IRC channel logs

<catonano>quiliro: what does happen exactly with those faulty characters in the xfce terminal ?

<quiliro>it does not even write them as other characters

<quiliro>catonano: it shows "arg0" on the bottom of the terminal window

<quiliro>when trying to start gnome-terminal, it does not start

<quiliro>this is the result of starting it from xfce4-terminal

<quiliro>$ gnome-terminal

<quiliro>(process:5006): Gtk-WARNING **: Locale not supported by C library.

<quiliro>Error constructing proxy for org.gnome.Terminal:/org/gnome/Terminal/Factory0: Error calling StartServiceByName for org.gnome.Terminal: GDBus.Error:org.freedesktop.DBus.Error.Spawn.ChildExited: Process org.gnome.Terminal exited with status 9

<quiliro>why is the locale not supported by the C library?

<catonano>ah, so yo're hahving a locale problem

<quiliro>catonano: how to solve it?

<catonano>I don't know, but i would take a look at the chapter in the manual

<catonano>thhere's a chapter about locales

<catonano>quiliro: then I'd write to the bugs mailing list

<catonano>quiliro: bug-guix@gnu.org

<catonano>or maybe to the help mailing list

<catonano>quiliro: help-guix@gnu.org

<mekeor>what's the value of your $LANG variable, quiliro ?

<quiliro>mekeor: ~$ echo $LANG

<quiliro>es_EC.UTF-8

<mekeor>can you try something like this in shell? LANG="en_US.UTF-8" gnome-terminal

<mekeor>do you have a (locale "es_EC.UTF-8") statement in your config.scm, quiliro ?

<lfam>efraim: I'm having trouble using sneek to save info about core-updates!

<lfam>sneek: what is core-updates?

<sneek>Its been said that core-updates is fontconfig, eudev and cups-minimal FTBFS, gnutls lacks IDN support unless it uses libidn2

<lfam>sneek: what is core-updates

<sneek>Core-updates is fontconfig, eudev and cups-minimal FTBFS, tcsh's test hangs

<quiliro>quiliro@portatil ~$ LANG="en_US.UTF-8"

<quiliro>quiliro@portatil ~$ echo $LANG

<quiliro>en_US.UTF-8

<quiliro>mekeor: i do ave that in config.scm

<methalo>quiliro: you use GNOME?

<quiliro>i have both but rather have xfce

<quiliro>i cannot enter gnome-terminal

<methalo>gnome-terminal crash?

<mekeor>you could open gnome-terminal with "LANG=en_US gnome-terminal"

<mekeor>i wonder if es_EC is just not supported by gnome-terminal or whether it's a Guix(SD) bug

<quiliro>same error as before

<mekeor>what about: LANG="en_US.UTF-8" gnome-terminal

<quiliro>$ LANG=en_US gnome-terminal

<quiliro>(process:5624): Gtk-WARNING **: Locale not supported by C library

<quiliro>same

<quiliro>$ LANG="en_US.UTF-8" gnome-terminal

<quiliro>(process:5654): Gtk-WARNING **: Locale not supported by C library.

<mekeor>when i run `locale -a` (which i think lists all supported locales), "es_EC" is not listed. only these spanish locales are listed: es_AR.utf8, es_CL.utf8, es_ES.utf8, es_MX.utf8. so, maybe you could try to reconfigure your system with one of those locales.

<quiliro>-bash: locale: command not found

<quiliro>i replaced es_EC.UTF-8 with es_ES.utf8 in config.scm

<lfam>efraim: freetype 2.7.1 broke fontconfig :)

<mekeor>i think the command `locale` is part of the package `gcc-toolchain`

<mekeor>but you don't neccessarily need it

<quiliro>mekeor: how to tell in what package a comman is?

<quiliro>s/comman/command/

<mekeor>i did it using `whereis` and `readlink`. but i don't know how to do that if the command isn't installed already. → http://sprunge.us/GHDI

<mekeor>i could see "gcc-toolchain" in output of readlink...

<quiliro>cool!

<quiliro>why:

<quiliro># realpath /var/guix/profiles/system/profile/bin/guix

<quiliro>/gnu/store/c77gfhkmnwjvgmnsafjxf1g9kkpiilbh-guix-0.12.0-4.d9da/bin/guix

<quiliro>and:

<quiliro># realpath /proc/$(pgrep guix-daemon | head -n1)/exe

<quiliro>/gnu/store/9nw4zglfl1dc8zc00rikkvln4kgjify9-guix-0.11.0/bin/guix-daemon

<lfam>quiliro: How did you start the daemon?

<quiliro>which guix version do i have 0.12 or 0.11?

<quiliro>herd restart guix-daemon

<quiliro>Service guix-daemon has been stopped.

<quiliro>Service guix-daemon has been started.

<quiliro>i have not done anything

<quiliro>it is automatic in guixsd...but just in case, i restarted it

<lfam>quiliro: What does `guix version` say? Run it as root

<lfam>Or, guix --versoin

<lfam>`guix --version`

<quiliro>is it supposed to be es_EC or ES_EC

<quiliro>?

<quiliro>uix (GNU Guix) 20170227.00

<quiliro>it is tomorow's version!

<lfam>How about ` realpath /run/current-system/profile/bin/guix-daemon`?

<lfam>On GuixSD, the guix-daemon is a "system" package, not a user package owned by root. You update the guix-daemon as root with `guix pull && guix system reconfigure $config.scm`.

<quiliro>i wonder why but guix system reconfigure works now

<quiliro>it did not work before

<quiliro>perhaps it will give an error

<quiliro>i will wait until it finishes

<quiliro>perhaps a newer guix version was downloaded today than was yesterday

<quiliro>substitute:

<quiliro>guix system: error: build failed: substituter `substitute' died unexpectedly

***Steap_ is now known as Steap

<efraim>sneek: what is core-updates

<sneek>Its been said that core-updates is fontconfig, eudev and cups-minimal FTBFS, tcsh's test hangs, util-linux needs a fix for CVE-2017-2616

***orly_owl_ is now known as orly_owl

<user10001012>I'm unable to install any packages, running the command ‘guix package -i $SOME_PACKAGE’ returns: guix package: error: unsupported manifest format

<user10001012>What's going on?

<user10001012>I'm running GuixSD.

<user10001012>Well, rolling back package generations seemed to fix it.

<user10001012>Now a new issue. I can't list generations from guix.el: down-list: Scan error: "Unbalanced parentheses", 1, 1

<user10001012>That's from M-x guix-generations

<jmd>How's the core-updates merge progressing?

<efraim>we've cleared all the issues that I know of

<jmd>That's good to hear.

<mvb>I think we should get the fix for #25873 and start the "core" evaluation again

<mvb>and then do a full evalation once #25177 is merged

<jmd>It sounds like there is at least a week to wait then ...

<catonano>mvb: as for #25873 I can't find it in the debbugs buffer

<catonano>mvb: where is it ?

<mvb>catonano: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=25873

<mvb>efraim: mind glancing over that?

<catonano>mvb: thhanks. I was looking fofr it among bugs, not patches

<efraim>I have my aarch64 bootstrap tarballs. If manolis says that it works for Hurd that's enough for me. Also being in make-bootstrap it shouldn't affect the rest of everything

<mvb>oh, good point

<quiliro>i have been 6 days with this guix system reconfigure problem

<quiliro>due to my old guix 0.12

<quiliro>due to my old guix 0.11

<catonano>quiliro: I would consider to install it from scratch

<quiliro>nooooo!

<thomasd>quiliro: did you try this fix rekado pasted? http://paste.lisp.org/display/339572

<quiliro>thomasd: sed -i "s|ExecStart=.*|ExecStart=/root/.guix-profile/bin/guix-daemon --build-users-group=guixbuild" /etc/systemd/system/guix-daemon.service

<quiliro>thomasd: will not work

<quiliro>thomasd: sed: -e expresión #1, carácter 90: orden `s' sin terminar

<quiliro>command `s' without end

<thomasd>quiliro: i, see there was a missing “|” in the sed command. Try this: sed -i "s|ExecStart=.*|ExecStart=/root/.guix-profile/bin/guix-daemon --build-users-group=guixbuild|" /etc/systemd/system/guix-daemon.service

<thomasd>(so, same command as in the paste, but with a closing “|” after group=guixbuild)

<quiliro>thomasd: /etc/systemd/system/guix-daemon.service does not exist

<thomasd>ah, you are using GuixSD, not Guix on another distribution?

<quiliro>i use GuixSD

<thomasd>but then it should work by reconfigureing after the pull of revision afd532337 (as root)

<thomasd>mew

<thomasd>(that should've been M-x mew :-) )

<quiliro>thomasd: it is working now on: guix system reconfigure /etc/config.scm

<thomasd>quiliro: ok, great! When that's done, you should be able to pull to a more recent version.

<rekado>wait, Leo already pushed a fix for this guix pull problem.

<rekado>so all you’d need to do is run “guix pull”

<thomasd>ha :-)

<quiliro>i ran the commands you told me:

<quiliro>guix pull --url=http://git.savannah.gnu.org/cgit/guix.git/snapshot/afd532337.tar.gz

<quiliro>guix system reconfigure /etc/config.scm

<quiliro>herd restart guix-daemon

<quiliro>with no errors

<quiliro>thomasd: now i am running

<quiliro>guix pull

<rekado>the afd53… commit is the commit right before the commit that broke “guix pull”.

<rekado>current master includes a commit that fixes “guix pull” again, so at this point you no longer need to update to that particular commit.

<quiliro>rekado: should i suspend guix pull ?

<quiliro>ping

<quiliro>realpath /proc/$(pgrep guix-daemon | head -n1)/exe

<quiliro>gives me

<quiliro>/gnu/store/9nw4zglfl1dc8zc00rikkvln4kgjify9-guix-0.11.0/bin/guix-daemon

<quiliro>but

<quiliro>realpath /var/guix/profiles/system/profile/bin/guix

<quiliro>gives me

<quiliro>/gnu/store/c77gfhkmnwjvgmnsafjxf1g9kkpiilbh-guix-0.12.0-4.d9da/bin/guix

<thomasd>(quiliro: I think at this point it doesn't matter, you can let guix pull run, just that the previous steps I told you to do wouldn't have been necessary anymore)

<quiliro>guix --version

<quiliro>gives:

<quiliro>guix (GNU Guix) 20170227.13

<thomasd>strange that the daemon wasn't updated after pullin adf53.. and reconfiguring... but anyway, as rekado said, now running guix pull should work again?

<quiliro>thomasd: it is not updated

<quiliro>?

<thomasd>the fact that it says ...-guix-0.11.0/... makes me think it wasn't? But anyway, did your last “guix pull” complete successfully?

<rekado_>quiliro: “guix system reconfigure” has to be run as root, so you need to make sure that “guix pull” is run for the root user.

<rekado_>ACTION has network problems

<quiliro>rekado_: i did reconfigure after guix pull --url=http://git.savannah.gnu.org/cgit/guix.git/snapshot/afd532337.tar.gz

<quiliro>but not after the simple guix pull

<quiliro>if they do the same job, then why is my guix ...-guix-0.11.0/...

<quiliro>?

<quiliro>now i am running the simple guix pull

<quiliro>sorry

<quiliro>now i am running reconfigure after the simple guix pull

<thomasd>but did you run guix pull as root? e.g. with “sudo guix pull”?

<quiliro>thomasd: not with sudo but with root login

<quiliro>reconfigure is downloading linux-libre-4.10-gnu.tar.xz now

<quiliro>7 minutes and 27 MB

<quiliro>my connection is very slow

<quiliro>64KiB/s

<rekado_>this means it will build the kernel

<rekado_>that’s going to take a long time.

<rekado_>after guix pull your guix version should not be 0.11.0

<rekado_>but 0.12.0

<catonano>quiliro: if you want to avoid buiding a new kernel, you should add ths line to your config file "(kernel linux-libre-4.9)"

<thomasd>catonano: I'm not sure it will help, because the 4.9 kernel was also updated 6 hours ago :-)

<catonano>ah :-/

<thomasd>anyway, depending on your machine, it doens't have take *that* long (I once let it build the guix kernel package, took ~1hr)

<quiliro>it would take a long time to build because i use a core duo processor

<quiliro>is there a way to use only substitutes?

<thomasd>quiliro: by pulling a version which is completely evaluated on hydra

<quiliro>download of linux Libre stopped at 38.4MiB

<quiliro>it is now on 25 MB

<quiliro>thomasd: how to do that?

<rekado_>quiliro: e.g. by “guix pull --url=…” with a somewhat older commit .

<thomasd>...and you can see which commits are evaluated at http://hydra.gnu.org/jobset/gnu/master

<nliadm>is there a way to turn off grafting for the guix-daemon?

<rekado_>nliadm: you can disable grafts for almost any command with “--no-grafts”; I don’t recommend making this a default option, though.

<rekado_>grafts are used for efficient security updates

<nliadm>right, but I don't want to run every command with --no-grafts, and packages I want are broken with grafts

<nliadm>I'd rather recompile and have things work

<rekado_>nliadm: is there a bug report for the packages that are broken with grafts?

<nliadm>yeah, it's the 'go incremental' report

<rekado_>ah

<ng0>I don't remember, did I ever sent in the urxvt,urxvtc,urxvtd .desktop patch or is it just in my local version here? Do we have this in master?

<efraim>its in the ML somewhere

<ng0>ah :)

<efraim>hasn't been commented on i believe

<efraim>:( sorry about that

<ng0>no problem :)

<ng0>in the past years I've had customized kernels with build-in modules rather than what guix does.. so with the kernel in guix, if something breaks I could just plug the harddrive and insert it somewhere else and everything including the store should work, right

<efraim>wingo: I don't know if you're worried about the guile-next failures of mips and armhf on guix, I tested 2.1.5 which is in core-updates and it built in about 6 hours

<wingo>hi

<wingo>i haven't been tracking them, no

<wingo>do you have a link to a failing build log?

<wingo>i can make a new guile release at some point

<efraim>nope, i think normally it was timeouts

<efraim>i haven't looked at it for forever, but might've just been a timeout in ice-9 or one of the other longer ones to compile

<wingo>at some point it would be nice to be able to use the prebuilt go files in the tarball :)

<wingo>maybe that's just not possible, dunno.

<wingo>but yeah the newer releases have more instructions which makes the assembler take longer to build which can lead to higher bootstrap times

<wingo>but if it's 20 hours that's probably not good, if 2.1.5 was able to do it in 6 hours...

***kelsoo1 is now known as kelsoo

<jmd>What are the current blockers for GuixSD on mips?

<bavier`>jmd: dev-power I'd assume

<rekado_> and related to that probably machine availability

<rekado_>I don’t have a mips device

<ng0>oh my. changed the graphic card to one which has an hdmi connector, and the login screen of sddm is suddenly 400% of its original size :D

<ng0>i think I can now safely log in from 6 meters distance

<jmd>What's the option to build guix without json ?

<mekeor>Hello Guix!

<catonano>mekeor: hi mekeor !

<quiliro>substitute: guix/ui.scm:1209:6: Bad Read-Header-Line header: #<eof>

<quiliro>guix system: error: build failed: substituter `substitute' died unexpectedly

<quiliro>is the error i get from

<quiliro>guix system reconfigure /etc/config.scm

<quiliro>In guix/ui.scm

<roelj>How can I modify the arguments passed to gpg-agent in GuixSD?

<lfam>roelj: Does it work to modify ~/.gnupg/gpg-agent.conf?

<roelj>lfam: Ehm, that file does not exist. Do you know what to put in?

<roelj>lfam: I'll duck that

<lfam>roelj: Check the gpg-agent man page. I think you can put any command-line arguments in the configuration file

<lfam>For example, a must on GuixSD: 'pinentry-program /home/leo/.guix-profile/bin/pinentry'

<roelj>Wow, perfect!

<roelj>Sometimes, the text on my screen turns into garbage on GuixSD.. Have you run into this as well?

<lfam>Me? No, but I'm not using a graphical system

<lfam>What kind of garbage?

<lfam>My GRUB menu is scrambled so I use a console interface for it: <https://bugs.gnu.org/24948>

<roelj>The text is supposed to be white, but is actually just black (like its background). It's funny, I tried to make a screenshot, but that re-renders the screen and therefore fixes the problem

<lfam>Hm, that's weird!

<roelj>It doesn't occur that often though..

<lfam>What interface are you using (console, DE (which one), etc?)

<roelj>GNOME3, on Intel IvyBridge.

<lfam>I'd file a bug report. Maybe someone else has the same issue

<roelj>Right, thank you.

<roelj>Now that I think about it.. it only happens inside Emacs! Anyway..

<jahboite>hi. i am using guix (on foreign distro) and see "warning: failed to install locale: Invalid argument" every time I use guix as a user, but the user has glibc-locales (and glibc-utf8-locales) installed and GUIX_LOCPATH is set to $GUIX_PROFILE/lib/locale for my user. The root user also has these things, but doesn't get the locale warning. Anyone know what I may be doing wrong?

<slyfox>does guix-daemon also have that env variable set?

<jahboite>no slyfox, I don't think it does. I just have "exec /path/to/guix-daemon --build-users-group=guixbuild" in an init script (upstart)

<roelj>What program is available in GuixSD to view a USB camera?

<remd123>hi, what is guix please

<slyfox>do you know anything about it?

<remd123>well iam using clear os by intel, need to get graphics.h to do some C, so need libgraph from gnu and just saw guix

<slyfox>guix is a nix-like package manager written in guile

<remd123>it is not a linux distribution ?

<slyfox>there is also a distribution based on this package manager called GuixSD

<mange>remd123: Have you read https://www.gnu.org/software/guix/manual/html_node/Introduction.html? It might be helpful if you read that and ask any questions you have after that?

<remd123>yes thats what im interested in

<remd123>sure dont wanna bother u guys just want to do some c, just need to know one thing, if i burn guixsd on my usb key i cant start using the libgraph and other gnu libs without head aches xd ?

<slyfox>should be easy to try: https://www.gnu.org/software/guix/manual/html_node/Installing-GuixSD-in-a-VM.html

<remd123>thanks guys, everytime i search some things i come across GNU so ill try in qemu

<lfam>remd123: You can use those programs from the USB image if you install the packages after booting. That image is a usable operating system and I use it sometimes. But, it's a very simple environment.

<remd123>simple gui and packet manager is enought right

<lfam>remd123: There is no GUI

<slyfox>libgraph is this thing, right? https://github.com/Prashant47/libgraph

<slyfox>i don't see it packaged for guix

<remd123>yes it is this thing

<remd123>so bad news for me right

<remd123>yes sorry i confused with guile because it needs guile to be installed

IRC channel logs

2017-02-27.log