Locales

Hello, In current ‘master’, libc builds all the available locales and carries them with itself (~110 MiB.) In ‘core-updates’ commit 97e1120, I changed libc to build and install zero locales. Instead, locale definitions that users may want must be specified in the ‘operating-system’ declaration,

gcl 2.6.12 breaks maxima

Hi, I noticed that after the update of gcl from 2.6.11 to 2.6.12, maxima fails to build. http://hydra.gnu.org/job/gnu/master/maxima-5.34.1.x86_64-linux Could we revert to the previous version until the incompatibility between the two is fixed? Regards, Fede

Re: gcl 2.6.12 breaks maxima

Hello, On Thu, Nov 27, 2014 at 10:54:49AM +0100, Federico Beffa wrote: > Could we revert to the previous version until the incompatibility > between the two is fixed? do you see a way of fixing it? That would be the best solution! Or is it a documented and known problem? Andreas

Re: gcl 2.6.12 breaks maxima

On Thu, Nov 27, 2014 at 10:56 AM, Andreas Enge wrote: > Hello, > > On Thu, Nov 27, 2014 at 10:54:49AM +0100, Federico Beffa wrote: >> Could we revert to the previous version until the incompatibility >> between the two is fixed? > > do you see a way of fixing it? That would be the best solution! >

Re: gcl 2.6.12 breaks maxima

On Thu, Nov 27, 2014 at 11:00 AM, Federico Beffa wrote: > On Thu, Nov 27, 2014 at 10:56 AM, Andreas Enge wrote: >> Hello, >> >> On Thu, Nov 27, 2014 at 10:54:49AM +0100, Federico Beffa wrote: >>> Could we revert to the previous version until the incompatibility >>> between the two is fixed? >> >>

Re: wip-arm

John Darrington skribis: > It contains the start of a port to an Arm architecture. Currently > it hasn't got very far unfortunately. It allow you to cross-build the bootstrap tarballs for ARM, right? That sounds like a good start, thanks for all the work! Ludo’.

Re: wip-arm

On Thu, Nov 27, 2014 at 02:19:44PM +0100, Ludovic Court??s wrote: John Darrington skribis: > It contains the start of a port to an Arm architecture. Currently > it hasn't got very far unfortunately. It allow you to cross-build the bootstrap tarballs for ARM, right?

use zsh as login shell

Currently, zsh doesn't work out of box. IMO, we should use `/etc/environment' for system variables: LANG=en_US.UTF-8 PATH=/run/current-system/profile/bin:/run/setuid-programs And add `pam_env.so' to /etc/pam.d/login, to make tty login work. With `/etc/profile', `/etc/zlogin' just: sou

Re: [PATCH] gnu: Add Tiled

l...@gnu.org (Ludovic Courtès) skribis: > Andreas Enge skribis: > >> The problem should now be fixed in 5dee80d. I tried to enter something >> in the file open dialog of tiled, and it worked. > > Nope, it doesn’t work on the standalone system. Actually, it does! Just tried with /gnu/store/4j4b9

[PATCH] gnu: Add matplotlib.

Hi, attached you find a patch for matplotlib. For python2-matplotlib I've noticed a behavior that I do not understand: matplotlib depends on numpydoc which I pushed yesterday and appears not to be ready yet on hydra. If I build python2-matplotlib with ./pre-inst-env guix build python2-matplotlib

Re: use zsh as login shell

宋文武 skribis: > Currently, zsh doesn't work out of box. On the standalone system, right? > IMO, we should use `/etc/environment' for system variables: > LANG=en_US.UTF-8 > PATH=/run/current-system/profile/bin:/run/setuid-programs > And add `pam_env.so' to /etc/pam.d/login, to make tty lo

Re: gcl 2.6.12 breaks maxima

On Thu, Nov 27, 2014 at 01:48:31PM +0100, Federico Beffa wrote: > After having a look at the maxima log I suspect that we are hitting > this problem: > https://www.ma.utexas.edu/pipermail/maxima/2008/009769.html Thanks for working on this! I tried dropping the 'strip' phase for gcl, and now maxima

[PATCH] Core sanity and taking build options from environment.

Perhaps GUIX_BUILD (suffix with _OPTS or some such?) is not the best name and I'm not certain if it shouldn't be stdout (why are we using error port for "normal" communication? Emacs?), but I think it might prove useful if user keeps being reminded his environment is polluted or not as his intent

Resolving Linux module dependencies

In light of , I realized that the naïve handling of kernel modules (à la ‘insmod’) in the initrd and at boot time did not scale: one had to specify the dependencies of the modules to be loaded from the initrd and do so in topological order. Fo