rather than trying to consolidate into a single fnmatch. There aren't that
many of them, and it will aid the integration of cwrappers which doesn't
support globs.
awk array iteration which could result in the wrong library being chosen.
Fixes issue seen with GCC 4.9 where the libgcc from the main GCC package
was found first and caused errors, even though the RPATH correctly had
the gcc49-libs version listed first.
Base xsrc on netbsd-5 has not really worked with pkgsrc for a while,
because various programs need newer versions of various X pieces. The
2014Q2 official bulk builds are missing about 1500 packages as a
result of this. Therefore, switch to modular on netbsd-5 (as netbsd-4
has been for a very long time), which should result in more useful
binary packages for netbsd-5 for 2014Q3.
(There is a notion of updating netbsd-5 base xsrc to more modern xorg.
If that happens, and there's a 5.x formal release, and builds show
that pkgsrc with native succeeds on it, this can perhaps be flipped
back. Odds are that's not going to happen, and it's overwhelmingly
unlikely to happen soon.)
Anyone who prefers to stay with native can set X11_TYPE=native in
mk.conf.
Note that this is about pkgsrc and specifically the default
dependencies for pkgsrc programs that use X11, so the native servers
are unaffected and can be run from /usr/X11R7, the same as they are
now, without any changes being necessary. (This message is in fact
being typed on a system with a native server, native xterm and modular
libs for pkgsrc.)
Discussed on tech-pkg, tech-x11 multiple times over the last 6 months
or so, and specifically encouraged by wiz@.
likely that the files in this code path do not exist, so this can save a
significant number of exec's, especially for packages with a large number of
dependencies, with a corresponding reduction in the run time for the 'wrapper'
phase.
In addition, remove a 'useless use of cat' instance to save more exec's.
a blanket removal of any long options, richard@ is concerned this may affect
packages which use the long options now available in newer SunOS ld.
Whilst here, transform --rpath to -R, used by a few packages.
Fixes unwanted execution of {aclocal,automake}-${latest_version} for the
case configure.ac (or something) is patched, but configure is also alredy
regenerated and aclocal and automake are not required (not in USE_TOOLS).
XXX: if {aclocal,automake}-${latest_version} is really required by somewhere,
XXX: it should be created below, same as plain aclocal and automake.
The script serves several purposes. Chief of them:
1. Fast track for those who just want to build their 1000 packages
and do not want to bother with optimizations.
2. Fast track for those who want to understand how pbulk is supposed
to be set up.
Because it serves as a documentation (providing working setup at the same time),
a lot of features are intentionally left out.
noticed by diger in pkgsrc-users@.
While here, enable useradd only for the case groupadd exists,
because former useradd is interactive command, not usable with pkgsrc framework.
* if p5-* in {,BUILD_,TOOL_}DEPENDS, use perl5 from pkgsrc.
* check whether builtin perl satisfies PERL5_REQD, or use perl5 from pkgsrc.
* if buitin perl satisfies all {perl>=n.m,p5-foo>=x.y} style dependencies,
get rid of such dependencies and use builtin perl.
* HASKELL_ENABLE_LIBRARY_PROFILING and HASKELL_ENABLE_HADDOCK_DOCUMENTATION
are "User-settable variables", not "Package-settable variables".
* Change to HASKELL_ENABLE_HADDOCK_DOCUMENTATION=yes by default.
* Add HASKELL_ENABLE_SHARED_LIBRARY("yes" by default), to enable shlib support.
* Add support for dynamically conditional PLIST entries handling for
HASKELL_ENABLE_SHARED_LIBRARY and HASKELL_ENABLE_LIBRARY_PROFILING.
discussed with pho@ and szptvlfn@.
and documented in mk/defaults/mk.conf. Both the "gpg" and "x509" methods
supported by pkg_admin(1) are supported. With package signing enabled, a
staging, unsigned copy of the package is always created, and its final copy
to the package repository is done with pkg_admin(1) instead of "ln || cp".
Proper operation should otherwise not be affected.
Tested both with and without user-destdir support in packages.
"can live with it" joerg@
From EdgeBSD.
SMF is the Service Management Facility, the default init system in
Solaris and derivatives since version 10. This adds "smf" to the list
of supported INIT_SYSTEM types, and makes it the default init system on
platforms where it is available.
Packages can introduce SMF support by providing a manifest file, by
default located in ${FILESDIR}/smf/manifest.xml but manifests under
${WRKSRC} can be used too if the package source includes one.
SMF method scripts are supported too if required, using SMF_METHODS in a
similar manner to RCD_SCRIPTS.
Many parts of the SMF infrastructure are configurable, see mk/smf.mk for
the full details.
This commit introduces an INIT_SYSTEM variable which will determine the
type of init system to be used on the target system, supporting "rc.d"
at this time.
The pkginstall infrastructure is changed to only install RCD_SCRIPTS if
INIT_SYSTEM is set to "rc.d", and PLIST entries for rc.d scripts are
now handled automatically based on RCD_SCRIPTS.
transformed before ${X11BASE}. On Red Hat Linux with native X11,
X11BASE=/usr and so any ${LOCALBASE} which is under that hierarchy
(e.g. the default /usr/pkg) will break in weird and wonderful ways.
In addition to this, use ${LIBABISUFFIX} where appropriate, and avoid
some extra transformations in the X11BASE=/usr case.
This along with other recent changes gets us to a very healthy >12k
packages with native X11 on Red Hat Linux 6 (and clones).
from devel/ncurses to the mk infrastructure.
FAKE_NCURSES=yes
Provides the system curses as ncurses.h and libncurses.
USE_CURSES=wide
Links to system curses if they provide wide support, otherwise ncursesw.
wide-curses in the package options also triggers this.
* dll may be in ${PREFIX}/bin instead of ${PREFIX}/lib.
* dll name may be cygXXX.dll instead of libXXX.dll.
* versioning name may be foo-X.Y.Z.dll instead of foo.X.Y.Z.dll.
figure out the target architecture based on the objects so we need to
explicitly set it.
This allows bootstrap --abi=32 to complete successfully on x86_64.
This was originally introduced to work around some behaviour in the
libtool build, however these days it is actively harmful for a number of
packages, where removing additional arguments when -v is present on the
command line can break ABI detection (notably in CMake packages).
Instead, filter out any references to BUILDLINK_DIR from the libtool
scripts, as that should do the same job.
Retain the ability to run the 'scan' wrapper script, as it can be useful
in certain cases, and is required to support the scan-libtool script
anyway.
