Re: guix environment guix doesn't provide guix dependencies?

[Chris Marusich]

John Darrington  writes:

> On Tue, Sep 27, 2016 at 09:18:30PM -0700, Chris Marusich wrote:
>  Hi,
>  
>  The manual says ((guix) Building from Git):
>  
>  --8<---cut here---start->8---
> The easiest way to set up a development environment for Guix is, of
>  course, by using Guix!  The following command starts a new shell where
>  all the dependencies and appropriate environment variables are set up to
>  hack on Guix:
>  
>   guix environment guix
>  --8<---cut here---end--->8---
>  
>  After running this command successfully, I get the following error while
>  running ./bootstrap:
>  
>  --8<---cut here---start->8---
>  + exec autoreconf -vfi
>  ./bootstrap: line 5: exec: autoreconf: not found
>  --8<---cut here---end--->8---
>  
>  This makes me wonder: has something changed, and the manual is now
>  incorrect?
>  
>
> I don't think so.   Running ./bootstrap is not part of the normal process for 
> *building*
> guix - only for bootstrapping.  But you are right - it is common to want to 
> do that.
> For this reason I often run:
>
>  guix environment guix --ad-hoc automake autoconf texinfo help2man
>
>
> But perhaps you are right - the manual might want to mention this.
>
> J'

That makes sense.  I feel like it used to provide all the dependencies,
including autoconf etc., but that was months ago, and I might be
mis-remembering.  It works fine when I add the needed bootstrap
dependencies with --ad-hoc.

-- 
Chris


signature.asc
Description: PGP signature

guix environment guix doesn't provide guix dependencies?

[Chris Marusich]

Hi,

The manual says ((guix) Building from Git):

--8<---cut here---start->8---
   The easiest way to set up a development environment for Guix is, of
course, by using Guix!  The following command starts a new shell where
all the dependencies and appropriate environment variables are set up to
hack on Guix:

 guix environment guix
--8<---cut here---end--->8---

After running this command successfully, I get the following error while
running ./bootstrap:

--8<---cut here---start->8---
+ exec autoreconf -vfi
./bootstrap: line 5: exec: autoreconf: not found
--8<---cut here---end--->8---

This makes me wonder: has something changed, and the manual is now
incorrect?

-- 
Chris


signature.asc
Description: PGP signature

Re: [PATCH 4/6] gnu: linux-libre-beagle-bone-black: Remove kernel variant.

[Mark H Weaver]

John Darrington  writes:

> On Mon, Sep 26, 2016 at 04:24:10PM +0200, Vincent Legoll wrote:
>  Hello,
>  
>  @David: the following question is nothing against you, I just
>  took the opportunity to ask, sorry for the thread hijacking.
>  
>  On Mon, Sep 26, 2016 at 3:23 PM, John Darrington wrote:
>  > Why are we removing this?
>  
>  Is it just me finding this changelog format usefulness to be
>  very low ?
>  
>
> I entirely agree.   
>
> Our current convention for commit messages is aesthetically pleasing,
> but useless.   It is redundant to put in the message WHAT has changed.
> That can be easily determined by running "git show".   What is much
> more usefull is the reason WHY we changed it.
>
>
> Recently, on another project,  I came across a line of code which I
> was pretty sure was pointless.  I was about to delete it - but I 
> decided to check first.  So I ran "git blame" to find out which 
> commit had added it.  Then I ran git show to get the full details
> of that commit.   Fortunately, the person who had committed it 5 years
> ago had mentioned exactly why it was needed.

Our conventions for commit logs, which follow the GNU Coding Standards
for Change Logs (see section 6.8 of the GNU Coding Standards), is that
explanations belong in the comments of the code itself, not in the
commit log.  If that had been done in the example you give above, you
would have known why the line was needed in a small fraction of the time
that it must have taken you to perform all the steps above.

However, I agree that commits that _remove_ code should include the
rationale in the commit log, if the reason is not obvious and if there's
no sensible place to put the explanation in the code.

What do you think?

  Mark

Re: fgrep fails to find grep during bootstrap

[Mark H Weaver]

csanchez...@gmail.com (Carlos Sánchez de La Lama) writes:

> Hi all,
>
> I am trying to bootstrap for powerpc-linux. I have generated the
> tarballs and patched gnu/packages/bootstrap.scm to use them.
>
> However, I get an error during when creating %bootstrap-coreutils
> package. It specifies "fgrep" as program to test, but fgrep is a shell
> script, and altought the shebang is correctly patched, it needs "grep"
> binary in the path (which is not at this stage). fgrep does:
>
> exec grep -F "$@"
>
> so this can not work unless grep is available in PATH.

I've fixed this problem on the core-updates branch:

http://git.savannah.gnu.org/cgit/guix.git/commit/?h=core-updates=1063d325ea76aa2b00dfcd3d436b16e412103df1

  Mark

[PATCH] gnu: Add lci.

[ng0]

I guess I'm really frustrated with being stuck for one month now on
trying to get a simple VM to outside world connection for testing
services I'm writing. So here's a lolcode interpreter, lci.

If someone reviews, please help me at this thread about the VM problems:
http://lists.gnu.org/archive/html/guix-devel/2016-09/msg02109.html

From b5dda0ba3b7d0432ab9d65e94aec4e28c02735b1 Mon Sep 17 00:00:00 2001
From: ng0 
Date: Tue, 27 Sep 2016 22:30:11 +
Subject: [PATCH] gnu: Add lci.

* gnu/packages/lolcode.scm: New file.
* gnu/local.mk (GNU_SYSTEM_MODULES): Add it.
---
 gnu/local.mk |  1 +
 gnu/packages/lolcode.scm | 60 
 2 files changed, 61 insertions(+)
 create mode 100644 gnu/packages/lolcode.scm

diff --git a/gnu/local.mk b/gnu/local.mk
index c2cdf8b..e34efaf 100644
--- a/gnu/local.mk
+++ b/gnu/local.mk
@@ -223,6 +223,7 @@ GNU_SYSTEM_MODULES =\
   %D%/packages/llvm.scm\
   %D%/packages/lout.scm\
   %D%/packages/logging.scm			\
+  %D%/packages/lolcode.scm  \
   %D%/packages/lsof.scm\
   %D%/packages/lua.scm\
   %D%/packages/lxde.scm\
diff --git a/gnu/packages/lolcode.scm b/gnu/packages/lolcode.scm
new file mode 100644
index 000..64dcb0c
--- /dev/null
+++ b/gnu/packages/lolcode.scm
@@ -0,0 +1,60 @@
+;;; GNU Guix --- Functional package management for GNU
+;;; Copyright © 2016 ng0 
+;;;
+;;; This file is part of GNU Guix.
+;;;
+;;; GNU Guix is free software; you can redistribute it and/or modify it
+;;; under the terms of the GNU General Public License as published by
+;;; the Free Software Foundation; either version 3 of the License, or (at
+;;; your option) any later version.
+;;;
+;;; GNU Guix is distributed in the hope that it will be useful, but
+;;; WITHOUT ANY WARRANTY; without even the implied warranty of
+;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+;;; GNU General Public License for more details.
+;;;
+;;; You should have received a copy of the GNU General Public License
+;;; along with GNU Guix.  If not, see .
+
+(define-module (gnu packages lolcode)
+  #:use-module (gnu packages)
+  #:use-module (gnu packages documentation)
+  #:use-module (gnu packages python)
+  #:use-module (gnu packages readline)
+  #:use-module (guix build-system cmake)
+  #:use-module (guix download)
+  #:use-module ((guix licenses) #:prefix license:)
+  #:use-module (guix packages))
+
+(define-public lci
+  (package
+(name "lci")
+(version "0.11.2")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "https://github.com/justinmeza/lci/archive/v;
+  version ".tar.gz"))
+  (sha256
+   (base32
+"1li7ikcrs7wqah7gqkirg0k61n6pm12w7pydin966x1sdn9na46b"))
+  (file-name (string-append name "-" version ".tar.gz"
+(build-system cmake-build-system)
+(native-inputs
+ `(("python-2" ,python-2)
+   ("readline" ,readline)))
+(native-inputs
+ `(("doxygen" ,doxygen)))
+(synopsis "LOLCODE interpreter written in C")
+(description
+ "lci is a LOLCODE interpreter written in C and is designed to be correct,
+portable, fast, and precisely documented.
+@enumerate
+@item correct: Every effort has been made to test lci's conformance to the
+LOLCODE language specification.  Unit tests come packaged with the lci source code.
+@item portable: lci follows the widely ported ANSI C specification allowing it
+to compile on a broad range of systems.
+@item fast: Much effort has gone into producing simple and efficient code
+whenever possible to the extent that the above points are not compromized.
+@end enumerate")
+(home-page "http://lolcode.org/;)
+(license license:gpl3+)))
-- 
2.10.0

-- 
  ng0


signature.asc
Description: PGP signature

Re: [PATCH 1/2] gnu: Add perl-net-psyc.

[ng0]

Thanks for reviewing,

Leo Famulari  writes:

> [ Unknown signature status ]
> On Wed, Sep 21, 2016 at 06:46:31PM +, ng0 wrote:
>> From 74a6c1e552a6ae8f438e91cbe318882401b440f8 Mon Sep 17 00:00:00 2001
>> From: ng0 
>> Date: Wed, 21 Sep 2016 18:08:42 +
>> Subject: [PATCH 1/2] gnu: Add psyclpc.
>> 
>> * gnu/packages/psyc.scm (psyclpc): New variable.
>
>> +(source (origin
>> +  (method url-fetch)
>> +  (uri (list (string-append "http://www.psyced.org/files/;
>> +name "-" version ".tar.xz")
>> + (string-append "http://krosos.sdf.org/static/unix/;
>> +"perlpsyc-" version ".zip")))
>> +  (sha256
>> +   (base32
>> +"0c2afcj8b2yr2vmy9sy0528iqs9sw01j6q35lvxicm42gs7vnik2"
>
> Do both of those URLs provide a file with the same hash?

They should. But the server of psyced.org recently had to move, like you
already noticed there's a mismatch now everywhere. I'll adjust again,
try tomorrow around 11:59 AM UTC if I don't run into a problem I can't
control. Updated patches will follow before this.

>> +(arguments
>> + `(#:tests? #f ; There are no tests/checks.
>> +   #:configure-flags (list
>> +  "--enable-use-tls=yes"
>> +  "--enable-use-mccp"
>
> The "Mud Client Compression Protocol"?

Probably. For the why/how/what refer to the source and the Gentoo
ebuild. I just do the integration/packaging.

>> +  (string-append "--with-openssl="
>> + (assoc-ref %build-inputs 
>> "openssl"))
>> +  (string-append "--prefix="
>> + (assoc-ref %outputs "out"))
>> +  (string-append "--libdir=" ; "-DMUD_LIB="
>
> What does this commented text mean?

References taken how I translated the buildsystem.

>> + (assoc-ref %outputs "out")
>> + "/opt/psyced/world")
>> +  (string-append "--bindir="
>> + (assoc-ref %outputs "out")
>> + "/opt/psyced/bin")
>> +  (string-append "--libexecdir=" ; "-DERQ_DIR="
>
> Same question here.
>
>> +(inputs
>> + `(("zlib" ,zlib)
>> +   ("openssl" ,openssl)))
>> +;; pcre is bundled to ensure the version is compatible. XXX: look into
>> +;; unbundling it. Upstream should update from pcre 4.5 to 8.38. For
>> +;; functionality reasons we can not unbundle it now.
>> +;; ("pcre" ,pcre)))
>
> That version of PCRE was released in 2003. We might want to add a
> warning to the package description...
>
> https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=pcre

Let's wait with the commit then, I've asked the people at our side who
are working on the code about a statement as I don't have a full
understanding of psyclpc. I've also pointed out that we (psyc/secushare)
are advised to update pcre (before you've sent this) to fix CVEs.

I have to send a new patchseries anyway, so I can adjust this. If
we'll add a comment, it's useful to point out that this is being fixed.
psyclpc in the wild is not used by anything other than psyced these
days, so I don't know if what's being used by psyclpc is worth to point
out.
-- 
  ng0

Re: CVE-2016-0634 code execution in Bash prompt when expanding hostname

[Leo Famulari]

On Thu, Sep 22, 2016 at 12:42:15AM +0900, Ludovic Courtès wrote:
> John Darrington  skribis:
> 
> > On Tue, Sep 20, 2016 at 04:55:30PM -0400, Leo Famulari wrote:
> >  Any advice on how we should handle CVE-2016-0634?
> >  
> >  http://seclists.org/oss-sec/2016/q3/534
> >
> > Like the comment there says, it is only a problem if the machine has
> > already been owned,
> 
> … or if a privilege application like a DHCP client can be made to set
> the host name to $(something bad), which was apparently possible at some
> point.
> 
> > so I don't see what the issue is.  If there is an issue it is for the
> > bash maintainers to patch.

Perhaps it's not the most critical bug, but I don't think we can
effectively anticipate the full impact of this (or any) bug. It's better
to just fix it now that we know about it.

> Chet proposed a patch:
> 
>   http://seclists.org/oss-sec/2016/q3/att-538/prompt-string-comsub.patch

I've asked Chet if he will add the patch to the bash-4.3-patches FTP
directory. If not, we can apply it the "normal" way.

> 
> IIUC, the just-released 4.4 isn’t affected, right?

Right.

> We should at least update it in core-updates, but core-updates won’t be
> merged until we have fixed that Binutils/MIPS issue (which shouldn’t be
> too hard, but we never know!).

I spent some time looking at the Bash package definition, but I'm stuck
on how to handle all the Bash %patch-series machinery. There are
currently no patches for Bash 4.4, nor is there an FTP directory
corresponding to a future patch series.

Does anyone have advice on how to proceed?

Re: remove patchworks. [notes on gitlab, go-build-system, ruby]

[ng0]

Ludovic Courtès  writes:

> ng0  skribis:
>
>> We have 22 pages of "open" patches. But they are not really open. I
>> believe almost no one uses patchworks. I did for some time, and so did
>> Ludovic and probably 3 more people but that's it.
>
> Yeah, I agree, this was experimental, and the experiment wasn’t quite
> successful.  We don’t need to formally “remove” the instance though; we
> can eventually email the Sourceware folks asking them to do so, but no
> rush IMO.
>
> Thanks,
> Ludo’.
>

I Agree.

Regarding gitlab: if we do not want the omnibus package (we will not be
able to use this easily on guixsd anyway), we need at least a
go-build-system. I have outlined most of what I need to achieve what I
want from GuixSD[0], and go-build-system based packages and items are not
on the list. I will not work on a go-build-system, in contrast to what
I've written in an earlier message.

We need go-build-system for:

Gogs
 - and its dependencies

Gitlab
 - gitlab-workhorse

There's also something ruby based in the gitlab suite (gitlab-shell,
gitlab-ce) I'm currently working on gitlab-shell, but I already shift
towards handing my wip over to our ruby experts.
-- 
  ng0

[PATCH] gnu: Add hdf-eos5.

[Thomas Danckaert]

>From f1a3bc9dcfbe9091110aacf353d8224a88788292 Mon Sep 17 00:00:00 2001
From: Thomas Danckaert 
Date: Fri, 17 Jun 2016 10:51:38 +0200
Subject: [PATCH] gnu: Add hdf-eos5.

* gnu/packages/maths.scm (hdf-eos5): New variable.
* gnu/packages/patches/hdf-eos5-build-shared.patch: New file.
* gnu/packages/patches/hdf-eos5-remove-gctp.patch: New file.
* gnu/packages/patches/hdf-eos5-fix-szip.patch: New file.
---
 gnu/packages/maths.scm   | 45 
 gnu/packages/patches/hdf-eos5-build-shared.patch | 88 
 gnu/packages/patches/hdf-eos5-fix-szip.patch | 32 +
 gnu/packages/patches/hdf-eos5-headers.patch  | 14 
 gnu/packages/patches/hdf-eos5-remove-gctp.patch  | 61 
 5 files changed, 240 insertions(+)
 create mode 100644 gnu/packages/patches/hdf-eos5-build-shared.patch
 create mode 100644 gnu/packages/patches/hdf-eos5-fix-szip.patch
 create mode 100644 gnu/packages/patches/hdf-eos5-headers.patch
 create mode 100644 gnu/packages/patches/hdf-eos5-remove-gctp.patch

diff --git a/gnu/packages/maths.scm b/gnu/packages/maths.scm
index 0401cd3..1cc041d 100644
--- a/gnu/packages/maths.scm
+++ b/gnu/packages/maths.scm
@@ -42,6 +42,7 @@
   #:use-module (guix build-system gnu)
   #:use-module (guix build-system r)
   #:use-module (gnu packages algebra)
+  #:use-module (gnu packages autotools)
   #:use-module (gnu packages bison)
   #:use-module (gnu packages boost)
   #:use-module (gnu packages check)
@@ -483,6 +484,50 @@ extremely large and complex data collections.")
 (license (license:x11-style
   "http://www.hdfgroup.org/ftp/HDF5/current/src/unpacked/COPYING;
 
+(define-public hdf-eos5
+  (package
+(name "hdf-eos5")
+(version "1.15")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "ftp://edhs1.gsfc.nasa.gov\
+/edhs/hdfeos5/latest_release/HDF-EOS5." version ".tar.Z"))
+  (sha256
+   (base32
+"1p8nzzy8rn5chxlm0hrkjjnhh2w1ji8ac0f9q4xzg838i58i"))
+  (patches (search-patches "hdf-eos5-build-shared.patch"
+   "hdf-eos5-remove-gctp.patch"
+   "hdf-eos5-fix-szip.patch"
+   "hdf-eos5-headers.patch"
+(native-inputs
+ `(("autoconf" ,autoconf)
+   ("automake" ,automake)
+   ("libtool" ,libtool)))
+(build-system gnu-build-system)
+(inputs
+ `(("hdf5" ,hdf5)
+   ("zlib" ,zlib)
+   ("gctp" ,gctp)))
+(arguments
+ `(#:configure-flags '("--enable-install-include")
+   #:phases
+   (modify-phases %standard-phases
+ (add-before
+ 'configure 'autogen
+   (lambda _  ; configure.ac was patched
+ (and (chmod "./configure" #o755)
+  (zero? (system* "autoreconf" "--install"))
+   #:parallel-tests? #f))
+(synopsis
+ "HDF-EOS5: HDF5-based data format for NASA's Earth Observing System")
+(description
+ "HDF-EOS5 is a software library built on HDF5 to support the construction
+of data structures used in NASA's Earth Observing System (Grid, Point and
+Swath).")
+(home-page "http://www.hdfeos.org/software/library.php#HDF-EOS5;)
+(license (license:bsd-style
+  "http://www.hdfeos.org/software/library.php#HDF-EOS5;
+
 (define-public hdf5-parallel-openmpi
   (package (inherit hdf5)
 (name "hdf5-parallel-openmpi")
diff --git a/gnu/packages/patches/hdf-eos5-build-shared.patch b/gnu/packages/patches/hdf-eos5-build-shared.patch
new file mode 100644
index 000..d1b4806
--- /dev/null
+++ b/gnu/packages/patches/hdf-eos5-build-shared.patch
@@ -0,0 +1,88 @@
+Allow shared build.
+
+We remove all references to the 'testdrivers' directory.  This directory is
+not included in the distributed source archive, and its absence trips up
+automake.
+
+---
+ Makefile.am |  6 --
+ configure.ac| 12 
+ include/HE5_HdfEosDef.h |  1 +
+ src/Makefile.am |  3 ++-
+ 4 files changed, 3 insertions(+), 19 deletions(-)
+
+diff --git a/Makefile.am b/Makefile.am
+index 363bcfb..01ed024 100644
+--- a/Makefile.am
 b/Makefile.am
+@@ -3,13 +3,7 @@
+ # Include boilerplate
+ include $(top_srcdir)/config/include.am
+ 
+-# List of subdirectories.
+-# Only build the testdrivers directory if configure detected that it's present.
+-if TESTDRIVERS_CONDITIONAL
+-TESTDRIVERS=testdrivers
+-else
+ TESTDRIVERS=
+-endif
+ 
+ if INSTALL_INCLUDE_CONDITIONAL
+ INCLUDE=include
+diff --git a/configure.ac b/configure.ac
+index 5d48b43..cfa9d4e 100644
+--- a/configure.ac
 b/configure.ac
+@@ -13,9 +13,6 @@ AC_CONFIG_AUX_DIR([config])
+ AM_INIT_AUTOMAKE([foreign])
+ AM_MAINTAINER_MODE
+ 
+-# Disable shared libraries (for now)
+-AC_DISABLE_SHARED
+-
+ AM_PROG_LIBTOOL
+ 
+ dnl --
+@@ -597,13 

Re: Network with QEMU generated image (guix system vm)?

[ng0]

Aljosha Papsch  writes:

> On 26.09.2016 19:30, Jan Nieuwenhuizen wrote:
>> What I don't understand: sometimes the clone works instantly, sometimes
>> I need to "wait a bit" until cloning or `ping gitlab.com' works.  It
>> seemed to be always immediately available when I added the mcron and
>> rottlog test services, which confuses me even more.  Might just be
>> coincidence.
>>
>> Greetings,
>> Jan
>>
> I'm guessing here but it sounds alot like neighbor discovery being slow 
> with ARP. I had the same issue few months ago with GNS3 where I created 
> a virtual topology. They use qemu VMs for each host. After creating 
> topology I wanted to ping a little and some packages would be dropped 
> before they went through.
>
> http://wiki.qemu.org/Documentation/Networking suggests in section 
> "Network backend types" that user networking is slow while tap 
> networking is fast. So ARP discovery performance might just suffer from 
> qemu user networking. Maybe it's worth trying tap networking. I used it 
> a few weeks ago (with lxc containers) and it's pretty flexible. Haven't 
> paid attention to performance though.

Afaik gnunet uses tun/tap when available[0]. My limited knowledge about
qemu I assumed that those modules in the way I can configure guix so far
do not exist in the way I might want them in the vm (maybe) (different
story, different problem).
As far as my goals are concerned, this try and and error of something as
simple as a stupid ping is slowing me down in developing guix services
for my roadmap items.

> I might be able to dig up notes on topology if you are interested. In 
> it, hypervisors form a VPN and integrate guests in the VPN using taps 
> and bridges. It's not a guix recipe unfortunately, just a Gnome Dia file.

I'm very much interested, throw anything at me which helps me to solve
this somehow.

> Best regards
>
>

[0] However, the lack of calling outside world is not gnunet exclusive.
-- 
  ng0

[PATCH] gnu: git: Add input for 'git-svn'.

[Thomas Danckaert]

Hi Guix-devel,

this adds a perl dependency for git-svn.

git-svn uses Term::ReadKey to read a user's password during http 
authentication with an svn server (and perhaps for other 
authentication methods as well).


best,

Thomas
>From d5e3bf1a9e26a0fce4485f9af67118674fa6f547 Mon Sep 17 00:00:00 2001
From: Thomas Danckaert 
Date: Tue, 27 Sep 2016 15:38:51 +0200
Subject: [PATCH] gnu: git: Add input for 'git-svn'.

---
 gnu/packages/version-control.scm | 8 +---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/gnu/packages/version-control.scm b/gnu/packages/version-control.scm
index 29559bd..ef65a03 100644
--- a/gnu/packages/version-control.scm
+++ b/gnu/packages/version-control.scm
@@ -143,6 +143,7 @@ as well as the classic centralized workflow.")
 
   ;; For 'git-svn'.
   ("subversion" ,subversion)
+  ("perl-term-readkey" ,perl-term-readkey)
 
   ;; For 'git-send-email'
   ("perl-authen-sasl" ,perl-authen-sasl)
@@ -227,14 +228,15 @@ as well as the classic centralized workflow.")
 (list gitk git-gui git-cit git-se git-svn)
 (list gitk* git-gui* git-cit* git-se* git-svn*))
 
-  ;; Tell 'git-svn' where Subversion is.
+  ;; Tell 'git-svn' where Subversion and perl-term-readkey are.
   (wrap-program git-svn*
 `("PATH" ":" prefix
   (,(string-append (assoc-ref inputs "subversion")
"/bin")))
 `("PERL5LIB" ":" prefix
-  (,(string-append (assoc-ref inputs "subversion")
-   "/lib/perl5/site_perl")))
+  ,(map (lambda (i) (string-append (assoc-ref inputs i)
+   "/lib/perl5/site_perl"))
+'("subversion" "perl-term-readkey")))
 
 ;; XXX: The .so for SVN/Core.pm lacks a RUNPATH, so
 ;; help it find 'libsvn_client-1.so'.
-- 
2.10.0

Re: [PATCH] gnu: boost: Update to 1.61.0.

[Andreas Enge]

On Thu, Sep 22, 2016 at 10:16:54PM +0100, Marius Bakke wrote:
> Should this not be pushed to core-updates? `guix refresh -l` reports:
> 
> Building the following 190 packages would ensure 395 dependent packages
> are rebuilt: [...]
> 
> That will take a while on Hydra, and in the mean time anyone trying to
> install any dependent (or sub-dependent) package will have to compile
> them locally.
> 
> Unless there is a security issue, isn't this exactly what core-updates
> is for?

Not quite. Initially, core-updates was for updates of the core packages,
that is, those that you can currently see being built on the core-updates
branch of hydra. With the lack of performance of hydra, we ended up putting
more and more packages requiring large rebuilds into that branch.

Often, we also just add a feature branch (something-update) and have it
built on hydra, especially also in cases where it is unclear how much we
break by the update, to make sure that master remains in a workable state.

After the hydra machine crashed, the FSF sysadmins switched to a more powerful
set-up, and right now, we can tolerate a larger package churn. In case of
doubt when updating a package requiring larger rebuilds, we can also have it
depend on the current load on hydra towards which branch we push.

Andreas

Re: Network with QEMU generated image (guix system vm)?

[Aljosha Papsch]

On 26.09.2016 19:30, Jan Nieuwenhuizen wrote:

What I don't understand: sometimes the clone works instantly, sometimes
I need to "wait a bit" until cloning or `ping gitlab.com' works.  It
seemed to be always immediately available when I added the mcron and
rottlog test services, which confuses me even more.  Might just be
coincidence.

Greetings,
Jan

I'm guessing here but it sounds alot like neighbor discovery being slow 
with ARP. I had the same issue few months ago with GNS3 where I created 
a virtual topology. They use qemu VMs for each host. After creating 
topology I wanted to ping a little and some packages would be dropped 
before they went through.


http://wiki.qemu.org/Documentation/Networking suggests in section 
"Network backend types" that user networking is slow while tap 
networking is fast. So ARP discovery performance might just suffer from 
qemu user networking. Maybe it's worth trying tap networking. I used it 
a few weeks ago (with lxc containers) and it's pretty flexible. Haven't 
paid attention to performance though.


I might be able to dig up notes on topology if you are interested. In 
it, hypervisors form a VPN and integrate guests in the VPN using taps 
and bridges. It's not a guix recipe unfortunately, just a Gnome Dia file.


Best regards

Re: [PATCH] gnu: Add sdcc.

[Leo Famulari]

On Tue, Sep 27, 2016 at 08:15:33PM +0200, David Craven wrote:
> > This URL returns 404.
> 
> thanks! missed that due to caching - again.
> 
> is it ok in it's own file? or should that file be renamed to
> microcontroller.scm?

If we don't already have an appropriate module, then it's a matter of
taste :)

microcontroller sounds like a useful module name.

Re: fgrep fails to find grep during bootstrap

[Efraim Flashner]

On Tue, Sep 27, 2016 at 03:00:30PM +0200, Carlos Sánchez de La Lama wrote:
> Hi all,
> 
> I am trying to bootstrap for powerpc-linux. I have generated the
> tarballs and patched gnu/packages/bootstrap.scm to use them.
> 
> However, I get an error during when creating %bootstrap-coreutils
> package. It specifies "fgrep" as program to test, but fgrep is a shell
> script, and altought the shebang is correctly patched, it needs "grep"
> binary in the path (which is not at this stage). fgrep does:
> 
> exec grep -F "$@"
> 
> so this can not work unless grep is available in PATH.
> 
> Options would be to patch those scripts to use `dirname $0`/grep, or
> just test another program in package-from-tarball (the program testes
> used to be "true"). I am doing the latter to go on bootstraping.
> 
> BR
> 
> Carlos
> 

I thought it was just me! I've been looking at that test error for a
while with my aarch64 machine(s) and trying to figure out what to do
with it.

have you tried it with reverting commit 
84cbe39c5a21495647ab6528715c42e4c5b94e83 ?

-- 
Efraim Flashner      אפרים פלשנר
GPG key = A28B F40C 3E55 1372 662D  14F7 41AA E7DC CA3D 8351
Confidentiality cannot be guaranteed on emails sent or received unencrypted


signature.asc
Description: PGP signature

Re: Call for volunteer(s) for Guix "security" web page

[Leo Famulari]

On Tue, Sep 27, 2016 at 10:58:09AM +0200, Ludovic Courtès wrote:
> > +   (h2 "Release signatures")
> > +   (p "Releases of Guix and GuixSD are signed using the 
> > OpenPGP "
> > +  "key with the fingerprint "
> > +  "3CE4 6455 8A84 FDC6 9DB4  0CFB 090B 1199 3D9A EBB5.  "
> > +  "This key can be obtained from XXX.")
> 
> Maybe link to
> 
> or copy/paste the text?  Though we should give a ‘gpg --recv-keys’
> command that uses the full fingerprint instead of just the 64-bit ID
> (which is still too small, some say.)

Here's a patch that uses the fingerprint in guix.texi. What do you
think? Also, please verify that I've got it right :)
From 64b1df0a9565154ac2a1bd5289a13572b00bb5e0 Mon Sep 17 00:00:00 2001
From: Leo Famulari 
Date: Tue, 27 Sep 2016 14:12:02 -0400
Subject: [PATCH] doc: Give the full key fingerprint instead of the long key
 ID.

* doc/guix.texi (OPENPGP-SIGNING-KEY-ID): Use fingerprint instead of
long key ID.
---
 doc/guix.texi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/doc/guix.texi b/doc/guix.texi
index c159e12..239428a 100644
--- a/doc/guix.texi
+++ b/doc/guix.texi
@@ -10,7 +10,7 @@
 @include version.texi
 
 @c Identifier of the OpenPGP key used to sign tarballs and such.
-@set OPENPGP-SIGNING-KEY-ID 090B11993D9AEBB5
+@set OPENPGP-SIGNING-KEY-ID 3CE464558A84FDC69DB40CFB090B11993D9AEBB5
 
 @copying
 Copyright @copyright{} 2012, 2013, 2014, 2015, 2016 Ludovic Courtès@*
-- 
2.10.0



signature.asc
Description: PGP signature

Re: Call for volunteer(s) for Guix "security" web page

[Leo Famulari]

On Tue, Sep 27, 2016 at 10:58:09AM +0200, Ludovic Courtès wrote:
> > +   (h2 "Release signatures")
> > +   (p "Releases of Guix and GuixSD are signed using the 
> > OpenPGP "
> > +  "key with the fingerprint "
> > +  "3CE4 6455 8A84 FDC6 9DB4  0CFB 090B 1199 3D9A EBB5.  "
> > +  "This key can be obtained from XXX.")
> 
> Maybe link to
> 
> or copy/paste the text?  Though we should give a ‘gpg --recv-keys’
> command that uses the full fingerprint instead of just the 64-bit ID
> (which is still too small, some say.)
> 
> > +   (h2 "Security updates")
> > +   (p "When security vulnerabilities are found in Guix or the "
> > +  "packages provided by Guix, we will provide "
> > +  (a (@ (href ,(base-url 
> > "manual/html_node/Security-Updates.html")))
> > + "security updates")
> > +  " quickly and with minimal disruption for users.")
> 
> Maybe also that Guix is a “rolling release”, so there’s currently no
> separate security-fix branch and all critical fixes go to master?

I tried to implement these suggestion in the attached patch.

> I wonder if it would make sense to add a note on reproducible builds,
> ‘guix challenge’ and all that; later maybe!

Yes, later. Volunteers still welcome :)

> Note that you’ll then need to commit the resulting HTML to CVS(!) to
> that the update pages show up, as per the instructions available on the
> Savannah project page.  If you’re unsure or anything, I can do that.

I'll try it if this new patch is okay.
From eeff071ec9fbe527a97e2c7487e79e4b843916a1 Mon Sep 17 00:00:00 2001
From: Leo Famulari 
Date: Sun, 25 Sep 2016 18:43:28 -0400
Subject: [PATCH] www: security: New page.

* website/www/security.scm: New file.
* website/www.scm (%web-pages): Add security-page.
* website/www/shared.scm (html-page-links): Add "Security".
---
 website/www.scm  |  2 ++
 website/www/security.scm | 55 
 website/www/shared.scm   |  1 +
 3 files changed, 58 insertions(+)
 create mode 100644 website/www/security.scm

diff --git a/website/www.scm b/website/www.scm
index f0465eb..244830b 100644
--- a/website/www.scm
+++ b/website/www.scm
@@ -28,6 +28,7 @@
   #:use-module (www about)
   #:use-module (www contribute)
   #:use-module (www help)
+  #:use-module (www security)
   #:use-module (sxml simple)
   #:use-module (sxml match)
   #:use-module (web client)
@@ -335,6 +336,7 @@ Distribution.")
 ("donate/index.html" ,donate-page)
 ("download/index.html" ,download-page)
 ("help/index.html" ,help-page)
+("security/index.html" ,security-page)
 ;; ("packages/index.html" ,packages-page) ; Need Guix
 ;; ("packages/issues.html" ,issues-page)
 ))
diff --git a/website/www/security.scm b/website/www/security.scm
new file mode 100644
index 000..efe8315
--- /dev/null
+++ b/website/www/security.scm
@@ -0,0 +1,55 @@
+;;; GuixSD website --- GNU's advanced distro website
+;;; Copyright © 2016 Leo Famulari 
+;;;
+;;; This file is part of GuixSD website.
+;;;
+;;; GuixSD website is free software; you can redistribute it and/or modify it
+;;; under the terms of the GNU Affero General Public License as published by
+;;; the Free Software Foundation; either version 3 of the License, or (at
+;;; your option) any later version.
+;;;
+;;; GuixSD website is distributed in the hope that it will be useful, but
+;;; WITHOUT ANY WARRANTY; without even the implied warranty of
+;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+;;; GNU Affero General Public License for more details.
+;;;
+;;; You should have received a copy of the GNU Affero General Public License
+;;; along with GuixSD website.  If not, see .
+
+(define-module (www security)
+  #:use-module (www utils)
+  #:use-module (www shared)
+  #:export (security-page))
+
+(define (security-page)
+  `(html (@ (lang "en"))
+ ,(html-page-header "Security")
+ ,(html-page-links)
+ (div (@ (id "content-box"))
+  (article
+   (h1 "Security")
+   (h2 "How to report security issues")
+   (p "To report sensitive security issues in Guix itself or the 
packages it "
+  "provides, you can write to the private mailing list "
+  (a (@ (href 
"https://lists.gnu.org/mailman/listinfo/guix-security;))
+ ("guix-secur...@gnu.org"))
+ ".  This list is monitored by a small team of Guix "
+ "developers.")
+   (h2 "Release signatures")
+   (p "Releases of Guix and GuixSD are signed using the OpenPGP "
+  "key with the fingerprint "
+  "3CE4 6455 8A84 FDC6 9DB4  0CFB 090B 1199 

Re: bug #18742

[ng0]

Leo Famulari  writes:

 > However you are replying to these bug reports is changing the "Subject"
 > line to only include the bug number. It's hard to know what the bug is
 > about...

I did not want to reply directly to these specific bugs, which is why I
asked on this list first.

 > On Tue, Sep 27, 2016 at 04:38:03PM +, ng0 wrote:
 > > http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18742
 > > 
 > > Logs are gone since hydra got reset at least once since October 2014. I
 > [ 2 more citation lines. Click/Enter to show. ]
 > > can not see but only assume that the tests I am currently helping to fix
 > > are different tests than the ones which failed in 2014, as the test
 > > suite for GNUnet was refactored.
 > > Can we close this bug or should I reference this bug in an update commit
 > > if/when the test suite is fixed for Guix builds?

 > It seems to be building or not building (depending on the architecture)
 > consistently. If you are happy with those results, please close the bug
 > :)

I think I can not do this, close bugs I did not create myself. If I am
wrong about this I advise that we write a short appendum for Guix
documentation (part: contributing, new subsection: reporting bugs) which
explains debbugs usage, because upstream or at least the version+online
handbook on gnu.org is simply terrible in getting to the point.


 > https://hydra.gnu.org/search?query=gnunet

Dependencies fail on architecture I do not have available (yet / in a
forseeable future).

I think I'm okay with referencing this bug when/if I can close the
current failing test suite bug.
-- 
  ng0

Re: [PATCH] gnu: Add sdcc.

[David Craven]

> This URL returns 404.

thanks! missed that due to caching - again.

is it ok in it's own file? or should that file be renamed to
microcontroller.scm?

Re: [PATCH] gnu: Add sdcc.

[Leo Famulari]

On Tue, Sep 27, 2016 at 07:40:26PM +0200, David Craven wrote:
> * gnu/packages/sdcc.scm (sdcc): New variable.

> +  (uri (string-append
> +"mirror://sourceforge/sdcc/files/sdcc"
> +"/" version "/sdcc-src-" version ".tar.bz2"
> +"/download"))

This URL returns 404.

Otherwise, not having built the package, the package defintion looks
good to me in general.

Re: Call for volunteer(s) for Guix "security" web page

[Leo Famulari]

On Thu, Sep 22, 2016 at 10:04:37AM +, ng0 wrote:
> In my opinion this could be extended later by something similar to
> https://security.gentoo.org/ and its subpages.
> As we don't have much on that topic currently, we can't write about
> it. If this would be too much for the website, an inclusion in the
> manual of which security measurements a vanilla Guix offers would be
> good.
> One example: I stumbled upon our /dev/mem configuration only when I
> wanted to use flashrom for internal flashing. This is not documented
> anywhere.

It's a good idea, but of course somebody needs to actually do the work
:)

My primary goal for this page right now is to make it easy to for anyone
to learn how to contact us about security issues. I'm hoping this page
is the first result on DuckDuckGo and Google for "guix security".

Re: bug #18742

[Leo Famulari]

On Tue, Sep 27, 2016 at 04:38:03PM +, ng0 wrote:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18742
> 
> Logs are gone since hydra got reset at least once since October 2014. I
> can not see but only assume that the tests I am currently helping to fix
> are different tests than the ones which failed in 2014, as the test
> suite for GNUnet was refactored.
> Can we close this bug or should I reference this bug in an update commit
> if/when the test suite is fixed for Guix builds?

https://hydra.gnu.org/search?query=gnunet

Re: bug #18742

[Leo Famulari]

On Tue, Sep 27, 2016 at 04:38:03PM +, ng0 wrote:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18742
> 
> Logs are gone since hydra got reset at least once since October 2014. I
> can not see but only assume that the tests I am currently helping to fix
> are different tests than the ones which failed in 2014, as the test
> suite for GNUnet was refactored.
> Can we close this bug or should I reference this bug in an update commit
> if/when the test suite is fixed for Guix builds?

It seems to be building or not building (depending on the architecture)
consistently. If you are happy with those results, please close the bug
:)

Re: bug #21879

[Leo Famulari]

On Tue, Sep 27, 2016 at 04:45:28PM +, ng0 wrote:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=21879
> 
> Can someone running Guix on a Debian based system check this if this
> weechat/python bug can be closed?

I still have the error.

patch a file which is created while a phase is running which otherwise fails?

[ng0]

How can I achieve this?

For multiple packages I have this happens:

Application setup launches into configure or build phase, some script
creates a file where the creation process can not be located easily
asuming that a "#! /bin/sh" as shebang and/or "/bin/sh" as executable
somewhere in the file will just work and our build system says "¡No
pasarán!"...

So I was thinking, Guile is flexible, it should be achievable to have
something which wraps itself around the phase and for a new file it just
patches in what you throw at it in advance. Of course I'm open for
whatever works before we get to this.

Examples I can think off my head are the mach buildsystem used by the
firefox deriviate palemoon, pike (as far as I know, been some time since
I made progress on this one).

Thanks
-- 
  ng0

Re: Expat regression fix for master branch

[Leo Famulari]

On Mon, Sep 26, 2016 at 11:21:07AM +0300, Efraim Flashner wrote:
> On Sun, Sep 25, 2016 at 07:18:11PM -0400, Leo Famulari wrote:
> > On Mon, Sep 12, 2016 at 05:35:15PM -0400, Leo Famulari wrote:
> > > This patch applies an upstream patch for a regression caused by the fix 
> > > for CVE-2016-0718.
> > > 
> > > Apparently, the bug only manifests when building with -DXML_UNICODE,
> > > which I don't think our package does.
> > 
> > Sebastian Pipping (the Expat maintainer) contacted me to recommend that
> > we apply the patch on the master branch.
> > 
> > He says that the faulty code path can be reached even when XML_UNICODE
> > is not defined. Apparently, building with -DXML_UNICODE merely makes it
> > easier to reach the faulty code.
> > 
> > I think we should take Sebastian's advice. What does everyone think?
> 
> 
> Seems to me that as the Expat maintainer he would know best.

I pushed the commit as b9bc6e842066b066ebdf9eaf75d41753598d75b5


signature.asc
Description: PGP signature

Re: [PATCH] Add gctp

[Leo Famulari]

On Sat, Sep 24, 2016 at 09:49:56PM +0200, Thomas Danckaert wrote:
> From: Leo Famulari 
> Subject: Re: [PATCH] Add gctp
> Date: Fri, 23 Sep 2016 21:08:25 -0400
> 
> >> +(version "2.0")
> >> +(source
> >> + (origin
> >> +   (method url-fetch)
> >> +   (uri "https://github.com/OkoSanto/GCTP/archive/v2.0.0.tar.gz;)
> > 
> > The only thing I would like to change is this mismatch between the
> > version and the tarball name. Can we call it version 2.0.0?
> > 
> > Also, we should add a (file-name) field so that the downloaded tarball
> > has "gctp" in the name.
> > 
> > I can make these changes locally if you think they are okay.
> 
> Of course, go ahead!
> 
> Thanks for reviewing the patch(-es).

Great, pushed as 32f70e1f86141ed30f1026ea0cf8a4bd377004b0 !

> 
> Thomas

Re: [PATCH] gnu: Add gloox.

[Leo Famulari]

On Sat, Sep 24, 2016 at 05:44:27PM +, ng0 wrote:
> From f6ea8a17772b7d3b80a603d7aef576bf13e13be5 Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Thu, 15 Sep 2016 21:46:46 +
> Subject: [PATCH] gnu: Add gloox.
> 
> * gnu/packages/messaging.scm (gloox): New variable.

Thanks! I made the home-page an HTTPS URL and pushed.


signature.asc
Description: PGP signature

Re: gnu: Add emacs-rudel.

[ng0]

Hi,

thanks for reviewing.

Alex Vong  writes:

> Hello,
>
> ng0  writes:
>
>> From b298b54bb732651daf7f7bb6a48e4749a4e391ed Mon Sep 17 00:00:00 2001
>> From: ng0 
>> Date: Tue, 27 Sep 2016 16:24:29 +
>> Subject: [PATCH] gnu: Add emacs-rudel.
>>
>> * gnu/packages/emacs.scm (emacs-rudel): New variabel.
>> ---
>>  gnu/packages/emacs.scm | 29 -
>>  1 file changed, 28 insertions(+), 1 deletion(-)
>>
>> diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
>> index 57fe8a9..e0df7c9 100644
>> --- a/gnu/packages/emacs.scm
>> +++ b/gnu/packages/emacs.scm
>> @@ -12,7 +12,7 @@
>>  ;;; Copyright © 2016 David Thompson 
>>  ;;; Copyright © 2016 Matthew Jordan 
>>  ;;; Copyright © 2016 Roel Janssen 
>> -;;; Copyright © 2016 ng0 
>> +;;; Copyright © 2016 ng0 
>>  ;;; Copyright © 2016 Alex Griffin 
>>  ;;; Copyright © 2016 Nicolas Goaziou 
>>  ;;; Copyright © 2016 Alex Vong 
>> @@ -,6 +,33 @@ perspective only its buffers are available by 
>> default.")
>>  ;; the Expat license.
>>  (license license:gpl3+)))
>>  
>> +(define-public emacs-rudel
>> +  (package
>> +(name "emacs-rudel")
>> +(version "0.3.1")
>> +(source
>> + (origin
>> +   (method url-fetch)
>> +   (uri (string-append "http://elpa.gnu.org/packages/rudel-;
>> +   version ".tar"))
>> +   (sha256
>> +(base32
>> + "0glqa68g509p0s2vcc0i8kzlddnc9brd9jqhnm5rzxz4i050cvnz"
>> +(build-system emacs-build-system)
>> +(home-page "http://rudel.sourceforge.net/;)
> Should https be used?

Do sf.net have https certs? I don't believe so:

Oops! Unable to display this website
The site at https://rudel.sourceforge.net/ seems to be unavailable.

It may be temporarily inaccessible or moved to a new address. You may wish to 
verify that your internet connection is working correctly.

▼ Technical information
The precise error was: Could not connect: Connection refused

Reload

>> +(synopsis "Collaborative editing framework")
>> +(description
>> + "Rudel is a collaborative editing environment for GNU Emacs.  Its 
>> purpose is to
>> +share buffers with other users in order to edit the contents of those 
>> buffers
>> +collaboratively.  Rudel supports multiple backends to enable communication 
>> with other
>> +collaborative editors using different protocols, though currently Obby (for 
>> use with
>> +the Gobby editor) is the only fully-functional one.
>> +Rudel mostly works in the background to change the behavior of the set of 
>> Emacs
>> +buffers for which it has been activated.  The user interface consists of a 
>> set of key
>> +bindings, a menu entry and some visual status indicators, which are added 
>> to the text,
>> +header line and/or mode line of buffers for which Rudel has been 
>> activated.")
>> +(license license:gpl3+)))
>> +
> I think the lines are overly long here.

Less than 90 or even 89 characters.

ng0@shadowwalker ~/src/guix/guix$ ./pre-inst-env guix lint emacs-rudel
;;; note: source file /home/ng0/src/guix/guix/gnu/packages/emacs.scm
;;;   newer than compiled /home/ng0/src/guix/guix/gnu/packages/emacs.go
gnu/packages/emacs.scm:2238:15: emacs-rudel-0.3.1: URI 
http://rudel.sourceforge.net/ unreachable: Connection refused
fetching CVE database for 2016.

uri problem can be ignored, loads in web browsers.

>>  (define-public emacs-hydra
>>(package
>>  (name "emacs-hydra")
>> -- 
>> 2.10.0
>
> Otherwise, the patch looks good and does build.

Yes. Functionality check is up to people who can use this, I was only
able to launch into rudel, no known server to connect to etc. I'd really
appreciate if someone would test it after this is commited.

> Some unrelated stuff to all: Right now, I only have free time on Tuesday
> and weekend, so I may not reply to email immediately.
>
> Cheers,
> Alex
>

-- 
  ng0

Re: [PATCH 3/3] gnu: Duplicity: Update to 0.7.10

[Leo Famulari]

On Thu, Sep 22, 2016 at 04:31:37PM +1000, Brendan Tildesley wrote:
> Eric Bavier 於 2016-09-21 05:25 寫道:
> > Let's leave out the nnecessary whitespace changes.
> >
> > Could you send an updated patch?
> 
> Ok, I remade it all to look like I didn't just rewrite everything from
> scratch, and added the incorrect indentation back in. Seems odd to
> prefer keeping the patch small over fixing indentation.

Having lots of code changes that don't change the function of the code
makes it difficult for reviewers and Git log readers to understand what
was changed. If a package definition has really bad indentation or style
problems, then I think it's best to fix them in their own commit where
reviewers can easily determine that nothing functional has changed.

Thank you for revising your patches!

> From 15721c677f20156071fe0efb150b3af63f64648c Mon Sep 17 00:00:00 2001
> From: Brendan Tildesley 
> Date: Tue, 20 Sep 2016 20:41:30 +1000
> Subject: [PATCH 1/3] gnu: duplicity: Update to 0.7.10.
> 
> * gnu/packages/backup.scm (duplicity): Update to 0.7.10.
> [source]: Remove duplicity-piped-password.patch.
> [source]: Remove duplicity-test_selection-tmp.patch.
> * gnu/local.mk (dist_patch_DATA): Removed above patch file entries.

In general, could you try following our conventions when writing the
changelogs? [0] Reviewers can rewrite them, of course, but we'd prefer
to not have to. For this patch, it would look something like what's
below.  The important thing is to mention all the changes. I look at
`git log` when I am unsure.

Subject: [PATCH 1/3] gnu: duplicity: Update to 0.7.10.

* gnu/packages/backup.scm (duplicity): Update to 0.7.10.
[source]: Remove obsolete patches.
[inputs]: Remove util-linux. Make python2-mock a native-input.
[arguments]: Also substitute path in 'testing/overrides/bin/lftp' in phase
'check-setup'.
* gnu/packages/patches/duplicity-piped-password.patch, 
gnu/packages/patches/duplicity-test_selection-tmp.patch: Delete files.
* gnu/local.mk (dist_patch_DATA): Remove them.

> From aa2ed1cc2e6863f26e675f77ff210b346b8bfd93 Mon Sep 17 00:00:00 2001
> From: Brendan Tildesley 
> Date: Thu, 22 Sep 2016 01:45:19 +1000
> Subject: [PATCH 2/3] gnu: duplicity: Use modify-phases.
> 
> * gnu/packages/backup.scm (duplicity): Use modify-phases instead of
>   alist-cons-before.

The patch LGTM, although I'd silently edit the commit message to this
when pushing:

* gnu/packages/backup.scm (duplicity)[arguments]: Use modify-phases instead of
alist-cons-before.

> From 95a4a505cf48627205cfd734b73c74913d88c36c Mon Sep 17 00:00:00 2001
> From: Brendan Tildesley 
> Date: Thu, 22 Sep 2016 01:03:32 +1000
> Subject: [PATCH 3/3] gnu: duplicity: Add various backends
> 
> * gnu/packages/backup.scm (duplicity): Add various backends.

* gnu/packages/backup.scm (duplicity)[inputs]: Add python2-lockfile,
python2-pexpect, python2-paramiko, python2-pycrypto, python2-botocore,
python2-dropbox, lftp, ncftp, par2cmdline.

This patch doesn't work, because we don't have a python2-dropbox
package. Also, I noticed that without this patch, the first patch (gnu:
duplicity: Update to 0.7.10.) failed because it needs pexpect.

So, whatever is needed for the update should go in that first patch, and
the python2-dropbox patch should be provided at the beginning of this
patch series.

Can you send a revised patch series?

[0] See the text about the ChangeLog format:
https://www.gnu.org/software/guix/manual/html_node/Submitting-Patches.html

[PATCH 1/3] gnu: Add m17n-db.

[Arun Isaac]

* gnu/packages/emacs.scm (m17n-db): New variable.
---
 gnu/packages/emacs.scm | 36 
 1 file changed, 36 insertions(+)

diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
index 57fe8a9..2506d3e 100644
--- a/gnu/packages/emacs.scm
+++ b/gnu/packages/emacs.scm
@@ -16,6 +16,7 @@
 ;;; Copyright © 2016 Alex Griffin 
 ;;; Copyright © 2016 Nicolas Goaziou 
 ;;; Copyright © 2016 Alex Vong 
+;;; Copyright © 2016 Arun Isaac 
 ;;;
 ;;; This file is part of GNU Guix.
 ;;;
@@ -74,6 +75,7 @@
   #:use-module (gnu packages statistics)
   #:use-module (gnu packages xiph)
   #:use-module (gnu packages mp3)
+  #:use-module (gnu packages gettext)
   #:use-module (guix utils)
   #:use-module (srfi srfi-1))
 
@@ -3104,3 +3106,37 @@ that allows users to concentrate more on their own work. 
 Its features are:
 a visual interface, reduce overhead of completion by using statistic method,
 extensibility.")
 (license license:gpl3+)))
+
+(define-public m17n-db
+  (package
+(name "m17n-db")
+(version "1.7.0")
+(source
+ (origin
+   (method url-fetch)
+   (uri (string-append
+ "http://download.savannah.gnu.org/releases/m17n/m17n-db-; version 
".tar.gz"))
+   (sha256
+(base32 "1w08hnsbknrcjlzp42c99bgwc9hzsnf5m4apdv0dacql2s09zfm2"
+(build-system gnu-build-system)
+(inputs
+ `(("gettext" ,gnu-gettext)))
+(arguments
+ `(#:configure-flags
+   (list (string-append "--with-charmaps="
+(assoc-ref %build-inputs "libc")
+"/share/i18n/charmaps"
+;; With `guix lint' the home-page URI returns a small page saying
+;; that your browser does not handle frames. This triggers the "URI
+;; returns suspiciously small file" warning.
+(home-page "http://www.nongnu.org/m17n/;)
+(synopsis "Multilingual text processing library (database)")
+(description "The m17n library realizes multilingualization of
+many aspects of applications.  The m17n library represents
+multilingual text as an object named M-text.  M-text is a string with
+attributes called text properties, and designed to substitute for
+string in C.  Text properties carry any information required to input,
+display and edit the text.
+
+This package contains the library database.")
+(license license:lgpl2.1+)))
-- 
2.10.0

[PATCH 2/3] gnu: Add m17n-lib.

[Arun Isaac]

* gnu/packages/emacs.scm (m17n-lib): New variable.
---
 gnu/packages/emacs.scm | 40 
 1 file changed, 40 insertions(+)

diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
index 2506d3e..7d10492 100644
--- a/gnu/packages/emacs.scm
+++ b/gnu/packages/emacs.scm
@@ -76,6 +76,9 @@
   #:use-module (gnu packages xiph)
   #:use-module (gnu packages mp3)
   #:use-module (gnu packages gettext)
+  #:use-module (gnu packages fribidi)
+  #:use-module (gnu packages gd)
+  #:use-module (gnu packages fontutils)
   #:use-module (guix utils)
   #:use-module (srfi srfi-1))
 
@@ -3140,3 +3143,40 @@ display and edit the text.
 
 This package contains the library database.")
 (license license:lgpl2.1+)))
+
+(define-public m17n-lib
+  (package
+(name "m17n-lib")
+(version "1.7.0")
+(source
+ (origin
+   (method url-fetch)
+   (uri (string-append
+ "http://download.savannah.gnu.org/releases/m17n/m17n-lib-;
+ version ".tar.gz"))
+   (sha256
+(base32 "10yv730i25g1rpzv6q49m6xn4p8fjm7jdwvik2h70sn8w3hm7f4f"
+(build-system gnu-build-system)
+(inputs
+ `(("fribidi" ,fribidi)
+   ("gd" ,gd)
+   ("libotf" ,libotf)
+   ("libxft" ,libxft)
+   ("libxml2" ,libxml2)
+   ("m17n-db" ,m17n-db)))
+(arguments
+ `(#:parallel-build? #f))
+;; With `guix lint' the home-page URI returns a small page saying
+;; that your browser does not handle frames. This triggers the "URI
+;; returns suspiciously small file" warning.
+(home-page "http://www.nongnu.org/m17n/;)
+(synopsis "Multilingual text processing library (runtime)")
+(description "The m17n library realizes multilingualization of
+many aspects of applications.  The m17n library represents
+multilingual text as an object named M-text.  M-text is a string with
+attributes called text properties, and designed to substitute for
+string in C.  Text properties carry any information required to input,
+display and edit the text.
+
+This package contains the library runtime.")
+(license license:lgpl2.1+)))
-- 
2.10.0

emacs m17n support

[Arun Isaac]

These are the patches to build emacs with m17n support. Please review and
consider for inclusion.

Thanks.

[PATCH 3/3] gnu: emacs: Build with m17n support.

[Arun Isaac]

* gnu/packages/emacs.scm (emacs): Build with m17n support.
---
 gnu/packages/emacs.scm | 6 +-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
index 7d10492..d311bcf 100644
--- a/gnu/packages/emacs.scm
+++ b/gnu/packages/emacs.scm
@@ -160,7 +160,11 @@
("libsm" ,libsm)
("alsa-lib" ,alsa-lib)
("dbus" ,dbus)
-   ("guix-src" ,(package-source guix
+   ("guix-src" ,(package-source guix))
+
+   ;; multilingualization support
+   ("libotf" ,libotf)
+   ("m17n-lib" ,m17n-lib)))
 (native-inputs
  `(("pkg-config" ,pkg-config)
("texinfo" ,texinfo)))
-- 
2.10.0

Re: gnu: Add emacs-rudel.

[Alex Vong]

Hello,

ng0  writes:

> From b298b54bb732651daf7f7bb6a48e4749a4e391ed Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Tue, 27 Sep 2016 16:24:29 +
> Subject: [PATCH] gnu: Add emacs-rudel.
>
> * gnu/packages/emacs.scm (emacs-rudel): New variabel.
> ---
>  gnu/packages/emacs.scm | 29 -
>  1 file changed, 28 insertions(+), 1 deletion(-)
>
> diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
> index 57fe8a9..e0df7c9 100644
> --- a/gnu/packages/emacs.scm
> +++ b/gnu/packages/emacs.scm
> @@ -12,7 +12,7 @@
>  ;;; Copyright © 2016 David Thompson 
>  ;;; Copyright © 2016 Matthew Jordan 
>  ;;; Copyright © 2016 Roel Janssen 
> -;;; Copyright © 2016 ng0 
> +;;; Copyright © 2016 ng0 
>  ;;; Copyright © 2016 Alex Griffin 
>  ;;; Copyright © 2016 Nicolas Goaziou 
>  ;;; Copyright © 2016 Alex Vong 
> @@ -,6 +,33 @@ perspective only its buffers are available by 
> default.")
>  ;; the Expat license.
>  (license license:gpl3+)))
>  
> +(define-public emacs-rudel
> +  (package
> +(name "emacs-rudel")
> +(version "0.3.1")
> +(source
> + (origin
> +   (method url-fetch)
> +   (uri (string-append "http://elpa.gnu.org/packages/rudel-;
> +   version ".tar"))
> +   (sha256
> +(base32
> + "0glqa68g509p0s2vcc0i8kzlddnc9brd9jqhnm5rzxz4i050cvnz"
> +(build-system emacs-build-system)
> +(home-page "http://rudel.sourceforge.net/;)
Should https be used?

> +(synopsis "Collaborative editing framework")
> +(description
> + "Rudel is a collaborative editing environment for GNU Emacs.  Its 
> purpose is to
> +share buffers with other users in order to edit the contents of those buffers
> +collaboratively.  Rudel supports multiple backends to enable communication 
> with other
> +collaborative editors using different protocols, though currently Obby (for 
> use with
> +the Gobby editor) is the only fully-functional one.
> +Rudel mostly works in the background to change the behavior of the set of 
> Emacs
> +buffers for which it has been activated.  The user interface consists of a 
> set of key
> +bindings, a menu entry and some visual status indicators, which are added to 
> the text,
> +header line and/or mode line of buffers for which Rudel has been activated.")
> +(license license:gpl3+)))
> +
I think the lines are overly long here.

>  (define-public emacs-hydra
>(package
>  (name "emacs-hydra")
> -- 
> 2.10.0

Otherwise, the patch looks good and does build.

Some unrelated stuff to all: Right now, I only have free time on Tuesday
and weekend, so I may not reply to email immediately.

Cheers,
Alex

Re: gnu: Add xonsh

[Leo Famulari]

On Wed, Sep 21, 2016 at 11:18:28PM +0200, Danny Milosavljevic wrote:
> Hi,
> 
> thanks for this package! With the recent 
> how-did-anyone-think-that-was-a-good-idea in bash (again), this is very 
> useful.
> 
> However, are you sure you want to have no inputs for xonsh?
> 
> I can see that it supports at least these:
> - python-ply
> - python-sphinx
> - python-numpydoc
> - python-prompt-toolkit
> 
> And maybe also these features (not that important):
> - jupyter
> - setproctitle
> - distro
> - Cloud
> - Sphinx-Theme
> 
> What do you think?

Should it block the package? Or perhaps we can push the patch as-is and
you can submit a follow-up to add the features you want?

Re: [PATCH 1/2] gnu: Add perl-net-psyc.

[Leo Famulari]

On Wed, Sep 21, 2016 at 06:46:31PM +, ng0 wrote:
> From 74a6c1e552a6ae8f438e91cbe318882401b440f8 Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Wed, 21 Sep 2016 18:08:42 +
> Subject: [PATCH 1/2] gnu: Add psyclpc.
> 
> * gnu/packages/psyc.scm (psyclpc): New variable.

> +(source (origin
> +  (method url-fetch)
> +  (uri (list (string-append "http://www.psyced.org/files/;
> +name "-" version ".tar.xz")
> + (string-append "http://krosos.sdf.org/static/unix/;
> +"perlpsyc-" version ".zip")))
> +  (sha256
> +   (base32
> +"0c2afcj8b2yr2vmy9sy0528iqs9sw01j6q35lvxicm42gs7vnik2"

Do both of those URLs provide a file with the same hash?

> +(arguments
> + `(#:tests? #f ; There are no tests/checks.
> +   #:configure-flags (list
> +  "--enable-use-tls=yes"
> +  "--enable-use-mccp"

The "Mud Client Compression Protocol"?

> +  (string-append "--with-openssl="
> + (assoc-ref %build-inputs "openssl"))
> +  (string-append "--prefix="
> + (assoc-ref %outputs "out"))
> +  (string-append "--libdir=" ; "-DMUD_LIB="

What does this commented text mean?

> + (assoc-ref %outputs "out")
> + "/opt/psyced/world")
> +  (string-append "--bindir="
> + (assoc-ref %outputs "out")
> + "/opt/psyced/bin")
> +  (string-append "--libexecdir=" ; "-DERQ_DIR="

Same question here.

> +(inputs
> + `(("zlib" ,zlib)
> +   ("openssl" ,openssl)))
> +;; pcre is bundled to ensure the version is compatible. XXX: look into
> +;; unbundling it. Upstream should update from pcre 4.5 to 8.38. For
> +;; functionality reasons we can not unbundle it now.
> +;; ("pcre" ,pcre)))

That version of PCRE was released in 2003. We might want to add a
warning to the package description...

https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=pcre


signature.asc
Description: PGP signature

Re: [PATCH 1/2] gnu: Add perl-net-psyc.

[Leo Famulari]

On Wed, Sep 21, 2016 at 10:34:15AM +, ng0 wrote:
> From 33d8584709a74e20924743b2606158cacd5ed0c6 Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Tue, 13 Sep 2016 10:20:31 +
> Subject: [PATCH 1/2] gnu: Add perl-net-psyc.
> 
> * gnu/packages/psyc.scm: New file.
> * gnu/local.mk (GNU_SYSTEM_MODULES): Add it.

The hash is incorrect:

Starting download of 
/gnu/store/7kl9swdb9215lzdg1f7z4y41z4jssb7w-perl-net-psyc-1.1.zip
From http://perl.psyc.eu/perlpsyc-1.1.zip...
 perlpsyc-1.1.zip  174KiB   282KiB/s 00:01 [] 100.0%
output path `/gnu/store/7kl9swdb9215lzdg1f7z4y41z4jssb7w-perl-net-psyc-1.1.zip' 
should have sha256 hash `1lw6807qrbmvzbrjn1rna1dhir2k70xpcjvyjn45y35hav333a42', 
instead has `02isxl14n7jb18s5qr482y5ndj3cy04xw3699wvcxrr98lxj7xim'
@ build-failed 
/gnu/store/6lfrf1q84v86l5bvh1sf9x6zv3k9gv4j-perl-net-psyc-1.1.zip.drv - 1 
output path `/gnu/store/7kl9swdb9215lzdg1f7z4y41z4jssb7w-perl-net-psyc-1.1.zip' 
should have sha256 hash `1lw6807qrbmvzbrjn1rna1dhir2k70xpcjvyjn45y35hav333a42', 
instead has `02isxl14n7jb18s5qr482y5ndj3cy04xw3699wvcxrr98lxj7xim'
cannot build derivation 
`/gnu/store/i4xfmwz4jiqw7jcs2njhgddq4a4l28ws-perl-net-psyc-1.1.tar.xz.drv': 1 
dependencies couldn't be built
cannot build derivation 
`/gnu/store/amg0r4pkn9pky2v931ipd4r2sh66kaj1-perl-net-psyc-1.1.drv': 1 
dependencies couldn't be built
guix build: error: build failed: build of 
`/gnu/store/amg0r4pkn9pky2v931ipd4r2sh66kaj1-perl-net-psyc-1.1.drv' failed


signature.asc
Description: PGP signature

bug #21879

[ng0]

http://debbugs.gnu.org/cgi/bugreport.cgi?bug=21879

Can someone running Guix on a Debian based system check this if this
weechat/python bug can be closed?

Thanks
-- 
  ng0

bug #18742

[ng0]

http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18742

Logs are gone since hydra got reset at least once since October 2014. I
can not see but only assume that the tests I am currently helping to fix
are different tests than the ones which failed in 2014, as the test
suite for GNUnet was refactored.
Can we close this bug or should I reference this bug in an update commit
if/when the test suite is fixed for Guix builds?
-- 
  ng0

Re: [PATCH] gnu: openldap: Use gnutls.

[Alex Vong]

Efraim Flashner  writes:

> On Sat, Sep 24, 2016 at 11:11:02AM +0900, Ludovic Courtès wrote:
>> Efraim Flashner  skribis:
>> 
>> > On Sat, Sep 17, 2016 at 12:21:10AM +0800, Alex Vong wrote:
>> >> Hello,
>> >> 
>> >> The following patch does what the title says. The link of the bug
>> >>  is no
>> >> longer accessible, so I assume the bug is fixd since I cannot find it on
>> >> the internet. In addition, I have checked debian's version, which is
>> >> slightly older (2.4.42), which also uses gnutls. So I think we are safe
>> >> here. Please object if there is a problem!
>> >> 
>> >> Please note that openldap is an input for curl, so I guess it is causing
>> >> the openssl dependency problem for octave I mentioned earlier. I haven't
>> >> test to build octave with curl though, which takes a while.
>> >> 
>> >
>> > I built it out. With this patch, octave will build with curl and
>> > cyrus-sasl added as inputs, and will retain references to gnutls and
>> > cyrus-sasl, as checked with `guix gc --references /gnu/store/...octave'
>> 
>> So I think this patch can be safely applied, at least in core-updates
>> due to the large number of rebuilds.
>> 
>> Efraim, could you take care of it?
>> 
>> Thanks!
>> 
>> Ludo’.
>
> pushed to core-updates!

Thank you!

Re: [PATCH] gnu: femtolisp: Remove support for mips and armhf.

[Leo Famulari]

On Tue, Sep 27, 2016 at 08:13:53AM +, ng0 wrote:
> Subject: [PATCH] gnu: femtolisp: Remove support for mips and armhf.
> 
> * gnu/packages/lisp.scm (femtolisp)[supported-systems]: New field,
> remove support for mips64el and armhf.

Thanks, pushed!


signature.asc
Description: PGP signature

gnu: Add emacs-rudel.

[ng0]

From b298b54bb732651daf7f7bb6a48e4749a4e391ed Mon Sep 17 00:00:00 2001
From: ng0 
Date: Tue, 27 Sep 2016 16:24:29 +
Subject: [PATCH] gnu: Add emacs-rudel.

* gnu/packages/emacs.scm (emacs-rudel): New variabel.
---
 gnu/packages/emacs.scm | 29 -
 1 file changed, 28 insertions(+), 1 deletion(-)

diff --git a/gnu/packages/emacs.scm b/gnu/packages/emacs.scm
index 57fe8a9..e0df7c9 100644
--- a/gnu/packages/emacs.scm
+++ b/gnu/packages/emacs.scm
@@ -12,7 +12,7 @@
 ;;; Copyright © 2016 David Thompson 
 ;;; Copyright © 2016 Matthew Jordan 
 ;;; Copyright © 2016 Roel Janssen 
-;;; Copyright © 2016 ng0 
+;;; Copyright © 2016 ng0 
 ;;; Copyright © 2016 Alex Griffin 
 ;;; Copyright © 2016 Nicolas Goaziou 
 ;;; Copyright © 2016 Alex Vong 
@@ -,6 +,33 @@ perspective only its buffers are available by default.")
 ;; the Expat license.
 (license license:gpl3+)))
 
+(define-public emacs-rudel
+  (package
+(name "emacs-rudel")
+(version "0.3.1")
+(source
+ (origin
+   (method url-fetch)
+   (uri (string-append "http://elpa.gnu.org/packages/rudel-;
+   version ".tar"))
+   (sha256
+(base32
+ "0glqa68g509p0s2vcc0i8kzlddnc9brd9jqhnm5rzxz4i050cvnz"
+(build-system emacs-build-system)
+(home-page "http://rudel.sourceforge.net/;)
+(synopsis "Collaborative editing framework")
+(description
+ "Rudel is a collaborative editing environment for GNU Emacs.  Its purpose is to
+share buffers with other users in order to edit the contents of those buffers
+collaboratively.  Rudel supports multiple backends to enable communication with other
+collaborative editors using different protocols, though currently Obby (for use with
+the Gobby editor) is the only fully-functional one.
+Rudel mostly works in the background to change the behavior of the set of Emacs
+buffers for which it has been activated.  The user interface consists of a set of key
+bindings, a menu entry and some visual status indicators, which are added to the text,
+header line and/or mode line of buffers for which Rudel has been activated.")
+(license license:gpl3+)))
+
 (define-public emacs-hydra
   (package
 (name "emacs-hydra")
-- 
2.10.0

-- 
  ng0


signature.asc
Description: PGP signature

Re: Kallithea - looking for someone to take over.

[ng0]

Ludovic Courtès  writes:

> Hello!
>
> ng0  skribis:
>
>> I worked on kallithea-scm, while I am very close to finishing it I'd
>> like to let someone else finish this if possible.
>
> Brave endeavor, looks like you went pretty far!
>
> If you haven’t done it yet, could you ‘git send-email’ the subset of
> those patches that you consider ready, to facilitate the reviewer’s work
> and also to make it easier for whoever would like to take over your
> work?

Ok. That will be most of the patches I've sent, excluding
Kallithea itself. I will send individual patches later this week.

> Thanks!
>
> Ludo’.
>

-- 
  ng0

Re: [PATCH] gnu: service: Add git-service.

[ng0]

So almost one month passed now.
To continue testing this, I need help on this first.

Thanks.

ng0  writes:

> I tried to address most of what you've written.
>
> While I was correcting the documentation I decided to add more options,
> now it doesn't work anymore, probably because of the ifs I added.
>
> Andy Wingo  writes:
>
>> On Tue 30 Aug 2016 13:45, ng0  writes:
>>
I also think that "path" might
 not be the right word, which in GNU manuals is only used for search
 paths.  See the "GNU Manuals" section of standards.texi for more.
 Anyway I suggest #:base-directory.  Make sure the port is an integer and
 not a string.
>>>
>>> See 'man git daemon'.
>>
>> I ran this and it did not work -- first showed me a page for git then
>> for daemon.  I believe you want "man git-daemon"?
>>
>>> The switch is called --base-path.  Looking at the openrc conf.d/git or
>>> what the config file was called again, they stick to this name too.
>>> It would just cause confusion if we go ahead and call it differently.
>>> Upstream should be fixed, but I'm not going there.  If you think we
>>> should break expectations, I can rename it.
>>
>> "Fixing" upstream is out of our remit :)  All I can ask is that we do
>> not introduce new uses of the word "path".
>>
> +Return a service to run the @uref{https://git-scm.com, git} daemon 
> version control
> +daemon.

 Extra "daemon" here.  Probably needs a sentence on what running the
 daemon will do (namely, expose local repositories for remote access).

 What about authentication?  Is this purely anonymous?
>>>
>>> Exactly, authentication is handled via other daemons, for example ssh or
>>> gitolite. git daemon supports no authentication and is read-only, as far
>>> as I know. At the servers I use and setup, I pull via
>>> git://,http://,https:// and push via ssh.
>>> Its selfdescription is:
>>> git-daemon - A really simple server for Git repositories.
>>
>> This needs to be documented in the manual, is what I was getting at :)
>> Mention that this is for anonymous read-only access please.
>
> read-only was wrong, anonymous write-access for all the world can be set
> up but it is not default.
>
>>
Use "file name" instead of path in general.
>>>
>>> Why?
>>
>> It is because it is standard in the GNU project.  I mentioned this
>> before.  See "info standards" and go to "GNU manuals".
>>
> +Furthermore it takes the parameter @var{port} which defaults to 9418.
> +Run @command{man git daemon} for information about the options.

 This man command does not work.
>>>
>>> Works for me. As far as I know man pages were merged into git package
>>> recently. When I run this on debian with guix, 'man git daemon' works
>>> too.
>>
>> It does not work for me on NixOS with Guix.  Maybe I am out of date
>> though.
>>
> +(define %git-accounts
> +  ;; User account and groups for git-daemon.
> +  ;; We can give it git-shell for now, otherwise we can switch to 
> /bin/sh.

 What does this comment mean?  Why would we switch?
>>>
>>> I am not sure about the limitations of git-shell compared to
>>> /bin/sh. If this turns out to be a mistake, it can be corrected. The
>>> only thing I know about git-shell is that it allows no logins.
>>
>> If you do not want a login then probably what you want is
>> #~(string-append #$shadow "/sbin/nologin").
>>
>> Andy
>
> From d1d7eb59ca53833098cea2d6eddaa59f1494b579 Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Fri, 8 Jul 2016 15:42:55 +
> Subject: [PATCH] gnu: services: Add git-service.
>
> * gnu/services/version-control.scm: New file, create it.
> (git-service): New Procedures.
> (git-service-type): New variable.
> * doc/guix.texi: Add documentation.
> ---
>  doc/guix.texi|  37 
>  gnu/local.mk |   1 +
>  gnu/services/version-control.scm | 196 
> +++
>  3 files changed, 234 insertions(+)
>  create mode 100644 gnu/services/version-control.scm
>
> diff --git a/doc/guix.texi b/doc/guix.texi
> index b22cf4a..78d7ee1 100644
> --- a/doc/guix.texi
> +++ b/doc/guix.texi
> @@ -7494,6 +7494,7 @@ declaration.
>  * Database Services::   SQL databases.
>  * Mail Services::   IMAP, POP3, SMTP, and all that.
>  * Web Services::Web servers.
> +* Version Control:: Git and others.
>  * Various Services::Other services.
>  @end menu
>  
> @@ -9910,6 +9911,42 @@ directories are created when the service is activated.
>  
>  @end deffn
>  
> +@node Version Control
> +@subsubsection Version Control
> +
> +The @code{(gnu services version-control)} module provides the following 
> services:
> +
> +@deffn {Scheme Procedure} git-service [#:git @var{git}] @
> +   [#:base-directory "/var/git/repositories"] @
> +   [#:user-directory? #f ""] [#:port 9418] @

Re: [PATCH 0/4] Add gobby 0.4

[Andy Wingo]

On Tue 27 Sep 2016 14:31, Andy Wingo  writes:

> These patches add a collaborative editor that we use at work.
>
> Andy Wingo (4):
>   gnu: Add libxml++-2.
>   gnu: Add libnet6.
>   gnu: Add obby.
>   gnu: Add obby.
>
>  gnu/local.mk   |   1 +
>  gnu/packages/gnome.scm |  14 ++
>  gnu/packages/gobby.scm | 132 
> +
>  3 files changed, 147 insertions(+)
>  create mode 100644 gnu/packages/gobby.scm

Clearly I am an idiot.  I was trying the technique where you can
git format-patch to a directory and then git send-email from those
files.  Never more!  Anyway I pushed the above after fixing the lint
errors for the `description' fields.

Apologies for the spam!

Andy

fgrep fails to find grep during bootstrap

[Carlos Sánchez de La Lama]

Hi all,

I am trying to bootstrap for powerpc-linux. I have generated the
tarballs and patched gnu/packages/bootstrap.scm to use them.

However, I get an error during when creating %bootstrap-coreutils
package. It specifies "fgrep" as program to test, but fgrep is a shell
script, and altought the shebang is correctly patched, it needs "grep"
binary in the path (which is not at this stage). fgrep does:

exec grep -F "$@"

so this can not work unless grep is available in PATH.

Options would be to patch those scripts to use `dirname $0`/grep, or
just test another program in package-from-tarball (the program testes
used to be "true"). I am doing the latter to go on bootstraping.

BR

Carlos

-- 
-Tiens! Mais vous êtes un homme de coeur! dit sir Francis Cromarty.
-Quelquefois, répondit simplement Phileas Fogg. Quand j'ai le temps.

Jules Verne, "Le Tour du monde en quatre-vingts jours" (1872)

Re: [PATCH 0/12]: Add asdf-build-system.

[James Richardson]

Andy Patterson writes:

> Hello,
>
> As promised, here's my work toward adding a build system for Common Lisp
> software. I still have some issues with it as follows:
>
Thank you!

> Getting things to work "out of the box": I'd like to set up an environment
> variable to allow implementations to find installed libraries, but it's a bit
> tricky:
>
> Essentially I'd like to have, in each profile, a setup something like
>
> export CL_SOURCE_FRAGMENT=#+sbcl \
>   (:directory \"${GUIX_PROFILE}/share/common-lisp/sbcl-bundle-systems/\") \
>   #+ecl ... \
>   ${CL_SOURCE_FRAGMENT}
>
> followed by
>
> export CL_SOURCE_REGISTRY=\
>   (:source-registry ${CL_SOURCE_FRAGMENT} \
>   (:directory \"${GUIX_PROFILE}/share/common-lisp/systems/\") \
>   :inherit-configuration)
>
> This would allow each implementation to pick up their own compiled libraries,
> as well as systems installed as source as a fallback. I don't know how to
> achieve that kind of a setup, so I'd like some help. I'm also worried about
> either blowing away user configuration, or having it shadow the desired
> configuration, since this isn't a simple search path which could be appended
> to. Should we try to patch the asdf implementations to handle a separate
> variable just for guix?
>
> I'd also like to hear your thoughts on how packages are laid out using the
> system, so I've included some examples. I was thinking that something like
> package-with-python2 could also be helpful here; is that the right way
> forward?
>

I don't really know enough to offer valid suggestions yet. There are
mostly the same questions/issues I hit when trying to package a lisp
program with dependencies. The Debian project came up with a thing
called common-lisp-controller
(http://ci-debian.alioth.debian.org/clid/clid.html/ch-clc.html) which
address some of the issues. It seems to build an abstraction layer over
asdf. I'm not advocating this as it does seem to presume a Debian-ish
infrastructure, just mentioning as a source of potential ideas.


> Thanks.

--
James Richardson

Re: [PATCH 1/2] gnu: Add and use gettext-minimal.

[Ludovic Courtès]

Alex Kost  skribis:

> * gnu/packages/gettext.scm (gnu-gettext): Rename to...
> (gettext-minimal): ... this.  Adjust synopsis and description.
> (gnu-gettext): Inherit from it.
> (po4a): Use 'gettext-minimal' instead of 'gnu-gettext'.
> * gnu/packages/acl.scm: Likewise.
> * gnu/packages/admin.scm: Likewise.
> * gnu/packages/apl.scm: Likewise.
> * gnu/packages/attr.scm: Likewise.
> * gnu/packages/audio.scm: Likewise.
> * gnu/packages/base.scm: Likewise.
> * gnu/packages/cdrom.scm: Likewise.
> * gnu/packages/commencement.scm: Likewise.
> * gnu/packages/crypto.scm: Likewise.
> * gnu/packages/databases.scm: Likewise.
> * gnu/packages/disk.scm: Likewise.
> * gnu/packages/documentation.scm: Likewise.
> * gnu/packages/education.scm: Likewise.
> * gnu/packages/engineering.scm: Likewise.
> * gnu/packages/enlightenment.scm: Likewise.
> * gnu/packages/fcitx.scm: Likewise.
> * gnu/packages/fontutils.scm: Likewise.
> * gnu/packages/freedesktop.scm: Likewise.
> * gnu/packages/games.scm: Likewise.
> * gnu/packages/gkrellm.scm: Likewise.
> * gnu/packages/glib.scm: Likewise.
> * gnu/packages/gnome.scm: Likewise.
> * gnu/packages/grub.scm: Likewise.
> * gnu/packages/gtk.scm: Likewise.
> * gnu/packages/guile.scm: Likewise.
> * gnu/packages/ibus.scm: Likewise.
> * gnu/packages/irc.scm: Likewise.
> * gnu/packages/iso-codes.scm: Likewise.
> * gnu/packages/kde-frameworks.scm: Likewise.
> * gnu/packages/kodi.scm: Likewise.
> * gnu/packages/linux.scm: Likewise.
> * gnu/packages/man.scm: Likewise.
> * gnu/packages/maths.scm: Likewise.
> * gnu/packages/mono.scm: Likewise.
> * gnu/packages/mp3.scm: Likewise.
> * gnu/packages/music.scm: Likewise.
> * gnu/packages/nano.scm: Likewise.
> * gnu/packages/networking.scm: Likewise.
> * gnu/packages/package-management.scm: Likewise.
> * gnu/packages/pdf.scm: Likewise.
> * gnu/packages/sawfish.scm: Likewise.
> * gnu/packages/statistics.scm: Likewise.
> * gnu/packages/terminals.scm: Likewise.
> * gnu/packages/version-control.scm: Likewise.
> * gnu/packages/vpn.scm: Likewise.
> * gnu/packages/w3m.scm: Likewise.
> * gnu/packages/webkit.scm: Likewise.
> * gnu/packages/wicd.scm: Likewise.
> * gnu/packages/wine.scm: Likewise.
> * gnu/packages/xdisorg.scm: Likewise.
> * gnu/packages/xorg.scm: Likewise.

LGTM!

Please make sure that the basics work on ‘core-updates’ with this patch,
such as:

  ./pre-inst-env guix build \
 -e '(@@ (gnu packages commencement) gettext-boot0)'

Thanks!

Ludo’.

Re: [PATCH] gnu: Add gnome-calculator.

[Ludovic Courtès]

Kei Kebreau  skribis:

> I've already pushed it! I guess my last message only went back to
> Leo. Thank you and Leo for your help. :-)

Well, thanks for reviewing and pushing it then!  Now you’re all set!
:-)

And thanks rennes for this patch.

Ludo’.

[PATCH 0/4] Add gobby 0.4

[Andy Wingo]

These patches add a collaborative editor that we use at work.

Andy Wingo (4):
  gnu: Add libxml++-2.
  gnu: Add libnet6.
  gnu: Add obby.
  gnu: Add obby.

 gnu/local.mk   |   1 +
 gnu/packages/gnome.scm |  14 ++
 gnu/packages/gobby.scm | 132 +
 3 files changed, 147 insertions(+)
 create mode 100644 gnu/packages/gobby.scm

-- 
2.9.3

[PATCH 1/4] gnu: Add libxml++-2.

[Andy Wingo]

* gnu/packages/gnome.scm (libxml++-2): New variable.
---
 gnu/packages/gnome.scm | 14 ++
 1 file changed, 14 insertions(+)

diff --git a/gnu/packages/gnome.scm b/gnu/packages/gnome.scm
index e47e435..e33f744 100644
--- a/gnu/packages/gnome.scm
+++ b/gnu/packages/gnome.scm
@@ -4531,6 +4531,20 @@ the available networks and allows users to easily switch 
between them.")
 libxml2.")
 (license license:lgpl2.1+)))
 
+(define-public libxml++-2
+  (package
+(inherit libxml++)
+(name "libxml++")
+(version "2.40.1")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "mirror://gnome/sources/" name "/"
+  (version-major+minor version) "/"
+  name "-" version ".tar.xz"))
+  (sha256
+   (base32
+"1sb3akryklvh2v6m6dihdnbpf1lkx441v972q9hlz1sq6bfspm2a"))
+
 (define-public gdm
   (package
 (name "gdm")
-- 
2.9.3

[PATCH 3/4] gnu: Add obby.

[Andy Wingo]

* gnu/packages/gobby.scm (obby): New variable.
---
 gnu/packages/gobby.scm | 29 +
 1 file changed, 29 insertions(+)

diff --git a/gnu/packages/gobby.scm b/gnu/packages/gobby.scm
index 7e90453..eac088a 100644
--- a/gnu/packages/gobby.scm
+++ b/gnu/packages/gobby.scm
@@ -67,3 +67,32 @@
  "net6 is a library which eases the development of network-based
 applications as it provides a TCP protocol abstraction for C++.")
 (license license:lgpl2.1)))
+
+(define-public obby
+  (package
+(name "obby")
+(version "0.4.8")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "http://releases.0x539.de/obby/obby-;
+  version ".tar.gz"))
+  (file-name (string-append name "-" version ".tar.gz"))
+  (sha256
+   (base32
+"0rwvp0kzsb8y6mq73rzb8yk4kvsrz64i2zf4lfqs3kh0x2k7n7bx"
+(build-system gnu-build-system)
+(native-inputs
+ `(("pkg-config" ,pkg-config)))
+(inputs
+ `(("libsigc++" ,libsigc++)
+   ("gnutls" ,gnutls)
+   ("libnet6" ,libnet6)))
+(arguments
+ ;; Required by libsigc++.
+ `(#:configure-flags '("CXXFLAGS=-std=c++11")))
+(home-page "https://gobby.github.io/;)
+(synopsis "Library for building collaborative editors")
+(description
+ "libobby is a library which provides synced document buffers. It supports
+multiple documents in one session.")
+(license license:gpl2+)))
-- 
2.9.3

[PATCH 2/4] gnu: Add libnet6.

[Andy Wingo]

* gnu/packages/gobby.scm: New file.
* gnu/local.mk (GNU_SYSTEM_MODULES): Add new file.
---
 gnu/local.mk   |  1 +
 gnu/packages/gobby.scm | 69 ++
 2 files changed, 70 insertions(+)
 create mode 100644 gnu/packages/gobby.scm

diff --git a/gnu/local.mk b/gnu/local.mk
index c2cdf8b..a690b6e 100644
--- a/gnu/local.mk
+++ b/gnu/local.mk
@@ -156,6 +156,7 @@ GNU_SYSTEM_MODULES =\
   %D%/packages/gnustep.scm \
   %D%/packages/gnuzilla.scm\
   %D%/packages/gnu-pw-mgr.scm  \
+  %D%/packages/gobby.scm   \
   %D%/packages/golang.scm  \
   %D%/packages/gperf.scm   \
   %D%/packages/gprolog.scm \
diff --git a/gnu/packages/gobby.scm b/gnu/packages/gobby.scm
new file mode 100644
index 000..7e90453
--- /dev/null
+++ b/gnu/packages/gobby.scm
@@ -0,0 +1,69 @@
+;;; GNU Guix --- Functional package management for GNU
+;;; Copyright © 2016 Andy Wingo 
+;;;
+;;; This file is part of GNU Guix.
+;;;
+;;; GNU Guix is free software; you can redistribute it and/or modify it
+;;; under the terms of the GNU General Public License as published by
+;;; the Free Software Foundation; either version 3 of the License, or (at
+;;; your option) any later version.
+;;;
+;;; GNU Guix is distributed in the hope that it will be useful, but
+;;; WITHOUT ANY WARRANTY; without even the implied warranty of
+;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+;;; GNU General Public License for more details.
+;;;
+;;; You should have received a copy of the GNU General Public License
+;;; along with GNU Guix.  If not, see .
+
+(define-module (gnu packages gobby)
+  #:use-module (guix packages)
+  #:use-module (guix download)
+  #:use-module (guix utils)
+  #:use-module ((guix licenses) #:prefix license:)
+  #:use-module (guix build-system gnu)
+  #:use-module (gnu packages)
+  #:use-module (gnu packages glib)
+  #:use-module (gnu packages gnome)
+  #:use-module (gnu packages gtk)
+  #:use-module (gnu packages pkg-config)
+  #:use-module (gnu packages tls)
+  #:use-module (gnu packages xml))
+
+(define-public libnet6
+  (package
+(name "libnet6")
+(version "1.3.14")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "http://releases.0x539.de/net6/net6-;
+  version ".tar.gz"))
+  (sha256
+   (base32
+"088yqq60wjx3jqjlhl12893p15gl9asjpavrbhh590qzpqndhp8m"
+(build-system gnu-build-system)
+(native-inputs
+ `(("pkg-config" ,pkg-config)))
+(arguments
+ `(#:configure-flags
+   '("CXXFLAGS=-std=c++11") ; required by libsigc++
+   #:phases
+   (modify-phases %standard-phases
+ (add-before 'configure 'update-gnutls-api
+   (lambda _
+ (substitute* "src/encrypt.cpp"
+   ;; The GnuTLS API to set authentication and other parameters
+   ;; and priorities changed in 3.4; update to allow ANON_DH via
+   ;; the new API.
+   (("gnutls_kx_set_priority\\(session, kx_prio\\)")
+(string-append "gnutls_priority_set_direct"
+   "(session, \"NORMAL:+ANON-DH\", NULL)"
+(inputs
+ `(("libsigc++" ,libsigc++)
+   ("gnutls" ,gnutls)))
+(home-page "https://gobby.github.io/;)
+(synopsis "Network access framework for IPv4/IPv6")
+(description
+ "net6 is a library which eases the development of network-based
+applications as it provides a TCP protocol abstraction for C++.")
+(license license:lgpl2.1)))
-- 
2.9.3

Re: remove patchworks.

[Ludovic Courtès]

ng0  skribis:

> We have 22 pages of "open" patches. But they are not really open. I
> believe almost no one uses patchworks. I did for some time, and so did
> Ludovic and probably 3 more people but that's it.

Yeah, I agree, this was experimental, and the experiment wasn’t quite
successful.  We don’t need to formally “remove” the instance though; we
can eventually email the Sourceware folks asking them to do so, but no
rush IMO.

Thanks,
Ludo’.

[PATCH 4/4] gnu: Add obby.

[Andy Wingo]

* gnu/packages/gobby.scm (gobby): New variable.
---
 gnu/packages/gobby.scm | 34 ++
 1 file changed, 34 insertions(+)

diff --git a/gnu/packages/gobby.scm b/gnu/packages/gobby.scm
index eac088a..c14e40e 100644
--- a/gnu/packages/gobby.scm
+++ b/gnu/packages/gobby.scm
@@ -96,3 +96,37 @@ applications as it provides a TCP protocol abstraction for 
C++.")
  "libobby is a library which provides synced document buffers. It supports
 multiple documents in one session.")
 (license license:gpl2+)))
+
+(define-public gobby
+  (package
+(name "gobby")
+(version "0.4.13")
+(source (origin
+  (method url-fetch)
+  (uri (string-append "http://releases.0x539.de/gobby/gobby-;
+  version ".tar.gz"))
+  (file-name (string-append name "-" version ".tar.gz"))
+  (sha256
+   (base32
+"0w8q01lf6bcdz537b29m7rwlbc7k87b12vnpm1h6219ypvzqkgcc"
+(build-system gnu-build-system)
+(native-inputs
+ `(("pkg-config" ,pkg-config)
+   ("intltool" ,intltool)))
+(inputs
+ `(("libxml++-2" ,libxml++-2)
+   ("gnutls" ,gnutls)
+   ("gtkmm-2" ,gtkmm-2)
+   ("gtksourceview-2" ,gtksourceview-2)
+   ("libnet6" ,libnet6)
+   ("obby" ,obby)))
+(arguments
+ ;; Required by libsigc++.
+ `(#:configure-flags '("CXXFLAGS=-std=c++11")))
+(home-page "https://gobby.github.io/;)
+(synopsis "Collaborative editor")
+(description
+ "Gobby is a collaborative editor supporting multiple documents in one
+session and a multi-user chat.  It allows multiple users to edit the same
+document together over the internet in real-time.")
+(license license:gpl2+)))
-- 
2.9.3

Re: [PATCH 2/2] gnu: gettext: Install Emacs tools.

[Ludovic Courtès]

Alex Kost  skribis:

> Fixes .
> Reported by Ivan Vilata i Balaguer .
>
> * gnu/packages/gettext.scm (gnu-gettext)[native-inputs]: Add 'emacs'.
> [arguments]: Add 'add-emacs-autoloads' phase.

LGTM, thank you!

Ludo’.

Re: [PATCH 4/6] gnu: linux-libre-beagle-bone-black: Remove kernel variant.

[Tobias Geerinckx-Rice]

Hullo,

On 26/09/16 18:49, John Darrington wrote:
> I vote that we make the reason for a commit to be compulsory.

+1, s/commit/non-trivial commit/

I agree that it should be more strongly encouraged, if just to make
people aware that it's actually allowed — plenty aren't.

If no-one submits a proposal, I will.

> All other details optional.

-1, but I think *that* would require more than a mere vote anyway ;-)


I found the strict commit format annoying at first. Then I started
following the guix-commit list for fun and edutainment. Then I tried
following another commit list. I have seen the light.

That curated list of changes might ‘duplicate’ the diff in certain
trivial cases, but it saves more people far more time than it cost me to
write.


Kind regards,

T G-R



signature.asc
Description: OpenPGP digital signature

Re: [PATCH 2/3] gnu: Add python-pyxb.

[Ben Woodcroft]

On 27/09/16 04:32, Marius Bakke wrote:

Ben Woodcroft  writes:


On 24/09/16 13:15, Marius Bakke wrote:

Ben Woodcroft  writes:


Htslib is propagated because it
is included by the installed samtools and bcftools headers; at least one
of pysams dependencies failed to build when it was a regular input.

Ah, I see. Can this be avoided with patchelf?

I don't think so. The htslib headers need to be propagated for anything
that uses header files from pysam, since they contain contain code such
as "include ". Perhaps you can update the comment to make
it more clear?

Patchelf is probably what made that work when htslib was bundled.

Ah, ok I see, so the problem is that python-plastid requires the .h
files during build. Given that this is the only package that has an
issue, and perhaps pysam doesn't intend to provide these .h files as
part of it's "interface", I think it might be best to simply put htslib
as an input to python-plastid.

I'm not sure, we would have to do this with every package that uses
the pysam cython interface.

Ah, you're right. I overlooked the cython interface.

I pushed the series propagating htslib with pysam, ending in 99caa6f.

Thanks for persevering.
ben

Re: Building Guix with Guile 2.1

[Ludovic Courtès]

taylanbayi...@gmail.com (Taylan Ulrich "Bayırlı/Kammer") skribis:

> I think we can keep the (compile 'dummy) hack.  That leaves two issues
> which may be solved in a cleaner manner than in this patch:
>
> - The (define foo (@@ (bar) foo)) parts.
> - Making %tty-gid public.  (The above didn't work for this one...)

Fixed in 3c185b24f593c982aeb33996324fa6878c6ed21b, thanks for reporting
it!

Ludo’.

[PATCH] gnu: New default Dovecot service postmaster_address

[Andy Wingo]

* gnu/services/mail.scm (dovecot-configuration): Change default for
  postmaster-address, as dovecot is now requiring a non-empty value and
  will fail to start up otherwise.
* doc/guix.texi (Mail Services): Update.
---
 doc/guix.texi | 4 ++--
 gnu/services/mail.scm | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/doc/guix.texi b/doc/guix.texi
index 808fbdc..c159e12 100644
--- a/doc/guix.texi
+++ b/doc/guix.texi
@@ -9964,8 +9964,8 @@ Defaults to @samp{""}.
 
 @deftypevr {@code{dovecot-configuration} parameter} string postmaster-address
 Address to use when sending rejection mails.
-Default is postmaster@@.  %d expands to recipient domain.
-Defaults to @samp{""}.
+%d expands to recipient domain.
+Defaults to @samp{"postmaster@@%d"}.
 @end deftypevr
 
 @deftypevr {@code{dovecot-configuration} parameter} string hostname
diff --git a/gnu/services/mail.scm b/gnu/services/mail.scm
index d94532c..cb0f119 100644
--- a/gnu/services/mail.scm
+++ b/gnu/services/mail.scm
@@ -1352,7 +1352,7 @@ regeneration entirely.")
"SSL crypto device to use, for valid values run \"openssl engine\".")
 
   (postmaster-address
-   (string "")
+   (string "postmaster@%d")
"Address to use when sending rejection mails.
 Default is postmaster@@.  %d expands to recipient domain.")
 
-- 
2.9.3

Re: [PATCH v2 0/4] Add GCC cross compiler for arm-none-eabi.

[Ludovic Courtès]

Hi Theodoros,

Theodoros Foradis  skribis:

> The original patch series was working correctly and producing working 
> binaries. Some
> flags (that I had been using with 6.2.0) are missing from that version of GCC 
> 4.9,
> so I added 6.2.0 as an extra option. I have tested it to produce working 
> binaries.
>
> Here are some modifications to Ricardo's patches for the arm-none-eabi
> bare metal cross compiler. The following changes have been made:
>
> - I have modified xbinutils to use binutils 2.25.1 from cross-base, as it 
> compiles
> correctly with it. The version from the svn commit that was used by Ricardo 
> is compiling
> correct binaries as well. Thus, if it is deemed appropriate, the source for 
> xbinutils can
> be swapped for the previous one, with (seemingly) no difference.
>
> - The xgcc of the original, was failing to find the headers that newlib 
> provided.
> I have set the native-cross-paths as a workaround. Not sure if there is a 
> better
> alternative, or if the failure was my mistake.
>
> - A package for cross GCC 6.2.0 is added, with appropriate patches for 
> multilib
> support.
>
> - Newlib-arm-none-eabi and newlib-nano-arm-none-eabi have been changed to
> procedures, taking an xgcc as argument, so as to facilitate building with
> either version of gcc.
>
> - An arm-none-eabi-toolchain procedure is declared, to create toolchain 
> packages
> for both gcc and newlib version. The four toolchain variables follow. Not sure
> if it's a mistake to include "nano" in the toolchain version.

This all sounds reasonable to me.  Ricardo was interested in using this
toolchain for one specific purpose, so maybe we’ll want to check that it
also works here.  Ricardo: could you comment?

FWIW I had commented on Ricardo’s patch series here:

  https://lists.gnu.org/archive/html/guix-devel/2016-09/msg01379.html

> This is the first patch that I send in guix-devel, so please bear with me.

Welcome!  Sorry that it already took a bit long, but hopefully we’ll
converge real soon!  :-)

Thanks,
Ludo’.

Re: Call for volunteer(s) for Guix "security" web page

[Ludovic Courtès]

Hi Leo,

Thanks a lot both for sending the call and replying to it!  :-)

> From 30699a5a8de5ac09c6fbba93be6b88a1d77bc039 Mon Sep 17 00:00:00 2001
> From: Leo Famulari 
> Date: Sun, 25 Sep 2016 18:43:28 -0400
> Subject: [PATCH] www: security: New page.
>
> * website/www/security.scm: New file.
> * website/www.scm (%web-pages): Add security-page.
> * website/www/shared.scm (html-page-links): Add "Security".

[...]

> +   (h2 "How to report security issues")
> +   (p "To report sensitive security issues in Guix itself or the 
> packages it "
> +  "provides, you can write to the private mailing list "
> +  (a (@ (href 
> "https://lists.gnu.org/mailman/listinfo/guix-security;))
> + ("guix-secur...@gnu.org"))
> + ".  This list is monitored by a small team of Guix "
> + "developers.")
> +   (h2 "Release signatures")
> +   (p "Releases of Guix and GuixSD are signed using the OpenPGP "
> +  "key with the fingerprint "
> +  "3CE4 6455 8A84 FDC6 9DB4  0CFB 090B 1199 3D9A EBB5.  "
> +  "This key can be obtained from XXX.")

Maybe link to

or copy/paste the text?  Though we should give a ‘gpg --recv-keys’
command that uses the full fingerprint instead of just the 64-bit ID
(which is still too small, some say.)

> +   (h2 "Security updates")
> +   (p "When security vulnerabilities are found in Guix or the "
> +  "packages provided by Guix, we will provide "
> +  (a (@ (href ,(base-url 
> "manual/html_node/Security-Updates.html")))
> + "security updates")
> +  " quickly and with minimal disruption for users.")

Maybe also that Guix is a “rolling release”, so there’s currently no
separate security-fix branch and all critical fixes go to master?

I guess you can already commit that!

I wonder if it would make sense to add a note on reproducible builds,
‘guix challenge’ and all that; later maybe!

Note that you’ll then need to commit the resulting HTML to CVS(!) to
that the update pages show up, as per the instructions available on the
Savannah project page.  If you’re unsure or anything, I can do that.

Thank you!

Ludo’.

Re: [PATCH] gnu: service: Add git-service.

[ng0]

So almost one month passed now.
To continue testing this, I need help on this first.

Thanks.

ng0  writes:

> I tried to address most of what you've written.
>
> While I was correcting the documentation I decided to add more options,
> now it doesn't work anymore, probably because of the ifs I added.
>
> Andy Wingo  writes:
>
>> On Tue 30 Aug 2016 13:45, ng0  writes:
>>
I also think that "path" might
 not be the right word, which in GNU manuals is only used for search
 paths.  See the "GNU Manuals" section of standards.texi for more.
 Anyway I suggest #:base-directory.  Make sure the port is an integer and
 not a string.
>>>
>>> See 'man git daemon'.
>>
>> I ran this and it did not work -- first showed me a page for git then
>> for daemon.  I believe you want "man git-daemon"?
>>
>>> The switch is called --base-path.  Looking at the openrc conf.d/git or
>>> what the config file was called again, they stick to this name too.
>>> It would just cause confusion if we go ahead and call it differently.
>>> Upstream should be fixed, but I'm not going there.  If you think we
>>> should break expectations, I can rename it.
>>
>> "Fixing" upstream is out of our remit :)  All I can ask is that we do
>> not introduce new uses of the word "path".
>>
> +Return a service to run the @uref{https://git-scm.com, git} daemon 
> version control
> +daemon.

 Extra "daemon" here.  Probably needs a sentence on what running the
 daemon will do (namely, expose local repositories for remote access).

 What about authentication?  Is this purely anonymous?
>>>
>>> Exactly, authentication is handled via other daemons, for example ssh or
>>> gitolite. git daemon supports no authentication and is read-only, as far
>>> as I know. At the servers I use and setup, I pull via
>>> git://,http://,https:// and push via ssh.
>>> Its selfdescription is:
>>> git-daemon - A really simple server for Git repositories.
>>
>> This needs to be documented in the manual, is what I was getting at :)
>> Mention that this is for anonymous read-only access please.
>
> read-only was wrong, anonymous write-access for all the world can be set
> up but it is not default.
>
>>
Use "file name" instead of path in general.
>>>
>>> Why?
>>
>> It is because it is standard in the GNU project.  I mentioned this
>> before.  See "info standards" and go to "GNU manuals".
>>
> +Furthermore it takes the parameter @var{port} which defaults to 9418.
> +Run @command{man git daemon} for information about the options.

 This man command does not work.
>>>
>>> Works for me. As far as I know man pages were merged into git package
>>> recently. When I run this on debian with guix, 'man git daemon' works
>>> too.
>>
>> It does not work for me on NixOS with Guix.  Maybe I am out of date
>> though.
>>
> +(define %git-accounts
> +  ;; User account and groups for git-daemon.
> +  ;; We can give it git-shell for now, otherwise we can switch to 
> /bin/sh.

 What does this comment mean?  Why would we switch?
>>>
>>> I am not sure about the limitations of git-shell compared to
>>> /bin/sh. If this turns out to be a mistake, it can be corrected. The
>>> only thing I know about git-shell is that it allows no logins.
>>
>> If you do not want a login then probably what you want is
>> #~(string-append #$shadow "/sbin/nologin").
>>
>> Andy
>
> From d1d7eb59ca53833098cea2d6eddaa59f1494b579 Mon Sep 17 00:00:00 2001
> From: ng0 
> Date: Fri, 8 Jul 2016 15:42:55 +
> Subject: [PATCH] gnu: services: Add git-service.
>
> * gnu/services/version-control.scm: New file, create it.
> (git-service): New Procedures.
> (git-service-type): New variable.
> * doc/guix.texi: Add documentation.
> ---
>  doc/guix.texi|  37 
>  gnu/local.mk |   1 +
>  gnu/services/version-control.scm | 196 
> +++
>  3 files changed, 234 insertions(+)
>  create mode 100644 gnu/services/version-control.scm
>
> diff --git a/doc/guix.texi b/doc/guix.texi
> index b22cf4a..78d7ee1 100644
> --- a/doc/guix.texi
> +++ b/doc/guix.texi
> @@ -7494,6 +7494,7 @@ declaration.
>  * Database Services::   SQL databases.
>  * Mail Services::   IMAP, POP3, SMTP, and all that.
>  * Web Services::Web servers.
> +* Version Control:: Git and others.
>  * Various Services::Other services.
>  @end menu
>  
> @@ -9910,6 +9911,42 @@ directories are created when the service is activated.
>  
>  @end deffn
>  
> +@node Version Control
> +@subsubsection Version Control
> +
> +The @code{(gnu services version-control)} module provides the following 
> services:
> +
> +@deffn {Scheme Procedure} git-service [#:git @var{git}] @
> +   [#:base-directory "/var/git/repositories"] @
> +   [#:user-directory? #f ""] [#:port 9418] @

Re: [PATCH] gnu: femtolisp: Remove support for mips and armhf.

[ng0]

Ludovic Courtès  writes:

> ng0  skribis:
>
>> Ludovic Courtès  writes:
>>
>>> Andreas Enge  skribis:
>>>
 femtolisp fails on mips and arm:
http://hydra.gnu.org:3000/build/1472987/nixlog/2/tail-reload
 with the message
 In file included from llt/llt.h:6:0,
  from string.c:16:
 llt/utils.h:27:4: error: #error "unknown architecture"
  #  error "unknown architecture"

 Should it be disabled there?
>>>
>>> Definitely.  It would be worth checking in that file the set of
>>> supported architectures, and putting them in ‘supported-platforms’.
>>
>> I'm not exactly sure what's supported, I found win32, osx, freebsd,
>> openbsd, and that's it.
>
> Sorry, I really meant supported architectures, which is what utils.h
> seems to be about (although I don’t have the source here to check).
> Could you check the #ifdefs in there?

#if defined( __amd64__ ) || defined( _M_AMD64 )
#  define ARCH_X86_64
#  define __CPU__ 686
#elif defined( _M_IX86 )//msvs, intel, digital mars, watcom
#  if ! defined( __386__ )
#error "unsupported target: 16-bit x86"
#  endif
#  define ARCH_X86
#  define __CPU__  ( _M_IX86 + 86 )
#elif defined( __i686__ )//gnu c
#  define ARCH_X86
#  define __CPU__ 686
#elif defined( __i586__ )//gnu c
#  define ARCH_X86
#  define __CPU__ 586
#elif defined( __i486__ )//gnu c
#  define ARCH_X86
#  define __CPU__ 486
#elif defined( __i386__ )//gnu c
#  define ARCH_X86
#  define __CPU__ 386
#else
#  error "unknown architecture"
#endif


>> +  ;; armhf and mips64el fail to build, it has been reported upstream:
>> +  ;; https://github.com/JeffBezanson/femtolisp/issues/25
>> +  (supported-systems
>> +   (and
>> +(delete "armhf-linux" %supported-systems)
>> +(delete "mips64el-linux" %supported-systems)
>> +#t))
>
> ‘supported-systems’ must be a list of strings, so:
>
>   (fold delete %supported-systems
> '("armhf-linux" "mips64el-linux"))
>
> Thanks,
> Ludo’.
>

From 32c4d64ada9ab2573a00385b15b2b0a03a6fa44f Mon Sep 17 00:00:00 2001
From: ng0 
Date: Thu, 22 Sep 2016 09:08:03 +
Subject: [PATCH] gnu: femtolisp: Remove support for mips and armhf.

* gnu/packages/lisp.scm (femtolisp)[supported-systems]: New field,
remove support for mips64el and armhf.
---
 gnu/packages/lisp.scm | 10 --
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/gnu/packages/lisp.scm b/gnu/packages/lisp.scm
index d1180a7..439433a 100644
--- a/gnu/packages/lisp.scm
+++ b/gnu/packages/lisp.scm
@@ -3,7 +3,7 @@
 ;;; Copyright © 2015 Taylan Ulrich Bayırlı/Kammer 
 ;;; Copyright © 2015 Mark H Weaver 
 ;;; Copyright © 2016 Federico Beffa 
-;;; Copyright © 2016 ng0 
+;;; Copyright © 2016 ng0 
 ;;; Copyright © 2016 Andy Patterson 
 ;;;
 ;;; This file is part of GNU Guix.
@@ -44,7 +44,8 @@
   #:use-module (gnu packages ed)
   #:use-module (gnu packages m4)
   #:use-module (gnu packages version-control)
-  #:use-module (ice-9 match))
+  #:use-module (ice-9 match)
+  #:use-module (srfi srfi-1))
 
 (define-public gcl
   (package
@@ -501,6 +502,11 @@ the InterLisp Standard.")
 (sha256
  (base32
   "04rnwllxnl86zw8c6pwxznn49bvkvh0f1lfliy085vjzvlq3rgja"
+  ;; See "utils.h" for supported systems. Upstream bug:
+  ;; https://github.com/JeffBezanson/femtolisp/issues/25
+  (supported-systems
+   (fold delete %supported-systems
+ '("armhf-linux" "mips64el-linux")))
   (build-system gnu-build-system)
   (arguments
`(#:make-flags '("CC=gcc" "release")
-- 
2.10.0


-- 
  ng0


signature.asc
Description: PGP signature

inox work in progress file

[ng0]

Here is the file for anyone who wants to take over or improve this a
bit. I have not continued to work on the patches to store method I
started, it's all work in progress.



inox.scm
Description: Binary data
-- 
  ng0


signature.asc
Description: PGP signature