Re: [Fink-devel] numpy 1.12.0 and nose-py36
Hi Kurt, > On 19 Jan 2017, at 9:18 pm, Kurt Schwehr wrote: > > yaml-py is updated. If you want to take over matplotlib (or any of my other > packages), be my guest > thanks; I have now uploaded the matplotlib-2.0 update. If you wish, I can take it over as maintainer, and cycler as well. Cheers, Derek -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] numpy 1.12.0 and nose-py36
On 17 Jan 2017, at 10:16 pm, Kurt Schwehr wrote: > > done > Thanks, worked! Could you do the same for yaml-py36; adding the Type is all that’s required as well. This would be the last missing Python3.6 dependency for an astropy update I am preparing. I have also just submitted two package requirements for matplotlib-2.0.0 to the tracker, which you can take over as maintainer or leave me in, whatever you prefer. Cheers, Derek > > On Tuesday, January 17, 2017 7:03 AM, Derek Homeier > wrote: > > > HI, > > when installing the py36 version of the new numpy version (thanks for the > swift update!), > it refuses to build with tests enabled, as there is no py36 version of nose > available yet. > nose-py-1.3.7 builds out of the box on 10.11 and 10.12 by just adding 3.6 to > the Type > field, and numpy-py36 passed building in maintainer mode when testing this > locally, > so it would be great to add the 3.6 type_pkg for nose. -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] numpy 1.12.0 and nose-py36
HI, when installing the py36 version of the new numpy version (thanks for the swift update!), it refuses to build with tests enabled, as there is no py36 version of nose available yet. nose-py-1.3.7 builds out of the box on 10.11 and 10.12 by just adding 3.6 to the Type field, and numpy-py36 passed building in maintainer mode when testing this locally, so it would be great to add the 3.6 type_pkg for nose. Cheers, Derek -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] boost1.63-nopython fails validation
On 14 Jan 2017, at 3:14 pm, Hanspeter Niederstrasser wrote: > > This has been fixed by disabling the boost_filter library. If someone > requests it in the future (or we stop supporting 10.9 or 10.10), I'll > re-enable it. Tested and found working here on 10.11 and 10.12. Thanks, Derek -- Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] boost1.63-nopython fails validation
On 11 Jan 2017, at 12:28 pm, Hanspeter Niederstrasser wrote: > >> + LIBDIR2= >> + ./b2 -d2 'cxxflags=-std=c++11 -MD -I/sw/include' -j4 --without-python >> --prefix=/sw --libdir=/sw/opt/boost-1_63/lib >> --includedir=/sw/opt/boost-1_63/include/boost --compatibility_version=1.63.0 >> --current_version=1.63.0 --build-type=complete --layout=tagged >> variant=release threading=single,multi link=shared >> Performing configuration checks >> >>- 32-bit : no >>- 64-bit : yes >>- arm : no >>- mips1: no >>- power: no >>- sparc: no >>- x86 : yes >> >> Building the Boost C++ Libraries. >> >> >>- symlinks supported : yes >>- C++11 mutex : yes >>- lockfree boost::atomic_flag : yes >>- Boost.Config Feature Check: cxx11_auto_declarations : yes >>- Boost.Config Feature Check: cxx11_constexpr : yes >>- Boost.Config Feature Check: cxx11_defaulted_functions : yes >>- Boost.Config Feature Check: cxx11_final : yes >>- Boost.Config Feature Check: cxx11_hdr_mutex : yes >>- Boost.Config Feature Check: cxx11_hdr_tuple : yes >>- Boost.Config Feature Check: cxx11_lambdas : yes >>- Boost.Config Feature Check: cxx11_noexcept : yes >>- Boost.Config Feature Check: cxx11_nullptr : yes >>- Boost.Config Feature Check: cxx11_rvalue_references : yes >>- Boost.Config Feature Check: cxx11_template_aliases : yes >>- Boost.Config Feature Check: cxx11_thread_local : yes > > The only difference in output is that my build doesn't find > cxx11_thread_local. > > I wonder if it's a clang version issue (I have Xcode 7.3). Does the command > > /usr/bin/clang++ -dM -E -x c++ -std=c++14 /dev/null | grep thread > > output anything? > Nope, nothing. I think I have Xcode 8 on the El Capitan machine, but not the command line tools (which are not available for 10.11 iirc); but I cannot connect to it to check right now. Cheers, Derek -- Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] boost1.63-nopython fails validation
> On 11 Jan 2017, at 11:46 am, Hanspeter Niederstrasser > wrote: > >> building the package with ‘-m’ fails on a missing Shlibs entry (the other >> boost1.63 variants build fine): >> >> Validating .deb dir >> /scratch.noindex/fink.build/root-boost1.63-nopython-1.63.0-1... >> Error: package contains the shared library >> /sw/opt/boost-1_63/lib/libboost_fiber-mt.dylib >> but the corresponding install_name and compatibility_version >> %p/opt/boost-1_63/lib/libboost_fiber-mt-1_63.dylib 1.63.0 >> are not listed in the Shlibs field. See the packaging manual. > > Interestingly, my configure output on 10.11 says the boost_fiber library is > going to be built but doesn't actually get built. However, the bindist > machines did build it. > > Can you send the build log output from the command > > './bootstrap.sh --without-libraries=python ...' through the first > '...patience...' string just before the actual compiling starts? I got the same validation error on both 10.12 and 10.11 (both case-sensitive FS, if that should matter). I had gcc5/gcc5-compiler installed at the time of the build (just thinking of the C++11 references in README.md and below)... Cheers, Derek + USEPYTHON=--with-python + UNDEFINED= + case nopython in + UNDEFINED= + USEPYTHON=--without-python + export PYTHON= + PYTHON= + LIBDIR=--libdir=/sw/opt/boost-1_63/lib + export ICONV_PATH=/sw + ICONV_PATH=/sw + export ICU_ROOT=/sw + ICU_ROOT=/sw + export PYTHON= + PYTHON= + ./bootstrap.sh --without-libraries=python --prefix=/sw/opt/boost-1_63 --includedir=/sw/opt/boost-1_63/include/boost --with-icu=/sw Building Boost.Build engine with toolset darwin... tools/build/src/engine/bin.macosxx86_64/b2 Unicode/ICU support for Boost.Regex?... /sw Generating Boost.Build configuration in project-config.jam... Bootstrapping is done. To build, run: ./b2 To adjust configuration, edit 'project-config.jam'. Further information: - Command line help: ./b2 --help - Getting started guide: http://www.boost.org/more/getting_started/unix-variants.html - Boost.Build documentation: http://www.boost.org/build/doc/html/index.html + LIBDIR2= + ./b2 -d2 'cxxflags=-std=c++11 -MD -I/sw/include' -j4 --without-python --prefix=/sw --libdir=/sw/opt/boost-1_63/lib --includedir=/sw/opt/boost-1_63/include/boost --compatibility_version=1.63.0 --current_version=1.63.0 --build-type=complete --layout=tagged variant=release threading=single,multi link=shared Performing configuration checks - 32-bit : no - 64-bit : yes - arm : no - mips1: no - power: no - sparc: no - x86 : yes Building the Boost C++ Libraries. - symlinks supported : yes - C++11 mutex : yes - lockfree boost::atomic_flag : yes - Boost.Config Feature Check: cxx11_auto_declarations : yes - Boost.Config Feature Check: cxx11_constexpr : yes - Boost.Config Feature Check: cxx11_defaulted_functions : yes - Boost.Config Feature Check: cxx11_final : yes - Boost.Config Feature Check: cxx11_hdr_mutex : yes - Boost.Config Feature Check: cxx11_hdr_tuple : yes - Boost.Config Feature Check: cxx11_lambdas : yes - Boost.Config Feature Check: cxx11_noexcept : yes - Boost.Config Feature Check: cxx11_nullptr : yes - Boost.Config Feature Check: cxx11_rvalue_references : yes - Boost.Config Feature Check: cxx11_template_aliases : yes - Boost.Config Feature Check: cxx11_thread_local : yes - Boost.Config Feature Check: cxx11_variadic_templates : yes - has_icu builds : yes warning: Graph library does not contain MPI-based parallel components. note: to enable them, add "using mpi ;" to your user-config.jam - zlib : yes - bzip2: yes warning: Skipping Boost.Locale library with threading=single. - native-atomic-int32-supported : yes - native-syslog-supported : yes - pthread-supports-robust-mutexes : no - compiler-supports-visibility : yes - compiler-supports-ssse3 : yes - compiler-supports-avx2 : no - gcc visibility : yes - long double support : yes warning: skipping optional Message Passing Interface (MPI) library. note: to enable MPI support, add "using mpi ;" to user-config.jam. note: to suppress this message, pass "--without-mpi" to bjam. note: otherwise, you can safely ignore this message. warning: Skipping Boost.Thread library with threading=single. warning: Skipping Boost.Wave library with threading=single. - zlib : yes (cached) - bzip2: yes (cached) - iconv (libc) : no - iconv (separate) : yes - icu : no - icu (lib64) : no Component configura
[Fink-devel] boost1.63-nopython fails validation
Hi, building the package with ‘-m’ fails on a missing Shlibs entry (the other boost1.63 variants build fine): Validating .deb dir /scratch.noindex/fink.build/root-boost1.63-nopython-1.63.0-1... Error: package contains the shared library /sw/opt/boost-1_63/lib/libboost_fiber-mt.dylib but the corresponding install_name and compatibility_version %p/opt/boost-1_63/lib/libboost_fiber-mt-1_63.dylib 1.63.0 are not listed in the Shlibs field. See the packaging manual. Cheers, Derek -- Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] OpenSSL 1.0.2 certificates
[cc’ing Daniel as OpenSSL maintainer] > On 4 Jan 2017, at 12:47 pm, Hanspeter Niederstrasser > wrote: > > On 1/2/17 9:19 AM, Derek Homeier wrote: >> Hi, >> >> I want to check if I am doing something very stupid here, since I am unable >> to properly >> use any apps linking to openssl100-shlibs (among others wget and python) >> since approximately >> the update to openssl-1.0.2, as it refuses to accept almost any host >> certificate: >> >> ariel:15579> curl -O https://www.openssl.org/source/openssl-1.0.2j.tar.gz >> % Total% Received % Xferd Average Speed TimeTime Time Current >>Dload Upload Total SpentLeft Speed >> 100 5183k 100 5183k0 0 985k 0 0:00:05 0:00:05 --:--:-- >> 1266k >> ariel:15580> wget https://www.openssl.org/source/openssl-1.0.2j.tar.gz >> --2017-01-02 15:03:01-- https://www.openssl.org/source/openssl-1.0.2j.tar.gz >> Resolving www.openssl.org... 2600:1406:1a:38f::c1e, 2600:1406:1a:38e::c1e, >> 104.91.180.27 >> Connecting to www.openssl.org|2600:1406:1a:38f::c1e|:443... connected. >> ERROR: cannot verify www.openssl.org's certificate, issued by ‘CN=Let's >> Encrypt Authority X3,O=Let's Encrypt,C=US’: >> Unable to locally verify the issuer's authority. >> To connect to www.openssl.org insecurely, use `--no-check-certificate'. > > according to 'fink info wget', you have to edit .wgetrc so that wget knows > about the ca-bundle certificates. > > 1. Install the 'ca-bundle' package. > 2. If you don't currently have $HOME/.wgetrc, generate it via > . > cp /sw/etc/wgetrc $HOME/.wgetrc > . > 3. Edit $HOME/.wgetrc with your favorite text editor and add the > following line to it: > . > ca_certificate = /sw/etc/ssl/certs/ca-bundle.crt > > I can confirm that wget fails here similarly to you before the edit, and > downloads find after the change. > Thanks for the info and sorry for not reading the docs myself; this does work - also with /etc/ssl/cert.pem. I was actually hoping for a solution that would work for openssl100-shlibs linked programs in general, since Python’s urllib failures were bothering me more. I had found some suggestions for fixing the problem with ruby, but nothing really useful for Python. However Pythons ssl module itself provided the following information: >>> import ssl >>> ssl.get_default_verify_paths() DefaultVerifyPaths(cafile=None, capath='/sw/etc/ssl/certs', openssl_cafile_env='SSL_CERT_FILE', openssl_cafile='/sw/etc/ssl/cert.pem', openssl_capath_env='SSL_CERT_DIR', openssl_capath='/sw/etc/ssl/certs’) And indeed, copying the system cert.pem to /sw/etc/ssl fixed the problem (could have sworn I had already tried that), or alternatively setting SSL_CERT_FILE=/etc/ssl/cert.pem in the program environment. It doesn’t seem there is an option to configure a different default openssl_cafile when building openssl with --openssldir=%p/etc/ssl or if copying or linking the system file into %p/etc/ssl is an option, or as a last resort having the ca-bundle package install an additional copy of ca-bundle.crt as %p/etc/ssl/cert.pem. I suggest in any case to add the info about the SSL_CERT_FILE environment variable to the DescDetail. Thanks, Derek -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] OpenSSL 1.0.2 certificates
Hi, I want to check if I am doing something very stupid here, since I am unable to properly use any apps linking to openssl100-shlibs (among others wget and python) since approximately the update to openssl-1.0.2, as it refuses to accept almost any host certificate: ariel:15579> curl -O https://www.openssl.org/source/openssl-1.0.2j.tar.gz % Total% Received % Xferd Average Speed TimeTime Time Current Dload Upload Total SpentLeft Speed 100 5183k 100 5183k0 0 985k 0 0:00:05 0:00:05 --:--:-- 1266k ariel:15580> wget https://www.openssl.org/source/openssl-1.0.2j.tar.gz --2017-01-02 15:03:01-- https://www.openssl.org/source/openssl-1.0.2j.tar.gz Resolving www.openssl.org... 2600:1406:1a:38f::c1e, 2600:1406:1a:38e::c1e, 104.91.180.27 Connecting to www.openssl.org|2600:1406:1a:38f::c1e|:443... connected. ERROR: cannot verify www.openssl.org's certificate, issued by ‘CN=Let's Encrypt Authority X3,O=Let's Encrypt,C=US’: Unable to locally verify the issuer's authority. To connect to www.openssl.org insecurely, use `--no-check-certificate'. ariel:15581> wget --ca-certificate=/etc/ssl/cert.pem https://www.openssl.org/source/openssl-1.0.2j.tar.gz --2017-01-02 15:03:10-- https://www.openssl.org/source/openssl-1.0.2j.tar.gz Resolving www.openssl.org... 2600:1406:1a:38f::c1e, 2600:1406:1a:38e::c1e, 104.91.180.27 Connecting to www.openssl.org|2600:1406:1a:38f::c1e|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 5307912 (5.1M) [application/x-gzip] Saving to: ‘openssl-1.0.2j.tar.gz.1’ openssl-1.0.2j.tar.gz.1 100%[===>] 5.06M 1.38MB/sin 3.7s 2017-01-02 15:03:15 (1.38 MB/s) - ‘openssl-1.0.2j.tar.gz’ saved [5307912/5307912] ariel:15582> wget --ca-certificate=/sw/etc/ssl/certs/ca-bundle.crt https://www.openssl.org/source/openssl-1.0.2j.tar.gz --2017-01-02 15:31:49-- https://www.openssl.org/source/openssl-1.0.2j.tar.gz Resolving www.openssl.org... 2600:1406:1a:38f::c1e, 2600:1406:1a:38e::c1e, 104.91.180.27 Connecting to www.openssl.org|2600:1406:1a:38f::c1e|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 5307912 (5.1M) [application/x-gzip] Saving to: ‘openssl-1.0.2j.tar.gz.2’ openssl-1.0.2j.tar.gz.2 100%[===>] 5.06M 1.40MB/sin 3.6s 2017-01-02 15:31:55 (1.40 MB/s) - ‘openssl-1.0.2j.tar.gz.3’ saved [5307912/5307912] This happens regardless of whether the certificate updates from ca-bundle are installed or not (which are almost a year old now anyway), but as the last two examples show, it obviously accepts both the system-provided certificates in /etc/ssl or the ca-certs one if directed explicitly to them. Since I’ve never seen any other message about this problem popping up on the list, I am still wondering if I am doing something blatantly wrong, but as it is, the CERTIFICATE_VERIFY_FAILED errors causes a number of package builds to fail tests that require downloading things. Is there some way to have openssl100 automatically recognise any of the installed certificate lists? There is a /sw/etc/ssl/openssl.cnf however belonging to openssl_1.1.0c, and its locations don’t seem to make any sense at all: [ CA_default ] dir = ./demoCA # Where everything is kept certs = $dir/certs# Where the issued certs are kept crl_dir = $dir/crl # Where the issued crl are kept as there is no demo* subdirectory present whatsoever (otherwise I would have suspected yet another bit of trouble due to a case-sensitive file system). Trying to modify this installed openssl.cnf to point to any of the actual certificate locations did not get me any further either. If I am really the only one having this problem I’m wondering what could possibly be broken here after rebuilding and reinstalling all openssl packages... Thanks for any help, Derek -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: https://sourceforge.net/p/fink/mailman/fink-devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Filename case error in openssl110-dev-1.1.0a-1 manpage (MD5.3)
Hi Daniel, subject already tells most of the story: the capitalisation workaround in openssl110-dev incorrectly copies the MD5.3 manpage over in lowercase, which fails on a case-sensitive FS. I think this should fix it: diff -Nurd /sw/fink/10.9-libcxx/{stable,local}/main/finkinfo/crypto/openssl110-dev.info --- /sw/fink/10.9-libcxx/stable/main/finkinfo/crypto/openssl110-dev.info 2016-09-27 00:54:14.0 +0200 +++ /sw/fink/10.9-libcxx/local/main/finkinfo/crypto/openssl110-dev.info 2016-09-27 16:45:56.0 +0200 @@ -80,7 +80,7 @@ ### for comptibility with libmd (don't know how to set up ## update-alternatives with original capitalisation). - mv %i/share/man/man3/md5.3 %i/share/man/man3/md5.3.openssl100 + mv %i/share/man/man3/MD5.3 %i/share/man/man3/md5.3.openssl100 for i in 2 4; do rm %i/share/man/man3/MD$i.3 cp %i/share/man/man3/md5.3.openssl100 %i/share/man/man3/md$i.3.openssl100 - for the record: on a Linux system, ‘man MD5’ gives me the same as ‘man 3 md5’, while on my Mac with case-sensitive HFS+, ‘man MD5’ does not exist, ‘man 3 md5’ works as expected, but I guess that’s as much as can be expected. Cheers, Derek -- Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Byte-compiled files in matplotlib-py-1.5.1-2
On 16 Feb 2016, at 2:35 pm, Kurt Schwehr wrote: > > If you have a working setup for matplotlib, can you commit it? > I’ve tested and built these on 10.10 and 10.11; can submit them to the package tracker. Cheers, Derek matplotlib-py.info Description: Binary data matplotlib-py.patch Description: Binary data -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Byte-compiled files in matplotlib-py-1.5.1-2
On 12 Feb 2016, at 8:42 pm, Derek Homeier wrote: > > On 11 Feb 2016, at 2:12 pm, Kurt Schwehr wrote: >> >> Sorry, I'm totally not following how this should be fixed. I'm feeling >> rather brain dead at the moment. >> > Seems we have two options - > change the TestScript as suggested to let python cleanup the byte-compiled > files from the test run: > > InfoTest: << > TestDepends: ffmpeg, inkscape, nose-py%type_pkg[python], ( %type_pkg[python] > <= 27 ) mock-py%type_pkg[python] > TestScript: << > #!/bin/bash -ev > export PYTHONPATH=$(ls -d %b/build/lib.macosx-*-%type_raw[python]) > %p/bin/python%type_raw[python] -B -c 'import matplotlib as m, os, sys; > r=m.test(verbosity=1); os.system("find > %b/build/lib.macosx-*-%type_raw[python] -name \*.py[ > oc] -exec rm {} \;"); sys.exit(1-r)' > << > TestSuiteSize: large > << > > or let the InstallScript do the cleanup afterwards: > > InstallScript: << > #!/bin/sh -ev > export > PKG_CONFIG_PATH="%p/lib/pango-ft219/lib/pkgconfig:%p/lib/fontconfig2/lib/pkgconfig:%p/lib/freetype219/lib/pkgconfig:$PKG_CONFIG_PATH" > find %b/build/lib.macosx-*-%type_raw[python] -name \*.py[oc] -exec rm {} \; > %p/bin/python%type_raw[python] setup.py install --root %d > mkdir -p %i/share/doc/%n > cp -R examples %i/share/doc/%n > << > > Both versions worked for me, so it’s mainly a decision on > aesthetic/philosophical grounds. > Commented on one of the errors in the other thread, but I found a different > one on another > machine... I just noticed another issue with the current setup: since the macosx backend is auto-detected, although it is later disabled due to the non-Framework build, it is still written as default into the system matplotlibrc. As a consequence, for users with a clean setup, like fink-bld when testing other packages, ‘import matplotlib.pyplot’ will fail. Explicitly disabling the backend in setup.cfg will let setup.py pick a default that’s actually working with the package. It then chose qt5agg however, which seems to work alright with a plain python, and I did not see any immediate problems with 'ipython —pylab’; nonetheless the more prudent solution might be to set this explicitly instead, e.g. add this to the patch --- a/setup.py +++ b/setup.py @@ -149,7 +149,7 @@ if __name__ == '__main__': package_dir = {'': 'lib'} install_requires = [] setup_requires = [] -default_backend = None +default_backend = 'qt4agg' # Go through all of the packages and figure out which ones we are # going to build/install. matplotlib-py.patch Description: Binary data Cheers, Derek -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Byte-compiled files in matplotlib-py-1.5.1-2
Hi Kurt, On 11 Feb 2016, at 2:12 pm, Kurt Schwehr wrote: > > Sorry, I'm totally not following how this should be fixed. I'm feeling > rather brain dead at the moment. > Seems we have two options - change the TestScript as suggested to let python cleanup the byte-compiled files from the test run: InfoTest: << TestDepends: ffmpeg, inkscape, nose-py%type_pkg[python], ( %type_pkg[python] <= 27 ) mock-py%type_pkg[python] TestScript: << #!/bin/bash -ev export PYTHONPATH=$(ls -d %b/build/lib.macosx-*-%type_raw[python]) %p/bin/python%type_raw[python] -B -c 'import matplotlib as m, os, sys; r=m.test(verbosity=1); os.system("find %b/build/lib.macosx-*-%type_raw[python] -name \*.py[ oc] -exec rm {} \;"); sys.exit(1-r)' << TestSuiteSize: large << or let the InstallScript do the cleanup afterwards: InstallScript: << #!/bin/sh -ev export PKG_CONFIG_PATH="%p/lib/pango-ft219/lib/pkgconfig:%p/lib/fontconfig2/lib/pkgconfig:%p/lib/freetype219/lib/pkgconfig:$PKG_CONFIG_PATH" find %b/build/lib.macosx-*-%type_raw[python] -name \*.py[oc] -exec rm {} \; %p/bin/python%type_raw[python] setup.py install --root %d mkdir -p %i/share/doc/%n cp -R examples %i/share/doc/%n << Both versions worked for me, so it’s mainly a decision on aesthetic/philosophical grounds. Commented on one of the errors in the other thread, but I found a different one on another machine... > On 11 Feb 2016, at 1:22 am, Alexander Hansen > wrote: > >> On Feb 10, 2016, at 15:55, Derek Homeier >> wrote: >> >> >>> On 11 Feb 2016, at 12:35 am, Alexander Hansen >>> wrote: > >> One could certainly call the find command _at the beginning_ of Install (and >> still on %b/build), >> which should do the job as well in perhaps somewhat simpler language. On the >> other hand I >> find it preferable for the TestScript to cleanup its own mess, and exactly >> that… >> >> Cheers, >> Derek >> > > Ah, OK, I understand the situation a bit better now. > > Interestingly enough, I didn’t trigger the failure when building > matplotlib-py35 or matplotlib-py27 here (10.11) using -m mode, possibly > because of test errors. I’ll send those in a separate thread. I wanted to check if the verification for some reason only fails on case-sensitive FS, so I tried the same on another 10.11 machine with standard HFS+ (otherwise nearly identical iMac), but instead got this segfault of the python interpreter itself (excerpts of huge coredump below). Happened with both python2.7 and python3.5, with identical OS and Xcode versions as my other iMac on 10.11; I rebuilt both python packages and tried to rebuild all matplotlib dependencies (all QT4 and QT5 libraries in particular, but to no avail. Might be a broken Xcode installation, but at least for libdyld.dylib I checked it is identical (by md5sum) on both machines… Cheers, Derek Feb 11 22:51:59 python3.5[10030] : CGSConnectionByID: 0 is not a valid connection ID. Feb 11 22:51:59 python3.5[10030] : Invalid Connection ID 0 2016-02-11 22:52:01.536 python3.5[10030:3820856] An uncaught exception was raised 2016-02-11 22:52:01.536 python3.5[10030:3820856] *** -[__NSArray0 objectAtIndex:]: index 0 beyond bounds for empty NSArray 2016-02-11 22:52:01.537 python3.5[10030:3820856] ( 0 CoreFoundation 0x7fff8510cae2 __exceptionPreprocess + 178 1 libobjc.A.dylib 0x7fff8fa9e73c objc_exception_throw + 48 2 CoreFoundation 0x7fff850a31f5 -[__NSArray0 objectAtIndex:] + 101 3 QtGui 0x00010f991d56 _ZN28QDesktopWidgetImplementation8onResizeEv + 150 4 QtGui 0x00010f991c56 _ZN28QDesktopWidgetImplementationC2Ev + 54 5 QtGui 0x00010f99215c _ZL28qdesktopWidgetImplementationv + 60 6 QtGui 0x00010f992659 _ZNK14QDesktopWidget13primaryScreenEv + 9 7 QtGui 0x00010f976f33 _ZN14QWidgetPrivate10create_sysElbb + 291 8 QtGui 0x00010fa26ddd _ZN7QWidget6createElbb + 701 9 QtGui 0x00010fe22b2c _ZN15QToolBarPrivate4initEv + 252 10 QtGui 0x00010fe23b65 _ZN8QToolBarC2EP7QWidget + 229 11 QtGui.so0x00010f4948ff _ZL18init_type_QToolBarP17_sipSimpleWrapperP7_objectS2_PS2_S3_S3_ + 495 12 sip.so 0x00010e23f7d4 sipSimpleWrapper_init + 187 13 libpython3.5m.dylib 0x0
Re: [Fink-devel] Test script errors in matplotlib-pyXY
On 11 Feb 2016, at 1:25 am, Alexander Hansen wrote: > > urllib.error.URLError: certificate verify failed (_ssl.c:645)> > > During handling of the above exception, another exception occurred: > > Traceback (most recent call last): > File "/sw/lib/python3.5/site-packages/nose/case.py", line 198, in runTest >self.test(*self.arg) > File > "/sw/src/fink.build/matplotlib-py35-1.5.1-2/matplotlib-1.5.1/build/lib.macosx-10.11-x86_64-3.5/matplotlib/tests/test_style.py", > line 63, in test_use_url >with style.context('https://gist.github.com/adrn/6590261/raw'): > File "/sw/lib/python3.5/contextlib.py", line 59, in __enter__ >return next(self.gen) > File > "/sw/src/fink.build/matplotlib-py35-1.5.1-2/matplotlib-1.5.1/build/lib.macosx-10.11-x86_64-3.5/matplotlib/style/core.py", > line 121, in context >use(style) > File > "/sw/src/fink.build/matplotlib-py35-1.5.1-2/matplotlib-1.5.1/build/lib.macosx-10.11-x86_64-3.5/matplotlib/style/core.py", > line 90, in use >raise IOError(msg % style) > OSError: 'https://gist.github.com/adrn/6590261/raw' not found in the style > library and input is not a valid URL or path. See `style.available` for list > of available styles. > > -- > Ran 5165 tests in 658.655s > > FAILED (KNOWNFAIL=22, SKIP=6, errors=1) > /sw/src/fink.build/matplotlib-py35-1.5.1-2/matplotlib-1.5.1/build/lib.macosx-10.11-x86_64-3.5/matplotlib/__init__.py:1350: > UserWarning: This call to matplotlib.use() has no effect > because the backend has already been chosen; > matplotlib.use() must be called *before* pylab, matplotlib.pyplot, > or matplotlib.backends is imported for the first time. > > warnings.warn(_use_error_msg) > > > I’m not a bit surprised by the SSL error on 10.11. > Isn’t that a fairly general problem with openssl and various sites like github, pypi etc.? I am getting the same error (on 10.10 as well) when trying to manually download the file with fink’s wget, while the system curl and ‘wget —no-check-certificate’ find it. I actually thought that was a real issue with the certificates on those sites, but since nothing has happened on that side in months, it may be a bug, but definitely somewhere further up with urllib/openssl, not really mpl’s problem. Cheers, Derek -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Byte-compiled files in matplotlib-py-1.5.1-2
> On 11 Feb 2016, at 12:35 am, Alexander Hansen > wrote: > >> when building the updated matplotlib package in maintainer mode, after the >> tests are run I am >> still getting the dreaded verification error for the .deb: >> >> Error: Compiled python module points to fink build dir. >> >> Uncommenting the cleanup line does not help, because after a few Fails and >> Errors in the >> test suite the script exits immediately. I have not found any way to exit >> the python command >> with the test with a nonzero return value that would not cause ‘bash -e’ to >> immediately exit, >> so the only way to get rid of the compiled files and still save the test >> result might be to do >> the cleanup within the python call itself: >> >> %p/bin/python%type_raw[python] -B -c 'import matplotlib as m, os, sys; >> r=m.test(verbosity=1); os.system("find >> %b/build/lib.macosx-*-%type_raw[python] -name \*.py[oc] -exec rm {} \;"); >> sys.exit(1-r)' >> >> lets the test phase finish with a warning for the current bunch of failures, >> and should pass if >> the tests all succeed. Bit lengthy perhaps, but it sees to do the job. >> >> Cheers, >> Derek >> > > How about putting the command (or perhaps a different one) at the end of the > InstallScript instead and apply it to the deb root directory (%i)? The > .deb-directory validation is run after the InstallScript. > > If the package is building the .deb differently depending on whether the test > suite is run or not, it breaks our fundamental premise of invariant .debs. The cleanup is actually there to make sure it does build the same way, as it only removes the files that were byte-compiled in the test run in the first place (since 'python -B' has no effect here for whatever reason). Putting it into the InstallScript to run on %i at the end would be entirely counter-productive, as it would just remove _all_ .pyc files that were precompiled by 'setup.py install' in the right place: byte-compiling /scratch.noindex/fink.build/root-matplotlib-py27-1.5.1-2/sw/lib/python2.7/site-packages/mpl_toolkits/tests/test_axes_grid.py to test_axes_grid.pyc byte-compiling /scratch.noindex/fink.build/root-matplotlib-py27-1.5.1-2/sw/lib/python2.7/site-packages/mpl_toolkits/tests/test_axes_grid1.py to test_axes_grid1.pyc byte-compiling /scratch.noindex/fink.build/root-matplotlib-py27-1.5.1-2/sw/lib/python2.7/site-packages/mpl_toolkits/tests/test_mplot3d.py to test_mplot3d.pyc running install_egg_info running egg_info … Writing control file... Writing md5sums file... Validating .deb dir /scratch.noindex/fink.build/root-matplotlib-py27-1.5.1-2... Package looks good! One could certainly call the find command _at the beginning_ of Install (and still on %b/build), which should do the job as well in perhaps somewhat simpler language. On the other hand I find it preferable for the TestScript to cleanup its own mess, and exactly that… Cheers, Derek -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Byte-compiled files in matplotlib-py-1.5.1-2
Hi, when building the updated matplotlib package in maintainer mode, after the tests are run I am still getting the dreaded verification error for the .deb: Error: Compiled python module points to fink build dir. Uncommenting the cleanup line does not help, because after a few Fails and Errors in the test suite the script exits immediately. I have not found any way to exit the python command with the test with a nonzero return value that would not cause ‘bash -e’ to immediately exit, so the only way to get rid of the compiled files and still save the test result might be to do the cleanup within the python call itself: %p/bin/python%type_raw[python] -B -c 'import matplotlib as m, os, sys; r=m.test(verbosity=1); os.system("find %b/build/lib.macosx-*-%type_raw[python] -name \*.py[oc] -exec rm {} \;"); sys.exit(1-r)' lets the test phase finish with a warning for the current bunch of failures, and should pass if the tests all succeed. Bit lengthy perhaps, but it sees to do the job. Cheers, Derek -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Filename case conflict in imagemagick6.q16-doc-6.9.3.2-2
Hi, on a case-sensitive HFS+ build I am getting the following error Validating .deb dir /scratch.noindex/fink.build/root-imagemagick6.q16-doc-6.9.3.2-2... Error: Pathname collision on case-insensitive filesystems Offending file: /sw/share/doc/ImageMagick-6/www/api/image.html which means that on case-insensitive HFS+, one of image.html or Image.html is destroyed. One contains C, the other C++ API documentation, so a rename will probably not work that easily without breaking an unknown number of links. Might be best to return to the upstream developers. Cheers, Derek -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] hdf5.10 build failure (invalid control file)
Hi, building hdf5-bin_1.8.16 fails with env LANG=C LC_ALL=C dpkg-deb -b root-hdf5-bin-1.8.16-1 /sw/fink/10.10/stable/main/binary-darwin-x86_64/sci /sw/bin/dpkg-deb: parse error, in file `root-hdf5-bin-1.8.16-1/DEBIAN/control' near line 10 package `hdf5-bin': `Replaces' field, syntax error after reference to package `hdf5-oldapi-bin’ It seems there are missing commas after hdf5-oldapi-bin, hdf5.8-oldapi as well as Conflicts: … hdf5.8-oldapi in the hdf5-bin SplitOff description. Cheers, Derek hdf5.10.info Description: Binary data -- Go from Idea to Many App Stores Faster with Intel(R) XDK Give your users amazing mobile app experiences with Intel(R) XDK. Use one codebase in this all-in-one HTML5 development environment. Design, debug & build mobile apps & 2D/3D high-impact games for multiple OSs. http://pubads.g.doubleclick.net/gampad/clk?id=254741551&iu=/4140___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Typo in graphicsmagick-q32 dependencies
Hi, this package does not update due to a non-existing dependency in the -dev Splitoff: < Depends: graphicsmagick1321-q32-shlibs (= %V) --- > Depends: graphicsmagick1322-q32-shlibs (= %V) Cheers, Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] openmpi-1.8.7-1 build error on case-sensitive FS
Hi Jack, the 1.8 versions of Open-MPI have a couple of new files with conflicting filenames, causing the build on case-sensitive HFS+ to fail (and probably some of those files to be lost on case-insensitive ones. The vtCC file versions should be added to the case-sensitive fix, I think: Validating .deb dir /scratch.noindex/fink.build/root-openmpi-1.8.7-1... Error: Pathname collision on case-insensitive filesystems Offending file: /sw/bin/vtcc Offending file: /sw/share/doc/vampirtrace/vtcc-wrapper-data.txt Cheers, Derek openmpi.info Description: Binary data -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] vtk510 missing Shlibs fields
On 9 Apr 2015, at 7:27 pm, Alexander Hansen wrote: > > It appears to need Apple’s Legacy Java-1.6, and thus system-java16*. You can > get that: > > https://support.apple.com/downloads/DL1572/en_US/JavaForOSX2014-001.dmg > > The build worked for me when I moved my Java-1.8 out of the way. > > The best way to handle this would probably be to put dependencies on > system-java16* and to ensure that vtk510 uses only that. Yes, that did it - so it certainly should have the system-java16[-dev] dependencies; obviously forcing it to ignore never Java installations during the build would be the ideal solution; even with the alternative of a Conflicts: system-java18-dev I don’t know if there is a canonical way to uninstall the jdk1.8.x… Thanks, Derek -- BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop your own process in accordance with the BPMN 2 standard Learn Process modeling best practices with Bonita BPM through live exercises http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual- event?utm_ source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] vtk510 missing Shlibs fields
On 6 Apr 2015, at 1:03 am, Hanspeter Niederstrasser wrote: > >> building vtk510-5.10.1-1 with -m fails the package check due to two entries >> missing in the Shlibs list: >> >> 147a148 >>> %p/lib/%N/libLSDyna.5.10.dylib 5.10.0 %n (>= >>> 5.10.0-1) >> 231a233 >>> %p/lib/%N/libvtkhdf5_hl.1.8.5.dylib 1.8.5 %n (>= 5.10.0-1) > > Thanks for the report. I've gone ahead and added the two files to the > Shlibs: field. I've also added missing *Depends to the various SplitOffs. > Thanks for the fix; unfortunately on 10.10 I was having another build problem with a rather non-informative error, after setting MaxBuildJobs to 1, this seems to point to a Java Problem (the control panel here said 8u31 was up-to-date, though initially the build reported an error with an include file in /Library/Java/JavaVirtualMachines/jdk1.8.0_26.jdk/). I reinstalled first the jdk1.8.0_31, then jdk1.8.0_40 from the Oracle site, but as it seems, producing the same error in all versions ca. 69% into the build (see first snippet of build log below). I finally removed all jdk versions from /Library/Java/JavaVirtualMachines, so no system-java is installed any more; with this the build completes to 96% but then complains about having to install a Java runtime and then fails with a java error (2nd piece of build log below). On the 10.9 machine where the build succeeded, all system-java1[4-6]-dev are installed. Since the package does not specify any java dependencies, do you have any idea what to use? Thanks, Derek +++ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:107:3: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? JAWT_MacOSXDrawingSurfaceInfo* dsi_mac; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:108:14: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? dsi_mac = (JAWT_MacOSXDrawingSurfaceInfo*)dsi->platformInfo; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:109:31: error: no member named 'cocoaViewRef' in 'jawt_DrawingSurfaceInfo' temp0->SetWindowId(dsi_mac->cocoaViewRef); ~~~ ^ 3 errors generated. make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/depend make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild && /sw/bin/cmake -E cmake_depends "Unix Makefiles" /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1 /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/DependInfo.cmake --color= Scanning dependencies of target VTKJavaCommon make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' /sw/bin/cmake -E cmake_progress_report /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/CMakeFiles [ 96%] Generating ../../java/vtk/vtkVTKJavaCommonDriver.class cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java && /usr/bin/javac -source 5 -classpath /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java/vtk/vtkVTKJavaCommonDriver.java No Java runtime present, requesting install. Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make:225: recipe for target 'java/vtk/vtkVTKJavaCommonDriver.class' failed make[2]: *** [java/vtk/vtkVTKJavaCommonDriver.class] Error 1 make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' CMakeFiles/Makefile2:8544: recipe for target 'Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/all' failed make[1]: *** [Wrapping/Java/CMakeFiles/VTKJavaComm
Re: [Fink-devel] vtk510 missing Shlibs fields
On 6 Apr 2015, at 1:03 am, Hanspeter Niederstrasser wrote: > >> building vtk510-5.10.1-1 with -m fails the package check due to two entries >> missing in the Shlibs list: >> >> 147a148 >>> %p/lib/%N/libLSDyna.5.10.dylib 5.10.0 %n (>= >>> 5.10.0-1) >> 231a233 >>> %p/lib/%N/libvtkhdf5_hl.1.8.5.dylib1.8.5 %n (>= >>> 5.10.0-1) > > Thanks for the report. I've gone ahead and added the two files to the > Shlibs: field. I've also added missing *Depends to the various SplitOffs. > Thanks for the fix; unfortunately on 10.10 I was having another build problem with a rather non-informative error, after setting MaxBuildJobs to 1, this seems to point to a Java Problem (the control panel here said 8u31 was up-to-date, though initially the build reported an error with an include file in /Library/Java/JavaVirtualMachines/jdk1.8.0_26.jdk/). I reinstalled first the jdk1.8.0_31, then jdk1.8.0_40 from the Oracle site, but as it seems, producing the same error in all versions ca. 69% into the build (see first snippet of build log below). I finally removed all jdk versions from /Library/Java/JavaVirtualMachines, so no system-java is installed any more; with this the build completes to 96% but then complains about having to install a Java runtime and then fails with a java error (2nd piece of build log below). On the 10.9 machine where the build succeeded, all system-java1[4-6]-dev are installed. Since the package does not specify any java dependencies, do you have any idea what to use? Thanks, Derek +++ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:107:3: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? JAWT_MacOSXDrawingSurfaceInfo* dsi_mac; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:108:14: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? dsi_mac = (JAWT_MacOSXDrawingSurfaceInfo*)dsi->platformInfo; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:109:31: error: no member named 'cocoaViewRef' in 'jawt_DrawingSurfaceInfo' temp0->SetWindowId(dsi_mac->cocoaViewRef); ~~~ ^ 3 errors generated. make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/depend make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild && /sw/bin/cmake -E cmake_depends "Unix Makefiles" /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1 /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/DependInfo.cmake --color= Scanning dependencies of target VTKJavaCommon make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' /sw/bin/cmake -E cmake_progress_report /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/CMakeFiles [ 96%] Generating ../../java/vtk/vtkVTKJavaCommonDriver.class cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java && /usr/bin/javac -source 5 -classpath /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java/vtk/vtkVTKJavaCommonDriver.java No Java runtime present, requesting install. Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make:225: recipe for target 'java/vtk/vtkVTKJavaCommonDriver.class' failed make[2]: *** [java/vtk/vtkVTKJavaCommonDriver.class] Error 1 make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' CMakeFiles/Makefile2:8544: recipe for target 'Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/all' failed make[1]: *** [Wrapping/Java
Re: [Fink-devel] vtk510 missing Shlibs fields
On 6 Apr 2015, at 1:03 am, Hanspeter Niederstrasser wrote: > >> building vtk510-5.10.1-1 with -m fails the package check due to two entries >> missing in the Shlibs list: >> >> 147a148 >>> %p/lib/%N/libLSDyna.5.10.dylib 5.10.0 %n (>= >>> 5.10.0-1) >> 231a233 >>> %p/lib/%N/libvtkhdf5_hl.1.8.5.dylib 1.8.5 %n (>= >>> 5.10.0-1) > > Thanks for the report. I've gone ahead and added the two files to the > Shlibs: field. I've also added missing *Depends to the various SplitOffs. > Thanks for the fix; unfortunately on 10.10 I was having another build problem with a rather non-informative error, after setting MaxBuildJobs to 1, this seems to point to a Java Problem (the control panel here said 8u31 was up-to-date, though initially the build reported an error with an include file in /Library/Java/JavaVirtualMachines/jdk1.8.0_26.jdk/). I reinstalled first the jdk1.8.0_31, then jdk1.8.0_40 from the Oracle site, but as it seems, producing the same error in all versions ca. 69% into the build (see first snippet of build log below). I finally removed all jdk versions from /Library/Java/JavaVirtualMachines, so no system-java is installed any more; with this the build completes to 96% but then complains about having to install a Java runtime and then fails with a java error (2nd piece of build log below). On the 10.9 machine where the build succeeded, all system-java1[4-6]-dev are installed. Since the package does not specify any java dependencies, do you have any idea what to use? Thanks, Derek +++ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:107:3: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? JAWT_MacOSXDrawingSurfaceInfo* dsi_mac; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:108:14: error: unknown type name 'JAWT_MacOSXDrawingSurfaceInfo'; did you mean 'JAWT_DrawingSurfaceInfo'? dsi_mac = (JAWT_MacOSXDrawingSurfaceInfo*)dsi->platformInfo; ^ JAWT_DrawingSurfaceInfo /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home/include/jawt.h:170:3: note: 'JAWT_DrawingSurfaceInfo' declared here } JAWT_DrawingSurfaceInfo; ^ In file included from /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Rendering/vtkRenderWindowJava.cxx:1405: /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Common/vtkJavaAwt.h:109:31: error: no member named 'cocoaViewRef' in 'jawt_DrawingSurfaceInfo' temp0->SetWindowId(dsi_mac->cocoaViewRef); ~~~ ^ 3 errors generated. make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/depend make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild && /sw/bin/cmake -E cmake_depends "Unix Makefiles" /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1 /scratch.noindex/fink.build/vtk510-5.10.1-2/VTK5.10.1/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/DependInfo.cmake --color= Scanning dependencies of target VTKJavaCommon make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' make -f Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build make[2]: Entering directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' /sw/bin/cmake -E cmake_progress_report /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/CMakeFiles [ 96%] Generating ../../java/vtk/vtkVTKJavaCommonDriver.class cd /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/Wrapping/Java && /usr/bin/javac -source 5 -classpath /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java /scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild/java/vtk/vtkVTKJavaCommonDriver.java No Java runtime present, requesting install. Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/build.make:225: recipe for target 'java/vtk/vtkVTKJavaCommonDriver.class' failed make[2]: *** [java/vtk/vtkVTKJavaCommonDriver.class] Error 1 make[2]: Leaving directory '/scratch.noindex/fink.build/vtk510-5.10.1-2/VTKBuild' CMakeFiles/Makefile2:8544: recipe for target 'Wrapping/Java/CMakeFiles/VTKJavaCommon.dir/al
[Fink-devel] vtk510 missing Shlibs fields
Hi, building vtk510-5.10.1-1 with -m fails the package check due to two entries missing in the Shlibs list: 147a148 > %p/lib/%N/libLSDyna.5.10.dylib5.10.0 %n (>= > 5.10.0-1) 231a233 > %p/lib/%N/libvtkhdf5_hl.1.8.5.dylib 1.8.5 %n (>= > 5.10.0-1) Derek -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] nightfall_1.86-2
On 18 Mar 2015, at 07:26 pm, Daniel Macks wrote: > On Wed, 18 Mar 2015 09:38:32 -0700, Alexander Hansen > wrote: > >> > On Mar 18, 2015, at 9:08 AM, Derek Homeier >> wrote: >> > > Hi, >> > > the nightfall update is still failing with name conflicts on >> case-sensitive HFS >> > (and is silently losing 2 of the files in question on insensitive ones). >> > This can be fixed with a >> > TarFilesRename: bh_virB.dat:bh_virBoldZ.dat bh_virV.dat:bh_virVoldZ.dat >> > > see also the last update to >> > https://sourceforge.net/p/fink/package-submissions/4423/ >> > > Cheers, >> > Derek >> > > Thanks! I’ve synced the relevant changes the tracker item over, >> so 1.86-4 should have everything, including having you as official >> maintainer. > > Thanks! Thanks as well! Maybe time to get CVS commit access after all… Derek -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] nightfall_1.86-2
Hi, the nightfall update is still failing with name conflicts on case-sensitive HFS (and is silently losing 2 of the files in question on insensitive ones). This can be fixed with a TarFilesRename: bh_virB.dat:bh_virBoldZ.dat bh_virV.dat:bh_virVoldZ.dat see also the last update to https://sourceforge.net/p/fink/package-submissions/4423/ Cheers, Derek -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] astropy-1.0
Hanspeter, Alexander, > On 25 Feb 2015, at 5:05 am, Hanspeter Niederstrasser > wrote: > >> following up on my recent question how to best get package updates into the >> distribution as a non-core developer, is the full package tree going to be >> migrated >> to github at some point? Then it would of course just take a pull request. >> In the meantime, I am attaching the update for Astropy the old-fashioned way; >> tested on 10.9 and 10.10 with all relevant versions. > > Checked in to the 10.7 tree (tested on 10.9). New/updated submissions are > better handled through the submissions tracker[1] or let us know if you want > CVS write access. thanks for checking in; it did seem more convenient to me to use the tracker as well, though I did not know how this was handled with a maintainer’s own packages - it might effectively force one of the core developers whom the ticket is assigned to to become a co-maintainer of the package. Could not login just know as SF seems to have decided to reset my password. But write access would be appreciated as well; if the github migration is foreseen to be completed within the next one or two months I could probably live without CVS rights (rather than having to relearn cvs so I won’t screw up anything...). Cheers, Derek -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] astropy-1.0
Hi, following up on my recent question how to best get package updates into the distribution as a non-core developer, is the full package tree going to be migrated to github at some point? Then it would of course just take a pull request. In the meantime, I am attaching the update for Astropy the old-fashioned way; tested on 10.9 and 10.10 with all relevant versions. Cheers, Derek astropy-py.info Description: Binary data -- Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] bottleneck-py-1.0.0-1
On 18 Feb 2015, at 12:44 am, Daniel Macks wrote: > >> as the entire test suite fails with an import error with 1.8 (apparently >> crashes the >> interpreter, thus the failure was not caught be the original Test field). > > I think the usual pattern for TestScript is: > > some command sequence || exit 2 > > rather than an intermediate TESTFAIL variable. If "some command sequence" > doesn't succeed, exit 2 is what fink wants to see to be fatal. If it does > succeed, then a normal success exit-code (e.g., not explicitly 2) is default. > But your package already looks like the test suite requires manual tricks, so > it's possible even the exit-handler needs special tricks too. The idea of first catching the return value of ‘some command sequence’ was a) to allow to return with 1 for a limited no. of (possibly expected) errors, thus only raising a warning b) to let the remainder of the test script run after an exit 1, in case some cleanup is necessary (not required in this package, but many python test suites leave byte-compiled files in the source tree behind). However in this case python seems to actually segfault rather than exiting with an exception, but with an exit value of only 1, so it was also only triggering a warning. The current setting takes care of failing the tests due to two failures when running bottleneck 0.8 + numpy 1.9 failing after the python crash showing up in the combination bottleneck 1.0 + numpy 1.8 It's supposed to give an exit 1 warning, should the test suite produce a single failure. Cheers, Derek -- Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] bottleneck-py-1.0.0-1
Hi Jack, > > The bottleneck-py package needs updated to the 1.0.0 source in > order to pass 'fink -m' against the newer numpy 1.9.x release… > thanks for the patch; the Depends then also need to be updated to Depends: python%type_pkg[python], numpy-py%type_pkg[python] (>= 1.9.0-1) as the entire test suite fails with an import error with 1.8 (apparently crashes the interpreter, thus the failure was not caught be the original Test field). Should probably also add the Distribution: << (%type_pkg[python] = 33) 10.7, (%type_pkg[python] = 33) 10.8, (%type_pkg[python] = 33) 10.9 << field on that occasion. BTW is there a preferred way for non-core package maintainers to get their updated .info files into the distribution? I am attaching the complete modified file as the Test needed a bit more trickery to catch the interpreter error as well, tested on 10.9+10.10, py27|py33|py34 Cheers, Derek > Index: bottleneck-py.info > === > RCS file: > /cvsroot/fink/dists/10.7/stable/main/finkinfo/libs/pythonmods/bottleneck-py.info,v > retrieving revision 1.1 > diff -r1.1 bottleneck-py.info > 3c3 > < Version: 0.8.0 > --- >> Version: 1.0.0 > 10c10 > < Source-MD5: 1a363fa35ce521eebb838e1bd6520e24 > --- >> Source-MD5: 380fa6f275bd24f27e7cf0e0d752f5d2 > bottleneck-py.info Description: Binary data -- Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] nose-py problems
On 6 Feb 2015, at 4:44 pm, Derek Homeier wrote: > > I found some issues with the build_tests run added to make the test suite for > nose 1.3.4 work: > > setup.py build_tests makes the test run with python3.x, but breaks it for > python2.7 (setup.py test actually works there). > > The same workaround seems to install an additional nose.egg-info in build/lib, > which is then installed as a second, non-versioned nose.egg-info dir. > The duplicate nose entry causes an error on all recent setuptools-tng > installations when setting up > further packages (both within and outside of the fink build system). > > I am attaching a modified info to let the tests run under python2.7 and > remove the extraneous nose.egg-info. > Uh, sorry, completely messed up the attachment, here comes the correct version! nose-py.info Description: Binary data -- Dive into the World of Parallel Programming. The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] nose-py problems
Hi Kurt, I found some issues with the build_tests run added to make the test suite for nose 1.3.4 work: setup.py build_tests makes the test run with python3.x, but breaks it for python2.7 (setup.py test actually works there). The same workaround seems to install an additional nose.egg-info in build/lib, which is then installed as a second, non-versioned nose.egg-info dir. The duplicate nose entry causes an error on all recent setuptools-tng installations when setting up further packages (both within and outside of the fink build system). I am attaching a modified info to let the tests run under python2.7 and remove the extraneous nose.egg-info. Cheers, Derek nose-py.info Description: Binary data -- Dive into the World of Parallel Programming. The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] pytables-py-3.1.1-2
Hi Jack, > > Derek, > The following changes to pytables-py.info aids in the migration > of the fink packages to the newer hdf5.9 and fixes the current thanks for taking care of this, had it on my todo list for a while. > validation error about "ompiled python module points to fink build > dir". The packaging now passes 'fink -m' on 10.10 as well as with > 'fink install pytables-py27' and 'fink remove pytables-py27 to confirm > the .pyo files created in the PostInstScript are removed in the > PreRmScript. Ah, did not notice that failed tests would break the cleanup! Though I actually had seen this in other packages before. But I am a bit wary about manipulating files directly in %i; this could be fixed alternatively well by making sure the test script is correctly intercepted even on returning an error: 21c27 < %p/bin/python%type_raw[python] setup.py install --cflags='-I%p/include' --lflags='-L%p/lib' --hdf5=%p/ --lzo=%p/ --bzip2=%p/ --root=%d --- > %p/bin/python%type_raw[python] setup.py install -O1 --cflags='-I%p/include' > --lflags='-L%p/lib' --hdf5=%p/ --lzo=%p/ --bzip2=%p/ --root=%d 50,52c56,60 < %p/bin/python%type_raw[python] -B -c 'import tables, sys; r=tables.test(); sys.exit(r)' < TESTFAIL=$? < find %b/build/lib.macosx-*-%type_raw[python] -name '*.py[oc]' -exec rm {} \; --- > TESTFAIL=0 > %p/bin/python%type_raw[python] -B -c 'import tables, sys; r=tables.test(); > sys.exit(r)' || TESTFAIL=$? > find %b/build -name '*.py[oc]' -exec rm {} \; > find %b/build -name __pycache__ -exec rm -r {} \; The .pyo files would also be taken care of that way. Built this with -m on 10.10 for py27 and py34, and tested the uninstall as well. And I’d also add the Distribution: field as Hanspeter had suggested a while ago. Cheers, Derek > Jack > > Index: pytables-py.info > === > RCS file: /cvsroot/fink/dists/10.7/stable/main/finkinfo/sci/pytables-py.info,v > retrieving revision 1.11 > diff -r1.11 pytables-py.info > 5c5 > < Revision: 1 > --- >> Revision: 2 > 8,9c8,9 > < Depends: python%type_pkg[python], numpy-py%type_pkg[python] (>= > 1.7.1-1), numexpr-py%type_pkg[python] (>= 2.2.2-1), bzip2-shlibs, > lzo2-shlibs, hdf5.8-shlibs > < BuildDepends: bzip2-dev, lzo2, cython-py%type_pkg[python] (>= 0.13-1), > hdf5.8 > --- >> Depends: python%type_pkg[python], numpy-py%type_pkg[python] (>= 1.7.1-1), >> numexpr-py%type_pkg[python] (>= 2.2.2-1), bzip2-shlibs, lzo2-shlibs, >> hdf5.9-shlibs >> BuildDepends: bzip2-dev, lzo2, cython-py%type_pkg[python] (>= 0.13-1), hdf5.9 > 25a26,27 >> >> /usr/bin/find %i/lib/python%type_raw[python]/site-packages/tables -name >> '*.pyc' -delete > 36a39,40 >> >> %p/bin/python%type_raw[python] -O >> %p/lib/python%type_raw[python]/compileall.py -q >> %p/lib/python%type_raw[python]/site-packages/tables > 42a47,48 >> >> /usr/bin/find %p/lib/python%type_raw[python]/site-packages/tables -name >> '*.pyo' -delete -- Derek Homeier Centre de Recherche Astrophysique de Lyon ENS Lyon 46, Allée d'Italie 69364 Lyon Cedex 07, France +33 47272-8894 -- Dive into the World of Parallel Programming. The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] hdf5.9-gfortran
Hi, I think the Usage Notes for the HDF5.9 Fortran libraries are outdated or mistakenly taken over from the -oldapi variant (pointing to the %p/opt/hdf5v1.6 subdirs). Also the InstallScript is removing all executables including the Fortran wrapper h5fc which is not provided by hdf5.9. I suggest to provide at least a basic example for how to compile and link without the wrapper with the correct paths, or include the wrapper in the installation, or both. Cheers, Derek hdf5.9-gfortran.info Description: Binary data -- New Year. New Location. New Benefits. New Data Center in Ashburn, VA. GigeNET is offering a free month of service with a new server in Ashburn. Choose from 2 high performing configs, both with 100TB of bandwidth. Higher redundancy.Lower latency.Increased capacity.Completely compliant. http://p.sf.net/sfu/gigenet___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] phase .deb validation: libwpg-0.2-shlibs-0.2.2-2 failed (linked using flat namespace) on Yosemite
Hi, the validation step fails when built with update -m: Validating .deb dir /scratch.noindex/fink.build/root-libwpg-0.2-shlibs-0.2.2-2... Error: /scratch.noindex/fink.build/root-libwpg-0.2-shlibs-0.2.2-2/sw/lib/libwpg-0.2.2.dylib appears to have been linked using a flat namespace. If this package BuildDepends on libtool2, make sure that you use BuildDepends: libtool2 (>= 2.4.2-4). and use autoreconf to regenerate the configure script. If the package doesn't BuildDepend on libtool2, you'll need to update its build procedure to avoid passing -Wl,-flat_namespace when linking libraries. Could it be that this block in configure:8260 darwin*) # darwin 5.x on # if running on 10.5 or later, the deployment target defaults # to the OS version, if on x86, and 10.4, the deployment # target defaults to 10.4. Don't you love it? case ${MACOSX_DEPLOYMENT_TARGET-10.0},$host in 10.0,*86*-darwin8*|10.0,*-darwin[91]*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; 10.[012]*) _lt_dar_allow_undefined='${wl}-flat_namespace ${wl}-undefined ${wl}suppress' ;; 10.*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; esac ;; does not account for 10.10 yet? Cheers, Derek -- Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] aquaterm-dev-1.1.1-3 case sensitivity issues
Hi Kevin, > I committed an aquaterm update which I think corrects the problem. But, the > original package validated for me, so I cannot be sure my fix will pass the > validator. yes, no way really to find if it breaks on a case-insensitive fs without building on a case-sensitive one… But everything successful here, thanks for the fix! Derek Validating .deb dir /scratch.noindex/fink.build/root-aquaterm-1.1.1-4... Package looks good! env LANG=C LC_ALL=C dpkg-deb -b root-aquaterm-1.1.1-4 /sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics dpkg-deb: building package `aquaterm' in `/sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics/aquaterm_1.1.1-4_darwin-x86_64.deb'. Reverting ownership of install dir to root - Depends line is: darwin (>= 13-1) Writing control file... Creating shlibs files... Writing md5sums file... Validating .deb dir /scratch.noindex/fink.build/root-aquaterm-shlibs-1.1.1-4... Package looks good! env LANG=C LC_ALL=C dpkg-deb -b root-aquaterm-shlibs-1.1.1-4 /sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics dpkg-deb: building package `aquaterm-shlibs' in `/sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics/aquaterm-shlibs_1.1.1-4_darwin-x86_64.deb'. Reverting ownership of install dir to root - Depends line is: aquaterm-shlibs (= 1.1.1-4), darwin (>= 13-1) Writing control file... Writing md5sums file... Validating .deb dir /scratch.noindex/fink.build/root-aquaterm-dev-1.1.1-4... Package looks good! env LANG=C LC_ALL=C dpkg-deb -b root-aquaterm-dev-1.1.1-4 /sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics dpkg-deb: building package `aquaterm-dev' in `/sw/fink/10.9/stable/main/binary-darwin-x86_64/graphics/aquaterm-dev_1.1.1-4_darwin-x86_64.deb'. -- Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] aquaterm-dev-1.1.1-3 case sensitivity issues
Hi Kevin, >>> Fink Developers - I’m not sure of the best way to fix the problem. I >>> searched man tar, but don’t see any options to deal with this. I could >>> create a new tarball with file names that don’t collide, load that on the >>> web somewhere, and change the source in the info to point to this new >>> tarball. Is there another option to fix this? >> >> This is a kludge, but if there are only a few offending files, you can use >> something like "mv .../ReadMe .../README" to de-duplicate them. (Note that >> the validation warning is *after* the files have been extracted.) >> >> On a case-sensitive filesystem, that would replace one file with another. (I >> assume the contents are the same-- if not, you probably want to repackage >> it.) >> >> On a case-insensitive filesystem, the original extraction process would have >> only left one file, and moving a file to itself fails silently. >> > > I see that there is a field for info files that sound like it solves this > problem. I’ll experiment with this on the weekend when I get back from a > business trip. yes, that’s exactly what it is designed for - I already had a vague feeling I’d been dealing with a similar problem not so long ago even when I reported this problem, sorry for not following up on this right away. Something like TarFilesRename: Readme:Readme.txt README:README.txt should probably do it (chances are the case variants have the same content and can be dumped anyway, at least they all have identical lengths…) HTH, Derek -- Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] aquaterm-dev-1.1.1-3 case sensitivity issues
Hi, FYI building aquaterm with package validation fails for me on a *case sensitive* file system: Validating .deb dir /scratch.noindex/fink.build/root-aquaterm-dev-1.1.1-3... Error: Pathname collision on case-insensitive filesystems Offending file: /sw/share/aquaterm/adapters/c/ReadMe Offending file: /sw/share/aquaterm/adapters/fortran/ReadMe Offending file: /sw/share/aquaterm/adapters/pgplot/ReadMe Offending file: /sw/share/aquaterm/adapters/plplot/Readme Cheers, Derek -- Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] Fwd: locale problems [Astropy v0.4]
Hi, forwarding this here in case there are other Fink packages where this has become an issue (the package in question is on https://sourceforge.net/p/fink/package-submissions/4396/): Begin forwarded message: >> I had a look at the new release and actually ran into a problem right away >> building it on OS X 10.9: >> File >> "/scratch.noindex/fink.build/astropy-py34-0.4-1/astropy-0.4/ah_bootstrap.py", >> line 435, in _check_submodule_using_git >>stderr = stderr.decode(stdio_encoding) >> TypeError: decode() argument 1 must be str, not None > > We've got a known issue for the issue of incorrectly set environment > variables (such as in the MacOS X terminal): > > > http://docs.astropy.org/en/stable/known_issues.html#utf-8-locale-error-on-macos-x > > I've now opened a pull request to add other errors such as the one above > which are due to the same issue: > > https://github.com/astropy/astropy/pull/2760 > > Erik B. has also been working on a workaround that will try and get > around issues of misconfigured locale variables. In the mean time, it > makes sense to set these correctly as shown in the 'known issue' above, > since this is an error that can occur with other packages, not just astropy. If this is a more common problem, maybe it makes sense to add something like SetLC_XXX: fields to the package description (or perhaps even have Fink always set sane defaults). But otoh in the fink build the issue was not LC_CTYPE being set to UTF-8 but rather being not set at all, and this problem might be more specific to astropy. In that case it’s perhaps better to just use the workaround until this is resolved upstream. Cheers, Derek -- Want fast and easy access to all the code in your enterprise? Index and search up to 200,000 lines of code with a free copy of Black Duck Code Sight - the same software that powers the world's largest code search on Ohloh, the Black Duck Open Hub! Try it now. http://p.sf.net/sfu/bds ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] trouble installing emacs24
On 23 Jun 2014, at 01:43 pm, Jesse Alama wrote: >> >> Emacs depends on the pkg 'passwd-games' so that user should normally be set >> up on the system. Try 'fink reinstall passwd-games' and select yes when >> asked to install the user. > > Thanks, that does seem to have made a difference. A fresh rebuild and > reinstall of emacs24 does lead to some curiosities, though: emacs24-24.3-3 now seems to be missing a dependency on gnutls28-shlibs. Cheers, Derek -- HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] [fink:package-submissions] #4416 h5py patch for hdf5 1.8.13
On 22.06.2014, at 1:54AM, Kurt Schwehr wrote: > Thanks for finding the issue. Patch files drive me crazy, so I added the two > changes as a patch script with two perl one liners. > Thanks for the quick fix. Strangely, it did not work on the quite fresh fink installation of my new MacBook; turned out that the C sources are rebuilt if and only if cython is installed (obviously they could not otherwise…). So either h5fd.c would have to be patched directly, or cython-py%type_pkg[python] be added to the BuildDepends. Cheers, Derek -- HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] h5py broken after hdf5.8 update
On 21.06.2014, at 5:44PM, Kurt Schwehr wrote: > :(. I saw the same issue recently, but have not yet had time to investigate > > -Kurt (maintainer of the h5py fink package) Apparently the H5FD_MPIPOSIX macro was not used anyway, so all it took was commenting it out; so hopefully the patch in https://sourceforge.net/p/fink/package-submissions/4416/ should be straightforward to merge. Cheers, Derek -- HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] h5py broken after hdf5.8 update
Hi, h5py-py no longer builds with hdf5.8_1.8.13 due to an obsolete macro (meaning that the current version no longer works either unless enforcing HDF5_DISABLE_VERSION_CHECK). I've submitted a version patched with the appropriate changeset from github master (#4416). Cheers, Derek -- HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] dateutil-py tests require *six*-py to run
On 06.04.2014, at 9:33PM, Derek Homeier wrote: > On 06.04.2014, at 5:58PM, Daniel Macks wrote: > >>> >>> found there is a missing >>> >>> TestDepends: six-py%type_pkg[python] >>> >>> in dateutil-py-2.2-1. >> >> Fixed (sip-pyXXX not six-pyXXX). > > hmm, I had gotten this error: > > sudo -u fink-bld [ENV] sh -c /tmp/fink.67Gmy > /tmp/fink.3vLAc > #!/bin/bash -ev > # -B prevents the test run from generating __pycache__ .pyc files. > /sw/bin/python2.7 -B test.py || exit 2 > Traceback (most recent call last): > File "test.py", line 5, in >from six import StringIO, BytesIO, PY3 > ImportError: No module named six Oh, but I should check my own subject line. I did get confused with the sip-pyXXX-bin update problems… Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] dateutil-py tests require sip-py to run
On 06.04.2014, at 5:58PM, Daniel Macks wrote: >> >> found there is a missing >> >> TestDepends: six-py%type_pkg[python] >> >> in dateutil-py-2.2-1. > > Fixed (sip-pyXXX not six-pyXXX). hmm, I had gotten this error: sudo -u fink-bld [ENV] sh -c /tmp/fink.67Gmy /tmp/fink.3vLAc #!/bin/bash -ev # -B prevents the test run from generating __pycache__ .pyc files. /sw/bin/python2.7 -B test.py || exit 2 Traceback (most recent call last): File "test.py", line 5, in from six import StringIO, BytesIO, PY3 ImportError: No module named six Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] PyQt-mac/x11 source/md5sum error
On 05.04.2014, at 11:44PM, Daniel Johnson wrote: > Well, not quite. The patch doesn't apply. Sigh. I'll have to work on it some > more. I don't use the -x11 package so it doesn't get a lot of attention and I > don't want to spend hours building qt4-x11. It builds with the revised patch now, thanks. I did not particularly care about the x11 version, but lately there were some issues when running the ipython --pylab environment with pyqt4-mac on Mavericks. I'll check if that's still the case with the ipython-2.0.0. Cheers, Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] version conflicts with sip-pyNN-bin
fink update sip-py33-bin Information about 9023 packages read in 1 seconds. The following package will be installed or updated: sip-py33-bin Reading buildlock packages... All buildlocks accounted for. /sw/bin/dpkg-lockwait -i /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/sip-py33-bin_4.15.5-1_darwin-x86_64.deb Selecting previously deselected package sip-py33-bin. (Reading database ... 463436 files and directories currently installed.) Unpacking sip-py33-bin (from .../sip-py33-bin_4.15.5-1_darwin-x86_64.deb) ... /sw/bin/dpkg: error processing /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/sip-py33-bin_4.15.5-1_darwin-x86_64.deb (--install): trying to overwrite `/sw/bin/sip', which is also in package sip-py32-bin /sw/bin/dpkg-deb: subprocess paste killed by signal (Broken pipe: 13) I think sip-py32-bin, sip-py33-bin, sip-py34-bin have to be added to the Conflicts: and Replaces: fields? Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] dateutil-py tests require sip-py to run
Hi, found there is a missing TestDepends: six-py%type_pkg[python] in dateutil-py-2.2-1. Cheers, Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] PyQt-mac/x11 source/md5sum error
Hi, there seems to be some mixup of the source files and checksums for pyqt4: The checksum of the file PyQt-mac-gpl-4.10.4.tar.gz of package pyqt4-py27-4.10.4-1 is incorrect. The most likely cause for this is a corrupted or incomplete download Expected: 20940f6b2c957269cdd02d0efe537515 Actual: MD5(a1d33b15808993055931083813f5f399) SHA1(ef3bb2a05a5c8c3ab7578a0991ef5a4e17c314c0) It is recommended that you download it again. How do you want to proceed? ... The checksum of the file PyQt-mac-gpl-4.10.4.tar.gz of package pyqt4-mac-py27-4.10.4-1 is incorrect. The most likely cause for this is a corrupted or incomplete download Expected: 20940f6b2c957269cdd02d0efe537515 Actual: MD5(a1d33b15808993055931083813f5f399) SHA1(ef3bb2a05a5c8c3ab7578a0991ef5a4e17c314c0) It is recommended that you download it again. How do you want to proceed? (1) Give up (2) Delete it and download again (3) Assume it is a partial download and try to continue (4) Don't download, use existing file Make your choice: [2] 4 Sourceforge also gives the latter (actual) MD5, pyqt4-mac-py27 builds correctly with that; but pyqt4-py27 fails in the patch phase; shouldn't it be using the PyQt-x11-gpl source? Cheers, Derek -- ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Killing python26
On 23.03.2014, at 9:50PM, Daniel Johnson wrote: >>> >>> Does anyone know btw if 3.4 already comes with full setuptools support >>> included? >>> It has a %i/bin/easy_install-3.4, and I was able to build and test >>> coverage, nose >>> and numpy without installing it, so maybe a change is in order to >>> >>> *Depends: ( %type_pkg[python] <= 33 ) setuptools-tng-py%type_pkg[python] >> >> It's not supposed to. :( I thought I'd disabled the installation of >> setuptools and pip but apparently the python build system isn't respecting >> the settings. Python 3.4 started bundling pip (which requires setuptools) as >> a standard packaging system. Unfortunately, that conflicts with fink's >> packaging since we can't know what packages users might install on their >> own. The configure script has a --without-ensurepip option which is supposed >> to not install pip, but it doesn't seem to work. A further problem is that >> installation of pip/setuptools is only performed if it isn't already present >> in site-packages. So the first time you build python34, they get installed. >> But if you rebuild python34 or install a newer version, they DON'T get >> installed but are, of course, overwritten by dpkg when the new package is >> installed. What a mess. >> >> I'm going to have to think about this some more. I either need to always >> install them or never install them but there doesn't seem to be any easy way >> to do that. :( Best to avoid using python34 for now. > > Of course, I could just spell the frickin' option correctly. > --without-ensurepip works a lot better than --without-ensureip. :p > Which did what - downloading packages from random IPs? ;-) > python34 3.4.0-2 now never installs setuptools/pip. I've also added py34 > variants to setuptools-tng-py and pip-py. Continue to depend on setuptools > like always. This way we retain control over the setuptools version separate > from python. Thanks, that seems to work now. Derek -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] Killing python26
On 20.03.2014, at 4:20PM, Daniel Johnson wrote: > I agree. There's no reason to keep 2.6 around as 2.7 is fully backward > compatible. > Just an aside, anyone aware of any major Linux distros that still only provide 2.6? This was an issue for many packages like numpy to keep support for older versions, since some places like large companies or computing centres would only upgrade on something like 5-year cycles; this in turn would be a potential incentive for developers to keep a 2.6 installation for testing purposes. Not saying that this has to be provided by fink though if it's too much maintenance. After all one could relatively easily maintain private copies in the local branch, if really needed. > It's also reasonable to kill 3.1 as it's now unsupported upstream. 3.3 and > the just released 3.4 will stay of course, but we might also consider killing > 3.2 even though it still gets patches. 3.3 introduced better backward > compatibility with 2.7 to ease porting to 3.x and vastly improved unicode > string handling. > > There are no language changes between 3.3 and 3.4, just new and updated > library modules, so if maintainers can add 3.4 variants to their packages > that would be awesome. Big +1 to get rid of 3.1! Does anyone know btw if 3.4 already comes with full setuptools support included? It has a %i/bin/easy_install-3.4, and I was able to build and test coverage, nose and numpy without installing it, so maybe a change is in order to *Depends: ( %type_pkg[python] <= 33 ) setuptools-tng-py%type_pkg[python] Cheers, Derek -- Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] astropy-0.3
Hi, the Fink astropy-py package is two minor versions out of date now; I could not wait to test it myself though. I do not remember the currently recommended procedure to suggest new packages or updates (did not find the link to the package tracker that once existed any more…), so I' just directly sending the files here. I made a few additional modifications vs. 0.2.4: - added Test section - explicitly added setuptools-tng to BuildDepends; with no setuptools/distribute installed, setup.py would download an older s'tools version on its own, and currently there is only a very indirect dependence via numpy->nose->setuptools-tng; but nose is not really required to use numpy, so this seemed safer - added sphinx and the documentation build under Python2.x only (tried unsuccessfully to build the docs with sphinx-1.2b3-py33); the html docs roughly double the installed size I am also throwing in a Python3 flavour of h5py, all built and tested (fink -m) with python2.6-3.3 under 10.5 and 10.8 (except for a couple of errors with python3.1). Cheers, Derek -- -------- Derek Homeier Centre de Recherche Astrophysique de Lyon ENS Lyon 46, Allée d'Italie 69364 Lyon Cedex 07, France +33 47272-8894 astropy-py.info Description: Binary data h5py-py.info Description: Binary data -- Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349351&iu=/4140/ostg.clktrk___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] perl5162-core build/install problem on case-sensitiveFS
On 20.11.2013, at 2:29AM, Hanspeter Niederstrasser wrote: > On 11/19/2013 8:04 PM, Derek Homeier wrote: >> But a lot of modules fail to build with this missing dependency on >> cc_runtime.h. >> There is no cc_runtime.h of either capitalisation in >> /sw/lib/perl5-core/5.16.2/darwin-thread-multi-2level/CORE >> nor anywhere in the perl-5.16.2 source. >> >> Thanks for the help so far! >> >> Derek >> >> ARCHFLAGS="" /usr/bin/arch -x86_64 perl5.16.2 Makefile.PL >> PERL="/usr/bin/arch -x86_64 perl5.16.2" PREFIX=/sw >> INSTALLPRIVLIB=/sw/lib/perl5/5.16.2 >> INSTALLARCHLIB=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level >> INSTALLSITELIB=/sw/lib/perl5/5.16.2 >> INSTALLSITEARCH=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level >> INSTALLMAN1DIR=/sw/share/man/man1 INSTALLMAN3DIR=/sw/share/man/man3 >> INSTALLSITEMAN1DIR=/sw/share/man/man1 INSTALLSITEMAN3DIR=/sw/share/man/man3 >> INSTALLBIN=/sw/bin INSTALLSITEBIN=/sw/bin INSTALLSCRIPT=/sw/bin >> Use of qw(...) as parentheses is deprecated at >> /sw/lib/perl5/ExtUtils/MM_Any.pm line 987. >> Testing if you have a C compiler >> Checking if your kit is complete... >> Looks good >> Writing Makefile for version::vxs >> Writing Makefile for version >> make CC=gcc CXX=g++ >> make[1]: Entering directory >> `/scratch.noindex/fink.build/version-pm5162-0.99.04-1/version-0.9904/vutil' >> cp lib/version/Internals.pod blib/lib/version/Internals.pod >> cp lib/version.pm blib/lib/version.pm >> cp lib/version.pod blib/lib/version.pod >> cp vperl/vpp.pm blib/lib/version/vpp.pm >> make[1]: *** No rule to make target >> `/sw/lib/perl5-core/5.16.2/darwin-thread-multi-2level/CORE/cc_runtime.h', >> needed by `vutil.o'. Stop. >> make[1]: *** Waiting for unfinished jobs > > Do you have extutils-makemaker-pm [1] installed? If yes, remove it and try > to rebuild your failing perlmods. > > Apparently the old unvarianted extutils-makemaker-pm conflicts with the > EU::MM that comes with perl5.16.2. Yep, worked! Thanks again, Derek -- Shape the Mobile Experience: Free Subscription Software experts and developers: Be at the forefront of tech innovation. Intel(R) Software Adrenaline delivers strategic insight and game-changing conversations that shape the rapidly evolving mobile landscape. Sign up now. http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] perl5162-core build/install problem on case-sensitiveFS
On 19.11.2013, at 8:36PM, Hanspeter Niederstrasser wrote: > On 11/19/2013 1:55 PM, Derek Homeier wrote: >> On 19.11.2013, at 7:47PM, Hanspeter Niederstrasser >> wrote: >> >>> The 'missing' perl5.16 should be fixed in fink-0.36.1, released two days >>> ago. Fink was incorrectly assuming that all perls were perlX.Y, but this >>> is only the case for system-perl, while fink's perls have always been >>> perlX.Y.Z. >> >> Hmm, could there be any delays with selfupdate-rsync? I did several over the >> weekend and another one just now, >> I've already been wondering about a bzr-git update that did not seem to show >> up: >> >> /usr/bin/rsync -az -v >> rsync://distfiles.master.finkmirrors.net/finkinfo//TIMESTAMP >> /sw/fink/TIMESTAMP.tmp >> receiving file list ... done >> TIMESTAMP > > Currently, the only mirror that happens to be updating is this one: > > rsync://hnd.jp.asi.finkmirrors.net/finkinfo/ > > You can edit the "Mirror-rsync" field in /sw/etc/fink.conf to use that. I really got to go out g my way here ;-) - OK, that did it, after waiting for an extended time for http://downloads.sourceforge.net/fink which seems also to be only one to have updated the fink source! A lot of the modules build now, though perl5162, which also includes /sw/bin/json-pp, still conflicts with json-pp-pm. But a lot of modules fail to build with this missing dependency on cc_runtime.h. There is no cc_runtime.h of either capitalisation in /sw/lib/perl5-core/5.16.2/darwin-thread-multi-2level/CORE nor anywhere in the perl-5.16.2 source. Thanks for the help so far! Derek ARCHFLAGS="" /usr/bin/arch -x86_64 perl5.16.2 Makefile.PL PERL="/usr/bin/arch -x86_64 perl5.16.2" PREFIX=/sw INSTALLPRIVLIB=/sw/lib/perl5/5.16.2 INSTALLARCHLIB=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level INSTALLSITELIB=/sw/lib/perl5/5.16.2 INSTALLSITEARCH=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level INSTALLMAN1DIR=/sw/share/man/man1 INSTALLMAN3DIR=/sw/share/man/man3 INSTALLSITEMAN1DIR=/sw/share/man/man1 INSTALLSITEMAN3DIR=/sw/share/man/man3 INSTALLBIN=/sw/bin INSTALLSITEBIN=/sw/bin INSTALLSCRIPT=/sw/bin Use of qw(...) as parentheses is deprecated at /sw/lib/perl5/ExtUtils/MM_Any.pm line 987. Testing if you have a C compiler Checking if your kit is complete... Looks good Writing Makefile for version::vxs Writing Makefile for version make CC=gcc CXX=g++ make[1]: Entering directory `/scratch.noindex/fink.build/version-pm5162-0.99.04-1/version-0.9904/vutil' cp lib/version/Internals.pod blib/lib/version/Internals.pod cp lib/version.pm blib/lib/version.pm cp lib/version.pod blib/lib/version.pod cp vperl/vpp.pm blib/lib/version/vpp.pm make[1]: *** No rule to make target `/sw/lib/perl5-core/5.16.2/darwin-thread-multi-2level/CORE/cc_runtime.h', needed by `vutil.o'. Stop. make[1]: *** Waiting for unfinished jobs cp lib/version/vxs.pm ../blib/lib/version/vxs.pm make[1]: Leaving directory `/scratch.noindex/fink.build/version-pm5162-0.99.04-1/version-0.9904/vutil' make: *** [subdirs] Error 2 ### execution of make failed, exit code 2 ### execution of /tmp/fink.RTgvH failed, exit code 2 Removing runtime build-lock... Removing build-lock package... /sw/bin/dpkg-lockwait -r fink-buildlock-version-pm5162-0.99.04-1 (Reading database ... 438233 files and directories currently installed.) Removing fink-buildlock-version-pm5162-0.99.04-1 ... Updating the list of locally available binary packages. Scanning dists/stable/main/binary-darwin-x86_64 New package: dists/stable/main/binary-darwin-x86_64/libs/perlmods/sub-install-pm5162_0.926-1_darwin-x86_64.deb New package: dists/stable/main/binary-darwin-x86_64/libs/perlmods/test-simple-pm5162_1.001002-1_darwin-x86_64.deb Failed: phase compiling: version-pm5162-0.99.04-1 failed -- Shape the Mobile Experience: Free Subscription Software experts and developers: Be at the forefront of tech innovation. Intel(R) Software Adrenaline delivers strategic insight and game-changing conversations that shape the rapidly evolving mobile landscape. Sign up now. http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] perl5162-core build/install problem on case-sensitiveFS
On 19.11.2013, at 7:47PM, Hanspeter Niederstrasser wrote: > The 'missing' perl5.16 should be fixed in fink-0.36.1, released two days ago. > Fink was incorrectly assuming that all perls were perlX.Y, but this is only > the case for system-perl, while fink's perls have always been perlX.Y.Z. Hmm, could there be any delays with selfupdate-rsync? I did several over the weekend and another one just now, I've already been wondering about a bzr-git update that did not seem to show up: /usr/bin/rsync -az -v rsync://distfiles.master.finkmirrors.net/finkinfo//TIMESTAMP /sw/fink/TIMESTAMP.tmp receiving file list ... done TIMESTAMP sent 59 bytes received 140 bytes 30.62 bytes/sec total size is 11 speedup is 0.06 I will now run the rsync command to retrieve the latest package descriptions. /usr/bin/rsync -rtz --delete-after --delete -v --include='10.7/' --include='10.7/stable/' --include='10.7/stable/main/' --include='10.7/stable/main/finkinfo/' --include='10.7/stable/main/finkinfo/*/' --include='10.7/stable/main/finkinfo/*' --include='10.7/stable/main/finkinfo/**/*' --include='VERSION' --include='DISTRIBUTION' --include='README' --exclude='**' 'rsync://distfiles.master.finkmirrors.net/finkinfo' '/sw/fink/' receiving file list ... done ./ 10.7/ sent 287 bytes received 128161 bytes 15111.53 bytes/sec total size is 13522177 speedup is 105.27 Scanning package description files.. Information about 8932 packages read in 2 seconds. No packages to install. The core packages have been updated. You should now update the other packages using commands like 'fink update-all'. miranda:3306> fink list fink Information about 8932 packages read in 2 seconds. i fink0.36.0-71Open-source software package manager fink-buildenv-modules 0.1.0-4 Script for getting common buildtime values i fink-mirrors0.34.4.1-1 Mirror infrastructure i fink-obsolete-packages 0.1-1Base package for obsolete packages fink-octave-scripts 0.2.0-2 Convenience scripts for Octave add-ons i fink-package-precedence 0.17-1 Check fink masking of system libs i passwd-fink-bld 20131113-1 OBSOLETE User/group for fink build system miranda:3307> fink list bzr-git Information about 8932 packages read in 2 seconds. bzr-git-py260.6.11-1 Clone git repositories with Bazaar i bzr-git-py270.6.11-1 Clone git repositories with Bazaar Thanks, Derek -- Shape the Mobile Experience: Free Subscription Software experts and developers: Be at the forefront of tech innovation. Intel(R) Software Adrenaline delivers strategic insight and game-changing conversations that shape the rapidly evolving mobile landscape. Sign up now. http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
Re: [Fink-devel] perl5162-core build/install problem on case-sensitiveFS
Hi Dan, thanks for the pointer, this seems to be a bigger can of worms than I've thought! >> It would seem, since the .deb contains both the lower- and uppercase version >> dirs with their content, >> it should install just fine on a case-sensitive filesystem, but apparently >> dpkg-deb is still refusing to >> install it (even in non-maintainer mode), because the package would be >> broken on a case-insensitive fs. >> I suppose the proper fix would be to manually rename either "version" or >> "Version" before/during/after >> the build process to have Requirements.pm and Internals.pod coexist in the >> same subdir? > > It's broken in several ways, including .deb portability and (in this case) > perl functionality relating to filename<->namespace mapping. Perl has fought > with this issue for years, and even this specific instance of it. Apparently > the .pod is the remnant documentation regarding the actual perl-module fix > (maybe there had been a version/Internals.pm that was renamed). See: > > http://www.nntp.perl.org/group/perl.perl5.porters/2011/06/msg173951.html > > that mentions the idea of moving the contents of the .pod into the > documentation for another module. That thread talks about it being finally > fixed by perl5.18ish, I wonder if perl more recent than 5.16.2 has any > improvements in this regard. Indeed, so renaming the directories in the source tree does not help, as perl immediately complains at the beginning of the build. So I tried the less nice approach of fixing things up afterwards, hoping, as the list mail above indicated, that this would not mess up namespace things, but I am certainly a bit over my head into this… With the following changes @@ -68,6 +68,7 @@ << InstallScript: << +#! /bin/sh -ev make install DESTDIR=%d mkdir -p %i/etc/profile.d @@ -77,6 +78,9 @@ echo "append_path MANPATH %p/lib/perl5/5.16.2/man" >> %i/etc/profile.d/%N-core.sh echo "export MANPATH" >> %i/etc/profile.d/%N-core.sh chmod 755 %i/etc/profile.d/%N-core.* + if [ ! -f %i/lib/perl5-core/%v/Version/Internals.pod ]; then + mv %i/lib/perl5-core/%v/version/Internals.pod %i/lib/perl5-core/%v/Version && rmdir %i/lib/perl5-core/%v/version + fi << SplitOff: << perl5162-core and perl5162 build and install in maintainer mode (though the latter has a name conflict with json-pp-pm - is that package unversioned on purpose?). But building any of the modules (e.g. test-harness-pm5162) fails at the step ARCHFLAGS="" /usr/bin/arch -x86_64 perl5.16 Makefile.PL PERL="/usr/bin/arch -x86_64 perl5.16" PREFIX=/sw INSTALLPRIVLIB=/sw/lib/perl5/5.16.2 INSTALLARCHLIB=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level INSTALLSITELIB=/sw/lib/perl5/5.16.2 INSTALLSITEARCH=/sw/lib/perl5/5.16.2/darwin-thread-multi-2level INSTALLMAN1DIR=/sw/share/man/man1 INSTALLMAN3DIR=/sw/share/man/man3 INSTALLSITEMAN1DIR=/sw/share/man/man1 INSTALLSITEMAN3DIR=/sw/share/man/man3 INSTALLBIN=/sw/bin INSTALLSITEBIN=/sw/bin INSTALLSCRIPT=/sw/bin arch: Can't find perl5.16 in PATH (there is /sw/bin/perl5.16.2 in PATH, and the equivalent works for "arch -x86_64 perl5.12" vs. /sw/bin/perl5.12.3; "/usr/bin/arch -x86_64 perl5.16.2" apparently works as well, but here not without the full versioned name). I don't know if this is somehow due to Internals.pod now possibly not being found - it seems to contain some versioning information, but I really don't know anything about perl internals at this level… Derek -- Shape the Mobile Experience: Free Subscription Software experts and developers: Be at the forefront of tech innovation. Intel(R) Software Adrenaline delivers strategic insight and game-changing conversations that shape the rapidly evolving mobile landscape. Sign up now. http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel
[Fink-devel] perl5162-core build/install problem on case-sensitive FS
It seems I am running into troubles with my fink installation on a case-sensitive volume again, but I don't quite understand the issue. perl5162-core builds ok so far, but the package fails to install with Validating package file /sw/fink/dists/stable/main/finkinfo/languages/perl5162.info... Package looks good! The package 'perl5162-core' will be installed. Reading dependency for perl5162-core-5.16.2-102... Reading runtime dependency for perl5162-core-5.16.2-102... The following package will be installed or updated: perl5162-core Reading buildlock packages... All buildlocks accounted for. /sw/bin/dpkg-lockwait -i /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/perl5162-core_5.16.2-102_darwin-x86_64.deb (Reading database ... 438260 files and directories currently installed.) Unpacking perl5162-core (from .../perl5162-core_5.16.2-102_darwin-x86_64.deb) ... /sw/bin/dpkg: error processing /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/perl5162-core_5.16.2-102_darwin-x86_64.deb (--install): unable to create `./sw/lib/perl5-core/5.16.2/version/Internals.pod': No such file or directory Errors were encountered while processing: /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/perl5162-core_5.16.2-102_darwin-x86_64.deb ### execution of /sw/bin/dpkg-lockwait failed, exit code 1 Updating the list of locally available binary packages. Scanning dists/stable/main/binary-darwin-x86_64 New package: dists/stable/main/binary-darwin-x86_64/libs/perlmods/package-generator-pm_1.105-1_darwin-x86_64.deb Failed: can't install package perl5162-core-5.16.2-102 > dpkg-deb -c > /sw/fink/dists/stable/main/binary-darwin-x86_64/languages/perl5162-core_5.16.2-102_darwin-x86_64.deb|grep > -i version/ > drwxr-xr-x root/admin0 2013-11-12 16:02 ./sw/lib/perl5-core/5.16.2/Version/ -r--r--r-- root/admin14605 2013-11-12 16:02 ./sw/lib/perl5-core/5.16.2/Version/Requirements.pm drwxr-xr-x root/admin0 2013-11-12 16:02 ./sw/lib/perl5-core/5.16.2/version/ -r--r--r-- root/admin25001 2013-11-12 16:02 ./sw/lib/perl5-core/5.16.2/version/Internals.pod Building the package in maintainer mode however results in an error in the validation phase: Validating .deb dir /Volumes/Oberon/scratch.noindex/fink.build/root-perl5162-5.16.2-102... Package looks good! env LANG=C LC_ALL=C dpkg-deb -b root-perl5162-5.16.2-102 /sw/fink/10.8/stable/main/binary-darwin-x86_64/languages dpkg-deb: building package `perl5162' in `/sw/fink/10.8/stable/main/binary-darwin-x86_64/languages/perl5162_5.16.2-102_darwin-x86_64.deb'. Reverting ownership of install dir to root - Depends line is: darwin (>= 12-1) Writing control file... Writing md5sums file... Validating .deb dir /Volumes/Oberon/scratch.noindex/fink.build/root-perl5162-core-5.16.2-102... Error: Pathname collision on case-insensitive filesystems Offending file: /sw/lib/perl5-core/5.16.2/version/ Removing runtime build-lock... Removing build-lock package... /sw/bin/dpkg-lockwait -r fink-buildlock-perl5162-5.16.2-102 (Reading database ... 438261 files and directories currently installed.) Removing fink-buildlock-perl5162-5.16.2-102 ... Failed: phase .deb validation: perl5162-core-5.16.2-102 failed It would seem, since the .deb contains both the lower- and uppercase version dirs with their content, it should install just fine on a case-sensitive filesystem, but apparently dpkg-deb is still refusing to install it (even in non-maintainer mode), because the package would be broken on a case-insensitive fs. I suppose the proper fix would be to manually rename either "version" or "Version" before/during/after the build process to have Requirements.pm and Internals.pod coexist in the same subdir? Cheers, Derek -- DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access Free app hosting. Or install the open source package on any LAMP server. Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native! http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk ___ Fink-devel mailing list Fink-devel@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.apple.fink.devel Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-devel