but the only difference is fixes for Windows IDE solutions Visual C 2008
and Visual C 2010, so prefer 1.3.4 to avoid dealing with the DISTVERSION)
- Move out .so.1-suffixed symlink creation to `post-install' target rather
than patching `src/Makefile' since the patch tends to get modified with
each update; and drop no longer needed patches that were merged upstream
- Define LICENSE_FILE as it is not just a standard boilerplate
PR: 220649 (modified)
Submitted by: sunpoet
Previously, binary packages would miss Xinerama and Xcursor dependencies and
thus were giving some odd behaviors as e.g. reported in the PR. While here,
switch to USES+=localbase instead of manually amending CPPFLAGS and LDFLAGS.
PR: 213294
and -fvisibility-inlines-hidden and uses it if available to reduce shared
library size. Unfortunately, this hides Fl_XFont_On_Demand::value, which
in turn causes undefined reference errors during linking. Fix by marking
Fl_XFont_On_Demand class as FL_EXPORT for the time being (Gentoo Linux is
doing the same)
- Do not strip libraries for now: FLTK uses the same command for installing
both dynamic and static libraries, but "fltk-config(1) --compile" prefers
static, which breaks programs like `graphics/iccexamin' (threading test)
- Pull another patch from Gentoo to fix segmentation fault in vncview(1) of
port `net/tigervnc' (Gentoo bug #529856 by Maxim Britov)
- Do not leak non-existent (build-time only) -Lfoo references in LDFLAGS as
it can potentially lead to various sporadic bugs in programs which obtain
these flags through fltk-config(1). To prevent build-time conflicts that
this gimmick was originally used for, we change shared linking command of
fluid(1) accordingly
Since these are one-line changes likely to be merged upstream sooner rather
than later, perform them with sed(1) instead of trivial patches.
Reported and reviewed by: antoine
knobs; resurrect documentation/Makefile patch; bring back recalculation of
${ETCDIR} length in Fl_Preferences.cxx, trim CONFIGURE_ARGS (again), etc.
- Restore correct LDFLAGS+= -L../src (previously -L${WRKSRC}/src), otherwise
the build fails with e.g. previous version of FLTK libraries installed
- Take maintainership to prevent Kato's monkeys jeopardizing it all over again
- Install code examples as PORTEXAMPLES (we do not install any documentation
because it requires LaTeX stack to build)
- Since `post-patch' target is a bit too verbose now, make it .SILENT
- Do not compile things with -Os by default (already honors user's CXXFLAGS)
- Do not blindly strip libraries and fluid(1) binary (respect ${STRIP} value)
- Sort the knobs, remove "default=yes" arguments from CONFIGURE_ARGS
- In post-staging era, conditional installation of documentation is no
longer necessary, so simplify the patch and drop post-install target
Specifically, newer autoconf (> 2.13) has different semantic of the
configure target. In short, one should use --build=CONFIGURE_TARGET
instead of CONFIGURE_TARGET directly. Otherwise, you will get a warning
and the old semantic may be removed in later autoconf releases.
To workaround this issue, many ports hack the CONFIGURE_TARGET variable
so that it contains the ``--build='' prefix.
To solve this issue, under the fact that some ports still have
configure script generated by the old autoconf, we use runtime detection
in the do-configure target so that the proper argument can be used.
Changes to Mk/*:
- Add runtime detection magic in bsd.port.mk
- Remove CONFIGURE_TARGET hack in various bsd.*.mk
- USE_GNOME=gnometarget is now an no-op
Changes to individual ports, other than removing the CONFIGURE_TARGET hack:
= pkg-plist changed (due to the ugly CONFIGURE_TARGET prefix in * executables)
- comms/gnuradio
- science/abinit
- science/elmer-fem
- science/elmer-matc
- science/elmer-meshgen2d
- science/elmerfront
- science/elmerpost
= use x86_64 as ARCH
- devel/g-wrap
= other changes
- print/magicfilter
GNU_CONFIGURE -> HAS_CONFIGURE since it's not generated by autoconf
Total # of ports modified: 1,027
Total # of ports affected: ~7,000 (set GNU_CONFIGURE to yes)
PR: 126524 (obsoletes 52917)
Submitted by: rafan
Tested on: two pointyhat 7-amd64 exp runs (by pav)
Approved by: portmgr (pav)
after the last amd64-7 build.
I'm doing this one myself because I was the one that marked it broken.
PR: ports/117843
Submitted by: Dmitry Marakasov <amdmi3 at amdmi3 dot ru>
Approved by: portmgr (self)
