Git Product home page Git Product logo

fink-distributions's Introduction

Fink ReadMe

This is Fink, a package management system that aims to bring the full world of Open Source software to Darwin and Mac OS X.

With the help of dpkg, it maintains a separate directory hierarchy. It downloads original source releases, patches them if necessary, configures them for Darwin and compiles and installs them. The information about available packages and the necessary patches (the "package descriptions") are maintained separately, but are usually included with this distribution. The actual source code is downloaded from the Internet as necessary.

Although Fink cannot be considered "mature" and it has some rough edges and lacking features, it is successfully used by a large number of people. Please read the instructions carefully and don't be surprised if something doesn't work as expected. There are good explanations for most failures; check the website if you need help.

Fink is released under the terms of the GNU General Public License. See the file COPYING for details.

Requirements

You need:

  • An installed Mac OS X system, version 10.9 or later. Earlier versions will not work with fink-0.39.2

  • The Xcode Command Line Tools are mandatory. This package can be installed either by downloading it directly via developer.apple.com, or by running the xcode-select --install command and choosing the Install button in the window that pops up. You may also need to use this command to update the tools, especially if you're having build problems.

    If you're doing a manual download, make sure that the tools you install match your Mac OS X version.

    Some packages require the full Xcode.

  • Java. Entering javac from a Terminal.app window should suffice to make the system download it for you, (10.7, 10.9) or open up a browser window from which you can downloadthe latest Java JDK (10.8).

  • Many other things that come with Mac OS X and the Developer Tools. This includes perl and curl.

  • Internet access. All source code is downloaded from mirror sites.

  • Patience. Compiling several big packages takes time. I'm talking hours or even days here.

Installation

The installation process is described in detail in the file INSTALL. Please read it first, the process is non-trivial. It also describes the upgrade procedure.

Using Fink

The file USAGE describes how to set your paths and how to install and remove packages. It also has a complete list of available commands.

Further Questions?

If the documentation included here doesn't answer your question, stroll over to the Fink website at https://www.finkproject.org/ and check out the Help page there: https://www.finkproject.org/help/. It will point you at the other documentation that is available and sources for support if you need it.

If you'd like to contribute to Fink, the Help page mentioned above also has a list of things you can do, like testing or creating packages.

Staying Informed

The project's website is at https://www.finkproject.org/.

To be informed of new releases, go to https://www.finkproject.org/lists/fink-announce.php and subscribe to the fink-announce mailing list. The list is moderated and low-traffic.

fink-distributions's People

Contributors

akhansen avatar babayoshihiko avatar bcully avatar beren12 avatar clepple avatar costabel avatar danielj7 avatar dhomeier avatar dmacks avatar dmalloc avatar eighthave avatar fangism avatar fingolfin avatar gecko2 avatar htodd avatar jwhowarth avatar kamischi avatar khorton avatar leighsmith avatar mommsen avatar nieder avatar pogma avatar povlabrahamsen avatar remkos avatar schwehr avatar smaret avatar sth0 avatar thesin- avatar wgscott avatar wodev avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

fink-distributions's Issues

Stable links to java's version-dependent chaos

I wonder if we should roll a set of system-javaXX-dev that make symlinks from stable %p locations to wherever apple/oracle decide to put it each time. Then that package can have a "fink-java-check" script that verifies the links point to actual places (or abort with "please install fink's system-javaXX-dev package that matches your system's java version" if not). This would centralize all the chaos and not require every package that uses java to have so much spaghetti and require an update each time there's a new java. Either those packages could all Provides:system-java-dev or there could be a real bundle by that name on top of them. Any package that wants java would just:

BDep:system-java-dev
SetCPPFLAGS: -I%p/include/system-java
SetLDFLAGS: -L%p/lib/system-java
CompileScript: <<
fink-java-check
...
<<

or if we need more variability (-framework vs -l flags?) could roll this into fink-buildenv-modules (@nieder) that could supply JAVA_{CPPFLAGS,LDFLAGS} or similar variables.

qgis17-py through qgis26-py are unbuildable with current PyQt4-4.11

@babayoshihiko

With PyQt4-4.11, PyQt4.pyqtconfig was removed http://pyqt.sourceforge.net/Docs/PyQt4/build_system.html#module-PyQt4.pyqtconfig. This causes problems for qgis{17,18,20,22,24,26) to find PyQt4 (qgis28-py works).

-- Found SIP version: 4.18
Traceback (most recent call last):
  File "/sw/src/fink.build/qgis18-nograss-py27-1.8.0-11/qgis-1.8.0/cmake/FindPyQt.py", line 5, in <module>
    import PyQt4.pyqtconfig
ImportError: No module named pyqtconfig
-- Found GSL: -L/sw/lib -lgsl -lgslcblas

And then leads to build errors:
qgis17-py: /sw/src/fink.build/qgis17-py27-1.7.4-9/qgis-1.7.4/src/core/spatialindex/tools/TemporaryFile.cc:52:3: error: use of undeclared identifier 'close' close( fd );
qgis18-py: sip: Unable to find file "QtCore/QtCoremod.sip"
qgis20-py: /sw/build.build/qgis20-py27-2.0.1-6/qgis-2.0.1/build/python/gui/sipguipart2.cpp:80202:30: error: 'hideEvent' is a private member of 'QgsMessageLogViewer'
Given that these 6 packages are marked as obsolete, and they're unbuildable with the current state of the 10.9 distribution, they should probably be removed from git. Nothing depends on them.

Scrap qgis less than 218

Following from Issue #99, we now have qgis28, qgis210, qgis212, qgis214, qgis216, qgis218 (up through qgis26 having been scrapped). On 10.12/10.13 platforms, qgis210, qgis212, qgis214, qgis216 are unbuildable because they depend on grass70, which is Distribution-restricted due to needing wxwidgets300-cocoa (which is Distribution-restricted due to needing QTKit). On the other hand, qgis218 seems available on all platforms because it uses grass72, which uses non-restricted wx* packages. The oldest we have, qgis28, uses the older grass64, which is still apparently available up through 10.13, but is one of only two packages still using grass64.

Rather than retaining but dist-restricting qgis210, qgis212, qgis214, qgis216 or hacking on them to use a newer widget-kit, I propose scrapping them altogether. And I also propose scrapping qgis28. No sense dragging along ancient versions and accumulating cruft in them, and having to maintain old versions of other packages as their dependencies.

brotli-py needs versioned dep on libbrotli

brotli-py27-1.0.5-1 builds ok with libbrotli1-shlibs-1.0.5 (even if the dev is not installed). If I have libbrotli1-shlibs-1.0.3 installed, I get this failure:

gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -Wstrict-prototypes -I/sw/include -DOS_MACOSX=1 -Ic/include -I/sw/include/python2.7 -c c/dec/bit_reader.c -o bin/temp.macosx-10.11-x86_64-2.7/c/dec/bit_reader.o
In file included from c/dec/bit_reader.c:9:
In file included from c/dec/./bit_reader.h:14:
c/dec/../common/platform.h:72:28: error: token is not a valid binary operator in a preprocessor subexpression
#if BROTLI_GNUC_HAS_BUILTIN(__builtin_expect, 3, 0, 0) || \
    ~~~~~~~~~~~~~~~~~~~~~~~^
c/dec/../common/platform.h:118:30: error: token is not a valid binary operator in a preprocessor subexpression
#if BROTLI_GNUC_HAS_ATTRIBUTE(always_inline, 4, 0, 0) ||                       \
    ~~~~~~~~~~~~~~~~~~~~~~~~~^
c/dec/../common/platform.h:126:32: error: token is not a valid binary operator in a preprocessor subexpression
#elif BROTLI_MSVC_VERSION_CHECK(12, 0, 0)
      ~~~~~~~~~~~~~~~~~~~~~~~~~^
c/dec/../common/platform.h:128:30: error: token is not a valid binary operator in a preprocessor subexpression
#elif BROTLI_TI_VERSION_CHECK(7, 0, 0) && defined(__cplusplus)
      ~~~~~~~~~~~~~~~~~~~~~~~^

fink-bld SHELL conflicts with numpy.distutils

Hi,

this problem has come up a couple of times in the past months, but tends to reappear (see links below)...
The issue, briefly, is that numpy's own distutils module, and any package using numpy.distutils, by default spawns all subprocesses in the shell $SHELL (in numpy/distutils/exec_command.py):

    if os.name == 'posix' and use_shell:
        # On POSIX, subprocess always uses /bin/sh, override
        sh = os.environ.get('SHELL', '/bin/sh')
        if is_sequence(command):
            command = [sh, '-c', ' '.join(command)]
        else:
            command = [sh, '-c', command]

Any suggestions how to deal with this in a more sustainable way (and if it's just adding a new SetSHELL field for convenience)? It's hard to track down if you are not aware of the problem, as even

DISTUTILS_DEBUG=1 %p/bin/python2.7 -d setup.py build

only gives you a vague pointer to the source of the failure.
I am considering reporting the issue upstream so numpy will at least test explicitly if os.environ.get('SHELL') produces a working command, but they might respond that building with SHELL=/usr/bin/false is a really bad idea™...

gnuplot-*-qtmac needs C++11 support

On 10.11/Xcode7.3 and 10.13/Xcode9, all the qtmac variants of gnuplot needs -std=c++11 added to CXXFLAGS. Adding this to CXXFLAGS needs to be variant restricted because the qtx11 (with qt4) variants fail when using the c++11 flag. The WX variant used doesn't seem to matter.

	c++ -DHAVE_CONFIG_H -I. -I..  -I../term -I../term -DBINDIR=\"/sw/bin\" -DX11_DRIVER_DIR=\"/sw/lib/gnuplot/5.0\" -DQT_DRIVER_DIR=\"/sw/lib/gnuplot/5.0\" -DGNUPLOT_SHARE_DIR=\"/sw/share/gnuplot/5.0\" -DGNUPLOT_PS_DIR=\"/sw/share/gnuplot/5.0/PostScript\" -DGNUPLOT_JS_DIR=\"/sw/share/gnuplot/5.0/js\" -DGNUPLOT_LUA_DIR=\"/sw/share/gnuplot/5.0/lua\" -DCONTACT=\"[email protected]\" -DHELPFILE=\"/sw/share/gnuplot/5.0/gnuplot.gih\" -DGNUPLOT_X11=\"`echo gnuplot_x11 | sed 's,x,x,'`\" -DXAPPLRESDIR=\"/sw/etc/app-defaults\" -DGNUPLOT_QT=\"`echo gnuplot_qt | sed 's,x,x,'`\" -DQTGNUPLOT_DATA_DIR=\"/sw/share/gnuplot/5.0/qt\" -F/sw/Library/Frameworks -I/sw/include   -I/opt/X11/include -I/sw/include -I/sw/include  -I/sw/include/glib-2.0 -I/sw/lib/glib-2.0/include -I/sw/include/pango-1.0 -I/sw/include/cairo -I/sw/include/freetype2 -I/opt/X11/include -I/sw/include/glib-2.0 -I/sw/lib/glib-2.0/include -I/sw/include/cairo -I/sw/include/pango-1.0 -I/sw/include/cairo -I/sw/include/freetype2 -I/opt/X11/include -F/sw/lib/qt5-mac/lib -I/sw/lib/qt5-mac/lib/QtCore.framework/Headers -I/sw/lib/qt5-mac/lib/QtGui.framework/Headers -I/sw/lib/qt5-mac/lib/QtWidgets.framework/Headers -I/sw/lib/qt5-mac/lib/QtNetwork.framework/Headers -I/sw/lib/qt5-mac/lib/QtSvg.framework/Headers  -g -O2 -I/sw/lib/wxwidgets300-gtk2/wx/include/gtk2-unicode-3.0 -I/sw/include/wx-3.0 -D_FILE_OFFSET_BITS=64 -DWXUSINGDLL -D__WXGTK__  -I/sw/include/glib-2.0 -I/sw/lib/glib-2.0/include -I/sw/include/pango-1.0 -I/sw/include/cairo -I/sw/include/freetype2 -I/opt/X11/include -I/sw/include/glib-2.0 -I/sw/lib/glib-2.0/include -I/sw/include/gtk-2.0 -I/sw/lib/gtk-2.0/include -I/sw/include/gtk-2.0 -I/sw/include/pango-1.0 -I/sw/include/atk-1.0 -I/sw/include/cairo -I/sw/include/pango-1.0 -I/sw/include/freetype2 -I/opt/X11/include -fPIC  -MT qtterminal/gnuplot_qt.o -MD -MP -MF $depbase.Tpo -c -o qtterminal/gnuplot_qt.o qtterminal/gnuplot_qt.cpp &&\
	mv -f $depbase.Tpo $depbase.Po
1 warning and 20 errors generated.
In file included from qtterminal/gnuplot_qt.cpp:44:
In file included from qtterminal/QtGnuplotApplication.h:47:
In file included from qtterminal/QtGnuplotEvent.h:47:
In file included from /sw/lib/qt5-mac/lib/QtCore.framework/Headers/QObject:1:
In file included from /sw/lib/qt5-mac/lib/QtCore.framework/Headers/qobject.h:46:
In file included from /sw/lib/qt5-mac/lib/QtCore.framework/Headers/qobjectdefs.h:48:
In file included from /sw/lib/qt5-mac/lib/QtCore.framework/Headers/qnamespace.h:43:
In file included from /sw/lib/qt5-mac/lib/QtCore.framework/Headers/qglobal.h:83:
/sw/lib/qt5-mac/lib/QtCore.framework/Headers/qcompilerdetection.h:562:6: error: Qt requires a
      C++11 compiler and yours does not seem to be that.
#    error Qt requires a C++11 compiler and yours does not seem to be that.
     ^
make[4]: *** [qtterminal/qt_term.o] Error 1

python37 build failures (conflict with pkg uuid)

python37 conflicts with the package uuid.

failure with uuid package installed:

checking uuid/uuid.h usability... yes
checking uuid/uuid.h presence... yes
checking for uuid/uuid.h... yes
checking uuid.h usability... no
checking uuid.h presence... yes
configure: WARNING: uuid.h: present but cannot be compiled
configure: WARNING: uuid.h:     check for missing prerequisite headers?
configure: WARNING: uuid.h: see the Autoconf documentation
configure: WARNING: uuid.h:     section "Present But Cannot Be Compiled"
configure: WARNING: uuid.h: proceeding with the compiler's result
configure: WARNING:     ## --------------------------------------- ##
configure: WARNING:     ## Report this to https://bugs.python.org/ ##
configure: WARNING:     ## --------------------------------------- ##
checking for uuid.h... no
checking for uuid_generate_time_safe... no
checking for uuid_create... yes
checking for uuid_enc_be... no
....
building '_uuid' extension
gcc -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -std=c99 -Wextra -Wno-unused-result -Wno-unused-parameter -Wno-missing-field-initializers -Wstrict-prototypes -Werror=implicit-function-declaration -I./Include -I. -I/sw/include -I/sw/build.build/python37-3.7.0-1/Python-3.7.0/Include -I/sw/build.build/python37-3.7.0-1/Python-3.7.0 -c /sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.c -o build/temp.macosx-10.11-x86_64-3.7/sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.o
/sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.c:23:5: error: implicit
      declaration of function 'uuid_create' is invalid in C99
      [-Werror,-Wimplicit-function-declaration]
    uuid_create(&uuid, &status);
    ^
/sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.c:23:5: note: did you mean
      'uuid_clear'?
/usr/include/uuid/uuid.h:53:6: note: 'uuid_clear' declared here
void uuid_clear(uuid_t uu);
     ^
1 error generated.

Success with uuid not present:

checking uuid/uuid.h usability... yes
checking uuid/uuid.h presence... yes
checking for uuid/uuid.h... yes
checking uuid.h usability... no
checking uuid.h presence... no
checking for uuid.h... no
checking for uuid_generate_time_safe... no
checking for uuid_create... no
checking for uuid_enc_be... no
....
building '_uuid' extension
gcc -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -std=c99 -Wextra -Wno-unused-result -Wno-unused-parameter -Wno-missing-field-initializers -Wstrict-prototypes -Werror=implicit-function-declaration -I/usr/include/uuid -I./Include -I. -I/sw/include -I/sw/build.build/python37-3.7.0-1/Python-3.7.0/Include -I/sw/build.build/python37-3.7.0-1/Python-3.7.0 -c /sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.c -o build/temp.macosx-10.11-x86_64-3.7/sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.o
gcc -bundle /sw/build.build/python37-3.7.0-1/Python-3.7.0/libpython3.7m.dylib -L/sw/lib -L/sw/lib -L/sw/lib -I/sw/include build/temp.macosx-10.11-x86_64-3.7/sw/build.build/python37-3.7.0-1/Python-3.7.0/Modules/_uuidmodule.o -L. -L/sw/lib -o build/lib.macosx-10.11-x86_64-3.7/_uuid.cpython-37m-darwin.so

freetype's debian diff has changed

Our .info looks for freetype_2.6.3-3.diff.gz, but upstream updated to freetype_2.6.3-3.2.diff.gz and our's has been deleted from their servers (obv still in our mirrors).

From the changelog, it fixed 3 CVEs

CVE-2016-10244-type1-Fix-heap-buffer-overflow.patch
CVE-2017-8105-psaux-Better-protect-flex-handling.patch
CVE-2017-8287-src-psaux-psobjs.c-t1_builder_close_contour-Add-safe.patch

libffi6-3.2.1-1 fails tests on xcode9.3

10.13/xcode9.3

		=== libffi tests ===

Schedule of variations:
    unix

Running target unix
Using /sw/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /sw/share/dejagnu/config/unix.exp as generic interface file for target.
Using ../../testsuite/config/default.exp as tool-and-target-specific interface file.
Running ../../testsuite/libffi.call/call.exp ...
ERROR: tcl error sourcing ../../testsuite/libffi.call/call.exp.
ERROR: clang: error: unsupported option '-dumpspecs'
clang: error: no input files
    while executing
"exec $compiler -dumpspecs"
    (procedure "get_multilibs" line 133)
    invoked from within
"get_multilibs"
    (procedure "g++_include_flags" line 9)
    invoked from within
"g++_include_flags"
    (procedure "default_target_compile" line 52)
    invoked from within
"default_target_compile $source $destfile $type $options"
    (procedure "target_compile" line 6)
    invoked from within
"target_compile $source $dest $type $options"
    (procedure "libffi_target_compile" line 57)
    invoked from within
"libffi_target_compile "$prog" "$output_file" "$compile_type" $options"
    (procedure "libffi-dg-test-1" line 52)
    invoked from within
"libffi-dg-test-1 target_compile $prog $do_what $extra_tool_flags"
    (procedure "libffi-dg-test" line 2)
    invoked from within
"${tool}-dg-test $prog [lindex ${dg-do-what} 0] "$tool_flags ${dg-extra-tool-flags}""
    (procedure "saved-dg-test" line 112)
    invoked from within
"saved-dg-test ../../testsuite/libffi.call/unwindtest.cc {-W -Wall -O0} {}"
    ("eval" body line 1)
    invoked from within
"eval saved-dg-test $args "
    (procedure "dg-test" line 6)
    invoked from within
"dg-test $test $options """
    (procedure "run-many-tests" line 36)
    invoked from within
"run-many-tests $tlist """
    (file "../../testsuite/libffi.call/call.exp" line 25)
    invoked from within
"source ../../testsuite/libffi.call/call.exp"
    ("uplevel" body line 1)
    invoked from within
"uplevel #0 source ../../testsuite/libffi.call/call.exp"
    invoked from within
"catch "uplevel #0 source $test_file_name""

		=== libffi Summary ===

# of expected passes		1780
make[3]: *** [check-DEJAGNU] Error 1

grep-3.1-1 needs standard gnulib fix

Building on 10.13 fails self-test:

PASS: uniwidth/test-uc_width2.sh
../build-aux/test-driver: line 107: 71717 Illegal instruction: 4  "$@" > $log_file 2>&1
FAIL: test-vasnprintf

Fixed by the standard patch: grep.patch (not sure why .patch isn't an accepted filetype, so I added a ".txt" extention to appease github)

ipython-genutils MIA

nbconvert-pyXX and nbformat-pyXX are unbuildable because they have a Depends:ipython-genutils-pyXX but no such package exists in our repo. While trying to see if some other package supplied a pymod with that name, I got the following in python3.6:

>>> import traitlets
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/sw/lib/python3.6/site-packages/traitlets/__init__.py", line 1, in <module>
    from .traitlets import *
  File "/sw/lib/python3.6/site-packages/traitlets/traitlets.py", line 56, in <module>
    from ipython_genutils import py3compat
ModuleNotFoundError: No module named 'ipython_genutils'

so our traitlets-pyXX package is missing a dependency on python-genutils-pyXX (and is also therefore broken because we have no such package). Could you please look into packaging that module? Or else, are there newer versions of the nb* and traitlets available that don't use what is seemingly a deprecated module?

libpulse uses orc if found

Copied from old tracker https://sourceforge.net/p/fink/bugs/1661/
On libpulse-11.1:

checking for LIBSOXR... yes
checking for ORC... yes
checking for usable orcc... 
checking for SYSTEMD... no
...
make[2]: Entering directory '/sw/src/fink.build/libpulse0-11.1-1/pulseaudio-11.1/src'
  ORCC   pulsecore/svolume-orc-gen.c
/bin/sh: --implementation: command not found
make[2]: *** [Makefile:11314: pulsecore/svolume-orc-gen.c] Error 127

http://bindist.finkproject.org/10.13/logs/pulseaudio-utils_11.1-1_657557fd53de16709e01bcc71c1d0154.log

libpulse uses orc if liborc-0.4-dev is present (headers found by f-p-p as well as linkage). If found, libpulse also needs the package orc to provide %p/bin/orcc at buildtime.

checking for ORC... yes
checking for usable orcc... /sw/bin/orcc

meep-mpi FTBFS on 10.10+ (mis-set compiler)

Building meep-mpi on 10.13 (and back to 10.10 according to bindist logs) fails:

  CXX      meep.o
clang: warning: argument unused during compilation: '-malign-double' [-Wunused-command-line-argument]
In file included from meep.cpp:1:
In file included from ./meep-ctl.hpp:10:
In file included from ./ctl-io.h:7:
/sw/include/ctl.h:270:8: error: reference to 'function' is ambiguous
extern function ctl_get_function(char *identifier);
       ^

That is consistent with our libctl package notes:

DescUsage: <<
The C++ binding for libctl's Guile commands is incompatible
with clang/libc++, because of the "function" type it introduces.
Programs using libctl therefore should be compiled with Fink's
gccXX compilers. On OSX 10.9, Fink's default g++/libc++ will
crash with "error: reference to 'function' is ambiguous".
<<

The non-mpi variant of meep builds fine, thanks to CXX=g++-fsf-5. But the mpi variant uses the mpic++ compiler wrapper, which ignores CXX and instead uses clang++ as its own preferred compiler. The openmpi docs list what variables can be used to override its own defaults. Adding:

export OMPI_CXX=g++-fsf-5

into CompileScript prior to ./configure gets the package buildable on my 10.13. Assigning to Martin, as maintainer of both meep and libctl.

Handling perl5.16 on 10.13

Apple has perl-5.16.2 on 10.9 and perl-5.18.2 on 10.10 and beyond. We supply "the other" of those two versions on 10.9-10.12 as our usual pattern of backward compatibility for all OS X that the dist supports. However, we do not currently have perl-5.16.2 on 10.13. However, we do have all the -pm5162 variants of modules on 10.13. Either we need to bring the interpretter up to support the modules, or stop bringing the modules along.

The only complication I can see is texi2html, which hardcodes perl5.16 because it relies on a deterministic ordering of a certain datastructure that is not inherently ordered but was ordered in perl up through 5.16. I don't know if the now-random ordering just causes test failures because some expected data isn't in the same order, or if the actual behavior is now non-deterministic.

pysocks-pyXX and socksipy-pyXX supply same files

site-packages/socks.py is present in both, which makes them collide during installation. Looks like one package is fork of other, or both are forks of some thing even older? See debian's thread diagnosing the history and way they planned to move forward (picking which is actually the correct current upstream and separately picking which name to use):

https://groups.google.com/forum/#!topic/linux.debian.maint.python/mXmRjEHK8Dc

pysocks-pyXX is @danielj7; socksipy-pyXX is @schwehr

discount-2.1.3 fails on 10.11 to 10.13

@hakamadare

looking for the getpwuid function (found)
looking for the srandom function (found)
looking for the bzero function (found)
looking for the random function (found)
looking for the strcasecmp function (not found)
looking for the stricmp function (not found)
markdown requires either strcasecmp() or stricmp()
### execution of ./configure failed, exit code 1

Upstream is several releases ahead (2.2.3a) and the release notes indicate clang fixes: http://www.pell.portland.or.us/~orc/Code/discount/

gdal2-shlibs-2.3.1-1 is linked against libcryptopp5-shlibs but is not in Depends

In the latest version of the gdal2 package (gdal2-shlibs-2.3.1-1), %p/lib/libgdal.20.dylib is linked against %p/lib/libcryptopp.5.dylib even though libcrytopp5-shlibs is not in the list of Required packages. Neither does libcrytopp5 appear in the BuildDepends.

In the configure step it is also not mentioned in the --with options. That means that the automatic detection of packages picked it about and it will need to be either:

  • explicitly be excluded with a --without- option
  • explicitly be included with a --with- option and added to the BuildDepends and Depends

Update dateutil-py to new version

I had added a python-dateutil-py package because I needed it for a dep without realizing that dateutil-py was the same package. I've removed mine but the current dateutil is too old for my needs. Can you update to the newest upstream? Note that tests won't work due to a circular dep: it needs freezegun to run tests but freezegun requires dateutil.

obspy-core tries to fetch distribute pymod in CompileScript

@schwehr

python2.7 setup.py build
Downloading http://pypi.python.org/packages/source/d/distribute/distribute-0.6.21.tar.gz
Traceback (most recent call last):
  File "setup.py", line 166, in <module>
    setupPackage()
  File "setup.py", line 123, in setupPackage
    distribute_setup.use_setuptools()
  File "/sw/build.build/obspy-core-py27-0.7.1-1/obspy.core-0.7.1/distribute_setup.py", line 145, in use_setuptools
    return _do_download(version, download_base, to_dir, download_delay)
  File "/sw/build.build/obspy-core-py27-0.7.1-1/obspy.core-0.7.1/distribute_setup.py", line 124, in _do_download
    to_dir, download_delay)
  File "/sw/build.build/obspy-core-py27-0.7.1-1/obspy.core-0.7.1/distribute_setup.py", line 193, in download_setuptools
    src = urlopen(url)
  File "/sw/lib/python2.7/urllib2.py", line 154, in urlopen
    return opener.open(url, data, timeout)
  File "/sw/lib/python2.7/urllib2.py", line 435, in open
    response = meth(req, response)
  File "/sw/lib/python2.7/urllib2.py", line 548, in http_response
    'http', request, response, code, msg, hdrs)
  File "/sw/lib/python2.7/urllib2.py", line 473, in error
    return self._call_chain(*args)
  File "/sw/lib/python2.7/urllib2.py", line 407, in _call_chain
    result = func(*args)
  File "/sw/lib/python2.7/urllib2.py", line 556, in http_error_default
    raise HTTPError(req.get_full_url(), code, msg, hdrs, fp)
urllib2.HTTPError: HTTP Error 403: SSL is required

distribute has been replaced by setuptools-tng, and I thought the two were interchangeable, but apparently the code might need some tweaking to find setutools instead of distribute.

This patch lets it build with setuptools:

	perl -pi -e 's|import distribute_setup|import setuptools|g' setup.py
	perl -pi -e 's|distribute_setup.use_setuptools|#distribute_setup.use_setuptools|g' setup.py

Lastly, latest upstream is %v=1.1.0.

PyQt needs newer versioned dep on sip and other failures

tgif without xmkmf

@akhansen this package of yours is not available in 10.13 because it uses xmkmf, which nobody has gotten working on 10.13. It looks like only the message-catalogs use imake, and this step is done manually in *Script, with the actual program instead using normal build tools. As a first-pass, could you have a separate "high revision" without message-catalogs (and therefore not needing xmkmf) for 10.13?

Looking at the build though, I see -D_NO_LOCALE_SUPPORT is hardcoded, which appears to disable all the internal support for translated messages via #ifndef and confirmed by checking the .libs/*.Po #include results. Without that, I'm not sure the message-catalogs are used even if xmkmf is present to enable them to be built and installed. Assuming that's true, you could drop xmkmf and the manually handled catalogs altogether on all platforms.

libnewt: whiptcl.so not installed on 10.13

On 10.13, whiptcl.so is not being installed and causes a failure when populating the newt-tcl SplitOff

mv /sw/src/fink.build/root-libnewt0-shlibs-0.52.19-1.1/sw/lib/whiptcl.so /sw/src/fink.build/root-newt-tcl-0.52.19-1.1/sw/lib/whiptcl/
mv: rename /sw/src/fink.build/root-libnewt0-shlibs-0.52.19-1.1/sw/lib/whiptcl.so to /sw/src/fink.build/root-newt-tcl-0.52.19-1.1/sw/lib/whiptcl/whiptcl.so: No such file or directory

whiptcl.so is built, just not installed. This is the relevant part of the install process:

ln -sf libnewt.0.52.19.dylib /sw/src/fink.build/root-libnewt0-shlibs-0.52.19-1.1//sw/lib/libnewt.0.dylib
[ -n "" ] && install -m 755 whiptcl.so /sw/src/fink.build/root-libnewt0-shlibs-0.52.19-1.1//sw/lib || :

On 10.11, that initial comparison is and it succeeds in putting whiptcl.so into %i/lib

[ -n "whiptcl.so" ] && install -m 755 whiptcl.so /sw/build.build/root-libnewt0-shlibs-0.52.19-1.1//sw/lib || :

moose-pm51*2 test failure

On 10.11, moose-pm 2.1600-1 fails tests as follows. Current moose upstream is 2.2010.

t/type_constraints/util_type_coercion.t .......................... Semicolon seems to be missing at t/type_constraints/util_type_coercion.t line 20.
syntax error at t/type_constraints/util_type_coercion.t line 21, near "coerce "
Execution of t/type_constraints/util_type_coercion.t aborted due to compilation errors.
t/type_constraints/util_type_coercion.t .......................... Dubious, test returned 255 (wstat 65280, 0xff00)
No subtests run 
t/type_constraints/util_type_constraints.t ....................... ok     
t/type_constraints/util_type_constraints_export.t ................ ok   
t/type_constraints/util_type_reloading.t ......................... ok   
t/type_constraints/with-specio.t ................................. 
    #   Failed test '[ 42.4, 84 ] is an acceptable array of ints'
    #   at t/type_constraints/with-specio.t line 67.
    #                   'Attribute (array_of_ints) does not pass the type constraint because: Validation failed for type named ArrayRef[Int] declared in package Foo (t/type_constraints/with-specio.t) at line 29 with value Specio::Constraint::Parameterized={ _ancestors: [ Specio::Constraint::Parameterizable=HASH(0x7f98... at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Moose/Object.pm line 24
    # 	Moose::Object::new('Foo', 'array_of_ints', 'ARRAY(0x7f98a34e1848)') called at t/type_constraints/with-specio.t line 66
    # 	main::__ANON__ at /sw/lib/perl5/Test/Fatal.pm line 129
    # 	Test::Fatal::try {...}  at /sw/lib/perl5/Try/Tiny.pm line 76
    # 	eval {...} at /sw/lib/perl5/Try/Tiny.pm line 72
    # 	Try::Tiny::try('CODE(0x7f98a34cd738)', 'Try::Tiny::Catch=REF(0x7f98a45142f8)') called at /sw/lib/perl5/Test/Fatal.pm line 136
    # 	Test::Fatal::exception('CODE(0x7f98a34a2b48)') called at t/type_constraints/with-specio.t line 67
    # 	main::__ANON__ at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	eval {...} at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	Test::Builder::subtest('Test::Builder=HASH(0x7f98a4093738)', 'Foo class', 'CODE(0x7f98a34a26c8)') called at /sw/lib/perl5/5.18.2/Test/More.pm line 807
    # 	Test::More::subtest('Foo class', 'CODE(0x7f98a34a26c8)') called at t/type_constraints/with-specio.t line 88
    # 	main::__ANON__(0) called at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 76
    # 	Test::Moose::with_immutable('CODE(0x7f98a34a25d8)', 'Foo') called at t/type_constraints/with-specio.t line 90
    # '
    #     doesn't match '(?^:does not pass the type constraint.+for anonymous type)'

    #   Failed test '{ foo => 42.4, bar => 84 } is an acceptable array of ints'
    #   at t/type_constraints/with-specio.t line 84.
    #                   'Attribute (hash_of_ints) does not pass the type constraint because: Validation failed for type named HashRef[Int] declared in package Foo (t/type_constraints/with-specio.t) at line 34 with value Specio::Constraint::Parameterized={ _ancestors: [ Specio::Constraint::Parameterizable=HASH(0x7f98... at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Moose/Object.pm line 24
    # 	Moose::Object::new('Foo', 'hash_of_ints', 'HASH(0x7f98a34e75e8)') called at t/type_constraints/with-specio.t line 81
    # 	main::__ANON__ at /sw/lib/perl5/Test/Fatal.pm line 129
    # 	Test::Fatal::try {...}  at /sw/lib/perl5/Try/Tiny.pm line 76
    # 	eval {...} at /sw/lib/perl5/Try/Tiny.pm line 72
    # 	Try::Tiny::try('CODE(0x7f98a34cd738)', 'Try::Tiny::Catch=REF(0x7f98a34e5d08)') called at /sw/lib/perl5/Test/Fatal.pm line 136
    # 	Test::Fatal::exception('CODE(0x7f98a34a2e60)') called at t/type_constraints/with-specio.t line 84
    # 	main::__ANON__ at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	eval {...} at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	Test::Builder::subtest('Test::Builder=HASH(0x7f98a4093738)', 'Foo class', 'CODE(0x7f98a34a26c8)') called at /sw/lib/perl5/5.18.2/Test/More.pm line 807
    # 	Test::More::subtest('Foo class', 'CODE(0x7f98a34a26c8)') called at t/type_constraints/with-specio.t line 88
    # 	main::__ANON__(0) called at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 76
    # 	Test::Moose::with_immutable('CODE(0x7f98a34a25d8)', 'Foo') called at t/type_constraints/with-specio.t line 90
    # '
    #     doesn't match '(?^:does not pass the type constraint.+for anonymous type)'
    # Looks like you failed 2 tests of 6.
t/type_constraints/with-specio.t ................................. 1/? 
#   Failed test 'Foo class'
#   at t/type_constraints/with-specio.t line 88.

    #   Failed test '[ 42.4, 84 ] is an acceptable array of ints'
    #   at t/type_constraints/with-specio.t line 67.
    #                   'Attribute (array_of_ints) does not pass the type constraint because: Validation failed for type named ArrayRef[Int] declared in package Foo (t/type_constraints/with-specio.t) at line 29 with value [ 42.4, 84 ] at constructor Foo::new (defined at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 77) line 54
    # 	Foo::new('Foo', 'array_of_ints', 'ARRAY(0x7f98a4514838)') called at t/type_constraints/with-specio.t line 66
    # 	main::__ANON__ at /sw/lib/perl5/Test/Fatal.pm line 129
    # 	Test::Fatal::try {...}  at /sw/lib/perl5/Try/Tiny.pm line 76
    # 	eval {...} at /sw/lib/perl5/Try/Tiny.pm line 72
    # 	Try::Tiny::try('CODE(0x7f98a4514ec0)', 'Try::Tiny::Catch=REF(0x7f98a4515898)') called at /sw/lib/perl5/Test/Fatal.pm line 136
    # 	Test::Fatal::exception('CODE(0x7f98a34a2b48)') called at t/type_constraints/with-specio.t line 67
    # 	main::__ANON__ at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	eval {...} at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	Test::Builder::subtest('Test::Builder=HASH(0x7f98a4093738)', 'Foo class (immutable)', 'CODE(0x7f98a34a26c8)') called at /sw/lib/perl5/5.18.2/Test/More.pm line 807
    # 	Test::More::subtest('Foo class (immutable)', 'CODE(0x7f98a34a26c8)') called at t/type_constraints/with-specio.t line 88
    # 	main::__ANON__(1) called at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 78
    # 	Test::Moose::with_immutable('CODE(0x7f98a34a25d8)', 'Foo') called at t/type_constraints/with-specio.t line 90
    # '
    #     doesn't match '(?^:does not pass the type constraint.+for anonymous type)'

    #   Failed test '{ foo => 42.4, bar => 84 } is an acceptable array of ints'
    #   at t/type_constraints/with-specio.t line 84.
    #                   'Attribute (hash_of_ints) does not pass the type constraint because: Validation failed for type named HashRef[Int] declared in package Foo (t/type_constraints/with-specio.t) at line 34 with value { bar: 84, foo: 42.4 } at constructor Foo::new (defined at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 77) line 84
    # 	Foo::new('Foo', 'hash_of_ints', 'HASH(0x7f98a3502140)') called at t/type_constraints/with-specio.t line 81
    # 	main::__ANON__ at /sw/lib/perl5/Test/Fatal.pm line 129
    # 	Test::Fatal::try {...}  at /sw/lib/perl5/Try/Tiny.pm line 76
    # 	eval {...} at /sw/lib/perl5/Try/Tiny.pm line 72
    # 	Try::Tiny::try('CODE(0x7f98a4514ec0)', 'Try::Tiny::Catch=REF(0x7f98a4515868)') called at /sw/lib/perl5/Test/Fatal.pm line 136
    # 	Test::Fatal::exception('CODE(0x7f98a34a2e60)') called at t/type_constraints/with-specio.t line 84
    # 	main::__ANON__ at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	eval {...} at /sw/lib/perl5/5.18.2/Test/Builder.pm line 337
    # 	Test::Builder::subtest('Test::Builder=HASH(0x7f98a4093738)', 'Foo class (immutable)', 'CODE(0x7f98a34a26c8)') called at /sw/lib/perl5/5.18.2/Test/More.pm line 807
    # 	Test::More::subtest('Foo class (immutable)', 'CODE(0x7f98a34a26c8)') called at t/type_constraints/with-specio.t line 88
    # 	main::__ANON__(1) called at /sw/build.build/moose-pm5182-2.1600-1/Moose-2.1600/blib/lib/Test/Moose.pm line 78
    # 	Test::Moose::with_immutable('CODE(0x7f98a34a25d8)', 'Foo') called at t/type_constraints/with-specio.t line 90
    # '
    #     doesn't match '(?^:does not pass the type constraint.+for anonymous type)'
    # Looks like you failed 2 tests of 6.

#   Failed test 'Foo class (immutable)'
#   at t/type_constraints/with-specio.t line 88.
# Looks like you failed 2 tests of 4.
t/type_constraints/with-specio.t ................................. Dubious, test returned 2 (wstat 512, 0x200)
Failed 2/4 subtests 
t/zzz-check-breaks.t ............................................. # Conflicts detected for Moose:
#   MooseX::NonMoose is version 0.24, but must be greater than version 0.24
t/zzz-check-breaks.t ............................................. 1/? # Breakages found with Moose:
# Installed version (0.24) of MooseX::NonMoose is in range '<= 0.24'
# 
# You should now update these modules!
t/zzz-check-breaks.t ............................................. ok   

Test Summary Report
-------------------
t/attributes/misc_attribute_coerce_lazy.t                      (Wstat: 65280 Tests: 0 Failed: 0)
  Non-zero exit status: 255
  Parse errors: No plan found in TAP output
t/todo_tests/role_insertion_order.t                            (Wstat: 0 Tests: 2 Failed: 0)
  TODO passed:   2
t/type_constraints/util_type_coercion.t                        (Wstat: 65280 Tests: 0 Failed: 0)
  Non-zero exit status: 255
  Parse errors: No plan found in TAP output
t/type_constraints/with-specio.t                               (Wstat: 512 Tests: 4 Failed: 2)
  Failed tests:  1-2
  Non-zero exit status: 2
Files=471, Tests=19498, 104 wallclock secs ( 2.53 usr  0.91 sys + 89.78 cusr  9.07 csys = 102.29 CPU)
Result: FAIL
Failed 3/471 test programs. 2/19498 subtests failed.
make: *** [test_dynamic] Error 255

What to do with xar

Fink has xar-1.5.2, which is from 2007. Upstream (googlecode) has since been forked (github) and the googlecode project set to archive status around the 1.5.3 release. The github project's last release is 1.6.1 from 5 years ago. Apple now supplies xar also, calling it "1.6dev" in 10.9, "1.7dev" in 10.10, and "1.8dev" in 10.11 and beyond. All three of these vendors have a libxar.1.dylib, but none of their latest releases are ABI-compatible with each other (I'm not even sure apple's all are). And only Apple's 1.8dev library is immune to a recent CVE. I think there's value in continuing to roll our own package, for security on older platforms and ABI sanity on all, so we need a new package-names for the library (shlibs policy vs libxar1-shlibs/-dev). Should we bump the version value in install_name (libxar.2.dylib) and call the package libxar2-shlibs/-dev? Or bury the .dylib in a subdir and call the package libxar1-apple-shlibs/-dev? Or some other option?

db53/db60 fail with unknown 'atomic_flag' errors on 10.13/Xcode9.3

This patch: https://gist.githubusercontent.com/LnL7/5153b251fd525fe15de69b67e63a6075/raw/7778e9364679093a32dec2908656738e16b6bdcb/clang.patch
from here: https://www.databaseusers.com/article/6047407/Bug%3A+Build+failure+on+Linux+with+clang+libc%2B%2B+(atomic+issues)

fixes the build for me.

Sorry I don't have a build log, but am having troubles transferring output from my VM. The error looks like that listed in that dbusers.com link above.

gcc7 build failing on 10.13.3

Compiling the package fails well into the build process, unfortunately with a (to me) not very helpful error message:

checking whether the /scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/./gcc/xgcc -shared-libgcc -B/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/./gcc -nostdinc++ -L/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/x86_64-apple-darwin17.4.0/i386/libstdc++-v3/src -L/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/x86_64-apple-darwin17.4.0/i386/libstdc++-v3/src/.libs -L/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/x86_64-apple-darwin17.4.0/i386/libstdc++-v3/libsupc++/.libs -B/sw/lib/gcc7/x86_64-apple-darwin17.4.0/bin/ -B/sw/lib/gcc7/x86_64-apple-darwin17.4.0/lib/ -isystem /sw/lib/gcc7/x86_64-apple-darwin17.4.0/include -isystem /sw/lib/gcc7/x86_64-apple-darwin17.4.0/sys-include  -m32 linker (/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir/./gcc/collect-ld) supports shared libraries... yes
checking dynamic linker characteristics... darwin17.4.0 dyld
checking how to hardcode library paths into programs... immediate
checking for gawk... (cached) gawk
checking size of void *... 4
checking for clock_getres... yes
checking for clock_gettime... yes
checking for clock_settime... yes
checking for shm_open in -lrt... no
checking for dlsym in -ldl... yes
checking for clock_gettime in -lrt... no
checking for necessary platform features... yes
checking __sync extensions... yes
checking __atomic extensions... yes
checking output filetype... 
configure: WARNING: could not determine output file type
looking for a compliant stdint.h in stdint.h, checking for uintmax_t... yes
checking for uintptr_t... yes
checking for int_least32_t... yes
checking for int_fast32_t... yes
checking for uint64_t... yes
checking what to include in gstdint.h... stdint.h (already complete)
checking sys/mman.h usability... yes
checking sys/mman.h presence... yes
checking for sys/mman.h... yes
checking alloca.h usability... yes
checking alloca.h presence... yes
checking for alloca.h... yes
checking link.h usability... no
checking link.h presence... no
checking for link.h... no
checking whether strnlen is declared... yes
checking rpc/xdr.h usability... no
checking rpc/xdr.h presence... yes
configure: WARNING: rpc/xdr.h: present but cannot be compiled
configure: WARNING: rpc/xdr.h:     check for missing prerequisite headers?
configure: WARNING: rpc/xdr.h: see the Autoconf documentation
configure: WARNING: rpc/xdr.h:     section "Present But Cannot Be Compiled"
configure: WARNING: rpc/xdr.h: proceeding with the compiler's result
configure: WARNING:     ## --------------------------- ##
configure: WARNING:     ## Report this to libsanitizer ##
configure: WARNING:     ## --------------------------- ##
checking for rpc/xdr.h... no
checking tirpc/rpc/xdr.h usability... no
checking tirpc/rpc/xdr.h presence... no
checking for tirpc/rpc/xdr.h... no
configure: updating cache ./config.cache
configure: creating ./config.status
config.status: creating Makefile
config.status: creating libsanitizer.spec
config.status: creating libbacktrace/backtrace-supported.h
config.status: creating interception/Makefile
Adding multilib support to interception/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating sanitizer_common/Makefile
Adding multilib support to sanitizer_common/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating libbacktrace/Makefile
Adding multilib support to libbacktrace/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating lsan/Makefile
Adding multilib support to lsan/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating asan/Makefile
Adding multilib support to asan/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating ubsan/Makefile
Adding multilib support to ubsan/Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: creating config.h
config.status: executing default-1 commands
Adding multilib support to Makefile in ../../../../gcc-7.3.0/libsanitizer
with_multisubdir=i386
config.status: executing depfiles commands
config.status: executing libtool commands
config.status: executing gstdint.h commands
config.status: executing depfiles commands
config.status: executing libtool commands
config.status: executing gstdint.h commands
make[1]: Leaving directory '/scratch.noindex/fink.build/gcc7-7.3.0-1/darwin_objdir'
make: *** [Makefile:21802: bootstrap] Error 2
### execution of /tmp/fink._dSHO failed, exit code 2
### execution of /tmp/fink.to25w failed, exit code 2
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-gcc7-7.3.0-1
(Reading database ... 378237 files and directories currently installed.)
Removing fink-buildlock-gcc7-7.3.0-1 ...
Failed: phase compiling: gcc7-7.3.0-1 failed

Before reporting any errors, please run "fink selfupdate" and try again.
Also try using "fink configure" to set your maximum build jobs to 1 and
attempt to build the package again.
If you continue to have issues, please check to see if the FAQ on Fink's 
website solves the problem.  If not, ask on one (not both, please) of
these mailing lists:

	The Fink Users List <[email protected]>
	The Fink Beginners List <[email protected]>.

Please try to include the complete error message in your report.  This
generally consists of a compiler line starting with e.g. "gcc" or "g++"
followed by the actual error output from the compiler.

Also include the following system information:
Package manager version: 0.42.99.git
Distribution version: selfupdate-rsync Fri Mar  9 14:26:44 2018, 10.13, x86_64
Trees: stable/main local/main
Xcode.app: 9.2
Xcode command-line tools: 9.2.0.0.1.1510905681
Max. Fink build jobs:  6

The command in darwin_objdir/Makefile:21802 afaict is the 'stage-bubble' in

bootstrap:
        echo stage3 > stage_final
        @r=`${PWD_COMMAND}`; export r; \
        s=`cd $(srcdir); ${PWD_COMMAND}`; export s; \
        $(MAKE) $(RECURSE_FLAGS_TO_PASS) stage3-bubble
        @: $(MAKE); $(unstage)
        @r=`${PWD_COMMAND}`; export r; \
        s=`cd $(srcdir); ${PWD_COMMAND}`; export s; \
        TFLAGS="$(STAGE3_TFLAGS)"; \
        $(MAKE) $(TARGET_FLAGS_TO_PASS) all-host all-target

building setuptools-scm-py 3.1.0 deadlocks if older version installed

If I have the old setuptools-scm-py-1.15.6 installed, building v3.1.0 deadlocks because of the BuildConflicts line in the new release. This blocks the bindist.

dpkg: regarding .../fink-buildlock-setuptools-scm-py27-3.1.0-1_2018.08.10-05.53.56_darwin-x86_64.deb containing fink-buildlock-setuptools-scm-py27-3.1.0-1:
 fink-buildlock-setuptools-scm-py27-3.1.0-1 conflicts with setuptools-scm-py27 (<= 2.0.0)
  setuptools-scm-py27 (version 1.15.6-1) is installed.

This was added in 9b2ebdf by @dhomeier. It's not immediately clear why BC: is needed for an old version of self.

Updating fink within stable tree only fails with 'unsatisfied dependency in fink: base-files'

Hi,

I wanted to rebuild some packages I had updated in my local tree using the stable tree only, to make sure I have all required dependencies pulled over before committing them to git. I.e.,

I have rsync'ed the set of packages required for the update from the local to the stable tree.

I then started the update as fink -t stable -m update pylint-py36.
This asks for updating the expected dependencies just added from local, but strangely also a bunch of packages that are already up-to-date in (and only in) stable.
I let it nonetheless proceed, but among the next steps fink wants to reinstall both base-files and fink itself, failing as follows:

/sw/bin/dpkg-lockwait -i /sw/fink/dists/stable/main/binary-darwin-x86_64/base/base-files_1.9.17-1_darwin-x86_64.deb
(Reading database ... 613747 files and directories currently installed.)
Preparing to replace base-files 1.9.17-1 (using .../base-files_1.9.17-1_darwin-x86_64.deb) ...
Unpacking replacement base-files ...
Setting up base-files (1.9.17-1) ...

Reading buildlock packages...
	All buildlocks accounted for.
/sw/bin/dpkg-lockwait -i /sw/fink/dists/stable/main/binary-darwin-x86_64/base/expat1-shlibs_2.2.5-1_darwin-x86_64.deb
(Reading database ... 613747 files and directories currently installed.)
Preparing to replace expat1-shlibs 2.2.5-1 (using .../expat1-shlibs_2.2.5-1_darwin-x86_64.deb) ...
Unpacking replacement expat1-shlibs ...
Setting up expat1-shlibs (2.2.5-1) ...

Reading buildlock packages...
	All buildlocks accounted for.
/sw/bin/dpkg-lockwait -i /sw/fink/dists/stable/main/binary-darwin-x86_64/base/base-files_1.9.17-1_darwin-x86_64.deb
(Reading database ... 613747 files and directories currently installed.)
Preparing to replace base-files 1.9.17-1 (using .../base-files_1.9.17-1_darwin-x86_64.deb) ...
Unpacking replacement base-files ...
Setting up base-files (1.9.17-1) ...

Reading buildlock packages...
	All buildlocks accounted for.

While trying to install:
  fink-0.43.0-111

The following inconsistencies found:
  Unsatisfied dependency in fink: base-files

Trying to resolve dependencies...
Could not resolve inconsistent dependencies!

Fink isn't sure how to install the above packages safely. You may be able to fix things by running:

  fink scanpackages
  sudo apt-get update
  sudo apt-get install fink=0.43.0-111

Failed: Fink::SysState: Could not resolve inconsistent dependencies

Needless to say, apt-get install just informs me that fink-0.43.0-111 is already the newest version...

Am I somehow missing a tree here, or is this way of using fink just not meant to work?

numexpr-py fails to build on 10.13

fink build numexpr-pyXX fails to build on 10.13. I tired with py27 and py36. Log:

...
/sw/bin/python3.6 setup.py build
running build
running config_cc
unifing config_cc, config, build_clib, build_ext, build commands --compiler options
running config_fc
unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options
running build_src
build_src
building py_modules sources
creating build
creating build/src.macosx-10.13-x86_64-3.6
creating build/src.macosx-10.13-x86_64-3.6/numexpr
building extension "numexpr.interpreter" sources
build_src: building npy-pkg config files
running build_py
creating build/lib.macosx-10.13-x86_64-3.6
creating build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/cpuinfo.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/version.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/expressions.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/__init__.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/utils.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying numexpr/necompiler.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
copying build/src.macosx-10.13-x86_64-3.6/numexpr/__config__.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr
creating build/lib.macosx-10.13-x86_64-3.6/numexpr/tests
copying ./numexpr/tests/__init__.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr/tests
copying ./numexpr/tests/test_numexpr.py -> build/lib.macosx-10.13-x86_64-3.6/numexpr/tests
Skipping optional fixer: buffer
Skipping optional fixer: idioms
Skipping optional fixer: set_literal
Skipping optional fixer: ws_comma
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
building 'numexpr.interpreter' extension
compiling C++ sources
C compiler: g++ -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -I/sw/include

creating build/temp.macosx-10.13-x86_64-3.6/numexpr
compile options: '-I/sw/lib/python3.6/site-packages/numpy/core/include -I/sw/include/python3.6m -c'
extra options: '-funroll-all-loops'
g++: numexpr/interpreter.cpp

Warning: Assuming default configuration (numexpr/tests/{setup_tests,setup}.py was not found)error: Command "g++ -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -I/sw/include -I/sw/lib/python3.6/site-packages/numpy/core/include -I/sw/include/python3.6m -c numexpr/interpreter.cpp -o build/temp.macosx-10.13-x86_64-3.6/numexpr/interpreter.o -MMD -MF build/temp.macosx-10.13-x86_64-3.6/numexpr/interpreter.o.d -funroll-all-loops" failed with exit status 1
### execution of /sw/bin/python3.6 failed, exit code 1
### execution of /tmp/fink.KfQGC failed, exit code 1
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-numexpr-py36-2.6.4-1
(Reading database ... 442337 files and directories currently installed.)
Removing fink-buildlock-numexpr-py36-2.6.4-1 ...
Failed: phase compiling: numexpr-py36-2.6.4-1 failed
...

I tried to find some things out, but could not find the issue → sorry, no patch.

valgrind FTBFS on 10.12+

I (and bindist) have a ./configure error on 10.13:

checking for a supported version of gcc... Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
no (9.1.0)
configure: error: please use gcc >= 3.0 or clang >= 2.9 or icc >= 13.0

According to the PatchScript, it only allows llvm up through 8.x. Upstream has a newer version and their git has a change to allow 9.x. That gets me further, but then a different error later:

checking for a supported version of gcc... Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
ok (Apple LLVM version 9.1.0)
checking build system type... x86_64-apple-darwin17.7.0
checking host system type... x86_64-apple-darwin17.7.0
checking for a supported CPU... ok (x86_64)
checking for a 64-bit only build... no
checking for a 32-bit only build... no
checking for a supported OS... ok (darwin17.7.0)
checking for the kernel version... unsupported (17.7.0)
configure: error: Valgrind works on Darwin 10.x, 11.x, 12.x, 13.x, 14.x and 15.x (Mac OS X 10.6/7/8/9/10/11)

There are a series of git changes that add support for 10.12 and preliminary for 10.13. I'm not sure which have actually made it into the current 3.13.0 release (fink is at 3.11.0).

chardet-py34-3.0.4-1

The package fails to install with the following error:

Preparing to unpack .../chardet-py34_3.0.4-1_darwin-x86_64.deb ...
Unpacking chardet-py34 (3.0.4-1) over (3.0.4-1) ...
Setting up chardet-py34 (3.0.4-1) ...
update-alternatives: error: alternative path /sw/bin/chardetect-34 doesn't exist
dpkg: error processing package chardet-py34 (--install):
 subprocess installed post-installation script returned error exit status 2
Errors were encountered while processing:
 chardet-py34
### execution of /sw/bin/dpkg-lockwait failed, exit code 1

idna-py 2.7 egg-info permissions still broken

idna-py-2.7-1 is installing files in %p/lib/python%type_raw[python]/site-packages/idna-2.7-py%type_raw[python].egg-info with permissions 700:

> ls -lF /sw/lib/python3.5/site-packages/idna-2.7-py3.5.egg-info/                                                                                      
total 48
-rw-r--r--  1 root  admin  10565 30 Jul 17:45 PKG-INFO
-rwx------  1 root  admin    529 30 Jul 17:45 SOURCES.txt*
-rwx------  1 root  admin      1 30 Jul 17:45 dependency_links.txt*
-rwx------  1 root  admin      5 30 Jul 17:45 top_level.txt*

This breaks e.g. pip:

> sudo -u fink-bld HOME=$PWD /sw/bin/pip-py3.5 freeze                                                                                              
Exception:
Traceback (most recent call last):
  File "/sw/lib/python3.5/site-packages/pip/_internal/basecommand.py", line 228, in main
    status = self.run(options, args)
  File "/sw/lib/python3.5/site-packages/pip/_internal/commands/freeze.py", line 93, in run
    for line in freeze(**freeze_kwargs):
  File "/sw/lib/python3.5/site-packages/pip/_internal/operations/freeze.py", line 42, in freeze
    dist.get_metadata_lines('dependency_links.txt')
  File "/sw/lib/python3.5/site-packages/pip/_vendor/pkg_resources/__init__.py", line 1408, in get_metadata_lines
    return yield_lines(self.get_metadata(name))
  File "/sw/lib/python3.5/site-packages/pip/_vendor/pkg_resources/__init__.py", line 1404, in get_metadata
    value = self._get(self._fn(self.egg_info, name))
  File "/sw/lib/python3.5/site-packages/pip/_vendor/pkg_resources/__init__.py", line 1516, in _get
    with open(path, 'rb') as stream:
PermissionError: [Errno 13] Permission denied: '/sw/lib/python3.5/site-packages/idna-2.7-py3.5.egg-info/dependency_links.txt'

The .info is already fixing this for PKG-INFO, but not for the other files.

Planning to kill libmsn-0.3

The package is "Library for connecting to MSN Messenger" but "MSN network, which was officially shutdown in Oc" according to why Debian removed this package from their distros in November 2016. I came across it while looking at unmaintained packages still using openssl100 (it FTBFS with openssl110), so it could be sent to a -shlibs stub and left to rot if it's useless.

The only reverse-dependency is kopete4-mac: @nieder is there a flag to disable its detection? Not having it means kopete loses WLM support, but Microsoft discontinued WLM a while ago also.

Installing from git checkout with inject.pl

When trying to install the latest distributions from my local git repo as per method 2 in the Readme, running ./inject.pl [/sw] fails as it looks for a distribution in the .git subdirectory. The script should be fixed to ignore that dir:

--- a/inject.pl
+++ b/inject.pl
@@ -34,7 +34,7 @@ my @contents = `ls`;
 my ($filename,@directories);
 foreach $filename (@contents) {
     chomp($filename);
-    if (-d $filename and not $filename eq "CVS") {
+    if (-d $filename and not ($filename eq "CVS" or $filename eq ".git")) {
        push(@directories, $filename);
     }
 }

Also it would be helpful to have an option to only install the distribution actually required for one's system, and not always having to copy over the entire content of (in my case) 10.[2-7]*.

mpg123-1.22.0 fails to link on 10.12.

Received the following output @danielj7 :

libtool: link: gcc  -o .libs/output_openal.so -bundle  .libs/output_openal_la-openal.o   -L/sw/lib -framework OpenAL /sw/lib/libltdl.dylib -lmx -lm  -O3   -framework OpenAL -Wl,-exported_symbols_list,.libs/output_openal-symbols.expsym
ld: warning: ignoring file /Library/Frameworks//OpenAL.framework/OpenAL, file was built for unsupported file format ( 0xFE 0xED 0xFA 0xCE 0x00 0x00 0x00 0x12 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x06 ) which is not the architecture being linked (x86_64): /Library/Frameworks//OpenAL.framework/OpenAL
Undefined symbols for architecture x86_64:
  "_alBufferData", referenced from:
      _write_openal in output_openal_la-openal.o
  "_alDeleteBuffers", referenced from:
      _flush_openal in output_openal_la-openal.o
      _close_openal in output_openal_la-openal.o
  "_alDeleteSources", referenced from:
      _close_openal in output_openal_la-openal.o
  "_alGenBuffers", referenced from:
      _write_openal in output_openal_la-openal.o
  "_alGenSources", referenced from:
      _open_openal in output_openal_la-openal.o
  "_alGetSourcei", referenced from:
      _flush_openal in output_openal_la-openal.o
      _write_openal in output_openal_la-openal.o
      _close_openal in output_openal_la-openal.o
  "_alIsExtensionPresent", referenced from:
      _get_formats_openal in output_openal_la-openal.o
  "_alSourcePlay", referenced from:
      _write_openal in output_openal_la-openal.o
  "_alSourceQueueBuffers", referenced from:
      _write_openal in output_openal_la-openal.o
  "_alSourceStop", referenced from:
      _flush_openal in output_openal_la-openal.o
  "_alSourceUnqueueBuffers", referenced from:
      _flush_openal in output_openal_la-openal.o
      _write_openal in output_openal_la-openal.o
      _close_openal in output_openal_la-openal.o
  "_alcCloseDevice", referenced from:
      _close_openal in output_openal_la-openal.o
  "_alcCreateContext", referenced from:
      _open_openal in output_openal_la-openal.o
  "_alcDestroyContext", referenced from:
      _close_openal in output_openal_la-openal.o
  "_alcMakeContextCurrent", referenced from:
      _open_openal in output_openal_la-openal.o
      _close_openal in output_openal_la-openal.o
  "_alcOpenDevice", referenced from:
      _open_openal in output_openal_la-openal.o
ld: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[3]: *** [output_openal.la] Error 1
make[2]: *** [all-recursive] Error 1
make[1]: *** [all] Error 2
make: *** [all-recursive] Error 1
### execution of /usr/bin/make failed, exit code 2
### execution of /tmp/fink._ndLr failed, exit code 2
Removing runtime build-lock...
Removing build-lock package...
/sw/bin/dpkg-lockwait -r fink-buildlock-mpg123-1.22.0-1
Unknown option: o
Unknown option
(Reading database ... 201030 files and directories currently installed.)
Removing fink-buildlock-mpg123-1.22.0-1 (2017.12.04-20.42.52) ...
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/crypto/db53-aes-bin_5.3.28-1_darwin-x86_64.deb
New package: dists/stable/main/binary-darwin-x86_64/crypto/db53-aes-doc_5.3.28-1_darwin-x86_64.deb
New package: dists/stable/main/binary-darwin-x86_64/crypto/db53-aes-shlibs_5.3.28-1_darwin-x86_64.deb
New package: dists/stable/main/binary-darwin-x86_64/crypto/db53-aes_5.3.28-1_darwin-x86_64.deb
Failed: phase compiling: mpg123-1.22.0-1 failed

Also include the following system information:
Package manager version: 0.42.0
Distribution version: selfupdate-rsync Mon Dec  4 20:37:24 2017, 10.12, x86_64
Trees: local/main remote/dak180 stable/main local/injected
Xcode.app: 9.1
Xcode command-line tools: 9.1.0.0.1.1508540944
Max. Fink build jobs:  1

Some small things for wget

  1. There are some obsolete items in the user-facing packaging notes of wget-1.19.2-2:
  • DescDetail talks about upstream versions 1.13.x, and especially mentions use of fink's openssl vs apple's because apple's is missing a symbol (vs now being missing altogether)--see also DescPort.
  • DescUsage talks about certificate support on OS X 10.4 and 10.5, and nothing else. Are these instructions still valid on 10.9 and beyond?
  1. I filed an upstream bug-report about the format-string warning in src/utils.c line 2822. Not critical to fix (just in a diagnostic output). The message is not appearing in their list archives yet:
-            logprintf (LOG_VERBOSE, _ ("Skipping key with wrong size (%d/%d): %s\n"),                  
+            logprintf (LOG_VERBOSE, _ ("Skipping key with wrong size (%lu/%d): %s\n"),                  
                        (strlen (begin_pos + 8) * 3) / 4, SHA256_DIGEST_SIZE,                            
                        quote (begin_pos + 8));                                                          

Update: see bug-wget message

  1. It currently uses openssl100. You can probably hold off testing with openssl110...upstream says at best this release only builds when openssl110 is built with certain options. But if you want to, they do have a recent git commit to support it (or could just wait for next release:)

.info/.patch files with +x permission

We have 136 files in the current distribution that have the executable bit set:

find /sw/fink/dists/stable/main/finkinfo -perm /111 -type f

which seems unusual for files that are not programs. It's made for some interesting (albeit confusing) shell autocompletions, and can raise a flag for paranoid sysadmins. In CVS, once a file is committed, its permissions are unchangeable. Does git let us change it?

R version support

Spawning a new topic from #12, wherein:

@dmacks:

We currently use a Distribution tag to limit r-base32 to 10.{9-12}, but we don't limit any other r-base*. Should we similarly restrict r-base31? Do we even need to keep dragging r-base33 along, or is r-base34 sufficient to keep going forward?

@nieder:

Personally, I don't know if there's any reason to keep r33 since I think most rmods that use r33 are available as r34. There are some in r31 that are not available in later releases (but I would imagine that's most likely because they're abandoned).

Here, let's figure out what rmods need to be brought forward vs be marked "do not bring forward" (so we don't have to have this conversation again in the future), and then whether to drop r31 and/or r33 on 10.13....

emboss-6.5.7 fails to find jni_md.h

At least on Java8, jni_md.h is no longer found in $java_home/include, but is in $java_home/include/darwin. It is also found in /System/Library/Frameworks/JavaVM.framework/Headers, but the logic in emboss.info::CompileScript precludes that from being found.

Java7 and java9 also fail to find jni_md.h (http://bindist.finkproject.org/10.13/logs/emboss_6.5.7-106_98838942d5268b7bfb05a44b9fa6da6f.log & http://bindist.finkproject.org/10.11/logs/emboss_6.5.7-106_98838942d5268b7bfb05a44b9fa6da6f.log), so I'm assuming the same situation applies there.

Adding export CPPFLAGS="$CPPFLAGS -I$java_includes/darwin" to CompileScript before autoreconf lets the header be found and the build finishes against java8 (tested on 10.11).

uritemplate/uritemplate.py confusion

We currently have:

   	uritemplate-py27	0.6-1	RFC6570 up to level 4
   	uritemplate-py34	0.6-1	RFC6570 up to level 4
   	uritemplate-py35	0.6-1	RFC6570 up to level 4
   	uritemplate.py-py27	3.0.2-1	Simple python lib to deal with URI Templates
   	uritemplate.py-py34	3.0.2-1	Simple python lib to deal with URI Templates
   	uritemplate.py-py35	3.0.2-1	Simple python lib to deal with URI Templates
   	uritemplate.py-py36	3.0.2-1	Simple python lib to deal with URI Templates

Their Desc* say they have a python namespace collision (which they do) and their packages Conflicts/Replaces each other, but both have reverse-dependencies:

  • uritemplate-pyXX: github3.py-pyXX, google-api-python-client-pyXX
  • uritemplate.py-pyXX: oauth2client-pyXX

so there are certain combinations of other packages that cannot be installed concurrently.

The Desc of uritemplate.py-pyXX says "both packages are now merged upstream so this package can serve as both", so does that mean one can simply switch the Depends from uritemplate-pyXX to uritemplate.py-pyXX, or are there other compatibility changes needed? Should uritemplate-pyXX become a fink-obsolete-packages wrapper bundle rather than a real library?

stamp-cvs- file

We have a %p/fink/stamp-cvs-$time file (or stamp-cvs-live in git) that is used by inject.pl (I can't find anywhere else via grep), but we're in git now, so that name is confusing. I don't know what the purpose of this file is, whether we should rename it to stamp-git-* for sanity vs deleting it and its associated code blocks as being a cvs-specific need that is not needed for git.

package-deprecationmanager-pm will cause circular dependencies

I'm building something that eventually needs package-deprecationmanager-pmXXXX in its dep tree. Currently, it fails due to Package::Stash not being present:

# The following REQUIRED prerequisites were not satisfied:
# 
# Package::Stash is not installed (any version required)
# 
t/00-report-prereqs.t .. ok   
t/basic.t .............. Can't locate Package/Stash.pm in @INC (@INC contains: /sw/build.build/package-deprecationmanager-pm5162-0.17-3/Package-DeprecationManager-0.17/blib/lib /sw/build.build/package-deprecationmanager-pm5162-0.17-3/Package-DeprecationManager-0.17/blib/arch /sw/lib/perl5/5.16.2/darwin-thread-multi-2level /sw/lib/perl5/5.16.2/darwin-thread-multi-2level /sw/lib/perl5/5.16.2 /sw/lib/perl5/darwin-thread-multi-2level /sw/lib/perl5/5.16.2/darwin-thread-multi-2level /sw/lib/perl5/5.16.2 /sw/lib/perl5/darwin-thread-multi-2level /sw/lib/perl5 /sw/lib/perl5/darwin /sw/lib/perl5/5.16.2/darwin-thread-multi-2level /sw/lib/perl5/5.16.2 /sw/lib/perl5/darwin-thread-multi-2level /sw/lib/perl5 /sw/lib/perl5/darwin /sw/lib/perl5/site_perl/5.16.2/darwin-thread-multi-2level /sw/lib/perl5/site_perl/5.16.2 /sw/lib/perl5-core/5.16.2/darwin-thread-multi-2level /sw/lib/perl5-core/5.16.2 .) at /sw/build.build/package-deprecationmanager-pm5162-0.17-3/Package-DeprecationManager-0.17/blib/lib/Package/DeprecationManager.pm line 10.

If I add package-stash-pmXXXX to the [Depends](http://deps.cpantesters.org/?module=Package%3A%3ADeprecationManager;perl=latest) field for package-deprecationmanager-pmXXXX, then I get a circular dependency warning by Fink. The circle includes just these two packages. But skimming through the package-stash source, I don't see Package::DeprecationManager used any more (0.21 tarball did).

Also, package-deprecationmanager needs TestDepends on test-warnings-pmXXXX. I can commit these 3 dependency changes if you want.

Missing biplist-py34 variant/confusing versioning

While looking at mismatches among our -pyXX package sets, found we have:

   	biplist-py27	1.0.3-1	
   	biplist-py34	1:0.8-2
   	biplist-py36	1.0.3-1

So the "middle" python-version has an older %v value but is epoch'ed to be newer and we have no -py35 at all. The packages are unmaintained. The 1:0.8 .info has never had -py2X or any -py3X higher than 34. The .info that now has 1.0.3 originally had version 0.9 and only -py27, but several updates by @nieder (bcd2587 and 044720f) brought up to 1.0.3 and added -py36.

Is 1.0.3 the current upstream? If so, it should become Epoch:1. Is it viable on python34, so we can scrap the -py34 special-case .info? And can a -py35 variant be added for completeness?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.