On Thu, Mar 10, 2016 at 07:35:53AM +1100, Jookia wrote:
> This should hopefully stop people from running in to issues when
> building Guix from source as they have in the past.
I still think we should make /var the default of localstatedir; this would
avoid most problems.
Andreas, doing his Cicer
On Wed, Mar 09, 2016 at 03:27:33PM -0500, Leo Famulari wrote:
> Thanks for the notification! Updated with 89e58e8e8c.
It is difficult to beat you!
There was also a libotr-3 without any dependent package, so I simply
removed it.
Andreas
On Tue, Mar 08, 2016 at 06:39:37PM -0500, Leo Famulari wrote:
> > + (delete 'check) ;; Don't run the 'make check' step of the
> > gnu-build-system
> When skippping the tests, we prefer to say why in the comment. It can be
> as simple as "no test suite" if that is the case.
And use in the a
On Tue, Mar 08, 2016 at 06:22:23PM +0100, Ricardo Wurmus wrote:
> > Guix should provide a stable experience, but if stable means too
> > old to deliver a useable experience of an still in development
> > network, what do we suggest? I would not replace gnunet and
> > gnunet-gtk, I would just make g
On Wed, Mar 09, 2016 at 10:50:46AM +0100, Andreas Enge wrote:
> Now I will try if libreoffice 5.1 builds without updating other libraries.
It does not, I think your patches are really necessary, Efraim. I pushed
the liblangtag one. Libreoffice 5.1 still checks for the old mdds, and
insists on
Hello,
I locally applied it, but changed the commit message as follows:
gnu: libetonyek: Update to 0.1.6.
* gnu/packages/libreoffice.scm (libetonyek)[source]: Update to 0.1.6.
[inputs]: Add required input liblangtag.
to record all modifications. This requires the 1.0 version o
On Wed, Mar 09, 2016 at 11:16:16AM +0200, Efraim Flashner wrote:
> Our current vim package is 2.5 years old, and the current patch set on top of
> it to bring it up to today is ~1500 patches. Interestingly, every. single.
> commit. is tagged in git, so updating to a more recent release is rather ea
On Tue, Mar 08, 2016 at 10:04:33AM +0100, Andy Wingo wrote:
> Right now hydra.gnu.org is in this weird situation where people who use
> it have to trust it, modulo "guix challenge" of course. But really all
> we have to trust is the mapping from the derivation (like the "foo"
> package) to a hash
On Mon, Mar 07, 2016 at 02:41:22PM -0500, Leo Famulari wrote:
> > - "http://kohei.us/files/mdds/src/mdds_"; version
> > ".tar.bz2"))
> > + "http://kohei.us/files/mdds/src/mdds-"; version
> > ".tar.bz2"))
> I can't figure out what the difference is between these
On Mon, Mar 07, 2016 at 12:42:02PM +0200, Efraim Flashner wrote:
> * gnu/packages/libreoffice.scm (liblangtag): New variable.
It compiles without problem for me.
> +(description "liblangtag is an interface library to access/deal with tags
> +for identifying languages, which is described in RF
On Mon, Mar 07, 2016 at 05:41:44PM +0100, Ludovic Courtès wrote:
> > Second, there is gettext with a native input texlive-bin.
> No, fortunately not! :-)
My bad! I was mixing it up with po4a in the gettext module.
It requires texlive for its tests; if we were to take texlive out, we would
need to
Hi Jan,
I am afraid that my digression on unison has made us forget laby. If I remember
well, you had a package more or less ready? Could you maybe send the patch
again?
Andreas
On Sun, Mar 06, 2016 at 06:42:11PM +0100, Nils Gillmann wrote:
> So mirror.gnu.org is not like rsync, but more like webcache
> servers.
Exactly!
Andreas
On Sun, Mar 06, 2016 at 05:17:59PM +0100, Nils Gillmann wrote:
> Okay, sounds odd but reasonable. But what about un-fonts-core and
> un-fonts-extra, if I add -extra (see packages.gentoo.org for
> example), with this naming scheme it should be then font-un
> which gets separated into font-un:core an
On Sat, Mar 05, 2016 at 10:48:33PM +0100, Ludovic Courtès wrote:
> This commit modifies some of the service documentation, which is also
> available in docstrings in gnu/services/*.scm. Could you update them
> accordingly?
Finally, I think this is unreasonable. I already spent 20 minutes just
try
On Sun, Mar 06, 2016 at 02:46:06PM +0100, Mathieu Lirzin wrote:
> It is painful but necessary:
>
> https://www.gnu.org/prep/standards/html_node/Doc-Strings-and-Manuals.html#Doc-Strings-and-Manuals
> A way to understand why this is important is to read some of the
> "documentation" used by the Ja
On Sun, Mar 06, 2016 at 02:49:11PM +0100, Ludovic Courtès wrote:
> > +(name "font-un-fonts")
> Shouldn’t it be “font-un-fonts-core” since that’s what the tarball is
> called?
I just read up on our font naming guidelines in the manual:
"The name of a package containing only one font family star
On Sat, Mar 05, 2016 at 10:48:33PM +0100, Ludovic Courtès wrote:
> This commit modifies some of the service documentation, which is also
> available in docstrings in gnu/services/*.scm. Could you update them
> accordingly?
Oh dear, correcting typos in the texinfo manual is already close to more
w
On Sat, Mar 05, 2016 at 05:32:07PM -0500, Leo Famulari wrote:
> I understood that message as "Python 3 could work, but we will have to
> do something special to help it link with libpython3.3m".
>
> When it built successfully and passed its test suite, I figured it was
> okay.
>
> Andreas, do you
On Sat, Mar 05, 2016 at 05:08:40PM -0500, Leo Famulari wrote:
> To make it more clear, what do you think about renaming the patch so it
> doesn't refer to cpio, and adding references to Rush and Dico in the
> patch?
Yes, I was thinking the same. But this is a very singular case, and now
everybody
On Sat, Mar 05, 2016 at 10:16:05PM +0100, Ricardo Wurmus wrote:
> We could ask drobilla for a new suil release. I don’t think we should
> package the development version as it’s not clear if dependent
> applications would work with the latest suil.
Sounds good. Would you like to do it, since you
On Thu, Mar 03, 2016 at 06:04:17PM +0100, Ludovic Courtès wrote:
> Andreas Enge skribis:
> > This was the original reason (plus easier debugging) that I kept the
> > internal packages -bin and -texmf public; I think it was a mistake,
> > since it seems to lead to a lot of con
On Sat, Mar 05, 2016 at 08:51:11PM +0200, Efraim Flashner wrote:
> I tried to apply this to check if libreoffice-5.1.1 would build successfully,
> but I got the following error:
> Applying: gnu: vigra: Update to a development snapshot.
> error: patch failed: gnu/packages/image.scm:1
> error: gnu/pa
On Fri, Feb 19, 2016 at 11:22:30PM -0500, Leo Famulari wrote:
> The patch file 'cpio-gets-undeclared.patch' was "de-applied" to the cpio
> sources when we upgraded cpio to 2.12 (92d0fcb6dc5). But, it was never
> deleted or removed from gnu-system.am.
I think it can be safely removed. The patch its
Hello,
I would like to suggest moving some chapters and sections around in the
manual. For keeping this short, I will refer to the section numbers.
My main motivation are the four digits now denoting subsections in 7;
this is an indication that we are putting too much into this chapter.
Notice tha
On Thu, Feb 25, 2016 at 10:06:42AM +0100, Andreas Enge wrote:
> Probably we can:
> "jalv.qt (...) This is a versionm of Jalv with a GUI implemented in Qt.
> It is mainly for developer testing purposes, for a production ready program
> use jalv.gtk."
Qt-4 support dropped in
On Sat, Mar 05, 2016 at 11:03:07AM +0100, Pjotr Prins wrote:
> And then the question, how can we roll back on a pull ;)
Not automatically, I think, but you can always delete the directory containing
it inside .local and install from git, where you can choose any commit that
works for you.
Andreas
Hi Nils,
thanks for the generous offer of a server donation!
So what could be done?
On Thu, Mar 03, 2016 at 11:48:11PM +0100, Nils Gillmann wrote:
> It's a 36€ / year server (I don't believe in the security of OVH,
> but others say it's okay, I personally favor in-berlin.de over
> most providers
On Fri, Mar 04, 2016 at 12:53:39PM +0100, Nils Gillmann wrote:
> What counts for us is what we deliver, therefore the version
> packaged, therefore GPL2 with no exception.
gpl2+, unless the "upgrade" is explicitly excluded in the package
(the license itself allows to upgrade it otherwise).
Andrea
:00 2001
From: Andreas Enge
Date: Sat, 5 Mar 2016 00:30:56 +0100
Subject: [PATCH] gnu: vigra: Update to a development snapshot.
* gnu/packages/image.scm (vigra): Update to a git snapshot to fix build
problems with the current python-numpy.
---
gnu/packages/image.s
Dear Guixers,
some of you have already spotted documents concerning a mysterious entity
called "Guix Europe" appear in the maintenance-git repo. So now I have a
pleasant announcement to make... But wait - I am not going to write it all up
a second time. Read on here:
https://enge.fr/blog/2016/0
On Wed, Mar 02, 2016 at 10:57:19PM +0100, Nils Gillmann wrote:
> It would be nice if we could have some automatized testing for
> failure with an libressl branch which just replaces every
> dependency on openssl with current libressl
Normally I would suggest to build such a branch on hydra, but we
On Wed, Mar 02, 2016 at 02:15:04PM -0500, Leo Famulari wrote:
> If a Scheme wizard can programatically replace all references of openssl
> to libressl in the code base, I would be interested in testing it
> locally.
Testing is easy: Just rename the "libressl" variable to "openssl", and the
"openss
On Wed, Mar 02, 2016 at 10:34:23AM +0100, Ludovic Courtès wrote:
> We don’t want people to have to do that, do we? :-)
> Personally, I don’t want to have to think beyond “guix package -i
> texlive”.
It is only about choice, you can always just install texlive if you are
ready to download gigabyte
On Tue, Feb 23, 2016 at 12:31:38PM +0100, Andreas Enge wrote:
> However, the first one is internally assigned to the variable "libwnck-1".
> I assume that someone got inspired by another case where the "-1" is actually
> a version number. Is it okay if I rename it to
On Sun, Feb 28, 2016 at 03:47:16PM +0100, Ricardo Wurmus wrote:
> Possibly, but the author would like it to run on more than one architecture:
> https://github.com/gmarcais/Jellyfish/issues/52#issuecomment-171319952
> I don’t know if this fixes the build for i686 and mips. Do you think we
> coul
...
On Sun, Feb 28, 2016 at 12:10:12PM +0200, Efraim Flashner wrote:
> * gnu/packages/weechat.scm (weechat): Move from here ...
> * gnu/packages/irc.scm (weechat): ... to here.
> * gnu-system.am (GNU_SYSTEM_MODULES): Remove weechat.scm.
Again, mention the deleted file.
Thanks!
Andreas
...
On Sun, Feb 28, 2016 at 12:10:11PM +0200, Efraim Flashner wrote:
> * gnu/packages/irc.scm (irssi)[arguments]: Use 'modify-phases'.
Nothing to say. I think it could be combined with the previous commit,
but this is also not a problem.
...
...
On Sun, Feb 28, 2016 at 12:10:10PM +0200, Efraim Flashner wrote:
> * gnu/packages/irssi.scm (irssi): Move from here ...
> * gnu/packages/irc.scm (irssi): ... to here.
> * gnu-system.am (GNU_SYSTEM_MODULES): Remove irssi.scm.
Also mention the deleted file in the commit message.
...
...
On Sun, Feb 28, 2016 at 12:10:09PM +0200, Efraim Flashner wrote:
> +(synopsis "distributed IRC client")
Start with a capital letter (this should be found by "guix lint").
> +(description "Quassel is a modern, cross-platform, distributed IRC
> client,
> +meaning that one or more clie
On Sun, Feb 28, 2016 at 12:10:08PM +0200, Efraim Flashner wrote:
> +Using a plugin system it is possible to create notifications with many
> +different notification systems")
Missing "." at the end.
...
On Sun, Feb 28, 2016 at 12:10:07PM +0200, Efraim Flashner wrote:
> +(define-public qca
Looks good. It could also go into the qt module, but I suppose we need
a space for kde applications anyway.
Andreas
Hello Jessica,
a quick reminder, it would be a pity to drop an almost finished package.
Andreas
On Wed, Feb 17, 2016 at 07:13:44PM +0100, Andreas Enge wrote:
> On Tue, Feb 09, 2016 at 11:55:51PM +0100, Andreas Enge wrote:
> > Since I tried out a few things, I realised I might as well
On Fri, Feb 26, 2016 at 11:11:22AM +0200, Efraim Flashner wrote:
> +(description
> + "A backport of recent additions to the 'collections.abc' module.")
This should be a full sentence; right now it is missing a verb, at least it
should be "foo is/constitutes/provides a backport". "recent ad
On Sat, Feb 27, 2016 at 12:20:33AM +0100, Ludovic Courtès wrote:
> Or simply:
> guix build -e '(@@ (gnu packages texlive) texlive-texmf)'
That is nifty! I suppose this also uses substitutes if present? In that case,
we could even make the texlive-bin package private as well, and a user could
exe
Hello,
on armhf, I get lots of errors such as these:
unit_tests/test_mer_dna.cc: In instantiation of ‘void
{anonymous}::MerDNA_GetBits_Test::TestBody() [with
gtest_TypeParam_ =
{anonymous}::VTC, 1>]’:
unit_tests/test_mer_dna.cc:529:1: required from here
unit_tests/test_mer_dna.cc:464:96: err
on Sep 17 00:00:00 2001
From: Andreas Enge
Date: Sat, 27 Feb 2016 17:21:04 +0100
Subject: [PATCH] gnu: Add einstein.
* gnu/packages/games.scm (einstein): New variable.
* gnu/packages/patches/einstein-install.patch: New file.
* gnu-system.am (dist_patch_DATA): Register pa
On Fri, Feb 26, 2016 at 07:07:23PM -0500, Thompson, David wrote:
> I install texlive-texmf into my profile because otherwise latex cannot
> find any extensions. In particular, I was trying to compile a beamer
> document and I spent hours wrestling with load paths and nothing
> worked until texlive
On Thu, Feb 25, 2016 at 09:32:22PM +0100, Ricardo Wurmus wrote:
> Other distributions have separate “*-devel” or “*-dev” packages (and I’m
> ambivalent about this) — would it be a bad idea if we provided “devel”
> or “dev” *outputs* so that users had more control over what ends up in
> their store?
Hello,
when installing xfce, there is a collision with xfce4-session for the file
startxfce4.
The reason is the following:
(alist-replace
'install
(lambda* (#:key outputs #:allow-other-keys)
(let* ((out (assoc-ref outputs "out"))
(bin (string-ap
On Thu, Feb 25, 2016 at 10:38:32AM +0100, Ricardo Wurmus wrote:
> > The suil package itself, however, depends on gtk+-2 _and_ qt-4.
> > Do you think we could drop the qt-4 input?
> We cannot know what toolkit the GUI of audio plugins will use. Suil
> supports the three most popular ones so plugin
Hello,
I would like to make the texlive-texmf package private, to avoid that people
erroneously install it into their profiles. One could still build the big
data part locally as follows:
guix build texlive-bin # This will use substitutes.
guix build --no-substitutes texlive
# This will build te
I would like to see an updated vigra. Presumably, in maybe a week there
should be a release that builds and that would allow us to build libreoffice.
I tried the most recent git snapshot and it compiles (well, I had problems
in the beginning, and I am not totally sure we do not need to disable
para
On Thu, Feb 25, 2016 at 02:35:25PM -0500, Leo Famulari wrote:
> You could modify it again by having us commit each package to master
> when it looks good, and then making adjustments later. The benefit of
> this is that it's very easy for other people to know how to help you
> with this project, fo
Probably we can:
"jalv.qt (...) This is a versionm of Jalv with a GUI implemented in Qt.
It is mainly for developer testing purposes, for a production ready program
use jalv.gtk."
Andreas
On Thu, Feb 25, 2016 at 09:35:45AM +0100, Andreas Enge wrote:
> I find it unlikely that an application would need both of gtk+ and qt.
Maybe I am wrong; jalv does depend on both... It creates binaries jalv.gtk
and jalv.qt. If nobody uses the qt version, we could remove the input qt-4.
Andreas
Hello,
looking at why ardour depends on qt-4, I came across suil:
(description
"Suil is a lightweight C library for loading and wrapping LV2 plugin UIs.
Suil makes it possible to load a UI of a toolkit in a host using another
toolkit. The API is designed such that hosts do not need to ex
On Wed, Feb 24, 2016 at 11:18:46PM +0100, Andreas Enge wrote:
> I am building "./pre-inst-env guix build guix" right now, and if it passes
> all its tests, I will push the update tomorrow.
It works and I pushed, thanks!
Andreas
On Wed, Feb 24, 2016 at 04:45:25PM -0500, Leo Famulari wrote:
> $ guix refresh -l guile-json
> Building the following 2 packages would ensure 2 dependent packages are
> rebuilt: guix-0.9.0 guix-0.9.0.c3f29bc
I am building "./pre-inst-env guix build guix" right now, and if it passes
all its tests,
On Thu, Feb 18, 2016 at 11:58:49AM -0500, myglc2 wrote:
> A patch to address issues raised in bug#22695.
Thanks, I modified mainly the commit message a bit and pushed.
This is quite useful, being used to the old init scripts, I would not
have guessed that additional action is needed apart from cop
This is incredible:
+"Due to a byte/bit confusion, the DH secret was too short. This file was
+completely reworked and will be commited in a future version."
I quickly looked over it and think that your approach (which also has been
discussed in detail on irc) looks good, please push.
Andreas
On Tue, Feb 23, 2016 at 10:53:57AM +0300, Alex Kost wrote:
> I would rather write: "The @code{(gnu services shepherd)} module ..."
> But I don't know if it's preferable or not, other opinions are needed.
> The rest fixes look good to me.
I agree and pushed like this.
Thanks, Chris and Alex!
Andr
On Wed, Feb 24, 2016 at 04:06:18PM +0100, Fabian Harfert wrote:
> the name libwnck-1 is right. The library files have this name and the
> pkg-config file refers to that. This is because it uses Gtk+ 2.x and
> has a different API than libwnck-3. I don't know why they chose this
> versioning, but it'
On Tue, Feb 23, 2016 at 10:18:45PM +0100, Jan Nieuwenhuizen wrote:
> (services
>(cons* (console-keymap-service "dvorak")
> ;;(dhcp-client-service)
> (lsh-service #:port-number
>#:allow-empty-passwords? #t
>#:root-login?
Hello,
we have the following in gnome.scm:
$ guix package -A libwnck
libwnck 2.30.7 out gnu/packages/gnome.scm:1464:2
libwnck 3.14.0 out gnu/packages/gnome.scm:1435:2
However, the first one is internally assigned to the variable "libwnck-1".
I assume that someone got inspired by another
On Tue, Feb 23, 2016 at 10:53:57AM +0300, Alex Kost wrote:
> I would rather write: "The @code{(gnu services shepherd)} module ..."
> But I don't know if it's preferable or not, other opinions are needed.
> The rest fixes look good to me.
I will have a look at the documentation again next weekend (
On Tue, Feb 23, 2016 at 10:00:24AM +0100, Ricardo Wurmus wrote:
> I’d like to keep it the way it is. In CRAN there are a number of
> packages with very similar names such as “graph” and “rgraph”, or “utils”
> and “R.utils”. There is also a great number of R packages that happen
> to start with “R
On Mon, Feb 22, 2016 at 09:17:56PM +0100, Danny Milosavljevic wrote:
> That said, on the general interest: guix is already slooow on a fairly modern
> X200. If that ran on an armv7, it would probably be much slower. Not fun.
> But might be worth a try. Did someone try the non-GuixSD armv7 thing?
On Sun, Feb 21, 2016 at 03:45:22PM -0500, Leo Famulari wrote:
> This updates MPD to 0.19.12 and ncmpcpp to 0.7.3.
> I'm presenting it for review since there are some changes that users
> should be aware of.
If there are no complaints, I would go ahead. People even survived to
new magit releases :-
On Sun, Feb 21, 2016 at 02:28:37AM -0500, Leo Famulari wrote:
> QJson is also used by libdbusmenu-qt.
The latter is also gone with the KDE 4 module. It had no followers outside
of it, but if it turns out to be needed somewhere else, we will have to
think again.
Andreas
On Thu, Feb 18, 2016 at 09:43:49PM +0100, Andreas Enge wrote:
> So while we are it it, I suggest to simply remove kde.scm (there is no use
> in keeping a lonely oxygen-icons around...).
> Also, python2-pyqt has no dependent package.
> If there is no outcry, I will remove the abo
Sorry, Chris, that I bothered you with the state of pumpa; I was so convinced
that you were the packager that I did not even check! I suppose that I have
read too many of your blog posts to planet gnu; whenever I hear "federation"
or "pumpsomething" now, I think of you.
On Sun, Feb 21, 2016 at 09:
Hello,
when tentatively calling "guix refresh", I saw a number of packages
called "r-r-...", for instance "r-r-oo". Should this not be simplified
to "r-oo"? We do not have separate packaging rules for R, but in analogy
with python and perl, it would make sense to preprend "r-" and at the
same time
On Mon, Feb 22, 2016 at 02:21:12PM -0500, Mark H Weaver wrote:
> As I recall, while I was trying to change our svg->png code to use
> simpler tools to avoid requiring 'inkscape', I guessed that the
> checkboard pattern might be a mere artifact, but Luis later confirmed
> that it was an intentional
On Sun, Feb 21, 2016 at 07:47:31PM +0100, Ludovic Courtès wrote:
> I meant to keep it until it was really really deprecated. Currently it
> is still used in a key place, in (guix packages), which is why I left
> it.
> Maybe we should reinstate it, WDYT?
Okay, no problem. I just thought that the d
Hello,
On Sun, Feb 21, 2016 at 07:43:25AM +1100, Jookia wrote:
> I brought up this topic a long time ago and as since have been thinking about
> this, reading the code base and getting accustomed to it. I've had the
> intention
> of porting GuixSD to use the U-Boot bootloader, which I think can b
Hello,
On Sat, Feb 20, 2016 at 06:17:28PM +0100, Fabian Harfert wrote:
> I carefully checked all the library source dependencies and so on. I
> think all these propagated inputs are needed.
okay, excellent.
> If that is necessary, I'll add these comments, but I think it's not too
> difficult to
Hello,
before considering more mate packages, I would like to get back to discussing
the existing ones; as I am not knowledgeable about desktop packages, I must
admit I did not look at them too closely before. So maybe my comments can
also be disregarded, but it would be good if someone more exper
On Thu, Feb 18, 2016 at 06:57:18PM -0500, Leo Famulari wrote:
> > +(inputs `(("gettext" ,gnu-gettext)))
> Can somebody with more experience comment on whether or not this should
> be propagated (assuming it's a runtime dependency, otherwise it should
> be native)?
It is most likely a native-in
On Thu, Feb 18, 2016 at 02:53:40PM -0800, Christopher Allan Webber wrote:
> I'm assuming this was directed at me (though I don't work on Pumpa, but
> I could ask the author things... I do *use* it every day though). What
> am I being asked here... I'm not sure? Whether Pumpa could be updated
> to
On Thu, Feb 18, 2016 at 09:32:52PM +, Christopher Baines wrote:
> A patch could be written as Andreas describes (to use a variable), and
> then this could be used for Guix. Once written, the patch could also be
> sent to the upstream maintainer to see if they want to include it.
This would be
On Thu, Feb 18, 2016 at 09:13:47PM +, Christopher Baines wrote:
> I am no perl expert, but I think it just comes with perl. Debian is easy
> to check [1].
Indeed, so I just put your patch. Thanks for your patience!
Andreas
Hi Leo,
thanks for your initiative!
On Sun, Feb 14, 2016 at 03:01:43PM -0500, Leo Famulari wrote:
> I think we need a Qt champion(s) for Guix.
I am not volunteering, mainly because I am not a Qt champion (but I am
interested in the package).
> $ guix refresh -l qt-4
> Bu
Hello,
On Sat, Feb 13, 2016 at 07:26:05PM +0100, Jan Nieuwenhuizen wrote:
> I am using this snippet in (operatating-system) to multi-boot
I have my /home in an encrypted lvm, so I cannot share it with a multi-booted
GuixSD. My solution will be to drop Debian and to move to GuixSD over the
weekend
I trust you tried them, they look good to me.
Andreas
On Thu, Feb 18, 2016 at 07:59:14PM +, Christopher Baines wrote:
> You can find it here [1], I have also attached it to this email.
Okay, thanks. I have added a "." to the tag line of the commit message
and dropped the line "needed for vcsh tests"; we can package it in its own
right. I also mod
On Wed, Feb 17, 2016 at 07:11:44PM -0500, Leo Famulari wrote:
> I missed this earlier. We don't usually propagate this sort of "end-user
> program" dependencies. For example rsync does not propagate ssh; the
> user must install ssh. [0]
I think the cleanest approach would be to patch all calls to
On Wed, Feb 17, 2016 at 07:15:23PM -0500, Leo Famulari wrote:
> The commit that removes the module only makes sense in the context of
> adding vcsh. It's analogous to adding a module that is required for a
> new package, and that is done in a single commit. So, I think it's
> better to squash the c
Hi Mark,
On Tue, Feb 16, 2016 at 05:51:06PM -0500, Mark H Weaver wrote:
> Without seeing your test-suite.log, I'm not sure if these are the same
> failures I'm getting, but network-manager consistently fails to build
> for me as well.
>
> Here's a patch I've used to enable building on my machine.
Hi Alex,
almost, perfect, expect for a little dot:
On Thu, Feb 18, 2016 at 11:06:34AM +0100, Alex Sassmannshausen wrote:
> Subject: [PATCH] gnu: guile-config: Update to 0.1.1
at the end of this line! I pushed.
Thanks,
Andreas
On Thu, Feb 18, 2016 at 12:46:50PM -0500, Leo Famulari wrote:
> I didn't realize that copy-file's return value is unspecified. I've
> found that it does tend to make the whole build fail when the source
> file doesn't exist, so it must be returning #f in that case, right?
Not necessarily. Many thi
On Thu, Feb 18, 2016 at 03:33:23PM +0100, Nils Gillmann wrote:
> Okay, understood. However, there's now the choice between a
> specific category (disassembler.scm) and a meta category
> (reverse-engineering.scm).
> Not having read every single line of the source of Guix yet, are
> meta-categories i
Just a quick reply to the first part of your message, as I do not have
time to read more right now ;-)
On Thu, Feb 18, 2016 at 03:45:36PM +0100, Ricardo Wurmus wrote:
> I think this might have to operate on the bag. I’m not sure if a Python
> package can (or should be able to expect to) import a
On Thu, Feb 18, 2016 at 01:21:48PM +0100, Ricardo Wurmus wrote:
> 2) avoid PYTHONPATH, patch all Python files invasively!
> import sys
> sys.path.append("/gnu/store/cabba9e...-numpy.../lib/...")
That sounds great! It appears to me as if you have found the equivalent of
a RUNPATH for python
Hi Nils,
On Wed, Feb 17, 2016 at 03:38:42PM +0100, Nils Gillmann wrote:
>
> As this doesn't do the job for me, I propose
> gnu/packages/disassembler.scm
> or
> gnu/packages/re.scm
> or
> gnu/packages/reverse-engineering.scm
> as it specifies an existing category of software and is not just
> anot
On Wed, Feb 17, 2016 at 05:50:31PM +, Chris Marusich wrote:
> I understand why it's
> nice to have the Guix package so that you can easily work with its
> dependencies, but isn't "guix pull" the recommended way to keep your Guix
> installation up to date?
As far as I know, the only ways of get
Hello, Jessica,
On Tue, Feb 09, 2016 at 11:55:51PM +0100, Andreas Enge wrote:
> Since I tried out a few things, I realised I might as well send you my
> modified version of the patch. Maybe you could continue from there and
> examine the tests. If you are not happy with some of my modi
Hello,
this looks good. I also added a copyright line for you.
On Sun, Feb 14, 2016 at 08:36:00PM +, Christopher Baines wrote:
> + "vcsh provides the @code{vcsh} command, which allows you to maintain
I would shorten to "vcsh allows ..."; that vcsh provides the vcsh command
is close to a
Just a quick "thank you" message to Taylan; your patch for compiling Guix
in parallel makes an amazing difference! I just noticed it again on my Novena
with its four relatively weak cores.
Andreas
On Wed, Feb 17, 2016 at 04:12:32AM +1100, Jookia wrote:
> I'm not sure how that would really fix the issue that people are hitting: Say
> we
> move some of the content to /gnu, things like profiles, the daemon socket, all
> the things that Guix looks for and symlinks to won't be found if Guix look
801 - 900 of 2012 matches
Mail list logo
