Re: LibreOffice doesn’t start after updating

[Martin Castillo]

Am 3. Mai 2023 17:31:40 MESZ schrieb Olivier Dion via :
>On Wed, 03 May 2023, Gottfried  wrote:
>
>> /gnu/store/5h2w4qi9hk1qzzgi1w83220ydslinr4s-glibc-2.33/lib/libpthread.so.0: 
>> undefined symbol: __libc_pthread_init, version GLIBC_PRIVATE
>
>Looks like something similar that I had with gnome-tweaks and evince.
>My system profile was updated but not my home profile.  I believe this
>results in two version of glibc being used.
>
>I fixed this by switching to my old generation for my system until I
>finish upgrading my home.
>

Shouldn't guix prevent this kind of error? Every profile is independent from 
eachother. Why does updating one profile break another?

I haven't looked into guix home though. Maybe those aren't real progiles?

Re: 2 Profiles with 2 different version of Musescore

[Martin Castillo]

Am 29. April 2023 18:18:26 MESZ schrieb Skyler :
>In order to avoid manually updating the alias each time the package is 
>upgraded, you can use a declaration in guix home:
>
>(home-environemnt
>  (services (list
>(service home-bash-service-type
>  (home-bash-configuration (guix-defaults? #t)
>   (aliases `(("mscore4" . ,(file-append musescore 
> "/bin/mscore")
>
>If you do not wish to use guix home, you could also use the extra-special-file 
>service in your system configuration:
>
>(extra-special-file "/usr/bin/mscore4" (file-append musescore "/bin/mscore"))
>
>Just make sure /usr/bin is in your $PATH.

My suggestion always points to the current version of the respective profile. I 
did not point to the store entry of musescore.

Re: 2 Profiles with 2 different version of Musescore

[Martin Castillo]

Am 28.04.23 um 17:07 schrieb Gottfried:

Hi,

I have 2 profiles, one with Musescore 3.6.2, the other with Musescore 
4.02 (the last version, ongoing).
When I enter "mscore" in the terminal it opens the version 3.6.2, which 
I want.


But what do I have to enter in the terminal to get the new version of 
Musescore?



You need to write down the absolute path to the musescore 4 executable.
Something like

/home/gfp/Projekte/pfad/zum/profil/mit/musescore4/bin/mscore

should work.

As a shorthand you could add an alias like this

alias mscore4="/home/gfp/Projekte/pfad/zum/profil/mit/musescore4/bin/mscore"

to .bash_aliases (if it is being 'source'd by bashrc) or.bashrc otherwise.

(Put everything in one line. My mailclient automatically wraps long 
lines even if it should not do so in this case.)


when you restart your shell (or type `source ~/.bashrc`) then
mscore4
should execute musescore 4.

Re: do old packages like Musescore 3.6.2 need updating?

[Martin Castillo]

Am 28.04.23 um 17:15 schrieb Gottfried:

Hi,

1.
I have the old version of Musescore 3.6.2 which I want in one profile.
When I update all profiles at once through a script, it also builds 
locally the old version of Musescore, which needs 1 hour on my laptop.

Firstly, why it always builds it on my laptop?


When you update guix, musescores dependencies get updated (like 
libraries it uses). Once that happens, your manifest then implicitly 
defines a new musescore 3.6.2, where it's dependencies are the newer 
ones. Therefore musescore gets rebuild. Since musescore 3.6.2 is not 
packaged by the guix distribution anymore, it won't be build by the 
official substitute servers and your laptop cannot just download the 
built version.


2.
Do old versions of a package also need to be upgraded?


Normally you'd want to have up-to-date software because of three reasons:
1. Security issues get fixed.
2. Other malfunctions/bugs get fixed.
3. New functionality.

1. becomes much less relevant if your software is never exposed to 
untrusted inputs (e.g. has not internet connection). I'd guess musescore 
never connects to the internet. Do you maybe open musescore project 
files that you got from someone you don't trust, like random forums on 
the web? Or do you get soundfont files from similar sources? In that 
case a malicious file opened by musescore might do bad things to your 
system, if it can exploit a vulnerability musescore 3.6.2 (or one of 
it's dependencies) has.


How would you prevent that? You can't update musescore, because version 
3 is any longer supported.
In case 3.6.2 has any security related vulnerability it would be best to 
not open any files with it, that you don't trust to be non-harmful.


Rebuilding with a newer guix version might get rid of vulnerabilities 
from musescore's dependencies, but not any problems in musescore itself.


2. Similar to 1. there might be bugs that make musescore crash, hang or 
something. These may stem from a bug in one of the dependencies. If you 
encounter such a problem, rebuilding musescore might help.


3. Does not apply here, because musescore 3 does not receive any updates.

So in summary, rebuilding musescore 3.6.2 might increase it's stability, 
but you still should not expose musescore to untrusted files.





3.
and why it want to build it always on my laptop locally?

I thought old versions of a package don’t need upgrading, because only 
new packages develop.

May be I am mistaken.
Do old versions of a package also need upgrading because of some 
dependencies?


4.
when I upgrade all profiles at once, but do not want to upgrade 
Musescore 3.6.2

what are the possibilities?

Should I put Musescore 3.6.2 as only package in one profile and
exclude this profile from updating?


That's a simple solution for your problem. If you don't notice any 
stability bugs of musescore, than you don't need to rebuild it.



or is it better sometimes also to upgrade this old package?

Kind regards

Gottfried

Re: Help-Guix Digest, Vol 89, Issue 35 (Martin Castillo)

[Martin Castillo]

Am 23.04.23 um 15:12 schrieb Gottfried:

Hi,

thanks very much for your help.

I did it, and when I use:
Alt + F2 and enter "icecat" it opens

or in the terminal I enter: "icecat" it opens.


Good.



But in the application menu there are no icon for "icecat"

So basically I get the packages through Alt + F2 and the terminal
but only the graphical icons in my application meny or the icons, 
shortcuts on the desktop are not there.




Of course it would also be fine to have the icons in the GUI, so that I 
know all the packages, which I have in all my profiles installed.




That's a problem with the application menu I think. I _think_ it's 
because guix has a different filesystem layout. So someone needs to 
write a patch for that software to search for the icons in the right 
directory.

Re: Help-Guix Digest, Vol 89, Issue 35 (Martin Castillo)

[Martin Castillo]

Hi,

Am 22.04.23 um 18:25 schrieb Gottfried:

Hi,

thanks for help.

1.

To test 1) add

echo reading xprofile on $(date) >>~/login.log
to your .xprofile and logout and back in.


I have a file .zprofile in my home directory
but not a file: .xprofile



I thought it was a typo in your other mail. zprofile is the startup file 
for the zsh shell. It's corresponds to the .bash_profile for bash.


But I really mean .xprofile. If it does not exist, create the file and write

# all Profile beim Start des Displays Managers öffnen
source ~/.bash_profile

into it.
Also, remove that line from the zprofile.


I tried with both to add it to my .zprofile file
did not help.

-

2.

If this file is really sourced on login, you should find the file
~/login.log with a line saying something like reading xprofile on Do 
20. Apr 16:13:21 CEST 2023.


I did not find a file: ~/login.log

I have  files:
.e-log.log
.e-log.log.old

or they are in a different directory?



Try again after doing what I wrote above.


--
in /etc/profile there is > I don’t know if that helps, so I copied it
No, it does not. Unless you changed it in your system config, it's the 
same every other guix user has on their system.




---
If login.log exists, then there seems to be something wrong with the 
lines that should activate the profiles in .bash_profile.

To test 2) start a login shell with a clean environment
env - bash -l

and check whether that shell has all the profiles activated. If not, 
there is something wrong with your .bash_profile. You should post that 
then.



gfp@Tuxedo ~$ env - $(which bash) -l
gfp@Tuxedo /home/gfp$ icecat
Error: no DISPLAY environment variable specified
gfp@Tuxedo /home/gfp$ chromium
Fontconfig error: No writable cache directories
Fontconfig error: No writable cache directories
Fontconfig error: No writable cache directories
Fontconfig error: No writable cache directories
Fontconfig error: No writable cache directories
[7535:7535:0422/182050.365759:ERROR:ozone_platform_x11.cc(238)] Missing 
X server or $DISPLAY
[7535:7535:0422/182050.366270:ERROR:env.cc(255)] The platform failed to 
initialize.  Exiting.

gfp@Tuxedo /home/gfp$



This looks good: You started a login shell with an empty environment, 
(it means when it started, no guix-profile was activated), and the 
commands icecat and chromium were found.


> I tried to open icecat and chromium in that shell but it doesn’t work.

The reason they printed errors is because the couldn't initialize the 
connection to the window manager, which is necessary for graphical 
applications. The couldn't connect because the environment variable 
DISPLAY was not set (since env - removed it for the shell and it's child 
processes).


So, if icecat and chromium are not installed system wide but only in one 
of your custom profiles, it means they must have been actived by the 
login shell you started. So your .bash_profile seems to work.


Martin Castillo

Re: Help-Guix Digest, Vol 89, Issue 35

[Martin Castillo]

Small correction to the mail I just sent:

Instead of
env - bash -l
you might need to use
env - $(which bash) -l

Re: Help-Guix Digest, Vol 89, Issue 35

[Martin Castillo]

Hi,

Am 20.04.23 um 14:52 schrieb Gottfried:

Hi,

gfp@Tuxedo ~$ ps $(ps   -p $(pidof Xorg) -o ppid=)
   PID TTY  STAT   TIME COMMAND
  1114 tty8 Ssl+   0:00 
/gnu/store/58hc6rh72z3r6zqazmavjnwbcyy6gkps-gdm-42.0/libexec/gd


it shows gdm-42.0 as my display manager
..

A quick search makes me think gdm sources .xprofile. So adding

source ~/.bash_profile
to ~/.xprofile should work.


this is in my .zprofile file, after adding the second sentence/your 
proposal


# Honor system-wide environment variables
source /etc/profile
# all Profile beim Start des Displays Managers öffnen
source ~/.bash_profile
..

but it didn’t help to enable all profiles at login.


Ok, I can think of 2 potential problems. 1) xprofile is not read on 
login or 2) something is wrong with the lines in .bash_profile that 
should activate the profiles.


To test 1) add
echo reading xprofile on $(date) >>~/login.log
to your .xprofile and logout and back in.

If this file is really sourced on login, you should find the file 
~/login.log with a line saying something like reading xprofile on Do 20. 
Apr 16:13:21 CEST 2023.


If login.log does not exist, then maybe ~/.xprofile is not executable?
Try chmod +x ~/.xprofile and re-login.

If login.log exists, then there seems to be something wrong with the 
lines that should activate the profiles in .bash_profile.

To test 2) start a login shell with a clean environment
env - bash -l

and check whether that shell has all the profiles activated. If not, 
there is something wrong with your .bash_profile. You should post that then.


Martin

Re: Help-Guix Digest, Vol 89, Issue 29

[Martin Castillo]

Hi,

sorry for the other mail, I just saw this one.

Am 19.04.23 um 18:05 schrieb Gottfried:

Hi,

thanks for help

I have MATE 1.24.1 Desktop
and GDM as display manager



A quick search makes me think gdm sources .xprofile. So adding
source ~/.bash_profile
to ~/.xprofile should work.

[1]: Somewhat helpful archwiki entry: 
https://wiki.archlinux.org/title/environment_variables#Graphical_environment



Martin

Re: to enable all profiles at login time

[Martin Castillo]

Hi,

Am 19.04.23 um 18:10 schrieb Gottfried:



but when starting MATE Desktop all my profiles are not enabled.


graphical sessions environment is not controlled by .bash_profile (or
.profile)


If sddm is used, this statement is false. Does source that file (if bash 
is the users shell) and therefore it influenced the environment 
variables of all applications started the graphical environment.


Gottfried, what display-manager do you use?

Maybe this command can tell you:
$ ps $(ps   -p $(pidof Xorg) -o ppid=)

Martin

Re: Alternatives to --emulate-fhs on foreign distros

[Martin Castillo]

Hi,



So everything is bash in the end, but I do not know whether bash behaves
differently if it is called under the name "sh". You could give it a try.



Yes. Search for \bsh\b in the manpage. It includes things like

If  bash is invoked with the name sh, it tries to mimic the startup be‐
havior of historical versions of sh as closely as possible, while  con‐
forming  to the POSIX standard as well.

Martin

Re: to enable all profiles at login time

[Martin Castillo]

Hi,


5.
is there also a possibility to enable all my profiles when I log in
to my MATE desktop?


So that all applications (including terminal emulators, regardless of
their configuration) open with them already enabled? There's no such
possibility I know of :/


There is a way and I think it may even be standardized.
On my non-guix distro I use sddm as display-manager. On login, it 
executes /usr/share/sddm/scripts/Xsession as my user.


It contains:
  9 case $SHELL in
 10   */bash)
 11 [ -z "$BASH" ] && exec $SHELL $0 "$@"
 12 set +o posix
 13 [ -f /etc/profile ] && . /etc/profile
 14 if [ -f $HOME/.bash_profile ]; then
 15   . $HOME/.bash_profile
 16 elif [ -f $HOME/.bash_login ]; then
 17   . $HOME/.bash_login
 18 elif [ -f $HOME/.profile ]; then
 19   . $HOME/.profile
 20 fi
...
 41   */fish)
 42 xsess_tmp=`mktemp /tmp/xsess-env-XX`
 43 $SHELL --login -c "/bin/sh -c 'export -p' > $xsess_tmp"
 44 . $xsess_tmp
 45 rm -f $xsess_tmp
 46 ;;
 47   *) # Plain sh, ksh, and anything we do not know.
 48 [ -f /etc/profile ] && . /etc/profile
 49 [ -f $HOME/.profile ] && . $HOME/.profile


So it does try to find shell specific config files (for those that it 
knows about).


You need to know what display-manager (the program where you log in to 
your user) you use and look at its documentation to see what files it 
may source before it logs you in.


In that case, you don't need to launch bash as login shell in your 
terminal, because all the profiles are activated when you login.


Martin

Re: Most used and least used build-system in Guix packages

[Martin Castillo]

==

   Does anyone know if there's a more accurate way of finding out this
   information? I'm asking because my method also counts build systems
   that are mentioned in comments. For example, if a file contains
   comments that mention build systems, they are also counted.

   ,
   | (... more sentences that mention dub-build-system...)
   | ;; One sentence that mentions dub-build-system
   | ;; Another sentence that mentions dub-build-system
   | (... more sentences that mention dub-build-system...)
   `


maybe filter out comment lines ?

grep -hREo '^[^;]+[a-zo-]+-build-system' $path | sort | uniq -c | sort -nr

Re: die Webseite für "Einstieg in Guix" ist nicht zu finden.

[Martin Castillo]

Hi,

Am 01.04.23 um 13:42 schrieb Gottfried:

Hallo,

im Handbuch, Kapitel 14 Dokumentation
gibt es ganz unten (vor den Fußnoten)

  Siehe "Einstieg in Guix" in Info


After comparing this with the english version it seems the two words "in 
Guix" are wrong and it should have been "Einstieg in Info", because (the 
english version) links to the "info" info-page:

https://www.gnu.org/software/emacs/manual/html_node/info/Getting-Started.html#Getting-Started

Der Text will da gar nicht auf einen "Einstieg in Guix" verlinken, 
sondern in einen Einstieg in info. Den korrekten Link habe ich oben 
hingepackt. Sonst im Terminal $ info info eingeben.



wenn man darauf klickt öffnet sich

https://www.gnu.org/software/emacs/manual/html_node/info/Einstieg-in-Guix.html#Einstieg-in-Guix:
An Introduction für eine Einführung in die Info-Navigation

Re: Newbie: How to see progress during 'build' phase?

[Martin Castillo]

Hi,

Am 28.03.23 um 02:38 schrieb Rodrigo Morales:

I'm downloading and building some heavy packages in Guix. Therefore, the
build phase usually takes more than 10 minutes. During that process, I
see the message a slash character that is rotating and the message
'build' phase. I wish I could see the progress of that phase?

Is it possible to show the progress or at least information of the build
phase? Is there a flag for the guix command that enables this behavior?


I know about guix build -v3 to enable printing the build progress. There 
might be other ways.


I don't know, whether one can enable it on the fly, after you already 
invoked the guix command without -v3.


Martin

Re: getting loginctl to work

[Martin Castillo]

Hi,

Am 27.03.23 um 12:21 schrieb SeerLite:

Hi! Is it maybe Polkit? Polkit was stopping me from suspending my laptop via 
SSH. The exact same behavior as you: no output and error code 1.

Since the only `loginctl` command I cared about was `suspend`, I added my user 
to the `power` group, added the `power` group to `operating-system`, defined 
the following extension:

 ; Rule template stolen from from gnu/services/desktop.scm: polkit-wheel
 ; and rule itself stolen from https://askubuntu.com/a/992878
 (define polkit-power-rules
   (file-union
"polkit-power"
`(("share/polkit-1/rules.d/power.rules"
   ,(plain-file
 "power.rules"
 "polkit.addRule(function(action, subject) {
 if (action.id == \"org.freedesktop.login1.suspend\" &&
 subject.isInGroup(\"power\")) {
 return polkit.Result.YES;
 }
 });")


and added the above to my services like:

 (simple-service 'polkit-power-rules polkit-service-type (list 
polkit-power-rules))

You'd have to do something but I'm guessing for `login1.poweroff` instead.


Thank you, that works. I think this kind of rule should be expected on 
most systems. It would be nice if this kind of polkit rule was 
predefined as a variable in guix, or the manual (or at least the 
cookbook) should give this as an example.

Re: creating a manifest

[Martin Castillo]

Hi,

Am 26.03.23 um 15:07 schrieb Gottfried:

Hi,
thank you very much for your help.

I am learning day by day how to use Guix.

1.
I understood till now that creating manifests mean several packages in 
one manifest, like in my music profile: ardour, audacity, audacious, 
musescore.


You mean `in one profile`? Correct.


2.
When I was opening that profile I used musescore, I had to close it and 
then open audacity and so on.

It seems to me I can’t open several packages at once.
Or I don’t know how to do it, if possible.


That's weird. What happens if you try to run a second program from the 
profile while the first one is still running?


Two ways that you can do it:
1) With two terminals:
   1. In the first one run:
  $ source ~/Projekte/Musik/guix-profil/etc/profile;
  $ mscore;
   2. In the the second terminal run:
  $ source ~/Projekte/Musik/guix-profil/etc/profile;
  $ vlc;

2) With one terminal:
  Run the first, (or all commands) in the background, so that the
  shell directly prompts you for the next command:
  $ source ~/Projekte/Musik/guix-profil/etc/profile;
  $ mscore &
  $ vlc &
  $



3.
Now I realised that to create a manifest means one package, only 
musescore, or only emacs with all its emacs packages.


No, that's not always right. The manifest you posted in this thread 
contains multiple packages and you created a profile with all those 
packages.



4.
Then upgrading the manifest/one profile
would not be a problem.


There can be multiple packages be in one profile, created by one 
manifest, and using this command it should be no problem to update all 
packages, but keeping musescore at version 3:

>> guix package -m /home/gfp/Projekte/Musik/musik.scm --profile
>> /home/gfp/Projekte/Musik/guix-profil



5.
What would be the difference to have several profiles in my main profile
without haveing guix home,


What do you mean by having several profiles in your main profile?
AFAIK there is no way to put one profile inside another.
Do you mean several packages in one profile?

or to have several profiles in Guix home? 


I never used guix home. I cannot answer that.

Re: creating a manifest

[Martin Castillo]

Hi,

Am 26.03.23 um 10:16 schrieb Gottfried:

please send the command you used to update the profile and the 
manifest file inside the profile, i.e. 
home/gfp/Projekte/Musik/guix-profil/manifest.scm


Martin


This was the command I used to upgrade the profile
(because I deleted one package: ardour
in my main profile, so I wanted to install it in this profile)

~$ guix package --upgrade --profile=/home/gfp/Projekte/Musik/guix-profil


Like you, I would expect that to work, but I read the info page again. 
It says about --upgrade:


  Note that this upgrades package to the latest version of packages
 found in the distribution currently installed.

And the latest version of musescore is 4.0.2, so it tries to install that.

So I guess the right command would be:

guix package -m /home/gfp/Projekte/Musik/musik.scm --profile 
/home/gfp/Projekte/Musik/guix-profil





-

here the musik.scm:



I actually meant /home/gfp/Projekte/Musik/guix-profil/manifest and not 
/home/gfp/Projekte/Musik/musik.scm, but now I don't need it anymore.


Martin

Re: getting loginctl to work

[Martin Castillo]

Hi,



Could it be that something is holding an inhibit lock on shutdown?  I
think ssh can do that.  It might even be automatic??
Try running elogind-inhibit with no arguments.


$ elogind-inhibit
WHOUID USER PID COMMWHAT 
WHY   MODE
NetworkManager 0   root 162 NetworkManager  sleep 
NetworkManager needs to turn off networks delay
UPower 0   root 250 .upowerd-real   sleep 
Pause device polling  delay
gdm983 gdm  456 .gsd-media-keys 
handle-power-key:handle-suspend-key:handle-hibernate-key GNOME handling 
keypresses block
gdm983 gdm  456 .gsd-media-keys sleep 
GNOME handling keypresses delay
gdm983 gdm  463 .gsd-power-real sleep 
GNOME needs to lock the screendelay


5 inhibitors listed.

Doesn't look like it.

getting loginctl to work

[Martin Castillo]

Hi,

I have guix system installed and want `loginctl poweroff` to work. I'm 
logged in via ssh. The vm is started using
qemu-system-x86_64   -drive media=disk,file=geex-unenc.img,format=raw 
-bios /usr/share/ovmf/x64/OVMF.fd  -m 2G   -smp 3 -nic 
user,model=virtio-net-pci,hostfwd=tcp::2223-:22 -enable-kvm



Currently loginctl does not work:
$ strace -o loginctl.str loginctl poweroff
returns 1 and prints no errors.
loginctl.str shows some partial error message. I don't know how to get 
loginctl or dbus to be more verbose and show me the error. 
/var/log/secure does not show anything regarding that error.


$ loginctl list
SESSION UID USER SEAT  TTY
 c1 983 gdm  seat0 tty7

1 sessions listed.


First I tried configuring %base-services plus elogind-service-type, but 
it didn't work, so I tried %desktop-services, and it still doesn't work, 
but I'm sure it should. My (approximate) configuration.scm is attached.



Martin;; This is an operating system configuration generated
;; by the graphical installer and adapted manually.

(use-modules (gnu))
(use-package-modules ssh)
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
;  (keyboard-layout (keyboard-layout "de" "neo")) ; commented out for help-guix
  (host-name "geex-unenc")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("wheel" "netdev" "audio" "video")))
%base-user-accounts))
  (packages
(append
  (list (specification->package "nss-certs"))
  %base-packages))
  (services
(append
  (list
	;(service xfce-desktop-service-type)

	(service dropbear-service-type (dropbear-configuration
			(root-login? #t)))
	)
;	(service dbus-root-service-type)
;	(service polkit-service-type) ;need use-modules services dbus
;	(service elogind-service-type)
;	(service dhcp-client-service-type)
  %desktop-services))
  (bootloader
(bootloader-configuration
  (bootloader grub-efi-bootloader)
  (targets '("/efi"))
  (keyboard-layout keyboard-layout)))
  (file-systems
(cons* (file-system
 (mount-point "/efi")
 (device (file-system-label "EFI"))
 (type "vfat"))
   (file-system
 (mount-point "/")
 (device
   (file-system-label "myroot"))
 (type "ext4")
	 (needed-for-boot? #t))
   %base-file-systems))
)


loginctl.str
Description: application/vnd.pg.format

Re: Musescore and other music notation packages like Lilypond, Frescobaldi, Denemo, Rosegarden

[Martin Castillo]

Hi,

I'm just leaving a simle note here.

Am 21.03.23 um 10:02 schrieb Gottfried:

I can’t create a document with:
lilypond THE_FILENAME.ly


And you are not supposed to. lilypond is not a program to create or edit 
.ly files. It is only there to read those files and produce pdfs and 
midi files.


The command
lilypond THE_FILENAME.ly
is used to create a pdf using an existing file named THE_FILENAME.ly
as input!

Here is a simple figure. Programs are inside bars (|program|) and take 
input from the left and output to the right.


With frescobaldi the data flow looked something like this
|frescobaldi| -> THE_FILENAME.ly -> |lilypond| -> THE_FILENAME.pdf
where the middle parts were invisible to you, because frescobaldi was 
made that way, to hide that part. You just interacted with frescobaldi 
and used the pdf at the end.


If you change away from frescobaldi, the middle part becomes visible 
(i.e. you have to "do more"):

|nano   | -> THE_FILENAME.ly -> |lilypond| -> THE_FILENAME.pdf

Now you have to create the file THE_FILENAME.ly explicitly yourself with 
some program, possibly nano, and then call lilypond THE_FILENAME.ly so 
that lilypond creates the pdf file from THE_FILENAME.ly.


Now you can replace nano by your favourite (working) editor, so once the 
gtk stuff is fixed, you may even use gedit in the future.


Hope that helps.

> with which command can I copy the songbook from this guix shell into 
my home folder?


Do you mean the pdf?
I'd guess it already is in your home folder: If you just started your 
terminal before executing

guix shell lilypond nano mercurial timidity++  and the other commands,
THE_FILENAME.ly should be in your home folder (/home/gfp I guess). And 
executing

lilypond THE_FILENAME.ly creates the pdf file in the same folder.

Here is how to copy a file:
cp source.file dest.file

Example:
cp THE_FILENAME.pdf ~/Dokumente/

This will copy THE_FILENAME.pdf to the Dokumente directory.
This assumes THE_FILENAME.pdf is in your current working directory of 
your shell.



(That just means you're in that directory with the shell. The command 
`pwd` prints the current working directory. With cd you can change 
between working directories:

$ cd   # this changes to your home directory
$ cd Dokumente  # go into ~/Dokumente, which is Dokumente inside your 
home directory

# you can swith down multiple directories at once
$ cd Musik/Noten # go into ~/Dokument/Musik/Noten
$ cd ..  # go one level up, back into ~/Dokument/Musik
$ cd ~/Desktop/Projekt/Musical # jump to Projekt Musical inside Desktop
$ cd ../.. # go two levels up to Desktop

Note, these examples assume that those directories even exist. If one of 
those doesn't, cd will print an Error:

bash: cd: Musik/Noten: Datei oder Verzeichnis nicht gefunden
)

Example 2: If THE_FILENAME.pdf was for example on your desktop, you could do
cp ~/Desktop/THE_FILENAME.pdf ~/Dokumente/songbook-v0.1.pdf
to copy it to Dokumente and give it a different name at once.

Martin

Re: Newbie: How to define a package that install files to ~/.config?

[Martin Castillo]

Hi,

Am 18.03.23 um 06:02 schrieb Rodrigo Morales:

Table of Contents
_

1. What I'm trying to do
2. What I've tried
3. The problem
4. The question


Newbie: How to define a package that creates files in ~/.config?


I just remembered there is something called guix home, which should be 
what you are looking for.

I haven't used it, so I can't give you anymore hints than
`info guix home`

Martin

Re: Help-Guix Digest, Vol 88, Issue 36

[Martin Castillo]

Am 19. März 2023 17:57:38 MEZ schrieb Gottfried :
>Hi,
>
>thanks for your help.
>
>I tested, tried and used AFAIU you the old version of Musescore,
>but it wanted to build Musescore 4, which I saw only at the end.
>
>but it failed:
>
>gfp@Tuxedo ~$ guix time-machine 
>--commit=6a54715988aee884bbb5f81e1b718a27ff771ec4 -- shell musescore
>Kanal „guix“ wird vom Git-Repository auf 
>„https://git.savannah.gnu.org/git/guix.git“ aktualisiert …
>substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
>aktualisiert … 100.0%
>substitute: Liste der Substitute von „https://bordeaux.guix.gnu.org“ wird 
>aktualisiert …  substitute: Liste der Substitute von 
>„https://bordeaux.guix.gnu.org“ wird aktualisiert … 100.0%
>substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
>aktualisiert … 100.0%
>substitute: Liste der Substitute von „https://bordeaux.guix.gnu.org“ wird 
>aktualisiert …  substitute: Liste der Substitute von 
>„https://bordeaux.guix.gnu.org“ wird aktualisiert … 100.0%
>substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
>aktualisiert … 100.0%
>substitute: Liste der Substitute von „https://bordeaux.guix.gnu.org“ wird 
>aktualisiert …  substitute: Liste der Substitute von 
>„https://bordeaux.guix.gnu.org“ wird aktualisiert … 100.0%
>/gnu/store/0y3w5sly6rw9yc0shqvyzkl8vjgr9gdf-module-import.drv wird erstellt …
>/gnu/store/whpg974y1nj7j1gddhfg0b15k4xpvx9n-module-import.drv wird erstellt …
>/gnu/store/nq8vmvriv2h095apywf8k4bb4z0jjpih-module-import-compiled.drv wird 
>erstellt …
>/gnu/store/6qd26vdgd11vjlv0lqp7h5z814qc82yi-module-import-compiled.drv wird 
>erstellt …
>/gnu/store/mi4klx1g21qwa4i5kjah6gbavncjgpm0-compute-guix-derivation.drv wird 
>erstellt …
>Computing Guix derivation for 'x86_64-linux'... /
>substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
>aktualisiert …   0.0%Backtrace:
>substitute: In ice-9/boot-9.scm:
>substitute:   1752:10 17 (with-exception-handler _ _ #:unwind? _ # _)
>substitute: In unknown file:
>substitute:   16 (apply-smob/0 #)
>substitute: In ice-9/boot-9.scm:
>substitute: 724:2 15 (call-with-prompt _ _ #default-prompt-handle…>)
>substitute: In ice-9/eval.scm:
>substitute: 619:8 14 (_ #(#(#)))
>substitute: In guix/ui.scm:
>substitute:2279:7 13 (run-guix . _)
>substitute:   2242:10 12 (run-guix-command _ . _)
>substitute: In ice-9/boot-9.scm:
>substitute:   1752:10 11 (with-exception-handler _ _ #:unwind? _ # _)
>substitute:   1752:10 10 (with-exception-handler _ _ #:unwind? _ # _)
>substitute: In guix/scripts/substitute.scm:
>substitute:844:18  9 (_)
>substitute:348:26  8 (process-query # _ #:cache-urls _ 
>#:acl _)
>substitute: In guix/substitutes.scm:
>substitute:382:27  7 (lookup-narinfos/diverse _ _ #…> …)
>substitute:339:31  6 (lookup-narinfos _ _ #:open-connection _ # _)
>substitute:262:26  5 (fetch-narinfos _ _ #:open-connection _ # _)
>substitute: In ice-9/boot-9.scm:
>substitute:   1685:16  4 (raise-exception _ #:continuable? _)
>substitute:   1685:16  3 (raise-exception _ #:continuable? _)
>substitute:   1780:13  2 (_ #< components: (#<> 
>#<…>)
>substitute:   1685:16  1 (raise-exception _ #:continuable? _)
>substitute:   1685:16  0 (raise-exception _ #:continuable? _)
>substitute:
>substitute: ice-9/boot-9.scm:1685:16: In procedure raise-exception:
>substitute: In procedure write_wait_fd: unimplemented
>guix time-machine: Fehler: 
>`/gnu/store/nn61n9c91vvxf95fs5n3ldxqf3klnsgf-guix-1.4.0-4.01fd830/bin/guix 
>substitute' died unexpectedly
>
>Kind regards
>
>Gottfried
>
>
>
>
>
>Am 19.03.23 um 16:06 schrieb Martin Castillo:
>> Hi,
>> 
>> Am 19.03.23 um 13:15 schrieb Gottfried:
>>> Hi,
>>> 
>>> thanks for trying...
>>> 
>>> I copied the
>>> "Commit 6a54715988aee884bbb5f81e1b718a27ff771ec4"
>>> into my musik.scm:
>>> ..
>>> ;; Manifest Musik Programme
>>> (use-modules (gnu packages music)
>>>   (guix packages))
>>> 
>>> (define musescore-3.6.2
>>>    (package
>>>  (inherit musescore)
>>>  (version "3.6.2")
>>>  (Commit 6a54715988aee884bbb5f81e1b718a27ff771ec4)))
>>> 
>>> ;; combine both manifest lists:
>>> (concatenate-manifests
>>>    (list
>>>  (packages->manifest (list musescore-3.6.2))
>>> (specifications->manifest '("ardour" "audacious" "audacity" "obs" "vlc"
>>> 
>>> 

Re: Help-Guix Digest, Vol 88, Issue 36

[Martin Castillo]

Hi,

Am 19.03.23 um 13:15 schrieb Gottfried:

Hi,

thanks for trying...

I copied the
"Commit 6a54715988aee884bbb5f81e1b718a27ff771ec4"
into my musik.scm:
..
;; Manifest Musik Programme
(use-modules (gnu packages music)
  (guix packages))

(define musescore-3.6.2
   (package
     (inherit musescore)
     (version "3.6.2")
     (Commit 6a54715988aee884bbb5f81e1b718a27ff771ec4)))

;; combine both manifest lists:
(concatenate-manifests
   (list
     (packages->manifest (list musescore-3.6.2))
(specifications->manifest '("ardour" "audacious" "audacity" "obs" "vlc"

.
and I had a go
I got this message:

gfp@Tuxedo ~$ guix package -p /home/gfp/Projekte/Musik/guix-profil -m 
/home/gfp/Projekte/Musik/musik.scm
/home/gfp/Projekte/Musik/musik.scm:6:2: Fehler: (package (inherit 
musescore) (version "3.6.2") (Commit 
#{6a54715988aee884bbb5f81e1b718a27ff771ec4}#)): extraneous field 
initializers (Commit)


.
So I have to add it somewhere else...


The commits I mentioned are from guix, not from the musescore project. 
You put it in the musescore definition, so guix downloaded musescore and 
tried to switch to that commit/version of musescore. But it is not a 
commit of musecore, so that is indeed the wrong place.


(And the (commit ) field belongs inside an (origin ) field...)

So my original suggestion was to copy the musescore package definition 
from one of those commits from the guix project

http://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/music.scm?id=6a54715988aee884bbb5f81e1b718a27ff771ec4#n4860

and hope that all dependencies of that package definition are still 
valid and compatible in the current guix. You might need to copy some of 
the use-modules lines at the top of the linked file into your manifest.


But I just had another idea: We want to install software from an older 
guix version: That's what guix time-machine is for!
guix time-machine --commit=6a54715988aee884bbb5f81e1b718a27ff771ec4 -- 
package -m manifest.scm -p /Pfad/zu/Musik/Musik

or to test it first:
guix time-machine --commit=6a54715988aee884bbb5f81e1b718a27ff771ec4 -- 
shell musescore


I tried it with the newer commit 
ced3be2b562866b82f97a530cd66610c84c381f7 to get musescore with newer 
dependencies. (That commit is the parent of 
9f93bcd1862c76d7ff30da6f712c9bd2912b8346, which changed to musescore 4.0.):
guix time-machine --commit=ced3be2b562866b82f97a530cd66610c84c381f7 -- 
shell musescore


The first time time-machine failed to build the older version of guix. 
It used up to much ram and was killed by the OOM-killer. (I'm using it 
inside a vm.) Otherwise it should work.


A downside of this approach is that you cannot mix the older musescore 
with the newer versions of vlc etc in the same profile.




Martin


kind regards

Gottfried



Date: Sat, 18 Mar 2023 13:44:08 +0100
From: Martin Castillo 
To: help-guix@gnu.org
Subject: Re: creating a manifest
Message-ID: 
Content-Type: text/plain; charset=UTF-8; format=flowed

I noticed an error.



For that you need to instead create a new package definition for
musescore version 3.6.2. The info page gives you an example, which you
just adjust. Using guix edit musescore you see how musescore is defined.
You can notice, there is a version field. So lets create a new package
with adjusted version:

(define musescore-3.6.2
    (package
      (inherit musescore)
      (version "3.6.2")))

;; We create this manifest from a _package_, and not from a
;; _specification_ (which is just something like a string "musescore")
(packages->manifest (list musescore-3.6.2))



While building guix says it's downloading musescore-4.0.1. Maybe it
switches to the respective branch before building.

I even tried explicitly writing the origin entry of
(define musescore-3.6.2
    (package

  (inherit musescore)

  (version "3.6.2")
  (source
   (origin
 (method git-fetch)
 (uri (git-reference
   (url "https://github.com/musescore/MuseScore;)
   (commit (string-append "v" "3.6.2"
 (file-name (git-file-name "musescore" "3.6.2"))
 (sha256
  (base32 
"0x2aahpbvss3sjydcq6xdh198fmslgypixmd2gckfwjqzady662y"))

 (modules '((guix build utils)))
 (snippet
  '(begin
 ;; Remove unused libraries...
 (for-each delete-file-recursively
   '("thirdparty/freetype"))
 ;; ... and precompiled binaries.
 (delete-file-recursively "src/diagnostics/crashpad_handler")
 (substitute* "src/diagnostics/CMakeLists.txt"
  

Re: creating a manifest

[Martin Castillo]

I noticed an error.



For that you need to instead create a new package definition for 
musescore version 3.6.2. The info page gives you an example, which you 
just adjust. Using guix edit musescore you see how musescore is defined. 
You can notice, there is a version field. So lets create a new package 
with adjusted version:


(define musescore-3.6.2
   (package
     (inherit musescore)
     (version "3.6.2")))

;; We create this manifest from a _package_, and not from a
;; _specification_ (which is just something like a string "musescore")
(packages->manifest (list musescore-3.6.2))



While building guix says it's downloading musescore-4.0.1. Maybe it 
switches to the respective branch before building.


I even tried explicitly writing the origin entry of
(define musescore-3.6.2
  (package 

(inherit musescore) 



(version "3.6.2")
(source
 (origin
   (method git-fetch)
   (uri (git-reference
 (url "https://github.com/musescore/MuseScore;)
 (commit (string-append "v" "3.6.2"
   (file-name (git-file-name "musescore" "3.6.2"))
   (sha256
(base32 "0x2aahpbvss3sjydcq6xdh198fmslgypixmd2gckfwjqzady662y"))
   (modules '((guix build utils)))
   (snippet
'(begin
   ;; Remove unused libraries...
   (for-each delete-file-recursively
 '("thirdparty/freetype"))
   ;; ... and precompiled binaries.
   (delete-file-recursively "src/diagnostics/crashpad_handler")
   (substitute* "src/diagnostics/CMakeLists.txt"
 (("install") "#install"

guix shell -m manifest.scm still says 4.0.1 and executing mscore fails:
[env]$ mscore --version
QEventLoop: Cannot be used without QApplication
QEventLoop: Cannot be used without QApplication
qt.qpa.xcb: could not connect to display
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" even 
though it was found.
This application failed to start because no Qt platform plugin could be 
initialized. Reinstalling the application may fix this problem.


Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, 
offscreen, vnc, xcb.


Abgebrochen

So seems like in this case it's not so easy. (Well the current guix 
packaged version is 4.0.1 so a major version change happened.)


One could try using the older package definition, which is available in 
the git history of guix. Commit 6a54715988aee884bbb5f81e1b718a27ff771ec4 
is the newest change to musescore with 3.6.2. 
9f93bcd1862c76d7ff30da6f712c9bd2912b8346 switched to 4.0. So it's parent 
commit should have the newest dependencies which might work with current 
guix.
I don't have time right now, but you might want to copy such an older 
package definition into your manifest.scm


Martin

Re: creating a manifest

[Martin Castillo]

Hi,

Am 18.03.23 um 11:50 schrieb Gottfried:
> Hi,
>
> I wanted to create a manifest for music,
> 
.

> my "musik.scm":
>
> ;; Manifest Musik Programme
> (use-modules (gnu packages version-control))If this is the whole 
file, this use-modules is unnecessary.


>
> (specifications->manifest ’("Musescore '--with-branch=musescore=v3.6.2"
> "Ardour" "Audacious" "Audacity" "OBS Studio" "VLC Media Player"))
The cryptic error message is about an unexpected character, namely
(specifications->manifest ’  <- this backtick
It needs to be the ascii >'<.

If you fix that, you get new errors.

If you read the info page "Writing manifests", you'll see you have to 
use the package names like when using `guix package -i musescore` etc. 
when using specifications->manifest.
guix search musescore would show you under `name:`, that the musescore 
package (in guix) is named in lowercase. Also OBS Studio is called obs, 
vlc is called vlc etc.


But you can not add cmdline args like --with-branch.

For that you need to instead create a new package definition for 
musescore version 3.6.2. The info page gives you an example, which you 
just adjust. Using guix edit musescore you see how musescore is defined. 
You can notice, there is a version field. So lets create a new package 
with adjusted version:


(define musescore-3.6.2
  (package
(inherit musescore)
(version "3.6.2")))

;; We create this manifest from a _package_, and not from a
;; _specification_ (which is just something like a string "musescore")
(packages->manifest (list musescore-3.6.2))


So now putting everything together:

;; Manifest Musik Programme
(use-modules (gnu packages music)
 (guix packages))


(define musescore-3.6.2
  (package
(inherit musescore)
(version "3.6.2")))

;; combine both manifest lists:
(concatenate-manifests
  (list
(packages->manifest (list musescore-3.6.2))
(specifications->manifest '("ardour" "audacious" "audacity" "obs" 
"vlc"



>
> 
..

>
> gfp@Tuxedo ~$ guix package -p /home/gfp/Projekte/Musik -m
> /home/gfp/Projekte/Musik/musik.scm
>
> 
..

> I got this message:
>
> /home/gfp/Projekte/Musik/musik.scm:4:26: error: #{\x2019;}#: unbound
> variable
> hint: Did you forget a `use-modules' form?
>
>
> I surely made a mistake in my "musik.scm"
> but playing around for hours
> I prefer to ask.
>
>
> Kind regards
>
> Gottfried
>

Re: Newbie: How to define a package that install files to ~/.config?

[Martin Castillo]

Am 18. März 2023 06:02:27 MEZ schrieb Rodrigo Morales 
:
>Table of Contents
>_
>
>1. What I'm trying to do
>2. What I've tried
>3. The problem
>4. The question
>
>
>Newbie: How to define a package that creates files in ~/.config?
>
>
>1 What I'm trying to do
>===
>
>  I want to define a package that downloads a file from a git repository
>  and puts the file under a directory in `~/.config/' (specifically,
>  `~/.config/ibus/rime/')
>
>
>2 What I've tried
>=
>
>  I've defined a module containing the definition of a package that
>  tries to accomplish that goal. It uses `install-file' to copy the
>  downloaded files to the desired location (please see code block below)
>
>  ,
>  | cat ~/my/packages/rime.scm
>  `
>
>  ,
>  | (define-module (rime)
>  |   #:use-module (guix licenses)
>  |   #:use-module (guix packages)
>  |   #:use-module (guix gexp)
>  |   #:use-module (guix build-system trivial)
>  |   #:use-module (guix git-download))
>  |
>  | (define-public rime-wubi
>  |   (package
>  |(name "rime-wubi")
>  |(version "1.0")
>  |(source (origin
>  | (method git-fetch)
>  | (uri (git-reference
>  |   (url "https://github.com/rime/rime-wubi;)
>  |   (commit "f1876f08f1d4a9696395be0070c0e8e4353c44cb")))
>  | (file-name (git-file-name name version))
>  | (sha256
>  |  (base32
>  |   "1d9y9rqssacria9d0hla96czsqv2wkfm6z926m1x269ryv96zxvk"
>  |(build-system trivial-build-system)
>  |(arguments
>  | (list
>  |  #:modules `((guix build utils))
>  |  #:builder
>  |  #~(begin
>  |  (use-modules (guix build utils))
>  |  (chdir (assoc-ref %build-inputs "source"))
>  |  (install-file "wubi86.dict.yaml" (string-append #$output
>"/.config/ibus/rime"))
>  |  (install-file "wubi86.schema.yaml" (string-append #$output
>"/.config/ibus/rime")
>  |(synopsis "Wubi86 schema and dictionary for RIME")
>  |(description "This package contains a dictionary and a schema
>definition for the 86
>  | version of Wubi, a shape-based input method for Chinese characters.")
>  |(home-page "https://github.com/rime/rime-wubi;)
>  |(license lgpl3)))
>  `
>
>  The package is installed without no problems.
>
>  ,
>  | export GUIX_PACKAGE_PATH="$HOME/my/packages"
>  | guix package -i rime-wubi
>  `
>
>  ,
>  | The following package will be installed:
>  |rime-wubi 1.0
>  |
>  `
>
>  ,
>  | echo $?
>  `
>
>  ,
>  | 0
>  `
>
>
>3 The problem
>=
>
>  However, the files `wubi86.dict.yaml' and `wubi86.schema.yaml' are
>  installed in `~/.guix-profile/.config/ibus/rime' (see code block
>  below)
>
>  ,
>  | find -L /home/rdrg/.guix-profile/.config
>  `
>
>  ,
>  | /home/rdrg/.guix-profile/.config
>  | /home/rdrg/.guix-profile/.config/ibus
>  | /home/rdrg/.guix-profile/.config/ibus/rime
>  | /home/rdrg/.guix-profile/.config/ibus/rime/wubi86.dict.yaml
>  | /home/rdrg/.guix-profile/.config/ibus/rime/wubi86.schema.yaml
>  `
>
>  Those files don't appear in `~/.config/ibus/rime' (the location I want
>  them to exist in)
>
>  ,
>  | find ~/.config/ibus/rime
>  `
>
>  ,
>  | /home/rdrg/.config/ibus/rime
>  | /home/rdrg/.config/ibus/rime/default.custom.yaml
>  `
>
>
>4 The question
>==
>
>  How to define a package that install files to any directory under
>  `~/.config/'?

Hi,

I think packages cannot install files outside the store.

Maybe you could try and see, whether ibus looks at a special environment 
variable to find configs in another directory, too. Then you could define that 
environment variable in your package and it will be added to your guix profile.

I dont know ibus, but:
Otherwise you may need to adjust the system-wide installed ibus package and add 
that configuration that you want to it.


Martin

Re: updating restic: go-build-system trying to resolve host names with nonexistent dnscache at [::1]:53

[Martin Castillo]

Am 13.03.23 um 01:26 schrieb Martin Castillo> Now the build fails trying 
to download dependencies:


Ok, naturally, we don't want go to access the internet to get the 
dependencies.


I tried adding (inputs (list go-github-com-minio-sha256-simd))
to the package, but go still tries to get that. Go doesn't care about 
the symlinked package at src/github.com/minio/sha256-simd.


(Someone added that symlinking to go-build-system once, so it may have 
worked with earlier (<1.18) go versions? 
https://issues.guix.gnu.org/50493 references 
https://issues.guix.gnu.org/50227 , which I don't understand, but seems 
to try to get rid of the symlinking and replace it with something better?)


If I run go run build.go manually in the failed dir, the dependencies 
get downloaded to


[...]
/pkg/mod/cache/download/google.golang.org/genproto
/pkg/mod/cache/download/google.golang.org/grpc
/pkg/mod/cache/download/google.golang.org/protobuf
/pkg/mod/cache/download/go.opencensus.io
/pkg/mod/cache/download/gopkg.in/
/pkg/mod/cloud.google.com
/pkg/mod/github.com
/pkg/mod/golang.org
/pkg/mod/google.golang.org
/pkg/mod/go.opencensus.io@v0.24.0
/pkg/mod/gopkg.in
etc.

and not to /src/github.com/ or such, where inputs get symlinked at.

My conclusion: go-build-system needs to be updated to account for the 
new way go handles dependencies.


Is this correct? And is this issue 50227?


Martin



Looks like go trys using ipv6 for the dns lookup. But
/etc/resolv.conf says
nameserver 10.0.2.3

I don't have a nameserver running on ::1. And this inside qemu, where I 
didn't set up anything specifically for ipv6. I have only site and link 
local ipv6 addresses. Why does go-build-system try to resolve the host 
names at [::1]:53?


Martin Castillo

Re: guix system image fails with: service 'swap-/swapfile' requires 'file-system-/'

[Martin Castillo]

Hi,

Am 07.03.23 um 16:50 schrieb Ludovic Courtès:

Hi Martin,

Sorry for the late reply.

Martin Castillo  skribis:


[...]


/mnt/root/man.config.scm:35:14: warning: the 'target' field is
deprecated, please use 'targets' instead
guix system: warning: Consider running 'guix pull' followed by
'guix system reconfigure' to get up-to-date packages and security updates.

guix system: error: service 'swap-/swapfile' requires 'file-system-/',
which is not provided by any service

I can't find anything regarding a file-system-/ service. I do have a
rootpartition in 'file-systems'. Why is that not sufficient?


You can definitely omit the ‘dependencies’ field of ‘swap-space’: it’s
implicit that the root file system is already mounted by the time swap
is activated.



What is the error message then about?


If I remove the dependency in swap-devices, it builds, though the
system gets stuck on boot shortly after generating ssh keys. Picture
is attached. Any hints for that problem?


How did you test?  With ‘guix system vm’?



I started the image with
 qemu-system-x86_64  -drive 
media=disk,file=dk1cc2q05zcw6rbc8l8vm7d6zb9rw771-image.qcow2 -m 1.5G 
-bios /usr/share/ovmf/x64/OVMF.fd -smp 3 -nic 
user,model=virtio-net-pci,hostfwd=tcp::2224-:22 -enable-kvm


After changing the resolution of the console, it does not continue the 
boot process.




The OS config you provided is truncated, so we can’t test it :-), but if
you provide the full thing we can give it a try.



Somehow I missed the last closing braces.

I used
   guix 73b5c95
 Repository-URL: https://git.savannah.gnu.org/git/guix.git
 Branch: master
 Commit: 73b5c955ae46af2548625fe31afe60193fb83e2a

to run
$ guix system image -t qcow2 min.swap.scm
/gnu/store/dk1cc2q05zcw6rbc8l8vm7d6zb9rw771-image.qcow2


Martin

;; This is an operating system configuration generated
;; by the graphical installer.

(use-modules (gnu))
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
  (keyboard-layout (keyboard-layout "de" "neo"))
  (host-name "minswap")
  (issue "minimal system with swap, without swap dependencies")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("wheel" "netdev" "audio" "video")))
%base-user-accounts))
  (packages
(append
  (list (specification->package "nss-certs"))
  %base-packages))
  (services
(append
  (list (service dhcp-client-service-type)
(service openssh-service-type)
)
  %base-services))
  (bootloader
(bootloader-configuration
  (bootloader grub-efi-bootloader)
  (target "/efi") ; deprecated
  (keyboard-layout keyboard-layout)))
  (mapped-devices
(list (mapped-device
	(source (uuid "e35658c9-6e31-431a-9397-2e988e32fe63"))
	(target "lukspart")
	(type luks-device-mapping))
	  (mapped-device
	(source "lvmSystem")
	(target "lvmSystem-volRoot")
	(type lvm-device-mapping
  (file-systems
(cons* (file-system
 (mount-point "/efi")
 (device (uuid "845E-533F" 'fat32))
 (type "vfat"))
   (file-system
 (mount-point "/")
 (device ; mapped device name: lvmSystem-volRoot, FS-label is guixsd-root
   (uuid "0584a348-341b-475d-b19f-93b5567ee02b"
 'ext4))
 (type "ext4")
	 (needed-for-boot? #t)
	 (dependencies mapped-devices))
   %base-file-systems))
  (swap-devices
(list
  (swap-space
	(target "/swapfile"
  (kernel-arguments
(cons* "resume=/swapfile"
	   %default-kernel-arguments)))

updating restic: go-build-system trying to resolve host names with nonexistent dnscache at [::1]:53

[Martin Castillo]

Hi,

the packaged restic is at 0.9.6 while upstream is at 0.15.1. I thought 
I'd try to update the package to learn how to package things for guix. 
I'm a complete noob with that, and I have zero experience with go.


I'm using guix system in qemu with
qemu-system-x86_64   -drive media=disk,file=geex-unenc.img,format=raw 
-bios /usr/share/ovmf/x64/OVMF.fd  -m 2G   -smp 3 -nic 
user,model=virtio-net-pci,hostfwd=tcp::2223-:22 -enable-kvm


I used guix edit restic as a starting point and saved it in 
~/packages/restic.scm. I removed all other programs, renamed restic to 
restic-new, adjusted the hash, removed a patch for an older go, and used 
a newer go version for the build system, as 1.17 is too old for restic 0.15.


Now the build fails trying to download dependencies:

$ guix build -L packages/ -K --fallback restic-new
...
go: downloading github.com/klauspost/compress v1.15.15
go: downloading golang.org/x/text v0.6.0
go: downloading github.com/pkg/xattr v0.4.10-0.20221120235825-35026bbbd013
go: downloading github.com/hashicorp/golang-lru/v2 v2.0.1
internal/backend/gs/gs.go:14:2: cloud.google.com/go/storage@v1.29.0: Get 
"https://proxy.golang.org/cloud.google.com/go/storage/@v/v1.29.0.zip": 
dial tcp: lookup proxy.golang.org on [::1]:53: read udp 
[::1]:36438->[::1]:53: read: connection refused
internal/backend/azure/azure.go:22:2: 
github.com/Azure/azure-sdk-for-go/sdk/azcore@v1.3.0: Get 
"https://proxy.golang.org/github.com/%21azure/azure-sdk-for-go/sdk/azcore/@v/v1.3.0.zip": 
dial tcp: lookup proxy.golang.org on [::1]:53: read udp 
[::1]:52528->[::1]:53: read: connection refused

...
build failed: exit status 1
exit status 1
error: in phase 'build': uncaught exception:
%exception #< program: "go" arguments: ("run" "build.go") 
exit-status: 1 term-signal: #f stop-signal: #f>

phase `build' failed after 2.4 seconds
command "go" "run" "build.go" failed with status 1

Looks like go trys using ipv6 for the dns lookup. But
/etc/resolv.conf says
nameserver 10.0.2.3

I don't have a nameserver running on ::1. And this inside qemu, where I 
didn't set up anything specifically for ipv6. I have only site and link 
local ipv6 addresses. Why does go-build-system try to resolve the host 
names at [::1]:53?


Martin Castillo;;; GNU Guix --- Functional package management for GNU

(define-module (restic)
  #:use-module (guix gexp)
  #:use-module (guix packages)
  #:use-module ((guix licenses) #:prefix license:)
  #:use-module (guix gexp)
  #:use-module (guix git-download)
  #:use-module (guix download)
  #:use-module (guix utils)
  #:use-module (guix build-system gnu)
  #:use-module (guix build-system go)
  #:use-module (gnu packages)
  #:use-module (gnu packages acl)
  #:use-module (gnu packages autotools)
  #:use-module (gnu packages bash)
  #:use-module (gnu packages base)
  #:use-module (gnu packages check)
  #:use-module (gnu packages compression)
  #:use-module (gnu packages crypto)
  #:use-module (gnu packages databases)
  #:use-module (gnu packages datastructures)
  #:use-module (gnu packages digest)
  #:use-module (gnu packages dbm)
  #:use-module (gnu packages dejagnu)
  #:use-module (gnu packages ftp)
  #:use-module (gnu packages gettext)
  #:use-module (gnu packages glib)
  #:use-module (gnu packages gnupg)
  #:use-module (gnu packages golang)
  #:use-module (gnu packages gperf)
  #:use-module (gnu packages gtk)
  #:use-module (gnu packages guile)
  #:use-module (gnu packages guile-xyz)
  #:use-module (gnu packages linux)
  #:use-module (gnu packages mcrypt)
  #:use-module (gnu packages ncurses)
  #:use-module (gnu packages nettle)
  #:use-module (gnu packages networking)
  #:use-module (gnu packages onc-rpc)
  #:use-module (gnu packages pcre)
  #:use-module (gnu packages perl)
  #:use-module (gnu packages pkg-config)
  #:use-module (gnu packages serialization)
  #:use-module (gnu packages ssh)
  #:use-module (gnu packages time)
  #:use-module (gnu packages tls)
  #:use-module (gnu packages valgrind)
  #:use-module (gnu packages xml))

(define-public restic-new
  (package
(name "restic-new")
(version "0.15.1")
;; TODO Try packaging the bundled / vendored dependencies in the 'vendor/'
;; directory.
(source (origin
  (method url-fetch)
  (uri (string-append
"https://github.com/restic/restic/releases/download/;
"v" version "/restic-" version ".tar.gz"))
  (file-name (string-append name "-" version ".tar.gz"))
  (sha256
   (base32
"0lhck49zzkdaya20naxf8aryg9p8smk0fr5abnimh0dcrpyq5qzw"))
  ;(patches
   ;(search-patches "restic-0.9.6-fix-tests-for-go1.15.patch"))
   ))
(build-system go-build-system)
(arguments
 `(#:import-path "github.com/res

guix system image fails with: service 'swap-/swapfile' requires 'file-system-/'

[Martin Castillo]

Hi,

I'm testing guix system image with the attachend config, and I get
this error:

/mnt/root/man.config.scm:35:14: warning: the 'target' field is 
deprecated, please use 'targets' instead

guix system: warning: Consider running 'guix pull' followed by
'guix system reconfigure' to get up-to-date packages and security updates.

guix system: error: service 'swap-/swapfile' requires 'file-system-/', 
which is not provided by any service


I can't find anything regarding a file-system-/ service. I do have a 
rootpartition in 'file-systems'. Why is that not sufficient?


If I remove the dependency in swap-devices, it builds, though the system 
gets stuck on boot shortly after generating ssh keys. Picture is 
attached. Any hints for that problem?



Martin



(use-modules (gnu))
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
  (keyboard-layout (keyboard-layout "de" "neo"))
  (host-name "minswap")
  (issue "minimal system with swap, swap dependencies")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("wheel" "netdev" "audio" "video")))
%base-user-accounts))
  (packages
(append
  (list (specification->package "nss-certs"))
  %base-packages))
  (services
(append
  (list (service dhcp-client-service-type)
(service openssh-service-type)
)
  %base-services))
  (bootloader
(bootloader-configuration
  (bootloader grub-efi-bootloader)
  (target "/efi") ; deprecated
  (keyboard-layout keyboard-layout)))
  (mapped-devices
(list (mapped-device
(source (uuid "e35658c9-6e31-431a-9397-2e988e32fe63"))
(target "lukspart")
(type luks-device-mapping))
  (mapped-device
(source "lvmSystem")
(target "lvmSystem-volRoot")
(type lvm-device-mapping
  (file-systems
(cons* (file-system
 (mount-point "/efi")
 (device (uuid "845E-533F" 'fat32))
 (type "vfat"))
   (file-system
 (mount-point "/")
 (device ; mapped device name: lvmSystem-volRoot, FS-label 
is guixsd-root

   (uuid "0584a348-341b-475d-b19f-93b5567ee02b"
 'ext4))
 (type "ext4")
 (needed-for-boot? #t)
 (dependencies mapped-devices))
   %base-file-systems))
  (swap-devices
(list
  (swap-space
(target "/swapfile")
(dependencies (filter
(file-system-mount-point-predicate "/")
file-systems)
  (kernel-arguments
(cons* "resume=/swapfile"
   %default-kernel-arguments

grub failure after fresh system installation with lvm on luks: disk lvmid/.. not found

[Martin Castillo]

Hi,

I'm trying to install guix (inside qemu) with "/" on lvm on luks on 
sda2. sda1 contains the efi partition:


NAMEMAJ:MIN RM   SIZE RO TYPE  MOUNTPOINTS
fd0   2:01 4K  0 disk
sda   8:0030G  0 disk
├─sda18:10   600M  0 part  /mnt/efi
└─sda28:20  29.4G  0 part
  └─lukspart253:00  29.4G  0 crypt
└─lvmSystem-volRoot 253:10  29.4G  0 lvm   /tmp
   /mnt
sr0  11:01 813.9M  0 rom
sr1  11:11  1024M  0 rom

The installation finishes with no problem. However on reboot grub errors 
out:
error: disk 
`lvmid/VyWL64-HaH6-Wnec-fiND-VZye-90JJ-LxWLuj/NvcAEV-u4y7-PAyi-6ctt-

   xlrL-Wbzt-eTfVq3' not found.
   Entering rescue mode...
   grub rescue> insmod lvmi
   error: disk 
`lvmid/VyWL64-HaH6-Wnec-fiND-VZye-90JJ-LxWLuj/NvcAEV-u4y7-PAyi-6ctt-

   xlrL-Wbzt-eTfVq3' not found.
   grub rescue> insmod lvm
   grub rescue> ls
   (hd0) (cd0) (cd1)
   grub rescue> insmod lvm
   grub rescue> cryptomount (hd0)
   Unknown command `cryptomount'.
   grub rescue> cryptomount
   Unknown command `cryptomount'.
   grub rescue> insmod luks2
   error: disk 
`lvmid/VyWL64-HaH6-Wnec-fiND-VZye-90JJ-LxWLuj/NvcAEV-u4y7-PAyi-6ctt-

   xlrL-Wbzt-eTfVq3' not found.
   grub rescue> lin
   Unknown command `lin'.
   grub rescue> insmod luks
   error: disk 
`lvmid/VyWL64-HaH6-Wnec-fiND-VZye-90JJ-LxWLuj/NvcAEV-u4y7-PAyi-6ctt-

   xlrL-Wbzt-eTfVq3' not found.
   grub rescue> insmod lvm
   grub rescue>

Looks like it cannot load the modules luks / luks2. Is it looking for 
those on the root partition?  That lvmid belongs to my root partition. 
Well, that's still encrypted. I attached the grub.cfg created by guix 
system init.


I don't know how this chicken-egg problem is generally solved by grub.
On my day-to-day system I have a similar setup (root on lvm on luks), 
but I don't know, why/how this works. I do have a 1M grub partition, 
though I believe I just followed a bad guide and it is only used on 
bios+gpt systems.


Is there something wrong with my config.scm, or is this a guix bug?

Martin Castillo;; This is an operating system configuration generated
;; by the graphical installer.

(use-modules (gnu))
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
  (keyboard-layout (keyboard-layout "de" "neo"))
  (host-name "geex")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("wheel" "netdev" "audio" "video")))
%base-user-accounts))
  (packages
(append
  (list (specification->package "nss-certs"))
  %base-packages))
  (services
(append
  (list (service xfce-desktop-service-type)
(service openssh-service-type)
(set-xorg-configuration
  (xorg-configuration
(keyboard-layout keyboard-layout
  %desktop-services))
  (bootloader
(bootloader-configuration
  (bootloader grub-efi-bootloader)
  (target "/efi") ; deprecated
  (keyboard-layout keyboard-layout)))
  (mapped-devices
(list (mapped-device
	(source (uuid "e35658c9-6e31-431a-9397-2e988e32fe63"))
	(target "lukspart")
	(type luks-device-mapping))
	  (mapped-device
	(source "lvmSystem")
	(target "lvmSystem-volRoot")
	(type lvm-device-mapping
  (file-systems
(cons* (file-system
 (mount-point "/efi")
 (device (uuid "845E-533F" 'fat32))
 (type "vfat"))
   (file-system
 (mount-point "/")
 (device ; mapped device name: lvmSystem-volRoot, FS-label is guixsd-root
   (uuid "0584a348-341b-475d-b19f-93b5567ee02b"
 'ext4))
 (type "ext4")
	 (dependencies mapped-devices))
   %base-file-systems))
  (swap-devices
(list
  (swap-space
	(target "/swapfile")
	(dependencies (filter
			(file-system-mount-point-predicate "/")
			file-systems)
  (kernel-arguments
(cons* "resume=/swapfile"
	   "resume_offset=34816" ;; got with: # filefrag -e /swapfile
	   %default-kernel-arguments)))
# This file was generated from your Guix configuration.  Any changes
# will be lost upon reconfiguration.
insmod luks
insmod luks2
cryptomount -u e35658c96e31431a93972e988e32fe63

# Set 'root' to the partition that contains /gnu/store.
sea

Re: older and newer Versions of IceCat and icedove

[Martin Castillo]

Hi,

Am 25.02.22 um 19:49 schrieb Gottfried:

My answer is under your email:

Am 25.02.22 um 16:40 schrieb Martin Castillo:



Am 25.02.22 um 15:06 schrieb Gottfried:

Sorry, I don't know any further.

Try executing 
/gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-icecat-91.6.0-guix0-preview1/lib/icecat/icecat 
--version and see what it reports. If it says 91.5, I cannot imagine, 
what causes this.





Hi,

Today I did a guix pull, guix -u and sudo guix system reconfigure 
etc/config.scm


If I open icecat in the graphical way through an icon it gives me the 
version 91.5


I'll repeat what I said earlier:
I use xfce4-panel, and when I added a firefox starter, it automatically 
saved the absolute path to the binary. In guix, this means it will 
always point to that one version.


This means, you need to change the path saved inside the icon.
I don't know whether you use a starter in xfce4-panel or another way, 
but try rightclicking it and see, if you find a menu where you can set a 
commandpath. Then set it to /home/gfp/.guix-profile/bin/icecat.
It is currently probably set to 
/gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-icecat-91.6.0-guix0-preview1/lib/icecat/icecat


This usability problem stems from the way guix works. One would need to 
touch these gui softwares to make them automatically run the version of 
the current guix profile.






Icedove starting in the graphical way through an icon it gave me the 
version 91.5, starting it from the menu and from the terminal it asked 
me to set up an email address and the following message: (today I 
downloaded 91.6, so it should open that one, but it didn’t)





With the channel and manifest from below I ran
$ guix time-machine -C g.channel -- package -m g.manifest
$ guix package -I
icedove 91.6.1  out 
/gnu/store/nmlxvxl2640r74fg9mlacmavpwg2n7xa-icedove-91.6.1

$ icedove --version
 Thunderbird 91.6.1

The installed version is correct. It's just the gui that launches an 
older version. You can fix that the way I described above.


Hth,
Martin

--

g.channel:
;; Diese Kanaldatei knnen Sie bei 'guix pull -C' oder
;; 'guix time-machine -C' angeben, um die Guix-Version zu bekommen, mit
;; der der Inhalt dieses Profils erzeugt wurde.

(list
 ;; Anmerkung: Manche der Pakete in diesem Profil wurden mit diesen 
anderen Commits installiert:

 ;;   "33ce3f1c866231a3015411fdce18a3e72649e2f6"
 ;;   "32ccbc3a1ebf62d90f618ababb206df57f7e767b"
 ;;   "dd823172d86f82dedf8b0a7eb65925db4b164d26"
 (channel
   (name 'guix)
   (url "https://git.savannah.gnu.org/git/guix.git;)
   (branch "master")
   (commit
;"33ce3f1c866231a3015411fdce18a3e72649e2f6") ; for hchyj*-icecat-*
 "ed8d236eb938bc123a9ca8a079fea95edf7df6cd") ;for nml*-icedove-*
   (introduction
 (make-channel-introduction
   "9edb3f66fd807b096b48283debdcddccfea34bad"
   (openpgp-fingerprint
 "BBB0 2DDF 2CEA F6A8 0D1D  E643 A2A0 6DF2 A33A 54FA")

g.manifest:
;; Diese "Manifest"-Datei kann an 'guix package -m' bergeben werden,
;; um den Inhalt Ihres Profils nachzubilden. Sie ist "symbolisch",
;; gibt also nur die Namen der Pakete an. Um genau das gleiche Profil
;; herauszubekommen, mssen Sie auch die verwendeten Kanle
;; nachbilden, wie "guix describe" sie anzeigt. Siehe den Abschnitt
;; "Guix nachbilden" im Handbuch.

(specifications->manifest
  (list
"icedove"
;"icecat"))

Re: older and newer Versions of IceCat

[Martin Castillo]

Am 25.02.22 um 15:06 schrieb Gottfried:

Hi,

 Last idea:
 check that the output of
 realpath /home/gfp/.guix-profile/bin/icecat

 is /gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-    icecat-
 91.6.0-    guix0-preview1/bin/icecat
 and that icecat --version is something like 91.6.

My output:

gfp@Tuxedo ~$ realpath /home/gfp/.guix-profile/bin/icecat
/gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-icecat-91.6.0-guix0-preview1/lib/icecat/icecat 


Sorry, I don't know any further.

Try executing 
/gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-icecat-91.6.0-guix0-preview1/lib/icecat/icecat 
--version and see what it reports. If it says 91.5, I cannot imagine, 
what causes this.


Maybe someone can reproduce that problem, if you give us your channel 
and profile manifest [1]. You can obtain those by running

guix package --export-manifest
guix package --export-channels

as documented in [1] and [2].

[1]: https://guix.gnu.org/manual/en/html_node/Replicating-Guix.html
[2]: 
https://guix.gnu.org/manual/en/html_node/Invoking-guix-package.html#index-profile-manifest





gottfried

Re: older and newer Versions of IceCat

[Martin Castillo]

Am 23.02.22 um 19:20 schrieb Gottfried:


 > How do you open it? Via a startmenu or another graphical way?

to open via startmenu or Alt+F2 or through the graphical way or through 
the terminal opens always the older version of icecat 91.5
That's weird. How do you determine the version? Via the graphical info 

menu or on the commandline?


Can you give us the output of
 > echo $PATH?

gfp@Tuxedo ~$ echo $PATH?
/run/setuid-programs:/home/gfp/.config/guix/current/bin:/home/gfp/.guix-profile/bin:/home/gfp/.guix-profile/sbin:/run/current-system/profile/bin:/run/current-system/profile/sbin? 


What's the output of type -a icecat?




If I run in root:
root@Tuxedo ~# guix package --list-installed
nothing comes up

gottfried



Am 22.02.22 um 23:20 schrieb Martin Castillo:

Hi,

Am 22.02.22 um 19:33 schrieb Gottfried:

Hi,
I upgraded my system and IceCat updated to 91.6 but when opening it, 
the version 91.5 opens.


How do you open it? Via a startmenu or another graphical way? I use 
xfce4-panel, and when I added a firefox starter, it automatically 
saved the absolute path to the binary. In guix, this means it will 
always point to that one version.


What version starts if you launch it from terminal (or Alt+F2)?
Can you give us the output of
echo $PATH?




In my generations there are + and - listed:

Generation 33    18. Februar 2022 15:35:27
  + gnucash  4.6 out 
/gnu/store/lxr7lq0hdwz77a1cisvz5qwkwznwxqkm-gnucash-4.6
  + hpcguix-web  0.2.0   out 
/gnu/store/4s9yx1dkmckx4h9zwn64s60npi4b64yg-hpcguix-web-0.2.0
  + icedove  91.6.1  out 
/gnu/store/przjhqp2bhkisb3c5vn4xrkcwq34slma-icedove-91.6.1
  + claws-mail   4.0.0   out 
/gnu/store/q0181iky0bnf1hs9g45v8972kh9swcgv-claws-mail-4.0.0
  + denemo   2.5.0   out 
/gnu/store/b63ilh11yy1i8vyfr66rrhdrwykvcd2s-denemo-2.5.0
  + calibre  5.21.0  out 
/gnu/store/y90mkqib721kh1fr0mhcc3kgarf7yqvd-calibre-5.21.0
  + vlc  3.0.16  out 
/gnu/store/6mbhbs0qna8bwrl7458cc5af8rcr6fkl-vlc-3.0.16
  + ungoogled-chromium    98.0.4758.102-1 out 
/gnu/store/zl70ybb266vrwnfhq8d86ml4lvnb3pfy-ungoogled-chromium-98.0.4758.102-1 

  + icecat   91.6.0-guix0-preview1    out 
/gnu/store/hchyj7faw7f98njsja27i89v3jk7bbqm-icecat-91.6.0-guix0-preview1
  + libreoffice  7.1.4.2 out 
/gnu/store/xpgm0m6vfhsk813fwvdin6hi49h6bykr-libreoffice-7.1.4.2
  + recutils 1.8 out 
/gnu/store/0xkqy1cqy3mwxij51fnxr3q1ij0fdmsx-recutils-1.8
  - gnucash  4.6 out 
/gnu/store/l5blqbqd48pmq9k2dqnryl3f6mxiqhd0-gnucash-4.6
  - hpcguix-web  0.2.0   out 
/gnu/store/98bxvpbqw727cglimkg4x4vlxglysp6n-hpcguix-web-0.2.0
  - icedove  91.5    out 
/gnu/store/rmly36prvd1jaxx91p8h7gwwg3jhdllh-icedove-91.5
  - claws-mail   4.0.0   out 
/gnu/store/lrv5i33rdjcliwmg7svclcw855w4yary-claws-mail-4.0.0
  - denemo   2.5.0   out 
/gnu/store/kh5vz41z5w5n7ixv6n0dwk65f1ibfxmx-denemo-2.5.0
  - calibre  5.21.0  out 
/gnu/store/nrc8qa4ah9v5lw9vvbr0p9ll10qm06yk-calibre-5.21.0
  - vlc  3.0.16  out 
/gnu/store/rgzzv23rwa2g4rfcrig1snpp2mw6x9hq-vlc-3.0.16
  - ungoogled-chromium    97.0.4692.99-1  out 
/gnu/store/3j6x96869ajiglrxhgv2yxrnnmlyvk68-ungoogled-chromium-97.0.4692.99-1 

  - icecat   91.5.0-guix0-preview1    out 
/gnu/store/r9n91hxqnsfvp5m55s40i3sxf4rpl8c2-icecat-91.5.0-guix0-preview1
  - libreoffice  7.1.4.2 out 
/gnu/store/kh9dff0ri6768r6msvnm0xajhcds3wzm-libreoffice-7.1.4.2
  - recutils 1.8 out 
/gnu/store/z53jx3vg2q1cszyakh8mmzrmx1zk4q0g-r


Did I make a mistake and install packages by chance in root and thatś 
why my versions stay on the installed ones without being able in my 
user account to upgrade to newer versions?


If you run guix pull as your user and not as root, it installs the 
programs in your profile.
Login as root (sudo -i) and let guix list the installed packages to 
see roots packages.



If so, whatś the best way to delete my packages in root, which don´t 
need to be there?, so that, if I upgrade in my user account, the 
packages get updated.

$ sudo -i
# guix package -r icecat ...



If there is an other mistake, I would be glad to know what to do.

gottfried


Martin

Guixsd boot stop

[Martin Castillo]

Hi,

I installed guixsd 1.3.0 inside qemu with efi. (On this list: "Create 
new EFI guixSD installaton from existing bios guixsd installation") That 
worked (though I noticed some problems inside the install process) and 
booted.
I then put the harddrive into a real machine (Asus X73B). It could not 
find the boot data and started the uefi setup. I then put the guixsd 
1.3. image onto a usbdrive and booted that. I followed the graphical 
installer until network was set up. Then switched to console:


mount /dev/sda3 /mnt
mount /dev/sda1 /mnt/boot/efi
cd /mnt
mount --bind /dev dev
mount --bind /sys sys
mount --bind /proc proc
chroot /mnt /bin/sh
bash
echo nameserver 192.168.178.1 >/etc/resolv.conf # NetworkManager inside 
qemu created a now invalid resolv.conf. fix that

# my user had an up to date guix, so I used that
/home/mcd/.config/guix/current/bin/guix-daemon 
--build-users-group=guixbuild --substitute-urls https://ci.guix.gnu.org &

/home/mcd/.config/guix/current/bin/guix system reconfigure /etc/config.scm

(There should be an official, documented way to chroot into a guixsd 
system!)


At the end of the installation I ran guix system build /etc/config.scm 
and got 2 dmesg lines about missing firmware [1]:

:03:00.0: Missing Free firmware (non-Free firmware loading is disabled)
r8169 000:03:00.0: Unable to load firmware /*(DEBLOBBED)*/ (-2)

I ignored it and rebooted. No it is stuck [1] just like in 
.


Even AltGr+SysRq+REISUB does not work.

I know you cannot recommend unfree software, but I'd like to use that 
machine instead of throwing it away, even if I need some binary blobs.

How can I install the necessary firmware?

TIA,
Martin

[1]: https://seafile.zfn.uni-bremen.de/d/08f98eca117f41ebaea4/

Re: How to work with npm from guix shell

[Martin Castillo]

Hi,

Am 18.02.22 um 13:27 schrieb Roland Everaert via:

Unfortunatelly I got the following error during the execution of 'npm install 
-g @solid/community-server'


The -g flag tells npm that you want to install the package globaly, 
which on guix, is inside the read-only gnu store.
You can either just omit -g to install the packages in the current 
directory, or use 
, 
which was proposed here: 





I am totally illiterate about node.js and its ecosystem. So, how to setup a 
proper environment for running node.js application with guix shell or guix in 
general?


Guix packages packages from other language specific packagers, like pip. 
But I don't think we have node modules tracked by guix.

AFAIK, the above solution is best practice.

Someone else might know more.

Martin

Re: Create new EFI guixSD installaton from existing bios guixsd installation

[Martin Castillo]

I'm back.

On 17.02.22 01:40, Martin Castillo wrote:
But here I did not know how to proceed? There is no grub-install in this 
chrooted /gnu/store(, which is weird, because a complete system should 
contain all files to update itself, right?).


What am I missing?
Or can you only create an EFI system installation when you're already 
running on one?


So I ran the install image in qemu with EFI firmware and ran through the 
graphical installer. The guided graphical install failed, manually 
running guix system init worked. You can find the outputs under [1].

Looks like a bug, but I have no idea how to give you more info.

To summarize:

1. I wish it would be possible to system init an efi system from a bios 
system.


2. The guided gui installer should not fail when installing inside qemu 
with EFI. The qemu cmdline was

$ qemu-system-x86_64 -m 2048 -smp 3 -enable-kvm \
-bios /usr/share/edk2-ovmf/x64/OVMF.fd \
  -nic user,model=virtio-net-pci -boot menu=on,order=d \
  -drive file=/dev/nbd1 \
  -drive media=cdrom,file=guix-system-install-1.3.0.x86_64-linux.iso


Martin

[1]: https://seafile.zfn.uni-bremen.de/d/e9acc264718f4ae7b703/


OpenPGP_signature
Description: OpenPGP digital signature

Create new EFI guixSD installaton from existing bios guixsd installation

[Martin Castillo]

[Sorry about the uncomplete message I (might?) have sent earlier.]

Hi,

I want to create a bootable guixsd installation for an EFI x86-64 
computer. But I don't want to use the installation disk, but create that 
from an existing guixsd system. The existing system is a BIOS system 
(and I'm running it inside qemu).


The new system will be installed on a black harddrive, so I will call it 
'black'. (But note, that I haven't connected it yet. I'm using a qcow 
image instead.) I manually partitioned the black drive, I created a fat, 
swap and ext4 partition with the same uuids as in the system 
configuration 'liteon.scm' and mounted the / and efi partition:

sdb 8:16   0 111,8G  0 disk
├─sdb1  8:17   0   549M  0 part  /mnt/boot/efi
├─sdb2  8:18   0   7,5G  0 part
└─sdb3  8:19   0 103,8G  0 part  /mnt


Inside the existing system I ran:
$ guix system init liteon.scm /mnt/
... downloaded everything. errord out because of missing rights
$ sudo guix system init liteon.scm  /mnt/
Passwort:
/gnu/store/69gsk4rzz7gfa73qpzj2rfw5hs9d3pl5-system
/gnu/store/ay76p7yzy90r0yi3fg4s8c2sg8c36j7y-grub.cfg

Betriebssystem unter „/mnt/“ wird initialisiert …
Nach „/mnt“ kopieren …
populating '/mnt'...
substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
aktualisiert … 100.0%

Folgende Ableitungen werden erstellt:
/gnu/store/42afgd18szcic0hiqmqqvnvnwl7gxkd1-install-bootloader.scm.drv
/gnu/store/akdppfv8bfyx3qzkjb2hzphlvp7ncq9r-module-import.drv
/gnu/store/bnjl98z303ax2qx8iwai3dhgzsfnsgsm-module-import.drv

1,4 MB werden heruntergeladen
  module-import-compiled  689KiB 2.3MiB/s 00:00 
[##] 100.0%
  module-import-compiled  689KiB 2.6MiB/s 00:00 
[##] 100.0%
/gnu/store/akdppfv8bfyx3qzkjb2hzphlvp7ncq9r-module-import.drv wird 
erstellt …
/gnu/store/bnjl98z303ax2qx8iwai3dhgzsfnsgsm-module-import.drv wird 
erstellt …
/gnu/store/42afgd18szcic0hiqmqqvnvnwl7gxkd1-install-bootloader.scm.drv 
wird erstellt …
guix system: error: 
'/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install 
--boot-directory /mnt/boot --bootloader-id=Guix --efi-directory 
/mnt/boot/efi' exited with status 1; output follows:



/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install: 
Fehler: 
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/lib/grub/i386-pc/modinfo.sh 
existiert nicht. Bitte geben Sie --target oder --directory an.


/gnu/store/mb7...-grub-efi-2.04/lib/grub/i386-pc/ does not exist,
/gnu/store/mb7...-grub-efi-2.04/lib/grub/x86_64-efi/ does exist.

grub-install seems to want an explicit target architecture. I didn't 
find anything in the guix-doc (the bootloader documentation). So I just 
tried to manually edit and run the failing command:
$ sudo 
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install 
--boot-directory /mnt/boot --bootloader-id=Guix --efi-directory 
/mnt/boot/efi --target=x86_64-efi

Passwort:
x86_64-efi wird für Ihre Plattform installiert.
EFI variables are not supported on this system.
EFI variables are not supported on this system.
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install: 
Fehler: efibootmgr schlug beim Registrieren des Boot-Eintrags fehl: 
Datei oder Verzeichnis nicht gefunden.


Then I thought: Maybe do everythung but installing the bootloader. And 
then try to chroot into /mnt and run guix system reconfigure.


$ sudo guix system init --no-bootloader liteon.scm  /mnt/
(Unfortunately, this copies everything again. Is there a way to prevent 
this? Or did the earlier guix system everything it would do with 
--no-bootloader, before it failed installing the bootloader?)



At this point, /mnt/{bin,run} were both empty. I expected /mnt/bin/sh 
and /run/current-system to exist!?

I checked the target of /mnt/var/guix/profiles/system and ran
$ sudo chroot /mnt/ /gnu/store/87kif...-bash/bin/sh
# . /var/guix/profiles/system/profile/etc/profile

But here I did not know how to proceed? There is no grub-install in this 
chrooted /gnu/store(, which is weird, because a complete system should 
contain all files to update itself, right?).


What am I missing?
Or can you only create an EFI system installation when you're already 
running on one?


Thanks,
Martin Castillo
;; This is an operating system configuration generated
;; by the graphical installer.

(use-modules (gnu))
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
  (keyboard-layout (keyboard-layout "de" "neo"))
  (host-name "geex")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("w

Create new EFI guixSD installaton from existing bios guixsd installation

[Martin Castillo]

Hi,

I want to create a bootable guixsd installation for an EFI x86-64 
computer. But I don't want to use the installation disk, but create that 
from an existing guixsd system. The existing system is a BIOS system 
(and I'm running it inside qemu).


The new system will be installed on a black harddrive, so I will call it 
'black'. (But note, that I haven't connected it yet. I'm using a qcow 
image instead.) I manually partitioned the black drive, I created a fat, 
swap and ext4 partition with the same uuids as in the system 
configuration 'liteon.scm' and mounted the / and efi partition:

sdb 8:16   0 111,8G  0 disk
├─sdb1  8:17   0   549M  0 part  /mnt/boot/efi
├─sdb2  8:18   0   7,5G  0 part
└─sdb3  8:19   0 103,8G  0 part  /mnt


Inside the existing system I ran:
$ guix system init liteon.scm /mnt/
... downloaded everything. errord out because of missing rights
$ sudo guix system init liteon.scm  /mnt/
Passwort:
/gnu/store/69gsk4rzz7gfa73qpzj2rfw5hs9d3pl5-system
/gnu/store/ay76p7yzy90r0yi3fg4s8c2sg8c36j7y-grub.cfg

Betriebssystem unter „/mnt/“ wird initialisiert …
Nach „/mnt“ kopieren …
populating '/mnt'...
substitute: Liste der Substitute von „https://ci.guix.gnu.org“ wird 
aktualisiert … 100.0%

Folgende Ableitungen werden erstellt:
   /gnu/store/42afgd18szcic0hiqmqqvnvnwl7gxkd1-install-bootloader.scm.drv
   /gnu/store/akdppfv8bfyx3qzkjb2hzphlvp7ncq9r-module-import.drv
   /gnu/store/bnjl98z303ax2qx8iwai3dhgzsfnsgsm-module-import.drv

1,4 MB werden heruntergeladen
 module-import-compiled  689KiB 2.3MiB/s 00:00 
[##] 100.0%
 module-import-compiled  689KiB 2.6MiB/s 00:00 
[##] 100.0%
/gnu/store/akdppfv8bfyx3qzkjb2hzphlvp7ncq9r-module-import.drv wird 
erstellt …
/gnu/store/bnjl98z303ax2qx8iwai3dhgzsfnsgsm-module-import.drv wird 
erstellt …
/gnu/store/42afgd18szcic0hiqmqqvnvnwl7gxkd1-install-bootloader.scm.drv 
wird erstellt …
guix system: error: 
'/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install 
--boot-directory /mnt/boot --bootloader-id=Guix --efi-directory 
/mnt/boot/efi' exited with status 1; output follows:



/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install: 
Fehler: 
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/lib/grub/i386-pc/modinfo.sh 
existiert nicht. Bitte geben Sie --target oder --directory an.


grub-install wants an explicit target architecture. I didn't find 
anything in the guix-doc (the bootloader documentation). So I just tried 
to manually edit and run the failing command:
$ sudo 
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install 
--boot-directory /mnt/boot --bootloader-id=Guix --efi-directory 
/mnt/boot/efi --target=x86_64-efi

Passwort:
x86_64-efi wird für Ihre Plattform installiert.
EFI variables are not supported on this system.
EFI variables are not supported on this system.
/gnu/store/mb7sg4q398ikdlx56whp2p8msa8pyakw-grub-efi-2.04/sbin/grub-install: 
Fehler: efibootmgr schlug beim Registrieren des Boot-Eintrags fehl: 
Datei oder Verzeichnis nicht gefunden.


Then I tried
;; This is an operating system configuration generated
;; by the graphical installer.

(use-modules (gnu))
(use-service-modules desktop networking ssh xorg)

(operating-system
  (locale "de_DE.utf8")
  (timezone "Europe/Berlin")
  (keyboard-layout (keyboard-layout "de" "neo"))
  (host-name "geex")
  (users (cons* (user-account
  (name "mcd")
  (comment "Martin")
  (group "users")
  (home-directory "/home/mcd")
  (supplementary-groups
'("wheel" "netdev" "audio" "video")))
%base-user-accounts))
  (packages
(append
  (list (specification->package "nss-certs"))
  %base-packages))
  (services
(append
  (list (service xfce-desktop-service-type)
(service openssh-service-type)
(set-xorg-configuration
  (xorg-configuration
(keyboard-layout keyboard-layout
  %desktop-services))
  (bootloader
(bootloader-configuration
  (bootloader grub-efi-bootloader)
  (target "/boot/efi")
  (keyboard-layout keyboard-layout)))
  (swap-devices
(list (uuid "30bbe6de-b824-4d8a-bc12-3dc8c7b46467")))
  (file-systems
(cons* (file-system
 (mount-point "/")
 (device
(uuid "8757ed03-b543-4416-8779-1fff91a4d027"
	'ext4))
 (type "ext4"))
   (file-system
 (mount-point "/boot/efi")
 (device
(uuid "257E-79BA" 'fat32))
 (type "vfat"))
   %base-file-systems)))


OpenPGP_signature
Description: OpenPGP digital signature

Re: `guix package -u` upgrades packages to themselves

[Martin Castillo]

On 11.04.2018 11:40, Pierre Neidhardt wrote:
>
> Sorry if I was confusing:  the output remains the same after upgrade.
>
> I can repeat the command indefinitely, it keep "upgrading" the same set
> of packages as per the output (but it does not actually do anything).
>

I see the same on 792ddfdc76147d494052fa93ecbfdbee7e72be68.


-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC
[13:56:45]pi@pi:~/code/updapr$ guix package -u
guile: warning: failed to install locale
warning: failed to install locale: Invalid argument
The following package will be upgraded:
   guile2.2.3 -> 2.2.3  
/gnu/store/nb39n4ia25fdmzcfm628il1jqdkipf7s-guile-2.2.3

substitute: guile: warning: failed to install locale
substitute: warning: failed to install locale: Invalid argument
nothing to be done
[14:19:21]pi@pi:~/code/updapr$ ll ~/.guix-profile/bin/guile
lrwxrwxrwx 5 root root 65 Jan  1  1970 /home/pi/.guix-profile/bin/guile -> 
/gnu/store/nb39n4ia25fdmzcfm628il1jqdkipf7s-guile-2.2.3/bin/guile*
[14:20:29]pi@pi:~/code/updapr$ guix package -u
guile: warning: failed to install locale
warning: failed to install locale: Invalid argument
The following package will be upgraded:
   guile2.2.3 -> 2.2.3  
/gnu/store/nb39n4ia25fdmzcfm628il1jqdkipf7s-guile-2.2.3

substitute: guile: warning: failed to install locale
substitute: warning: failed to install locale: Invalid argument
nothing to be done
[14:21:18]pi@pi:~/code/updapr$ ll ~/.guix-profile/bin/guile
lrwxrwxrwx 5 root root 65 Jan  1  1970 /home/pi/.guix-profile/bin/guile -> 
/gnu/store/nb39n4ia25fdmzcfm628il1jqdkipf7s-guile-2.2.3/bin/guile*


signature.asc
Description: OpenPGP digital signature

Re: VLC could not decode the format "ssa " (SubStation Alpha subtitles)

[Martin Castillo]

I think he means that you need to add libass to the inputs of the vlc
package. look a few lines under where $ guix edit vlc; takes you. try
adding it there.

On 13.03.2018 03:01, Dang Duong wrote:
> My video.scm already has libass. I tried removing and reinstalling VLC
> but it doesn't work.
>  Am I missing something?
> 
> On Mon, Mar 12, 2018 at 5:14 PM, Ludovic Courtès  > wrote:
> 
> Hello,
> 
> Dang Duong >
> skribis:
> 
> > I have installed the latest version of VLC (2.2.8) on Trisquel 7.0 using
> > Guix. It works normally with .srt files but cannot open .ass files. I 
> tried
> > installing libass but it still doesn't work.
> 
> It seems that our vlc package lacks libass as an input:
> 
>  
> 
> https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/video.scm#n810
> 
> 
> 
> Could you try adding libass in the vlc package definition, then install
> it, and see if it works?
> 
> If you’re not familiar with Guix packaging, see
>  
> >,
> and do not hesitate to ask for guidance here or on #guix on IRC!
> 
> HTH,
> Ludo’.
> 
> 

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC



signature.asc
Description: OpenPGP digital signature

Re: Dual-booting solution with only one actual Grub installation

[Martin Castillo]

Hi,

On 12.03.2018 19:43, Thorsten Wilms wrote:
> There's a convenient dual-booting solution that I think hasn't been
> mentioned in these circles. You can let Grub2 offer an item that will
> load another grub.cfg.

I mentioned it in [0], and I think I got the idea from some other post
in guix-devel or help-guix.

Martin

[0]: https://lists.gnu.org/archive/html/guix-devel/2017-12/msg00134.html
-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC



signature.asc
Description: OpenPGP digital signature

export openssh's AcceptEnv option in openssh-configuration

[Martin Castillo]

Hi,
not sure whether this belongs here, guix-devel or bug-guix.

I'd like to be able to set AcceptEnv in sshd_config from config.scm.
Currently, this option is not exported in openssh-configuration.

General question: How does guixsd handle this? I think this is a general
problem. Do we expect everyone who wants to configure something, that is
not yet configurable, to patch guix themself or ask the others to do it?
What about adding a field 'extraOptions' which is a simple string that
is appended to the configuration file?


Martin

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC



signature.asc
Description: OpenPGP digital signature

Re: ssh-daemon service fails to auto start

[Martin Castillo]

On 06.02.2018 21:01, Leo Famulari wrote:
> On Sat, Jan 27, 2018 at 11:08:23PM +0100, Martin Castillo wrote:
>> my configuration is (similar to) desk,nossh.scm.
> 
> Can you share the QEMU command line you use to run desk,nossh.scm?
> 

qemu-system-x86_64 -m 2048 -smp 2 -enable-kvm \
-net nic,model=virtio \
-net user,hostfwd=tcp::5561-: \
-drive file=guixsd.img

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC



signature.asc
Description: OpenPGP digital signature

Re: ssh-daemon service fails to auto start

[Martin Castillo]

On 29.01.2018 18:00, Ludovic Courtès wrote:
> Hi Martin,
> 
> Martin Castillo <casti...@uni-bremen.de> skribis:
> 
>> i have a problem with the ssh-service. I want it to autostart on system
>> boot.
>>
>> my configuration is (similar to) desk,nossh.scm.
> 
> I tried this config in ‘guix system vm’ and sshd is automatically
> started on boot, as can be seen with ‘herd status ssh-daemon’ etc.
> 
> Does this config work for you in ‘guix system vm’?

No.


> Could you check /var/log/shepherd.log?  It will tell you which services,
> besides sshd, failed to start.
> 
> ‘ssh-daemon’ depends only on ‘syslogd’, so as long as ‘syslogd’ is
> started, ‘ssh-daemon’ should start.
> 

I checked /var/log/shepherd.log. Only user-homes and ssh-daemon could
not be started.

Has this maybe to do with missing randomness to create keys? (Even
though there are already server keys.)

Maybe shepherd should store the output of launched programs? Is that
really not possible? I can't be the first one to need that.

Martin

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

ssh-daemon service fails to auto start

[Martin Castillo]

hey guix,

i have a problem with the ssh-service. I want it to autostart on system
boot.

my configuration is (similar to) desk,nossh.scm. sshd does start on
system boot with the no-desk,ssh-works.scm and no-desk,ssh-semi.scm
configuration, but is not reachable from outside the vm with the latter
configuration.

now, one difference between the latter two is the dhcp-client-service.

For some reason, I can not use that service with xfce-service. But I
don't know, how related that is.

Can someone with more insight tell me, why sshd won't autostart?
shepherd tells me only 'ssh-daemon could not be started'. Thank you
shepherd, very helpful!


Martin

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC
;; Running in Qemu.
;; This config has no gui. ssh-daemon starts and is reachable from outside

(use-modules (gnu) (gnu system nss))
(use-service-modules ssh)
(use-service-modules networking)
(use-service-modules base desktop)
(use-service-modules xorg)

(use-package-modules wm certs linux)

(define cyborg-quirk
"Section \"InputClass\"
	Identifier \"Saitek Cyborg R.A.T.5 Mouse\"
	MatchIsPointer \"on\"
	MatchProduct \"Saitek Cyborg R.A.T.5 Mouse\"
	MatchVendor \"Saitek|SAITEK\"
	MatchDevicePath \"/dev/input/event*\"
	Option \"ButtonMapping\" \"1 2 3 4 5 6 7 2 9 4 5 12 0 0 0\"
	Option \"ZAxisMapping\" \"10 11\"
#	Option \"AutoReleaseButtons\" \"12 13 14 15\"
EndSection")

(define %my-xorg-conf
	(xorg-configuration-file #:extra-config (list cyborg-quirk)))

(define %my-startx
	(xorg-start-command #:configuration-file %my-xorg-conf))

(define %my-services
   ;; My very own list of services.
   (modify-services %base-services
 (guix-service-type config =>
(guix-configuration
 (inherit config)
 (substitute-urls (cons "https://berlin.guixsd.org;
		(guix-configuration-substitute-urls config)
	 (slim-service-type config =>
			(slim-configuration
			  (inherit config)
			  (startx %my-startx)
			  (default-user "mcd")

(operating-system
  (host-name "komputilo")
  (timezone "Europe/Berlin")
  (locale "de_DE.UTF-8")

  ;; Assuming /dev/sdX is the target hard disk, and "my-root" is
  ;; the label of the target root file system.
  (bootloader (bootloader-configuration (bootloader grub-bootloader)
	(target "/dev/sda")
  ;(menu-entries '((menu-entry
;		(label "Grub Nixos")
;		(multiboot "(hd0)") ; still unsupported
;   )))
	))
  (file-systems (cons (file-system
(device "muh-guix")
(title 'label)
(mount-point "/")
(type "ext4"))
  %base-file-systems))

  ;; This is where user accounts are specified.  The "root"
  ;; account is implicit, and is initially created with the
  ;; empty password.
  (users (cons (user-account
(name "mcd")
(comment "Martin Castillo")
(group "users")

;; Adding the account to the "wheel" group
;; makes it a sudoer.  Adding it to "audio"
;; and "video" allows the user to play sound
;; and access the webcam.
(supplementary-groups '("wheel" "audio" "video"
	"netdev"))
(home-directory "/home/mcd"))
   %base-user-accounts))

  ;; Globally-installed packages.
  ;(packages (cons tcpdump %base-packages))
  ;; Add a bunch of window managers; we can choose one at
  ;; the log-in screen with F1.
  (packages (cons* ;dmenu ;window managers
   nss-certs  ;for HTTPS access
   kbd-neo
   %base-packages))

  ;; Add services to the baseline: a DHCP client and
  ;; an SSH server.
  (services (cons* (dhcp-client-service)
   (console-keymap-service (file-append kbd-neo "/share/keymaps/neo.map"));;
   (service openssh-service-type
(openssh-configuration
  (port-number )))

   (simple-service 'store-my-config etc-service-type
   `(("config.scm" ,(local-file (assoc-ref
  (current-source-location)
  'filename)
;   (xfce-desktop-service)
   %my-services))

  ;; Use the "desktop" services, which include the X

Re: adding files to xorg.conf.d

[Martin Castillo]

hi

On 22.01.2018 07:30, Chris Marusich wrote:
> Martin Castillo <casti...@uni-bremen.de> writes:
> 
>> hi,
>>
>> I want to add one file to xorg.conf.d, but can't find the right way to
>> do so. my current solution just appends my xsettings to xorg.conf:
>>
>> (define cyborg-quirk
>> "Section \"InputClass\"
>> #yadayada
>> EndSection")
>>
>> (define %my-xorg-conf
>> (xorg-configuration-file #:extra-config (list cyborg-quirk)))
>>
>> (define %my-startx
>> (xorg-start-command #:configuration-file %my-xorg-conf))
>>
>> (define %my-services
>>;; My very own list of services.
>>(modify-services %desktop-services
>>  (slim-service-type config =>
>> (slim-configuration
>>   (inherit config)
>>   (startx %my-startx)
>>
>>
>> currently, xorg.conf.d is created by a function that gets a list of
>> X-modules or so. but it doesn't seem like there is any interface to
>> pass any of these functions a gexp that evaluates to a file or so. Is my
>> observation correct? If yes, is this wanted?
> 
> This is the only way I know of right now to customize the Xorg config.
> For example, this same solution was recommended here, also:
> 
> https://lists.gnu.org/archive/html/guix-devel/2017-06/msg00264.html
> 
> Does this solution not solve your problem?
> 

okay. It does solve my problem but I think it is not so 'clean'. So I
wondered, if the clean method is possible.
-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC



signature.asc
Description: OpenPGP digital signature

adding files to xorg.conf.d

[Martin Castillo]

hi,

I want to add one file to xorg.conf.d, but can't find the right way to
do so. my current solution just appends my xsettings to xorg.conf:

(define cyborg-quirk
"Section \"InputClass\"
#yadayada
EndSection")

(define %my-xorg-conf
(xorg-configuration-file #:extra-config (list cyborg-quirk)))

(define %my-startx
(xorg-start-command #:configuration-file %my-xorg-conf))

(define %my-services
   ;; My very own list of services.
   (modify-services %desktop-services
 (slim-service-type config =>
(slim-configuration
  (inherit config)
  (startx %my-startx)


currently, xorg.conf.d is created by a function that gets a list of
X-modules or so. but it doesn't seem like there is any interface to
pass any of these functions a gexp that evaluates to a file or so. Is my
observation correct? If yes, is this wanted?

Martin

ghost command: suspend

[Martin Castillo]

hey,

when I run `suspend` on my guixsd installation in qemu, bash
freezes/sleeps and the system does not suspend. `whereis suspend` does
not find any path, but bash completes `susp[Tab]` to suspend.

can some people test this on their machines? and can someone maybe
explain, what's going on here?

Martin
-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

Re: correct 'guix system reconfigure config.scm' files

[Martin Castillo]

Hi,

On 11.01.2018 01:33, Quiliro Ordonez Baca wrote:
> 
> I have 3 different config files:
> pastebin.com/tmLvPAYG
> pastebin.com/jmYHb6us
> pastebin.com/aGHcY3Ey
> I do not remember which one was good.
good in what sense?

> Please criticize the sintax where
> appropriate. Thank you.
I think guile can critisize syntax errors, too.
try `guix system build config.scm`

Martin

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

Re: problem with Icecat in GuixSD 0.14.0

[Martin Castillo]

I wasn't asking for help. I just wanted to say that it may be a problem
outside of guix.

On 21.12.2017 13:04, Adonay Felipe Nogueira wrote:
> Perhaps you should tell the developers of GNU IceCat (if you're using
> GNU IceCat) or GNU LibreJS (if you're using GNU LibreJS in a browser
> other than GNU IceCat).
> 
> I don't know if help-guix mailing list will be able to answer as
> precisely as the original projects.
> 
> 2017-12-14T00:06:56+0100 Martin Castillo wrote:
>> I have problems with librejs on android. Even if I deactivate it, some
>> sites don't work. They work in firefox, though. This might be (part of)
>> the problem.
>>
>> Martin
> 

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

xorg documentation only available when installing xorg-server

[Martin Castillo]

I have a graphical system running and wanted to man xorg.conf.
But that failed. I have to install xorg-server to get the manuals.

Why is that so? the system wide manpath was included in my
environment(as is default on guixsd).


Martin
-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

Re: problem with Icecat in GuixSD 0.14.0

[Martin Castillo]

Hi,

> 
> It says that it cannot play because JavaScript is not found, and that I
> need to enable it in the settings of my webbrowser.  But it should
> already be: when I click on the LibreJS button, it gives me the option
> to re-"Block all nonfree/nontrivial scripts from this page".
> 
> Of course, this may quite likely be an upstream problem.
> 
> —Marco
> 

I have problems with librejs on android. Even if I deactivate it, some
sites don't work. They work in firefox, though. This might be (part of)
the problem.

Martin

-- 
GPG: 7FDE 7190 2F73 2C50 236E  403D CC13 48F1 E644 08EC

Re: I cannot install bootloader in Virtualbox

[Martin Castillo]

Hi,

it looks like sda consists of a single filesystem, i.e. you did something like 
mkfs.ext4 /dev/sda instead of formatting the disk with fdisk /dev/sda, creating 
one or more partitions an then creating a filesystem on one partition: 
mkfs.ext4 /dev/sda1. 

However, this was maybe intended from you. In this case you need to 
grub-install --force --boot-directory=/mnt/boot /dev/sda. As far as I know, 
this kind of setup might not be bootable, because there may be no space at the 
first 512 bytes of the disk. (Maybe I'm wrong, if ext4 leaves the first 512 
bytes free.)

If you are doing the first solution, use bios and not uefi, there are 2 ways to 
format a disk. Gpt and msdos/mbr. If you choose gpt, you'll need a second 1mb 
partition for grub.

HTH.

Martin

Lilypond version

[Martin Castillo]

I currently have no access to a system with guix installed, but did you try
$ guix package install lilypond@2.18.2
?

Martin