolfa>i want to test it, but it requires root
<dstolfa>i've been digging through docs but i think i'm missing something probably obvious...
<parnikkapore>Hi, where is the "/etc/config.scm
<parnikkapore>" file in the QEMU image of Guix System?
<parnikkapore>I can't find it at that location.
<roptat>parnikkapore, maybe /run/current-system/configuration.scm
<roptat>dstolfa, whatever option you used with guix build to build that package, you can use with guix install to install it
<dstolfa>roptat: so i followed the online documentation with guix environment --pure guix, built everything, then i modified the package and installed it. however, in order to test the package i need to run it as root, rather than as a user
<dstolfa>installing sudo causes weird setuid problems
<dstolfa>so i'm not sure how to test something as root
<roptat>you might be able to run the package directly from its store path, without installing it
<dstolfa>hm, let me try
<roptat>it doesn't always work, especially if there are propagated inputs, but it might be enough for a quick test
<dstolfa>yeah things are a bit weird
<dstolfa>roptat: so for context, i'm basically trying to make strongswan usable on guix (the default options in strongswan from upstream are minimal to the point where it won't connect to almost anything)
<dstolfa>the way guix does it right now is that in networking.scm there's a definition for a strongswan package and it relies on upstream flags to be sensible
<dstolfa>unfortunately they aren't
<dstolfa>so all it does is --disable-blowfish and --disable-des, but you need to actually manually enable about half of the options in strongswan for it to connect to any "common" VPN setup
<dstolfa>for example, --enable-md4 --enable-eap-mschapv2 --enable-eap-identity to connect to a very common eap-mschapv2 based setup
<dstolfa>and the only way i can test this is by providing my actual VPN config file and run this as root (the former is not an issue, the latter is proving to be a bit of an issue)
<parnikkapore>roptat: Thanks, found it! One more thing: I'm trying to change the resolution of the VM's display, but it keeps resetting itself. How should I change it?
<roptat>parnikkapore, no idea, sorry
<parnikkapore>roptat: It's OK, thanks!
<bone-baboon>How can I find out if an entire project is "GPL 3+"? The example I am looking at coreutils. The coreutils repository has a COPYING file that says it is using GPL version 3 <>. Looking at an individual source code file I can see that it says "either version 3 of the License, or (at your option) any later version."
<bone-baboon> <>.
<roptat>then it's gpl3+, I think the default text of COPYING says that too
<dstolfa>so... i've got a patch for a package, but i still am unable to test it as root (all tests pass though)
<dstolfa>am i just dumb for not being able to find this, or is it actually difficult
<solene>dstolfa: that may sound stupid, but does it work if you run it with sudo?
<ruffni>is it possible that the handy hacker-text-tool `ack` is not yet packaged for guix yet? it's kinda hard to search for...
<dstolfa>solene: sadly no, it was the first thing i tried :-(
<dstolfa>i run into issues with setuid
<roptat>is it the software you package that tries to run setuid?
<jorge-65>Hola,tengo un error guix pull:error git: error inflating zlib stream
<dstolfa>nope, the software itself needs to create (and connect to) a socket which is root only roptat
<dstolfa>so it needs to be run as root, and one attempt of that was with sudo, but sudo needed to setuid
<roptat>ah, are you on the guix system?
<roptat>you can't install programs that need setuid as user (or root for that matters), you need to add them to your os config, in setuid-programs
<bone-baboon>roptat: Thanks
<roptat>like this:
<dstolfa>right, but how would i do that for a package i've modified and built locally in a guix environment?
<roptat>it contains sudo by default
<dstolfa>basically the situation i'm in is one where i've modified a package in my local git checkout, built it and i can install it in that environment, but i can't actually do a test run of it outside of the pre-defined tests because it complains about needing to be root
<roptat>does this package need to be setuid, or only sudo?
<dstolfa>only sudo
<dstolfa>the program itself just assumes it will be run as root and doesn't do any setuid magic
<roptat>then you need to make sure it's part of the setuid-programs, reconfigure and then you should be able to run the program with sudo
<roptat>(and remove sudo from your user's profile)
<roptat>I guess you can also create a VM with that program inside, then you won't need root to run the VM, and can be root inside
<roptat>but yeah, it's not easy to run that software as root if you can use sudo
<dstolfa>roptat: is there maybe a way to install the software from the local git checkout?
<roptat>maybe as root, cd to the checkout, and ./pre-inst-env guix install the-software?
<dstolfa>hmm, let me try
<dstolfa>oh that falls over really badly
<dstolfa>roptat: oh. i'm stupid. it's actually really simple
<dstolfa>basically su as root, guix environment --pure guix and use the pre-install-env to install it to that profile
<dstolfa>then you can just run it as root itself
<ruffni>what does "no updater for foo" mean in `guix lint foo`?
<thorwil>hi! i though after a `guix environment fftw` that dependency should be dealt with, but i get a /media/hd/src/eq10q-2.2/eq4qs.c:29:10: fatal error: fftw3.h: No such file or directory
<thorwil>what am i missing here, regarding header files?
<zap>thorwil: hello. 'guix environment fftw' will create environment with all deps needed for building fftw not fftw package itself
<zap>you need to invoke it with magic argument --ad-hoc
<zap>like so "guix environment --ad-hoc fftw" -- everithing that goes after --ad-hoc added as package to your environment
<zap>/s/to your environment/to environment you about to switch to/
<thorwil>zap: thanks, but i still get `fftw3.h: No such file or directory`
<thorwil>actually, i wanted to build eq10q directly as guix package. managed to write the package description well enough for it to get going, but unfortunately, it failed with `/tmp/guix-build-eq10q-2.2.drv-0/eq10q-2.2/gui/bassup_ui.cpp:41:49: error: ‘_LV2UI_Descriptor’ does not name a type; did you mean ‘LV2UI_Descriptor’?`
<ecbrown>apteryx: i've posted patches to bug list that update qt6.
<ecbrown>if i had better patch-fu i should have applied your patches first, i am sorry about this. maybe this can be retro-patched since i tried to be robotic about it. :-)
<thorwil>so it appears eq10q has to be patched. before figuring out how to do that in a guix package, i wanted to test it in an old fashioned build. except, it would be preferable to build againts guix libs, not what ubuntu provides ...
<roptat>ruffni, the "updater" is used by guix refresh, it can be used to tell you if a package is up to date or not, give you the latest version, and help updating the package definition
***TheCreeper1 is now known as TheCreeper
<roptat>for instance, there's an updater for github that checks release tags, an updater for gnu software that check, etc
<zap>thorwil: maybe export C_INCLUDE_PATH="$GUIX_ENVIRONMENT/include/fftw3/include:$C_INCLUDE_PATH"?
<ruffni>roptat: thanks! could you hint me to the appropriate section in the docs?
<roptat>see the section of --type
<thorwil>zap: that’s not it, either. maybe i should skip right ahead to figuring out the patching business
<zap>thorwil: oops I meant C_INCLUDE_PATH="$GUIX_ENVIRONMENT/include:$C_INCLUDE_PATH"
<zap>thorwil: yep its a good idea :)
<thorwil>zap: works ... to the next error. thanks!
<zap>lfam: hellou
<zap>hey roptat, do you have frama-c package published anywhere?
<roptat>zap, yes on the bug tracker
<zap>oh cool
<roptat>see the latest message:
<roptat>oh it's been almost a month now, I'll push it
<dstolfa>hum, just making sure... submitting a patch is done with sending git format-patch to guix-patches@ right?
<dstolfa>i got an ACK, but it's not on the archive
<dstolfa>(been about 30 minutes now)
<roptat>dstolfa, yes
<dstolfa>alright, great. thanks roptat
<zap>roptat: thanks for working on it!
<ruffni>ok, so how do i link my package to an updater?
<roptat>there needs to be an updater that makes sense for it
<ruffni>there is: it's a github repo
<roptat>then, the updater needs to be able to recognize your package as something it can update
<roptat>for github, you need to use the git repository, not download a tarball from github, I think
<ruffni>do i need to add the version variable to the commit tag?
<roptat>ah maybe
<ruffni>nope, that wasn't it
<roptat>are you sure you're using https and not http?
<roptat>looks like the updater is looking for "" specifically
<ruffni>the url looks like: "git://"
<roptat>github supports http/https clones
<roptat>so, just use https instead
<ruffni>that's it! thanks!
<roptat>this is a bit silly, but we could have a linter warning for git:// urls
<lfam>roptat: I have a patch on the mailing list for that :)
<lfam>It warns about all uses of the Git protocol
<roptat>ah, great :)
*lfam searches for it
<lfam> <>
<lfam>The git:// protocol has no advantages over HTTP and is less likely to work on restrictive networks (corporate / enterprise etc)
<lfam>I guess it's waiting for somebody to write a test case
<lfam>Help wanted :)
<raghavgururajan>Hello Guix!
<raghavgururajan>cbaines: you around?
<thorwil>i managed to patch the lv2 descriptor issue myself and found a patch for "‘pow10’ was not declared in this scope". opensuse has another patch for eq10q:
<thorwil>can anyone familiar with build flags tell me if that patch should be included, too?
<Allan>Hello, I have a very interesting problem whenever I run guix pull. It computes the hash for about a minute and a half before it logs me out. As a result I can't update any of my packages. What might be happening?
<Allan>Uploaded file:
<Allan>Here is a video of it
<nckx>Welcome Allan. To resume: yow! That's bad. Anything appear in /var/log/messages when that happens?
<Allan>I have not checked the log messages yet. I'll try it again later when I get home. I did try running "guix pull
<Allan> "guix pull | tee guix_pull.txt" to see if anything would come up but the file remained empty
<nckx>Try ‘guix pull 2>&1 | tee guix_pull.txt’ or errors won't be logged.
<nckx>(So no Matrix bridge yet? Pity.)
<Allan>So far, none that I've seen. Can't wait til its up. I will try the command later today though, hopefully I can get a better idea of why its happening
<drakonis>nckx: soon
*dstolfa thinks he has everything accounted for to run guix on bare metal if the strongswan patch gets accepted
<Noisytoot>nckx, What character is that ("🦇")? I do not have the correct font
<nckx>It's a spoooky bat.
<nckx>dstolfa: It's worrying that ‘sane strongswan default’ are so far from the actual strongswan defaults‌… :-/
<dstolfa>nckx: they are unfortunately :(
<dstolfa>actual strongswan defaults is basically no plugins
<dstolfa>so no auth works
<nckx>Two side nitpicks: all comments can go (except maybe the last one: I wouldn't gather that it merely sets a default, not the only choice) and keep inputs alphabetical.
<nckx>dstolfa: ‘That is suboptimal.’
<drakonis>speaking of which
<drakonis>once there's parametric packages, it'll be possible to set up those flags, yeah?
<dstolfa>also nckx, just a question about AESNI: by default, strongswan doesn't enable it, but a lot of distros do for side channel concerns. however, i would argue that AESNI is basically non-free software and should be explicitly disabled in guix for something like an ipsec-based VPN?
<dstolfa>would you agree?
<nckx>dstolfa: Leaving aside that… interesting discussion (really!): we simply shouldn't enable it for compatibility with older CPUs, anyway.
<dstolfa>yeah, that's a fair point about older CPUs
*nckx disables CONFIG_RANDOM_TRUST_CPU in their kernel too.
<nckx>Strongswan supporting both instruction sets in one binary *and* giving the user that choice would be ideal.
<nckx>Do you know if it can?
<dstolfa>hmm, the only thing i could find is an autoconf option, nothing at runtime
<nckx>That's plausible.
<dstolfa>nckx: strongswan is something i use on a daily basis so i follow their stuff (as a user) quite closely, as i literally can't do my work without it. if at some point i notice it creeping in, i will happily change the package to do that
<nckx>Sure, I trust and defer to your judgment as a user.
<nckx>It's the Internet so y is out of the question.
<dstolfa>anyway, let me fix up the stuff you've said and send it off. for subsequent diffs, do i just email the ACK i got on the first email with the output of git format-patch?
<dstolfa>nckx: heh, never trust the internet!
<nckx>dstolfa: Yup. To: Don't send a diff to the diff (it happens). I hope another Strongswan user reviews it.
<lfam>I don't use IPSEC so don't care very much, but it's going pretty far to claim that CPU instructions are not free software
<dstolfa>nckx: yeah, unfortunately i can only base the patch on how i use strongswan and running the test
<lfam>They aren't software in any meaningful sense
<nckx>Great. Let's finally ship microcode in Guix.
<lfam>I don't think you need to use microcode to use AESNI?
<nckx>You need to use microcode to use CPU.
<Noisytoot>nckx, No, it's nonfree!
<dstolfa>lfam: well this was just a thought which mostly originated from the fact that it's replacing a part of free software with a bit of a black box that claims to do the right thing under you
<raghavgururajan>cbaines: Oops, missed your message. Was substitute service disabled on bayfront?
<nckx>Noisytoot: That was m'point.
<lfam>Well, yeah, but you don't need to do the "bad" thing of updating the microcode. The computer comes with it
<efraim>Wow, Debian has about 35 configure flags for strongswan
<lfam>I mean, if AESNI only "claimed" to do the right thing, then it wouldn't work
<nckx>‘CPU instructions’ are software now and have been for ages. Your x86 doesn't run in hardware.
<raghavgururajan>Folks, is there a way to use multiple values for `#:import-path` in go-build-system?
<lfam>In reality, it works
<Noisytoot>nckx, That depends on the architecture, in ARM they are hardware, in x86 they are hardware
<nckx>Default firmware good, upgrades bad, standard FSF line and we toe it.
<lfam>I just think it's reaching very far to claim that it isn't "free" and thus we shouldn't use it
<cbaines>raghavgururajan, substitutes from bayfront have changed, but it should be a lot better now
<nckx>Noisytoot: <in x86 they are hardware> No. Less familiar with ARM.
<Noisytoot>nckx, s/hardware/software/ for x86
<cbaines>at least for x86_64-linux, bayfront had more substitutes available for recent revisions than
<efraim>raghavgururajan: here's what I did
<dstolfa>lfam: i mean i'm not opposed to enabling it, but it's probably best kept off (for now) due to old CPUs and defaults by strongswan as nckx mentioned
<dstolfa>(hence why i asked for some thoughts here)
<lfam>Can't the software adapt at runtime?
<nckx>Apparently not lfam.
<dstolfa>not that i'm aware of, i haven't been able to find anything on it
<dstolfa>though, maybe `--enable-aesni` means "enable if possible and adapt if cpu doesn't support it"?
<lfam>That would be typical
<dstolfa>let me take a look at the code around that option real quick
<Noisytoot>So will it not run on computers that don't support AES-NI?
<raghavgururajan>cbaines: I see. I'll retry and get back to you.
<dstolfa>Noisytoot: i'm trying to figure out if it has a fallback in case AES-NI isn't a thing
<raghavgururajan>efraim: Thanks!
<dstolfa>if it doesn't, then it should stay disabled. if it does, we might as well enable it
<Noisytoot>dstolfa, You could compile it with AES-NI support and try to run it on a computer without AES-NI support
<dstolfa>i don't have such a computer :(
<drakonis>if you have an intel as old as 2010, it supports aesni
<Noisytoot>My computer doesn't
<Noisytoot>It's a ThinkPad X200
<dstolfa>do we have a volunteer?
<drakonis>i have that :V
<drakonis>i will volunteer
<cbaines>raghavgururajan, note that the substitutes bayfront is now serving aren't coming through guix publish, so substituting arbitrary store items won't work
<nckx>cbaines: That's very unfortunate but I'm glad it's at least confirmed. Can it be fixed?
<drakonis>i'm installing strongswan atm
<nckx>You should ideally use dstolfa's binary. But if your binary is identical, that's good too.
<drakonis>actually, are we looking for someone to test what happens with or without aesni?
<drakonis>i have a computer with support for it
<drakonis>so let's see
<nckx>What happens when a binary built --with is run on a CPU without, with the caveat that --with might silently omit the instructions when built on a CPU without them.
<nckx>That would be stupid but...
<Noisytoot>What is the hash of dstolfa's binary?
<dstolfa>i think running test would be sufficient, right?
<dstolfa>at least on a computer with no AES-NI
<drakonis>run it on a VM maybe?
<cbaines>nckx, guix publish could be run on a different port, or behind nginx through a different domain name. Not using the name for general substitutes going forward, and instead using something machine independent has come up so once a name is settled on for the general substitutes could be served through that, and can go back to pointing at guix
<cbaines>publish on bayfront
<drakonis>tell the vm that the cpu doesnt support aesni
<dstolfa>but, from what i can see is that this is being built as a plugin. strongswan plugins are loaded at runtime, and i'm looking to see if the loading code does a check at load time if AES-NI is supported, and if so simply doesn't load itself
<nckx>Whoo, thanks for the lengthy answer cbaines. Will read that.
<dstolfa>okay, confirmed
<dstolfa>nckx: i think we can enable it
***jess_ is now known as jess
<nckx>User control and old-CPU compatibility, life is good.
<dstolfa>i'll send another patch with the things you said addressed and aesni compiled in as a plugin
***jpoiret0 is now known as jpoiret
<nckx>dstolfa: Thanks!
***ulfvonbe` is now known as ulfvonbelow
<Noisytoot>Are there any 100% free systems with CPUs that support AES-NI?
***irfus- is now known as irfus
***Server sets mode: +Ccntz
<dstolfa>nckx: i... think i sent it in?
<dstolfa>and it should be formatted alphabetically this time (hopefully)
<nckx>I don't see it in the list interface. So either it hasn't arrived yet or it already has.
*nckx .oO yep those are the two options.
<dstolfa>nckx: i guess it takes some time? i haven't used debbugs before
<nckx>Yup there it is. And approved.
<nckx>dstolfa: You can expect everything from instant delivery to 8-hour delays depending on the mood of the moonmen ¯\_(ツ)_/¯
<dstolfa>heh :)
<drakonis>oh, hm, gzdoom works fine if i use an environment instead of installing it to my profile
<drakonis>in its full glory
<drakonis>who'd think.
<dstolfa>nckx: if there's anything else please just ping me and i'll address it ASAP :)
<jorge-34>Hola, tengo el siguiente error guix pull:error git: error inflating zlib stream
<drakonis>except of course i had to change its settings to find the required data it needs to work
<drakonis>will fix though
<pineapples>drakonis: gzdoom problems, too? I tried running it today but it'd not launch for me
<pineapples>(what a coincidence, though)
<nckx>drakonis: That is decidedly odd.
<nckx>Have you identified the baddie?
*nckx is on a minetest binge personally.
<nckx>Hola jorge-34. It's possible that a network interruption caused an incomplete stream (so the zlib decompressor complains). Try again?
<solene>for gzdoom users, I recommend brutal doom and brutal hexen mods :D
<drakonis>yeah i have
<drakonis>solene: oh please, there's a lot of nice mods
<drakonis>a whole world to see!
<pineapples>drakonis: How do I work around it? I'm impatient, and I'd like to play it now :')
<drakonis>okay so
<solene>drakonis: please, introduce me to this world :D
<drakonis>the current workaround is to add to gzdoom's configs the path to the missing folders that it is expecting on /usr/share
<drakonis>point to .guix-profile/share
<drakonis>there should be 8 lines for soundfonts and fm_banks
<solene>I have a gaming peertube channel, I would like to make more gzdoom videos ^^
<drakonis>it doesnt see these and its crashing it right now
<drakonis>i'm going to patch it up to install them to the same directory as the binary
<drakonis>the zdoom forums has quite a lot of fun mods and doomworld should net you a decent amount of maps
<solene>good night dear guix community
<solene>drakonis: I'll take a look tomorrow
<dstolfa>solene: good night!
<drakonis>good night
<drakonis>it is a temporary fix mind you, just while i update gzdoom and fix the definition
<nckx>o/ solene.
<zap>do I need to to delete no longer needed patches when updating a package?
<NieDzejkob_>yeah, that's what you're supposed to do. Make sure to delete them from the makefile's list
<zap>NieDzejkob_: thanks I completely forgot about makefile
<drakonis>this assumes you have it in your profile btw
<pineapples>drakonis: Where do I find `fm_banks'?
<drakonis>same directory you'd find soundfonts on
<nckx>dstolfa: If nobody actually using SS reviews your patch in ~2 weeks, ping me up.
<dstolfa>nckx: alright, will do. thanks :)
<drakonis> for reference
<pineapples>drakonis: I can't find it in either `/gnu/store/gzdoom/share/*' or `/gnu/store/fluid-3/share/*'
<drakonis>its in share/gzdoom i think?
<drakonis>if you look inside .config/gzdoom.ini you should be able to find the default patch these files live in
<drakonis>okay so its still missing
<drakonis>i think i have it because i installed 4.5.0 and cleaned up the definition
<pineapples>This may explain it
<pineapples>drakonis: By the by, why 4.5.0 specifically?
<drakonis>new version
<pineapples>4.6.0 is out as of the last 2 days
<drakonis>4.6.0 is the current one
<drakonis>i did write it before the new update
<pineapples>oh, hahah, I see