- Disable hdf5 support (science/hdf-java is broken after recent update) and
remove BROKEN
- Take maintainership. Not that I do have time to properly maintain the port,
but I can assist with updates.
a zeising, kwm production, with help from dumbbell, bdrewery:
NEW XORG ON FREEBSD 9-STABLE AND 10-STABLE
This update switches over to use the new xorg stack by default on FreeBSD 9
and 10 stable, on osversions where vt(9) is available.
It is still possible to use the old stack by specifying WITHOUT_NEW_XORG in
/etc/make.conf .
FreeBSD 8-STABLE and released versions of FreeBSD still use
the old version.
A package repository with binary packages for new xorg will
be available soon.
This patch also contains updates of libxcb and related ports, pixman, as well
as some drivers and utilities.
Bump portrevisions for xf86-* ports, as well as virtualbox-ose-additions due
to xserver version change.
Apart from these updates, the way shared libraries are handled has been
changed for all xorg ports, as well as libxml2 and freetype, which means
ltverhack is gone and as a consequence shared libraries have been bumped.
The plan is that this change will make library bumps less likely in the
future.
All affected ports have had their portrevisions bumped as a consequence of
this.
Fix some issues where WITH_NEW_XORG weren't detected properly on CURRENT.
Update instructions, hardware support, and more notes can be found on
https://wiki.freebsd.org/Graphics
Thanks to: all testers, bdrewery and the FreeBSD x11@ team
exp-run by: bdrewery [1]
PR: ports/187602 [1]
Approved by: portmgr (bdrewery), core (jhb)
checking minimal version (1.4.1.0) of jgraphx... yes
checking jhdf5... /usr/local/share/java/classes//jhdf5.jar
checking for h5JNIFatalError in -ljhdf5... no
configure: error: libjhdf5: Library missing (Cannot find symbol h5JNIFatalError). Check if libjhdf5 - C/Java (JNI) interface for HDF5 - is installed and if the version is correct. Note that you might have to update etc/librarypath.xml to provide the actual path the the JNI libraries.
===> Script "configure" failed unexpectedly.
Reported by: pkg-fallout
- bump PORTREVISION after math/matio update
math/dynare:
- fix configure check for matio
- while here, remove needless patch for configure.ac, and convert to USES=tar:xz
Approved by: Fernando Apesteguia (maintainer)
math/scilab:
- mark BROKEN when building with MATIO option (doesn't support new matio)
science/kst2:
- fix dependency type for matio
- add dependence on pkgconfig
GCC 4.6.4 to GCC 4.7.3. This entails updating the lang/gcc port as
well as changing the default in Mk/bsd.default-versions.mk.
Part II, Bump PORTREVISIONs.
PR: 182136
Supported by: Christoph Moench-Tegeder <cmt@burggraben.net> (fixing many ports)
Tested by: bdrewery (two -exp runs)
- Switched to automake 1.11.6, see CVE-2012-3386.
- #14669: Fixed extraction of CC from gmp.h.
- Fixed case of intermediate zero real or imaginary part in mpc_fma,
found by hydra with GMP_CHECK_RANDOMIZE=1346362345.
This is on top of the following changes from version 1.0
- Licence change towards LGPLv3+ for the code and GFDLv1.3+ (with no
invariant sections) for the documentation.
- 100% of all lines are covered by tests
- Renamed functions
. mpc_mul_2exp to mpc_mul_2ui
. mpc_div_2exp to mpc_div_2ui
- 0^0, which returned (NaN,NaN) previously, now returns (1,+0).
- Removed compatibility with K&R compilers, which was untestable due
to lack of such compilers.
- New functions
. mpc_log10
. mpc_mul_2si, mpc_div_2si
- Speed-ups
. mpc_fma
- Bug fixes
. mpc_div and mpc_norm now return a value indicating the effective
rounding direction, as the other functions.
. mpc_mul, mpc_sqr and mpc_norm now return correct results even if
there are over- or underflows during the computation.
. mpc_asin, mpc_proj, mpc_sqr: Wrong result when input variable has
infinite part and equals output variable is corrected.
. mpc_fr_sub: Wrong return value for imaginary part is corrected.
Convert to the new LIB_DEPENDS standard and remove hard-coded
.so versions from a couple of dependent ports.
Bump PORTREVISIONS of all dependent ports.
PR: 183141
Approved by: portmgr (bdrewery)
vendors:
Remove any reference to bsdjava and freebsd Java Vendors, as the both are
outdated, have security issues and expired.
Mark icedtea6-stubs as broken as it doesn't build with any of the left supported
vendors.
Switch tuxguitar to be able to use openjdk16 [1]
Approved by: gahr [1]
unzip dependency.
Add a post-extract target to copy the .jar files from distfiles to the
thirdparty directory instead of doing this as part of post-patch.
Slightly re-order the contents of Makefile to get rid of some new
portlint warnings introduced in rev 1.89.
Reported by: pav@ and bapt@ from pointyhat error logs
Approved by: Rob Farmer <rfarmer AT predatorlabs.net> (maintainer)
the thirdparty build directory for the scilab port. This allows the
removal of the textproc/saxon build dependency from the scilab port so
that scilab can be built on a host that has apache-openoffice-3.4.*
installed. The latter port has a run dependency on textproc/saxon-devel,
which conflicts with textproc/saxon.
PR: 168398
Approved by: Rob Farmer <rfarmer AT predatorlabs.net> (maintainer)
- make use of optimizations by default, where possible
(quad flavor to follow shortly, and mpi at a later date)
Suggestions and patches from: ache, C. Burns, wen
- Relocate JGraphx substitution to allow running Scilab Java components
from WRKDIR before install (necessary to build docs)
- DATADIR/.atoms shouldn't be GUI only
PR: ports/153402
Submitted by: Rob Farmer (maintainer)
Upstream changes:
http://www.scilab.org/content/download/1791/19327/file/changes5.2.2-5.3.0.pdf
Port changes:
-Drop dependency on readline from ports - this was never removed from the 4.x
port and isn't used
-Make UMFPACK off by default - the only reason it was on by default was for
consistency among 5.2.x versions, since I didn't expose the knob through
OPTIONS properly until 5.2.2
-Update JGraphX to compatible version for Scilab 5.3.0
-Scilab does not start java in the normal way (with javavmwrapper) so the
JAVA_HOME environment variable isn't set properly. Explictly set JAVA_HOME
to whatever was used to build the port. This seems to correct occasional
crashes on startup.
-Use batik-all.jar instead of batik.jar - has some extra classes in it
-Additional dependency: java/jakarta-commons-beanutils
-"bsdjava" (java/jdk16) doesn't work with this version. Require diablo or
openjdk.
-Use COMMENT in DESKTOP_ENTRIES
-Switch to a dynamic plist for include/scilab, lib/scilab, and share/scilab.
These directories vary quite a bit depending on what options are set but
should be exclusive to scilab.
-Reorganize post-patch section so java related substitutions aren't done
unless building with java
-Fix NOPORTDOCS to actually do something
-Silence post-patch output
-Remove rename of Visual Basic demos - BSD make couldn't handle the unquoted
spaces in the path names, but this port now uses GNU make
-Remove unnecessary find statements in pre-install
-Remove empty directories in DATADIR/modules - toolboxes don't like them
-Fix where threads incorrectly detected as Linux style
-A couple patches are no longer necessary
-ATOMS (package manager for downloading toolboxes in binary form) is not
available for FreeBSD. It tries to download a list of available packages
from the Scilab website (based on uname). There is no list for FreeBSD and
it HTTP 404's. Scilab misinterprets this and says "Please connect to the
Internet", which is a bit misleading. Patch to indicate that this will never
work. I considered faking the uname to Linux, but there are no RPMs in ports
for math-type things, like fortran, blas/lapack, etc. so I doubt anything
would run. Porting any desired toolboxs is probably the best route.
-Fix ATOMS to use fetch instead of requiring wget or curl (Due to the
previous issue, this will only be useful for 3rd party repositories)
-Fix toolbox build macros to not write to PREFIX/share/scilab when building
as root
PR: ports/153274
Submitted by: Rob Farmer (maintainer)
