Hello Jack,
That did the trick! I rebuilt llvm-gc42 and after that the fink update-all went
through without any problems. I still don't understand where the problem really
came from as I don't know where to choose to install a deb from either the
10.10 or 10.9 bindist. But it works...
Cheers,
Jacques,
Try rebuilding the llvm-gcc42 package on your 10.10 and see if that
eliminates the build problems with xfig? I didn't revision bump llvm-gcc42
when I patched it for 10.10 as I assumed any bindist dist installs would
come off a 10.10 fink build bot. My guess is that you have installe
Hi Jack,
Xcode's command line tools were already installed and the checks you asked me
to make give the correct results. That is:
$ which cc
/usr/bin/cc
cc -v
Apple LLVM version 6.0 (clang-600.0.54) (based on LLVM 3.5svn)
Target: x86_64-apple-darwin14.0.0
Thread model: posix
Cheers,
Jacques
Jacques,
I strongly suspect you have legacy compilers installed on your system
rather than those from Xcode 6's Command Line Tools. You would have...
% which cc
/usr/bin/cc
% cc -v
Apple LLVM version 6.0 (clang-600.0.54) (based on LLVM 3.5svn)
Target: x86_64-apple-darwin14.0.0
Thread mo
Alexander,
I find that using the change
-- java-oct.info.orig 2014-10-23 14:52:30.0 -0400
+++ java-oct.info 2014-10-23 15:02:08.0 -0400
@@ -40,6 +40,8 @@
CompileScript: <<
#!/bin/sh -ev
+export JAVA_HOME=/System/Library/Frameworks/JavaVM.framework/Home
+
pkgsrc
Alexander,
Building specfun-oct364-1.1.0-6 exposed one last thinko in the
octave-3.6.4 and octave-3.8.2 packaging on 10.0.
In file included from ellipj.cc:40:
In file included from /sw/include/octave-3.6.4/octave/oct.h:33:
In file included from /sw/include/octave-3.6.4/octave/Matrix.h:30:
That’s what was _supposed to happen_, but I missed pushing my changes to github
before tagging the release so nobody else got them.
version 0.3.2.1 actually does the right thing. Sorry about that.
On Oct 23, 2014, at 9:56 AM, Jack Howarth wrote:
> Alexander,
> Of course, the installed /sw
Alexander,
Of course, the installed /sw/bin/oct-cxx also needs to be changed from
case $osversion in
13)
COMPILER=/sw/var/lib/fink/path-prefix-libcxx/c++ ;;
to
case $osversion in
13|14)
COMPILER=/sw/var/lib/fink/path-prefix-libcxx/c++ ;;
Alexander,
This problem is due to the /sw/bin/oct-cc installed by the
current fink-octave-scripts
package having...
case $osversion in
1[123])
COMPILER=/sw/var/lib/fink/path-prefix-clang/cc ;;
instead of
case $osversion in
1[1234])
COMPILE
On 10/23/14, 7:32 AM, Jacques Bloch wrote:
> Error in fink update-all
>
> /checking for a BSD-compatible install... /sw/bin/ginstall -c
> checking whether build environment is sane... yes
> checking for a thread-safe mkdir -p... /sw/bin/gmkdir -p
> checking for gawk... gawk
> checking whether make
On 10/23/14, 5:50 AM, Jacques Bloch wrote:
> I found a different solution, namely install "Java for OS X 2014-001" from
> http://support.apple.com/kb/DL1572, which somehow install a Mac version of
> Java 6. The "fink update-all" seems to proceed now.
>
> For some reason, which I do not understan
I found a different solution, namely install "Java for OS X 2014-001" from
http://support.apple.com/kb/DL1572, which somehow install a Mac version of Java
6. The "fink update-all" seems to proceed now.
For some reason, which I do not understand, the main java version on my system
is still the
On 10/23/14, 2:00 AM, Jacques Bloch wrote:
> When making a "fink update-all" I get:
>
> /Can't resolve dependency "system-java16-dev" for package
> "octave-3.8.2-1" (no matching packages/versions found)/
> /
> /
> I tried installing the latest java sdk (1.8.0_25) but still get this
> message, proba
Hi John,
That indeed seems to have solved the problem!
Thanks for you help,
Jacques
> On 15 Sep 2014, at 21:49, John Lillibridge - NOAA Federal
> wrote:
>
> Hi Jacques,
>
> I ran into similar problem with a few .info files when trying to resolve
> dependencies under Yosemite. I was able
After downloading the latest fink version from halfdome, running inject.pl and
fink selfupdate (see the thread Yosemite selfupdate) the fink update-all
command also runs fine now. The problem to compile gdb seems to have
evaporated, probably due to the fact that it now installs
gdb_7.8-2_darwi
On 8/22/14, 6:09 AM, Jacques Bloch wrote:
> I believe that gdb should compile when changing
> by in
> /sw/src/fink.build/gdb-7.8-1/gdb-7.8/gdb/darwin-nat.c
> (courtesy of https://trac.macports.org/ticket/43973)
>
> HOWEVER, when I change this line in the source file and then run "fink update
>
On 8/22/14, 6:07 AM, Jacques Bloch wrote:
> I am the second public Yosemite release which came out last night. The
> perl version is:
>
> /$ perl -v
>
> This is perl 5, version 18, subversion 2 (v5.18.2) built for
> darwin-thread-multi-2level/
> /
> /
> /
> /
>
>
>>> WARNING! This version of Perl
I am the second public Yosemite release which came out last night. The perl
version is:
$ perl -v
This is perl 5, version 18, subversion 2 (v5.18.2) built for
darwin-thread-multi-2level
> On 22 Aug 2014, at 14:55, Alexander Hansen
> wrote:
>
> On 8/22/14, 5:07 AM, Jacques Bloch wrote:
>>
On 8/22/14, 5:07 AM, Jacques Bloch wrote:
> I know got fink to work on Yosemite but, as might have been expected, I
> am getting some errors here and there. The fink selfupdate gives the
> following problem related to perl:
>
> /$ fink selfupdate
> /usr/bin/rsync -az -q
> rsync://distfiles.master.f
I think I somehow got things tho work thanks to Alexander.
Here is the order of things I did:
- changed the Distribution: in /sw/etc/fink.conf back from 10.10 to 10.9,
- ./inject.pl from the halfdome beta distribution
- "fink reinstall fink" (I had to run this twice before getting the next
comm
Hmm…maybe try setting the Distribution: to 10.9.
On Aug 21, 2014, at 1:33 PM, Jacques Bloch wrote:
> /sw/etc/fink.conf has "Distribution: 10.10"
>
> $ ls -l /sw/fink
> total 48
> lrwxr-xr-x 1 root wheel 4 Aug 21 00:58 10.10 -> 10.9
> drwxr-xr-x10 root wheel 340 Aug 2
/sw/etc/fink.conf has "Distribution: 10.10"
$ ls -l /sw/fink
total 48
lrwxr-xr-x 1 root wheel 4 Aug 21 00:58 10.10 -> 10.9
drwxr-xr-x10 root wheel 340 Aug 21 14:07 10.7
lrwxr-xr-x 1 root wheel 4 Nov 11 2013 10.9 -> 10.7
-rw-r--r--1 root wheel
On 8/21/14, 10:20 AM, Jacques Bloch wrote:
> No luck...
>
> /$ fink reinstall fink
> Scanning package description files..
> Information about 10010 packages read in 23 seconds.
> Failed: no package found for specification 'fink'!
> /
>
>
>
Sorry about that. I wasn't 100% certain whether "
No luck...
$ fink reinstall fink
Scanning package description files..
Information about 10010 packages read in 23 seconds.
Failed: no package found for specification 'fink'!
> On 21 Aug 2014, at 18:57, Alexander Hansen
> wrote:
>
> On 8/21/14, 9:37 AM, Jacques Bloch wrote:
>> Ah well
On 8/21/14, 9:37 AM, Jacques Bloch wrote:
> Ah well... that may have something to do with it. Here is the outcome:
>
> /$ fink dumpinfo -fallversions fink
> Scanning package description files..
> Information about 10010 packages read in 26 seconds.
> Failed: no package found for specificati
Ah well... that may have something to do with it. Here is the outcome:
$ fink dumpinfo -fallversions fink
Scanning package description files..
Information about 10010 packages read in 26 seconds.
Failed: no package found for specification 'fink'!
> On 21 Aug 2014, at 17:00, Alexander Ha
On 8/21/14, 12:09 AM, Jacques Bloch wrote:
>>
>> inject.pl is creating a fink.info file for fink with a version
>> fink-0.37.99.git-20140820.2302 (the number is based on the time at which you
>> run the script) in /sw/etc/fink.conf
>>
>> Make sure your /sw/etc.fink.conf has local/injected in its
>
> inject.pl is creating a fink.info file for fink with a version
> fink-0.37.99.git-20140820.2302 (the number is based on the time at which you
> run the script) in /sw/etc/fink.conf
>
> Make sure your /sw/etc.fink.conf has local/injected in its Trees: line.
>
> --
> Alexander Hansen, Ph.D.
On 8/20/14, 4:11 PM, Jacques Bloch wrote:
> After creating the symboling link from /sw/fink/10.10 to /sw/fink/10.9
> the ./inject.pl runs with less error messages, but is still not
> successful. It finishes with:
> /
> /
> //sw/bin/fink install fink-0.37.99.git-20140820.2302
> Scanning package desc
After creating the symboling link from /sw/fink/10.10 to /sw/fink/10.9 the
./inject.pl runs with less error messages, but is still not successful. It
finishes with:
/sw/bin/fink install fink-0.37.99.git-20140820.2302
Scanning package description files..
Information about 10010 packages r
On 8/20/14, 6:27 AM, Jacques Bloch wrote:
> Hi Alexander,
>
> Not sure if I should send this to you or on some mailing list. I did as
> you suggested: downloaded the development branch at halfdome, modified
> the Engine.pm file and ran inject.pl with the following unsuccessful ouput:
>
> /> sudo ./
Hi Alexander,
Not sure if I should send this to you or on some mailing list. I did as you
suggested: downloaded the development branch at halfdome, modified the
Engine.pm file and ran inject.pl with the following unsuccessful ouput:
> sudo ./inject.pl
Creating fink tarball...
tar -cf /sw/src/fi
Thanks Daniel,
Hopefully this new version of openSSL works better. However, I have seen
pod2man issues before, and in the interest of making progress compiling
packages without being held up by the man pages I worked around the problem by
linking /usr/bin/pod2man5.16 -> pod2man. There is defini
On Fri, 8 Aug 2014 16:18:57 -0400, John Lillibridge - NOAA Federal
wrote:
Hi Alexander,
>
> Sorry I wasn't more explicit. The failure occurred in the
> openssl100-dev package that CVS depends upon. I had seen similar
> pod2man errors "Expected text after =item, not a number" and thought
> it w
Hi Alexander,
Sorry I wasn't more explicit. The failure occurred in the openssl100-dev
package that CVS depends upon. I had seen similar pod2man errors "Expected text
after =item, not a number" and thought it was a more generic problem. Perhaps a
newer version of Perl under Yosemite 10.10? In a
On 8/8/14, 9:50 AM, John Lillibridge - NOAA Federal wrote:
> Dear Alexander & Fink developers,
>
> I wanted to share my experience based on your advice below, to build Fink
> under Yosemite 10.10 (public beta) from
> https://github.com/fink/fink/tree/HalfDome";.
>
> After installing the XCode 6.0
in the bootstrap build in
the HalfDome branch.
All the best,
John Lillibridge
On Aug 8, 2014, at 7:37 AM, fink-devel-requ...@lists.sourceforge.net wrote:
> Date: Fri, 08 Aug 2014 04:37:21 -0700
> From: Alexander Hansen
> Subject: Re: [Fink-devel] Yosemite
> To: Jacques Bloc
On 8/8/14, 7:38 AM, Juan M Courcoul wrote:
> Given that the builds released to the Beta and Seed communities are different
> (abundantly highlighted in the NDA'd communications sent to the Seed
> participants), will that require additional work by Fink's dedicated
> developers? Besides, of cours
Given that the builds released to the Beta and Seed communities are different
(abundantly highlighted in the NDA'd communications sent to the Seed
participants), will that require additional work by Fink's dedicated
developers? Besides, of course, requiring two separate test benches and going
That was a small data set. :-) And our mailing lists aren't set up by
default to have the reply go to the list, and that can admittedly be
misleading.
Actually, I get a fair amount of direct email about packages that I
don't maintain or use because I happened to be involved in a mailing
list
Hi Alexander,
sorry to have upset you with my mail. Last time we wrote with each other (last
year) it seemed to me that you preferred being contacted by email.
Apologies,
Jacques
---
Dr. Jacques C.R. Bloch
Institute for Theoretical Physics
University of Regensburg
93040 Regensburg, Germany
tel:
On 8/8/14, 12:50 AM, Jacques Bloch wrote:
> Dear Alexander,
>
> I moved to Yosemite and can no longer use fink, as it does not support
> OSX 10.10. Following something that I found on a fink-forum I added the
> following lineIn the file /sw/lib/perl5/Fink/Engine.pm:
> $valid_upgrade = 1 if ($osver
42 matches
Mail list logo
