[UPDATE] lang/python/3.6
Hi, attached is the diff to update to python 3.6.7. Please test this diff in a bulk build. Cheers, Remi. Index: Makefile === RCS file: /cvs/ports/lang/python/3.6/Makefile,v retrieving revision 1.13 diff -u -p -u -p -r1.13 Makefile --- Makefile18 Sep 2018 16:43:12 - 1.13 +++ Makefile22 Oct 2018 05:03:40 - @@ -6,10 +6,9 @@ # Python itself. VERSION = 3.6 -PATCHLEVEL = .6 +PATCHLEVEL = .7 SHARED_LIBS = python3.6m 0.0 VERSION_SPEC = >=3.6,<3.7 -REVISION = 1 CONFIGURE_ARGS += --with-ensurepip=no CONFIGURE_ARGS += --enable-loadable-sqlite-extensions Index: distinfo === RCS file: /cvs/ports/lang/python/3.6/distinfo,v retrieving revision 1.7 diff -u -p -u -p -r1.7 distinfo --- distinfo2 Jul 2018 04:56:07 - 1.7 +++ distinfo22 Oct 2018 05:03:40 - @@ -1,2 +1,2 @@ -SHA256 (Python-3.6.6.tgz) = fVba32x9kqI4cCOJ6Az+Zvv65z5YQYntb4nHW78+2lg= -SIZE (Python-3.6.6.tgz) = 22930752 +SHA256 (Python-3.6.7.tgz) = t8Nvftj3FDssRhU7czLbIidmn1g+oMznU/rPVJ0aQjk= +SIZE (Python-3.6.7.tgz) = 22969142 Index: pkg/PLIST-idle === RCS file: /cvs/ports/lang/python/3.6/pkg/PLIST-idle,v retrieving revision 1.5 diff -u -p -u -p -r1.5 PLIST-idle --- pkg/PLIST-idle 3 Apr 2018 17:53:48 - 1.5 +++ pkg/PLIST-idle 22 Oct 2018 05:03:40 - @@ -55,9 +55,6 @@ lib/python3.6/idlelib/__pycache__/browse lib/python3.6/idlelib/__pycache__/calltip_w.cpython-36.opt-1.pyc lib/python3.6/idlelib/__pycache__/calltip_w.cpython-36.opt-2.pyc lib/python3.6/idlelib/__pycache__/calltip_w.cpython-36.pyc -lib/python3.6/idlelib/__pycache__/calltips.cpython-36.opt-1.pyc -lib/python3.6/idlelib/__pycache__/calltips.cpython-36.opt-2.pyc -lib/python3.6/idlelib/__pycache__/calltips.cpython-36.pyc lib/python3.6/idlelib/__pycache__/codecontext.cpython-36.opt-1.pyc lib/python3.6/idlelib/__pycache__/codecontext.cpython-36.opt-2.pyc lib/python3.6/idlelib/__pycache__/codecontext.cpython-36.pyc @@ -199,9 +196,6 @@ lib/python3.6/idlelib/__pycache__/tree.c lib/python3.6/idlelib/__pycache__/undo.cpython-36.opt-1.pyc lib/python3.6/idlelib/__pycache__/undo.cpython-36.opt-2.pyc lib/python3.6/idlelib/__pycache__/undo.cpython-36.pyc -lib/python3.6/idlelib/__pycache__/windows.cpython-36.opt-1.pyc -lib/python3.6/idlelib/__pycache__/windows.cpython-36.opt-2.pyc -lib/python3.6/idlelib/__pycache__/windows.cpython-36.pyc lib/python3.6/idlelib/__pycache__/zoomheight.cpython-36.opt-1.pyc lib/python3.6/idlelib/__pycache__/zoomheight.cpython-36.opt-2.pyc lib/python3.6/idlelib/__pycache__/zoomheight.cpython-36.pyc @@ -214,7 +208,6 @@ lib/python3.6/idlelib/autocomplete_w.py lib/python3.6/idlelib/autoexpand.py lib/python3.6/idlelib/browser.py lib/python3.6/idlelib/calltip_w.py -lib/python3.6/idlelib/calltips.py lib/python3.6/idlelib/codecontext.py lib/python3.6/idlelib/colorizer.py lib/python3.6/idlelib/config-extensions.def @@ -267,9 +260,6 @@ lib/python3.6/idlelib/idle_test/__pycach lib/python3.6/idlelib/idle_test/__pycache__/test_browser.cpython-36.opt-1.pyc lib/python3.6/idlelib/idle_test/__pycache__/test_browser.cpython-36.opt-2.pyc lib/python3.6/idlelib/idle_test/__pycache__/test_browser.cpython-36.pyc -lib/python3.6/idlelib/idle_test/__pycache__/test_calltips.cpython-36.opt-1.pyc -lib/python3.6/idlelib/idle_test/__pycache__/test_calltips.cpython-36.opt-2.pyc -lib/python3.6/idlelib/idle_test/__pycache__/test_calltips.cpython-36.pyc lib/python3.6/idlelib/idle_test/__pycache__/test_colorizer.cpython-36.opt-1.pyc lib/python3.6/idlelib/idle_test/__pycache__/test_colorizer.cpython-36.opt-2.pyc lib/python3.6/idlelib/idle_test/__pycache__/test_colorizer.cpython-36.pyc @@ -381,7 +371,6 @@ lib/python3.6/idlelib/idle_test/mock_tk. lib/python3.6/idlelib/idle_test/test_autocomplete.py lib/python3.6/idlelib/idle_test/test_autoexpand.py lib/python3.6/idlelib/idle_test/test_browser.py -lib/python3.6/idlelib/idle_test/test_calltips.py lib/python3.6/idlelib/idle_test/test_colorizer.py lib/python3.6/idlelib/idle_test/test_config.py lib/python3.6/idlelib/idle_test/test_config_key.py @@ -445,6 +434,5 @@ lib/python3.6/idlelib/textview.py lib/python3.6/idlelib/tooltip.py lib/python3.6/idlelib/tree.py lib/python3.6/idlelib/undo.py -lib/python3.6/idlelib/windows.py lib/python3.6/idlelib/zoomheight.py lib/python3.6/idlelib/zzdummy.py Index: pkg/PLIST-main === RCS file: /cvs/ports/lang/python/3.6/pkg/PLIST-main,v retrieving revision 1.8 diff -u -p -u -p -r1.8 PLIST-main --- pkg/PLIST-main 2 Jul 2018 04:56:07 - 1.8 +++ pkg/PLIST-main 22 Oct 2018 05:03:41 - @@ -2219,10 +2219,98 @@ lib/python3.6/http/client.py lib/python3.6/http/cookiejar.py lib/python3.6/http/cookies.py
Re: NEW: math/mlpack (and dependency math/armadillo)
Jeremie Courreges-Anglas [2018-10-21, 17:25:19]: > On Sun, Oct 21 2018, Marc Espie wrote: > > Here's a revised version of the armadillo port, adding arpack and hdf5 > > as a dependency. > > > > And the missed COMPILER line, oops. > > > > Notes: > > - I definitely prefer to be explicit for CONFIGURE_STYLE > > grep says that most ports rely on the cmake module setting up > CONFIGURE_STYLE. Just sayin'... > > > - with arpack added, no need for explicit depends on blas/lapack, > > as arpack needs them anyway. > > Not sure what's the point of trimming the list of deps, especially as > armadillo explicitely references code from blas, lapack and arpack. > I don't think it makes maintenance any easier, rather the opposite. > That said... > > I like this updated version, ok jca@ > > Regarding libgfortran, > > > - the library definitely requires linking with the gfortran library > > because some of lapack blas want symbols in there > > Yeah I see. IIUC blas, apack and arpack don't register the dep on > libgfortran because of a build system quirk (linking is done with with > cc instead of gfortran). A simple diff like below would register the > dep on libgfortran, and remove the need for patch-CMakeLists_txt. > > cc'ing steven@ (maintainer) > > Thoughts, ok? > > > Index: blas/Makefile > === > RCS file: /cvs/ports/math/blas/Makefile,v > retrieving revision 1.27 > diff -u -p -r1.27 Makefile > --- blas/Makefile 13 Nov 2017 06:56:38 - 1.27 > +++ blas/Makefile 21 Oct 2018 14:31:16 - > @@ -4,6 +4,7 @@ COMMENT= Basic Linear Algebra Subprogram > > VERSION= 3.7.1 > DISTNAME=blas-${VERSION} > +REVISION=0 > > SHARED_LIBS= blas2.1 > > Index: blas/files/Makefile > === > RCS file: /cvs/ports/math/blas/files/Makefile,v > retrieving revision 1.3 > diff -u -p -r1.3 Makefile > --- blas/files/Makefile 13 Nov 2017 06:56:38 - 1.3 > +++ blas/files/Makefile 21 Oct 2018 14:31:16 - > @@ -25,6 +25,7 @@ SRCS = caxpy.f ccopy.f cdotc.f cdotu. > zhpmv.f zhpr.f zhpr2.f zrotg.f zscal.f zswap.f zsymm.f \ > zsyr2k.f zsyrk.f ztbmv.f ztbsv.f ztpmv.f ztpsv.f ztrmm.f \ > ztrmv.f ztrsm.f ztrsv.f xerbla_array.f > +LDADD = -lgfortran > > printsrc: > @echo ${SRCS} How about just linking with gfortran by passing CC to make? (blas diff below, but similar for the other ports) I didn't check if dependent ports are happy with that in terms of WANTLIB etc. Index: Makefile === RCS file: /cvs/ports/math/blas/Makefile,v retrieving revision 1.27 diff -u -p -u -r1.27 Makefile --- Makefile13 Nov 2017 06:56:38 - 1.27 +++ Makefile21 Oct 2018 21:40:59 - @@ -4,6 +4,7 @@ COMMENT=Basic Linear Algebra Subprogram VERSION= 3.7.1 DISTNAME= blas-${VERSION} +REVISION= 0 SHARED_LIBS= blas2.1 @@ -30,7 +31,8 @@ BUILD_DEPENDS=${MODFORTRAN_BUILD_DEPEND MAKE_FILE= ${FILESDIR}/Makefile MAKE_ENV= SHLIB_MAJOR=${LIBblas_VERSION:R} \ SHLIB_MINOR=${LIBblas_VERSION:E} \ - FC="${MODFORTRAN_COMPILER} -cpp" + FC="${MODFORTRAN_COMPILER} -cpp" \ + CC=${MODFORTRAN_COMPILER} FAKE_FLAGS=LIBDIR=${LOCALBASE}/lib DEBUGLIBS=no USE_GROFF= Yes
amd64: build failures with lld (2018-10-20)
kettenis@'s latest lld fix has allowed building gcc49, which has unlocked numerous ports. Here is the current fallout from a test build with lld; newly revealed failures are marked with '+': +audio/audacity undefined symbol: g_signal_connect_data benchmarks/wrk string table non-null terminated databases/pgmodelerundefined symbol: backtrace devel/ffcall can't create dynamic relocation R_X86_64_64 editors/emacs21segfault +games/flightgear/base undefined symbol: XOpenDisplay +games/pingus undefined symbol: XFree +games/pokerth undefined symbol: SSL_library_init games/tome4undefined symbol: _Unwind_GetCFA +games/valyriatear undefined symbol: libiconv_open games/warzone2100 undefined symbol: ogg_sync_init lang/crystal undefined symbol: OPENSSL_add_all_algorithms_noconf lang/fpc invalid alignment of section headers +mail/evolution-rss edbus-private not found misc/rocrail undefined symbol: operator new(unsigned long) +net/bitcoin,no_x11 -shared and -pie may not be used together net/telepathy/folksedbus-private not found productivity/glabels edbus-private not found +productivity/ledgereditline/readline.h not found security/xca undefined symbol: _Unwind_Resume sysutils/firmware/vmm ld "does not properly handle alignments" sysutils/memtest86+unknown argument: --warn-constructors x11/gnustep/terminal undefined symbol: libiconv The error logs are available under http://build-failures.rhaalovely.net/amd64-clang/2018-10-20/ To try this yourself, do "ln -f /usr/bin/ld.lld /usr/bin/ld" and tweak ports/infrastructure/mk/arch-defines.mk like this: -LLD_ARCHS = aarch64 arm +LLD_ARCHS = aarch64 amd64 arm -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: Update of p5-Test-Base, p5-Test-YAML, p5-YAML, p5-YAML-Tiny and p5-YAML-XS
On Sun, Oct 21, 2018 at 04:05:48AM +, wen heping wrote: > Hi, ports@: > > Here is a patch to update : > devel/p5-Test-Base --> 0.89 I think this needs a new TEST_DEPENDS of p5-Algorithm-Diff, see below > devel/p5-Test-YAML --> 1.07 OK afresh1@ > devel/p5-YAML --> 1.26 This needed some TEST_DEPENDS fixes for me, see below > devel/p5-YAML-Tiny --> 1.73 Looks like most of the test depends moved to the xt/ directory so we don't need them, but it did need p5-JSON-MaybeXS, see below. > devel/p5-YAML-XS --> 0.74 OK afresh1@ (and I can take MAINTAINER on it, as I'm already p5-YAML) >All patch build and test well on my amd64 system. > >OK ? I didn't run reverse dependency tests on these, but may get to trying that later, however I expect it'll be OK and we can solve any fallout before the next release. Index: Makefile === RCS file: /cvs/ports/devel/p5-Test-Base/Makefile,v retrieving revision 1.11 diff -u -p -r1.11 Makefile --- Makefile20 Mar 2016 19:56:36 - 1.11 +++ Makefile21 Oct 2018 19:49:02 - @@ -5,15 +5,16 @@ COMMENT= data driven test framework MODULES= cpan PKG_ARCH= * -DISTNAME= Test-Base-0.88 +DISTNAME= Test-Base-0.89 CATEGORIES=devel # perl PERMIT_PACKAGE_CDROM= Yes -RUN_DEPENDS= devel/p5-Spiffy +RUN_DEPENDS= devel/p5-Spiffy>=0.40 -TEST_DEPENDS= devel/p5-Test-Deep \ - devel/p5-YAML +TEST_DEPENDS= devel/p5-Algorithm-Diff>=1.15 \ + devel/p5-Test-Deep \ + textproc/p5-Text-Diff>=0.35 .include Index: distinfo === RCS file: /cvs/ports/devel/p5-Test-Base/distinfo,v retrieving revision 1.7 diff -u -p -r1.7 distinfo --- distinfo19 Apr 2015 12:41:24 - 1.7 +++ distinfo21 Oct 2018 19:49:02 - @@ -1,2 +1,2 @@ -SHA256 (Test-Base-0.88.tar.gz) = UjaMxanLvE6roz7YIGcvkgAbc9i8ugu5XV/bHTcLkDk= -SIZE (Test-Base-0.88.tar.gz) = 52032 +SHA256 (Test-Base-0.89.tar.gz) = J5Shqq6x06KH3SxyhiWGY3llYvfbnMxrQkvE8d6K0BQ= +SIZE (Test-Base-0.89.tar.gz) = 52105 Index: Makefile === RCS file: /cvs/ports/devel/p5-YAML/Makefile,v retrieving revision 1.20 diff -u -p -r1.20 Makefile --- Makefile23 Nov 2016 00:47:06 - 1.20 +++ Makefile21 Oct 2018 20:11:30 - @@ -5,7 +5,7 @@ COMMENT = YAML Ain't Markup Language MODULES = cpan PKG_ARCH = * -DISTNAME = YAML-1.19 +DISTNAME = YAML-1.26 CATEGORIES = devel MAINTAINER = Andrew Fresh @@ -13,11 +13,15 @@ MAINTAINER =Andrew Fresh =1.05 \ - devel/p5-Test-Pod -MAKE_ENV +=RELEASE_TESTING=1 +TEST_DEPENDS = devel/p5-Test-Deep \ + devel/p5-Test-YAML>=1.05 + +# Additional depends to avoid skipping tests +TEST_ENV +=AUTHOR_TESTING=1 +TEST_DEPENDS +=devel/p5-Test-Pod .include Index: distinfo === RCS file: /cvs/ports/devel/p5-YAML/distinfo,v retrieving revision 1.12 diff -u -p -r1.12 distinfo --- distinfo23 Nov 2016 00:47:06 - 1.12 +++ distinfo21 Oct 2018 20:05:15 - @@ -1,2 +1,2 @@ -SHA256 (YAML-1.19.tar.gz) = MBD2ULDxehIKNMKj3N/J7iux/hRapSMFk266nAEgAFw= -SIZE (YAML-1.19.tar.gz) = 81305 +SHA256 (YAML-1.26.tar.gz) = +i+Z1UxK+8WvnaYyJgnStpfQIAtrzm/fZEr/QkKVf7w= +SIZE (YAML-1.26.tar.gz) = 86698 Index: Makefile === RCS file: /cvs/ports/devel/p5-YAML-Tiny/Makefile,v retrieving revision 1.13 diff -u -p -r1.13 Makefile --- Makefile4 Dec 2017 17:56:27 - 1.13 +++ Makefile21 Oct 2018 19:43:08 - @@ -4,17 +4,13 @@ COMMENT = read/write YAML files with as MODULES = cpan PKG_ARCH = * -DISTNAME = YAML-Tiny-1.70 +DISTNAME = YAML-Tiny-1.73 CATEGORIES = devel -FIX_EXTRACT_PERMISSIONS= Yes # Perl PERMIT_PACKAGE_CDROM = Yes -TEST_DEPENDS = devel/p5-YAML \ - devel/p5-YAML-Perl \ - devel/p5-YAML-Syck \ - devel/p5-YAML-XS +TEST_DEPENDS = converters/p5-JSON-MaybeXS>=1.001000 MAKE_ENV +=AUTOMATED_TESTING=1 Index: distinfo === RCS file: /cvs/ports/devel/p5-YAML-Tiny/distinfo,v retrieving revision 1.7 diff -u -p -r1.7 distinfo --- distinfo29 Apr 2017 13:26:12 - 1.7 +++ distinfo21 Oct 2018 19:43:08 - @@ -1,2 +1,2 @@ -SHA256 (YAML-Tiny-1.70.tar.gz) = u85LUrXq/bBOMEOXWgjb85TQC30slYrbnQPZ9+kpElU= -SIZE (YAML-Tiny-1.70.tar.gz) = 72663 +SHA256 (YAML-Tiny-1.73.tar.gz) = vDFfoS6PHj7l4vQw2Qtwil3H5HyGfbqNzjprj74ld0Q= +SIZE (YAML-Tiny-1.73.tar.gz) = 73708
[UPDATE] devel/p5-MRO-Compat to 0.13
This is a fairly minor update, mostly to deal with the lack of "." in @INC. It also moves the Pod tests to an "xt" directory so they don't even try to run normally. It does remove the dependency on p5-Class-C3 as that is ignored if the perl version >= 5.9.5 or greater (which ours is). There is one difference in the reverse dependency test results, p5-Catalyst-View-TT went from failing to passing, so yay! (I'm not quite sure why, but didn't dig too deep) There are also still some reverse dependency failures: * p5-Catalyst-Plugin-Authentication which seems to need an of a dependency * p5-Catalyst-Plugin-Unicode which is the same issue with Class::MOP::load_class being deprecated * p5-Catalyst-Controller-HTML-FormFu which is again the same * p5-Catalyst-Plugin-PageCache same again * p5-Catalyst-Plugin-Session and again * p5-Catalyst-Runtime-5.90006 and again, probably should look here * p5-HTML-FormFu which apparently uses qw() in a for loop without parens and so needs an update. And some similar failures that were dependencies of p5-Devel-GlobalDestruction: * p5-Moose * p5-MooseX-Daemonize * p5-MooseX-Getopt * p5-MooseX-POE https://marc.info/?l=openbsd-ports=154005860118583=2 (Catalyst and Moose are both mine, so will need to get to looking at an update for those) OK? Index: Makefile === RCS file: /cvs/ports/devel/p5-MRO-Compat/Makefile,v retrieving revision 1.12 diff -u -p -r1.12 Makefile --- Makefile20 Mar 2016 19:56:27 - 1.12 +++ Makefile21 Oct 2018 18:56:27 - @@ -2,10 +2,10 @@ COMMENT = mro::* interface compatibility for Perl < 5.9.5 -DISTNAME = MRO-Compat-0.12 +DISTNAME = MRO-Compat-0.13 CATEGORIES = devel -CPAN_AUTHOR = BOBTFISH +CPAN_AUTHOR = HAARG MAINTAINER = Andrew Fresh @@ -16,11 +16,5 @@ MODULES =cpan PKG_ARCH = * CONFIGURE_STYLE = modinst - -RUN_DEPENDS = devel/p5-Class-C3>=0.20 - -# Optional depends to avoid skipping tests -TEST_DEPENDS = devel/p5-Test-Pod \ - devel/p5-Test-Pod-Coverage .include Index: distinfo === RCS file: /cvs/ports/devel/p5-MRO-Compat/distinfo,v retrieving revision 1.4 diff -u -p -r1.4 distinfo --- distinfo10 May 2014 17:15:50 - 1.4 +++ distinfo21 Oct 2018 18:56:27 - @@ -1,2 +1,2 @@ -SHA256 (MRO-Compat-0.12.tar.gz) = u6W5OGmqU3oziZSWadaC8EfTAU1TvDotcgnGgZ5QFdY= -SIZE (MRO-Compat-0.12.tar.gz) = 24230 +SHA256 (MRO-Compat-0.13.tar.gz) = iiw7bMwZMo1VedAqfZEoXir9hdgB9J1COo6xbzI9pPg= +SIZE (MRO-Compat-0.13.tar.gz) = 8711
[PATCH] update duplicity to 0.7.18.2
Hey, I just upgrade to OpenBSD 6.4 on my server and noticed my duplicity backup job fails. The current port installs version 0.7.18.1, which broke remote backups to at least SFTP locations[1]. The update to 0.7.18.2 was released to fix this problem. Attached is a patch to bump the port version. I don't know the exact protocol for this, but it may be desirable to backport this change to 6.4 STABLE. Best regards, Henrik Index: Makefile === RCS file: /cvs/ports/sysutils/duplicity/Makefile,v retrieving revision 1.48 diff -u -p -r1.48 Makefile --- Makefile 3 Sep 2018 01:16:06 - 1.48 +++ Makefile 21 Oct 2018 19:03:29 - @@ -5,7 +5,7 @@ COMMENT = encrypted backup using rsync algorithm -MODPY_EGG_VERSION = 0.7.18.1 +MODPY_EGG_VERSION = 0.7.18.2 DISTNAME = duplicity-${MODPY_EGG_VERSION} CATEGORIES = sysutils Index: distinfo === RCS file: /cvs/ports/sysutils/duplicity/distinfo,v retrieving revision 1.30 diff -u -p -r1.30 distinfo --- distinfo 3 Sep 2018 01:16:06 - 1.30 +++ distinfo 21 Oct 2018 19:03:29 - @@ -1,2 +1,2 @@ -SHA256 (duplicity-0.7.18.1.tar.gz) = yTUBntlT5HZ9+NOXZcTdQRmHCaFGaOgj4unj4gcQgJ0= -SIZE (duplicity-0.7.18.1.tar.gz) = 1726064 +SHA256 (duplicity-0.7.18.2.tar.gz) = wjaIj0MSjpbNMwF7AaKFXA4kc4GV/tXK2tCMKP1rZ0g= +SIZE (duplicity-0.7.18.2.tar.gz) = 1726157
devel/sdl2: fix for macppc init crash
Hi, George Koehler ( kernigh () gmail DOT com ) reports that all devel/sdl2 crash on init on macppc. The reason seems to be an inappropriate bool type. The patch below for SDL_x11keyboard.c fixes the issue according to George. This has been merged by upstream along with another change to the types for XkbSetDetectableAutoRepeat. The diff below includes both patches, along with formatting of the Makefile to align it with more commonly used formatting (space before '='). While this may be harder to review, the Makefile only contains REVISION bump and a change of MAINTAINER email address. I tested sdl2 with these changes on amd64 with cataclysm-dda, milkytracker, citra, dolphin, endless-sky. All of them run on brief testing without issues. ok to import? Index: Makefile === RCS file: /cvs/ports/devel/sdl2/Makefile,v retrieving revision 1.22 diff -u -p -r1.22 Makefile --- Makefile4 Sep 2018 12:46:11 - 1.22 +++ Makefile21 Oct 2018 18:11:03 - @@ -1,52 +1,52 @@ # $OpenBSD: Makefile,v 1.22 2018/09/04 12:46:11 espie Exp $ -COMMENT= cross-platform multimedia library -BROKEN-hppa= src/atomic/SDL_spinlock.c:101:2: error: \ +COMMENT = cross-platform multimedia library +BROKEN-hppa = src/atomic/SDL_spinlock.c:101:2: error: \ #error Please implement for your platform. -V= 2.0.8 -DISTNAME= SDL2-${V} -PKGNAME= sdl2-${V} -CATEGORIES=devel -MASTER_SITES= https://www.libsdl.org/release/ -REVISION= 0 +V =2.0.8 +DISTNAME = SDL2-${V} +PKGNAME = sdl2-${V} +CATEGORIES = devel +MASTER_SITES = https://www.libsdl.org/release/ +REVISION = 1 -SHARED_LIBS= SDL20.5 # 0.8 +SHARED_LIBS = SDL20.5 # 0.8 -HOMEPAGE= https://www.libsdl.org/ +HOMEPAGE = https://www.libsdl.org/ -MAINTAINER=Thomas Frohwein +MAINTAINER = Thomas Frohwein # zlib -PERMIT_PACKAGE_CDROM= Yes +PERMIT_PACKAGE_CDROM = Yes -WANTLIB= m pthread sndio usbhid samplerate +WANTLIB += m pthread sndio usbhid samplerate # GL/X11/Xext/Xrender/Xrandr are dlopen'd by SDL -WANTLIB+= GL X11 Xau Xdmcp Xext Xrandr Xrender xcb +WANTLIB += GL X11 Xau Xdmcp Xext Xrandr Xrender xcb -LIB_DEPENDS= audio/libsamplerate +LIB_DEPENDS = audio/libsamplerate -USE_GMAKE= Yes -SEPARATE_BUILD=Yes -CONFIGURE_STYLE= gnu -MODGNU_CONFIG_GUESS_DIRS= ${WRKSRC} ${WRKSRC}/build-scripts -CONFIGURE_ENV+=CPPFLAGS="-I${LOCALBASE}/include" \ - LDFLAGS="-L${LOCALBASE}/lib" - -CONFIGURE_ARGS+= --disable-alsa \ ---disable-arts \ ---disable-dbus \ ---disable-esd \ ---disable-ibus \ ---disable-jack \ ---disable-libudev \ ---disable-nas \ ---disable-oss \ ---disable-pulseaudio +USE_GMAKE =Yes +SEPARATE_BUILD = Yes +CONFIGURE_STYLE = gnu +MODGNU_CONFIG_GUESS_DIRS = ${WRKSRC} ${WRKSRC}/build-scripts +CONFIGURE_ENV += CPPFLAGS="-I${LOCALBASE}/include" \ + LDFLAGS="-L${LOCALBASE}/lib" + +CONFIGURE_ARGS += --disable-alsa \ + --disable-arts \ + --disable-dbus \ + --disable-esd \ + --disable-ibus \ + --disable-jack \ + --disable-libudev \ + --disable-nas \ + --disable-oss \ + --disable-pulseaudio # in case devel/usb is installed, don't pick it up. -CONFIGURE_ENV+= ac_cv_lib_usb_hid_init=no \ - ac_cv_header_usb_h=no +CONFIGURE_ENV += ac_cv_lib_usb_hid_init=no \ + ac_cv_header_usb_h=no -NO_TEST= Yes +NO_TEST = Yes .include Index: patches/patch-src_video_x11_SDL_x11keyboard_c === RCS file: patches/patch-src_video_x11_SDL_x11keyboard_c diff -N patches/patch-src_video_x11_SDL_x11keyboard_c --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_video_x11_SDL_x11keyboard_c 21 Oct 2018 18:11:03 - @@ -0,0 +1,18 @@ +$OpenBSD$ + +fix incorrect function signature that breaks macppc platform (likely any big +endian platform): "stack overflow in function X11_InitKeyboard" +included in upstream changeset 9dd351b26dcb (future versions) + +Index: src/video/x11/SDL_x11keyboard.c +--- src/video/x11/SDL_x11keyboard.c.orig src/video/x11/SDL_x11keyboard.c +@@ -266,7 +266,9 @@ X11_InitKeyboard(_THIS) + int best_distance; + int best_index; + int distance; +-BOOL xkb_repeat = 0; ++Bool xkb_repeat = 0; + + X11_XAutoRepeatOn(data->display); + Index: patches/patch-src_video_x11_SDL_x11sym_h
Re: UPDATE: KDE5
On Sun, Oct 21, 2018 at 06:05:20PM +0200, Rafael Sadowski wrote: > On Fri Oct 19, 2018 at 07:07:45AM +0200, Landry Breuil wrote: > > On Wed, Oct 17, 2018 at 11:20:07PM +0200, Landry Breuil wrote: > > > On Tue, Oct 16, 2018 at 09:42:34PM +0200, Rafael Sadowski wrote: > > > > Hi All. > > > > > > > > please find below a diff to rule them all. > > > > > > > > - Update KDE Frameworks to 5.51.0 > > > > -- Change examples handling and use @sample for all of them. Idea by > > > >ajacoutot@. Discussed with ajacoutot@,sthen@,naddy@. > > > > > > > > - Update our KDE Applications to 18.08.2. > > > > -- Nothing special except okteta. They use there own version pattern > > > >now. That's why I set EPOCH. > > > > > > > > - Update all devel/kf5 consumers there are effected by the update. > > > > > > Will put it in my next bulk. > > Thanks a lot! > > > - missing REVISION bump on libkface > > Why? libkface is KDE4. Well, PLISTDB complained that its PLIST changed wrt depends: -@depend graphics/opencv,-main:opencv-*:opencv-2.4.13 -@depend lang/gcc/4.9,-libs:gcc-libs->=4.9,<4.10:gcc-libs-4.9.4p4 -@depend x11/kde4/libs,-main:kdelibs->=4.14,<5:kdelibs-4.14.10p7 -@depend x11/qt4,-main:qt4-*:qt4-4.8.7p11 +@depend graphics/opencv,-main:opencv-*:opencv-2.4.13.4p0 +@depend x11/kde4/libs,-main:kdelibs->=4.14,<5:kdelibs-4.14.10p15 +@depend x11/qt4,-main:qt4-*:qt4-4.8.7p17 *shrug* afaict, your diff removes REVISION and HOMEPAGE, and it seems x11/kde-applications/libkface doesnt follow whatever version is globally set in this subdir, as its version is libkface-15.08.3. > > - /bin/sh: python: not found in qtcreator (already fixed in cvs) > > -Error: > > The changes below have no effect on qtcreator. > > > /usr/obj/ports/krita-4.1.3/fake-amd64/usr/local/share/examples/xdg/kritarc > > does not exist in krita > > > > Thanks for the hint. New diff bewlow (Without phonon-4.10.1 bits -- > sorry for that) More issues as dpb crashes and goes..: konsole fails with: Could NOT find KF5 (missing: NewStuff NewStuffCore) (found suitable version "5.51.0", minimum required is "5.6.0") calligra failed with the same error as krita: Error: /usr/obj/mfs/calligra-3.1.0/fake-amd64/usr/local/share/examples/xdg/calligra_stencils.knsrc does not exist Error: /usr/obj/mfs/calligra-3.1.0/fake-amd64/usr/local/share/examples/xdg/calligrasheetsrc does not exist Error: /usr/obj/mfs/calligra-3.1.0/fake-amd64/usr/local/share/examples/xdg/calligrastagerc does not exist Error: /usr/obj/mfs/calligra-3.1.0/fake-amd64/usr/local/share/examples/xdg/calligrawordsrc does not exist Error: /usr/obj/mfs/calligra-3.1.0/fake-amd64/usr/local/share/examples/xdg/karbonrc does not exist Landry
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: es...@cvs.openbsd.org 2018/10/21 10:14:47 Modified files: math : Makefile Log message: link armadillo to the build
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: es...@cvs.openbsd.org 2018/10/21 10:13:58 Log message: Armadillo is a linear algebra library (matrix maths) for the C++ language, aiming towards a good balance between speed and ease of use. Provides high-level syntax and functionality deliberately similar to Matlab (dependency for mlpack) okay jca@ Status: Vendor Tag: espie Release Tags: espie_20181021 N ports/math/armadillo/distinfo N ports/math/armadillo/Makefile N ports/math/armadillo/pkg/PLIST N ports/math/armadillo/pkg/DESCR N ports/math/armadillo/patches/patch-CMakeLists_txt No conflicts created by this import
Re: UPDATE: devel/cmake (i386 build wanted)
On Sat Oct 20, 2018 at 10:56:35PM +0200, Rafael Sadowski wrote: > Hi All. > > Please find below diff to update cmake to the latest stable version. The > last time it failed on i386 so it would be great if someone could build > it on i386 and report to me/ports@. Please ignore the wrong CVS tags, > the diff based on an old diff from sthen@. > > Rafael Sadowski > Successful i386 build by Charlene Wendling and Jan Vlach. Thanks! The following tests failed on amd64: 28 - FindModulesExecuteAll (Failed) 230 - Java.Javah (Failed) 253 - CMakeOnly.AllFindModules (Failed) 399 - RunCMake.install (Failed) New in cmake 3.12.2: 231 - Java.NativeHeaders (Failed) We still have to work on this. New clean diff below: Index: Makefile === RCS file: /cvs/ports/devel/cmake/Makefile,v retrieving revision 1.171 diff -u -p -u -p -r1.171 Makefile --- Makefile13 Jul 2018 03:00:16 - 1.171 +++ Makefile21 Oct 2018 09:57:28 - @@ -4,9 +4,8 @@ DPB_PROPERTIES =parallel COMMENT = portable build system -VER = 3.10.2 +VER = 3.12.2 EPOCH =0 -REVISION = 0 DISTNAME = cmake-${VER} CATEGORIES = devel Index: distinfo === RCS file: /cvs/ports/devel/cmake/distinfo,v retrieving revision 1.53 diff -u -p -u -p -r1.53 distinfo --- distinfo14 Jun 2018 17:04:45 - 1.53 +++ distinfo21 Oct 2018 09:57:28 - @@ -1,2 +1,2 @@ -SHA256 (cmake-3.10.2.tar.gz) = gND6rUq1beB6ohp/xpLIjEzmFW1CsFecaWIASnCjIYs= -SIZE (cmake-3.10.2.tar.gz) = 7824452 +SHA256 (cmake-3.12.2.tar.gz) = D5dIV5nlGnBwzBFJTz4CNJsPw6JMwSsILnN79noFgaQ= +SIZE (cmake-3.12.2.tar.gz) = 8388114 Index: patches/patch-CMakeLists_txt === RCS file: /cvs/ports/devel/cmake/patches/patch-CMakeLists_txt,v retrieving revision 1.26 diff -u -p -u -p -r1.26 patch-CMakeLists_txt --- patches/patch-CMakeLists_txt14 Jun 2018 17:04:45 - 1.26 +++ patches/patch-CMakeLists_txt21 Oct 2018 09:57:28 - @@ -2,7 +2,7 @@ $OpenBSD: patch-CMakeLists_txt,v 1.26 20 Index: CMakeLists.txt --- CMakeLists.txt.orig +++ CMakeLists.txt -@@ -317,6 +317,15 @@ macro (CMAKE_BUILD_UTILITIES) +@@ -337,6 +337,15 @@ macro (CMAKE_BUILD_UTILITIES) CMAKE_SET_TARGET_FOLDER(${KWSYS_NAMESPACE}TestSharedForward "${kwsys_folder}") endif() @@ -18,7 +18,7 @@ Index: CMakeLists.txt #- # Setup third-party libraries. # Everything in the tree should be able to include files from the -@@ -350,7 +359,8 @@ macro (CMAKE_BUILD_UTILITIES) +@@ -370,7 +379,8 @@ macro (CMAKE_BUILD_UTILITIES) message(FATAL_ERROR "CMAKE_USE_SYSTEM_LIBRHASH is ON but LibRHash is not found!") endif() @@ -28,7 +28,7 @@ Index: CMakeLists.txt else() set(CMAKE_LIBRHASH_LIBRARIES cmlibrhash) add_subdirectory(Utilities/cmlibrhash) -@@ -516,7 +526,8 @@ macro (CMAKE_BUILD_UTILITIES) +@@ -536,7 +546,8 @@ macro (CMAKE_BUILD_UTILITIES) message(FATAL_ERROR "CMAKE_USE_SYSTEM_JSONCPP is ON but a JsonCpp is not found!") endif() @@ -38,7 +38,7 @@ Index: CMakeLists.txt else() set(CMAKE_JSONCPP_LIBRARIES cmjsoncpp) add_subdirectory(Utilities/cmjsoncpp) -@@ -531,7 +542,8 @@ macro (CMAKE_BUILD_UTILITIES) +@@ -551,7 +562,8 @@ macro (CMAKE_BUILD_UTILITIES) message(FATAL_ERROR "CMAKE_USE_SYSTEM_LIBUV is ON but a libuv is not found!") endif() Index: patches/patch-Modules_CMakeCInformation_cmake === RCS file: /cvs/ports/devel/cmake/patches/patch-Modules_CMakeCInformation_cmake,v retrieving revision 1.8 diff -u -p -u -p -r1.8 patch-Modules_CMakeCInformation_cmake --- patches/patch-Modules_CMakeCInformation_cmake 14 Jun 2018 17:04:45 - 1.8 +++ patches/patch-Modules_CMakeCInformation_cmake 21 Oct 2018 09:57:28 - @@ -2,7 +2,7 @@ $OpenBSD: patch-Modules_CMakeCInformatio Index: Modules/CMakeCInformation.cmake --- Modules/CMakeCInformation.cmake.orig +++ Modules/CMakeCInformation.cmake -@@ -165,7 +165,7 @@ include(CMakeCommonLanguageInclude) +@@ -142,7 +142,7 @@ include(CMakeCommonLanguageInclude) # create a C shared library if(NOT CMAKE_C_CREATE_SHARED_LIBRARY) set(CMAKE_C_CREATE_SHARED_LIBRARY Index: patches/patch-Modules_CMakeCXXInformation_cmake === RCS file: /cvs/ports/devel/cmake/patches/patch-Modules_CMakeCXXInformation_cmake,v retrieving revision 1.7 diff -u -p -u -p -r1.7 patch-Modules_CMakeCXXInformation_cmake --- patches/patch-Modules_CMakeCXXInformation_cmake 14 Jun 2018 17:04:45 - 1.7 +++ patches/patch-Modules_CMakeCXXInformation_cmake 21 Oct 2018 09:57:28 - @@ -1,6 +1,7 @@ $OpenBSD:
Re: NEW: math/mlpack (and dependency math/armadillo)
On Sun, Oct 21 2018, Marc Espie wrote: > Here's a revised version of the armadillo port, adding arpack and hdf5 > as a dependency. > > And the missed COMPILER line, oops. > > Notes: > - I definitely prefer to be explicit for CONFIGURE_STYLE grep says that most ports rely on the cmake module setting up CONFIGURE_STYLE. Just sayin'... > - with arpack added, no need for explicit depends on blas/lapack, > as arpack needs them anyway. Not sure what's the point of trimming the list of deps, especially as armadillo explicitely references code from blas, lapack and arpack. I don't think it makes maintenance any easier, rather the opposite. That said... I like this updated version, ok jca@ Regarding libgfortran, > - the library definitely requires linking with the gfortran library > because some of lapack blas want symbols in there Yeah I see. IIUC blas, apack and arpack don't register the dep on libgfortran because of a build system quirk (linking is done with with cc instead of gfortran). A simple diff like below would register the dep on libgfortran, and remove the need for patch-CMakeLists_txt. cc'ing steven@ (maintainer) Thoughts, ok? Index: blas/Makefile === RCS file: /cvs/ports/math/blas/Makefile,v retrieving revision 1.27 diff -u -p -r1.27 Makefile --- blas/Makefile 13 Nov 2017 06:56:38 - 1.27 +++ blas/Makefile 21 Oct 2018 14:31:16 - @@ -4,6 +4,7 @@ COMMENT=Basic Linear Algebra Subprogram VERSION= 3.7.1 DISTNAME= blas-${VERSION} +REVISION= 0 SHARED_LIBS= blas2.1 Index: blas/files/Makefile === RCS file: /cvs/ports/math/blas/files/Makefile,v retrieving revision 1.3 diff -u -p -r1.3 Makefile --- blas/files/Makefile 13 Nov 2017 06:56:38 - 1.3 +++ blas/files/Makefile 21 Oct 2018 14:31:16 - @@ -25,6 +25,7 @@ SRCS =caxpy.f ccopy.f cdotc.f cdotu. zhpmv.f zhpr.f zhpr2.f zrotg.f zscal.f zswap.f zsymm.f \ zsyr2k.f zsyrk.f ztbmv.f ztbsv.f ztpmv.f ztpsv.f ztrmm.f \ ztrmv.f ztrsm.f ztrsv.f xerbla_array.f +LDADD = -lgfortran printsrc: @echo ${SRCS} Index: lapack/Makefile === RCS file: /cvs/ports/math/lapack/Makefile,v retrieving revision 1.26 diff -u -p -r1.26 Makefile --- lapack/Makefile 13 Nov 2017 06:57:36 - 1.26 +++ lapack/Makefile 21 Oct 2018 14:31:16 - @@ -4,6 +4,7 @@ COMMENT=library of Fortran linear algeb VERSION= 3.7.1 DISTNAME= lapack-${VERSION} +REVISION= 0 SHARED_LIBS= lapack 6.0 Index: lapack/files/Makefile === RCS file: /cvs/ports/math/lapack/files/Makefile,v retrieving revision 1.7 diff -u -p -r1.7 Makefile --- lapack/files/Makefile 13 Nov 2017 06:57:36 - 1.7 +++ lapack/files/Makefile 21 Oct 2018 14:31:53 - @@ -384,6 +384,7 @@ ZLASRC = \ SRCS=${SLASRC} ${DLASRC} ${DSLASRC} ${CLASRC} ${ZLASRC} ${ZCLASRC} \ ${SCLAUX} ${DZLAUX} ${ALLAUX} +LDADD= -lgfortran printsrc: @echo ${SRCS} Index: arpack/Makefile === RCS file: /cvs/ports/math/arpack/Makefile,v retrieving revision 1.16 diff -u -p -r1.16 Makefile --- arpack/Makefile 13 Nov 2017 07:09:06 - 1.16 +++ arpack/Makefile 21 Oct 2018 14:31:16 - @@ -4,7 +4,7 @@ COMMENT=solve large scale eigenvalue pr DISTNAME= arpack96 PKGNAME= arpack-96 -REVISION= 5 +REVISION= 6 SHARED_LIBS= arpack 1.0 CATEGORIES=math Index: arpack/files/Makefile === RCS file: /cvs/ports/math/arpack/files/Makefile,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 Makefile --- arpack/files/Makefile 19 Oct 2006 16:19:54 - 1.1.1.1 +++ arpack/files/Makefile 21 Oct 2018 14:31:16 - @@ -18,5 +18,6 @@ SRCS= sgetv0.f slaqrb.f sstqrb.f ssortc. zgetv0.f zsortc.f zstatn.f \ zvout.f zmout.f \ icnteq.f icopy.f iset.f iswap.f ivout.f second.f +LDADD= -lgfortran .include -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: chrooting package build process possible?
Thanks Marc I will check those out. And I port new and updated applications as there are missing or outdated packages for my needs. On 2018-10-21 10:05, Marc Espie wrote: On Sun, Oct 21, 2018 at 09:50:24AM -0500, Edward Lopez-Acosta wrote: I have noticed that when building packages I am required to install dependencies globally which leads to a messy system if I don't remember to remove them. This is an issue when building ports that may not be installed on the same system. RTFM, See proot(1) and dpb(1) and bulk(8) Why do you build packages yourself, btw ? It takes time and space, and you can just use the trusted binary packages unless you have very good reasons to avoid them. Is it possible to somehow have the make process use a chroot and avoid installing packages globally? Slight tangent, but why do pkg_add/pkg_delete not prompt to confirm before taking actions which to me would make much more sense than just doing the action right away. This is Unix. Commands do stuff by default. Observe rm. By default, it removes files.
Re: chrooting package build process possible?
On Sun, Oct 21, 2018 at 09:50:24AM -0500, Edward Lopez-Acosta wrote: > I have noticed that when building packages I am required to install > dependencies globally which leads to a messy system if I don't remember to > remove them. This is an issue when building ports that may not be installed > on the same system. RTFM, See proot(1) and dpb(1) and bulk(8) Why do you build packages yourself, btw ? It takes time and space, and you can just use the trusted binary packages unless you have very good reasons to avoid them. > Is it possible to somehow have the make process use a chroot and avoid > installing packages globally? Slight tangent, but why do pkg_add/pkg_delete > not prompt to confirm before taking actions which to me would make much more > sense than just doing the action right away. This is Unix. Commands do stuff by default. Observe rm. By default, it removes files.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: es...@cvs.openbsd.org 2018/10/21 08:57:55 Modified files: databases/sqlports/files: TreeWalker.pm Log message: make sure we don't catch SUBDIRLIST on later calls
chrooting package build process possible?
I have noticed that when building packages I am required to install dependencies globally which leads to a messy system if I don't remember to remove them. This is an issue when building ports that may not be installed on the same system. Is it possible to somehow have the make process use a chroot and avoid installing packages globally? Slight tangent, but why do pkg_add/pkg_delete not prompt to confirm before taking actions which to me would make much more sense than just doing the action right away. Thank you.
Re: UPDATE: net/onionshare
On 10/21/18 1:51 PM, Grégoire Jadi wrote: fredl writes: Hey, Hello, Both cli and -gui works on amd64. Hey, thanks for your feedback! Thanks for the update. I think you need to drop the REVISION-gui (but I'm just a port-newb). You are right, please find the new diff attached to this mail! -- fredl Index: Makefile === RCS file: /cvs/ports/net/onionshare/Makefile,v retrieving revision 1.2 diff -u -p -r1.2 Makefile --- Makefile27 Jun 2018 21:04:00 - 1.2 +++ Makefile21 Oct 2018 13:25:00 - @@ -6,11 +6,10 @@ COMMENT-gui = graphical user interface GH_ACCOUNT = micahflee GH_PROJECT = onionshare GH_TAGNAME = v${MODPY_EGG_VERSION} -MODPY_EGG_VERSION =1.3 +MODPY_EGG_VERSION =1.3.1 PKGNAME-main = onionshare-${MODPY_EGG_VERSION} PKGNAME-gui = onionshare-gui-${MODPY_EGG_VERSION} -REVISION-gui = 0 CATEGORIES = net @@ -37,5 +36,15 @@ RUN_DEPENDS-gui =${RUN_DEPENDS} \ # XXX: not yet working NO_TEST = Yes + +DOCDIR=${PREFIX}/share/doc/onionshare +LICENSEDIR=${WRKSRC}/install/licenses + +post-install: + ${INSTALL_DATA_DIR} ${DOCDIR} + ${INSTALL_DATA} ${LICENSEDIR}/license-obfs4.txt ${DOCDIR} + ${INSTALL_DATA} ${LICENSEDIR}/license-onionshare.txt ${DOCDIR} + ${INSTALL_DATA} ${LICENSEDIR}/license-tor.txt ${DOCDIR} + ${INSTALL_DATA} ${WRKSRC}/README.md ${DOCDIR} .include Index: distinfo === RCS file: /cvs/ports/net/onionshare/distinfo,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 distinfo --- distinfo9 Mar 2018 23:36:42 - 1.1.1.1 +++ distinfo21 Oct 2018 13:25:00 - @@ -1,2 +1,2 @@ -SHA256 (onionshare-1.3.tar.gz) = AIkUctiW5AWg9y36jq+D2uyj3DG+Uz1uLIbyxTkUz+0= -SIZE (onionshare-1.3.tar.gz) = 431352 +SHA256 (onionshare-1.3.1.tar.gz) = h+H6llSCBk6MfIvDIyb1dI/QOPfSr29RBWRE2yLvA1c= +SIZE (onionshare-1.3.1.tar.gz) = 436741 Index: pkg/PLIST-main === RCS file: /cvs/ports/net/onionshare/pkg/PLIST-main,v retrieving revision 1.1.1.1 diff -u -p -r1.1.1.1 PLIST-main --- pkg/PLIST-main 9 Mar 2018 23:36:42 - 1.1.1.1 +++ pkg/PLIST-main 21 Oct 2018 13:25:00 - @@ -20,6 +20,11 @@ lib/python${MODPY_VERSION}/site-packages lib/python${MODPY_VERSION}/site-packages/onionshare/strings.py lib/python${MODPY_VERSION}/site-packages/onionshare/web.py share/applications/ +share/doc/onionshare/ +share/doc/onionshare/README.md +share/doc/onionshare/license-obfs4.txt +share/doc/onionshare/license-onionshare.txt +share/doc/onionshare/license-tor.txt share/onionshare/ share/onionshare/html/ share/onionshare/html/404.html @@ -42,7 +47,6 @@ share/onionshare/images/server_working.p share/onionshare/images/settings.png share/onionshare/images/web_file.png share/onionshare/images/web_folder.png -share/onionshare/license.txt share/onionshare/locale/ share/onionshare/locale/cs.json share/onionshare/locale/da.json
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: e...@cvs.openbsd.org2018/10/21 07:29:48 Modified files: lang/pypy : Makefile distinfo lang/pypy/patches: patch-rpython_jit_backend_x86_detect_feature_py patch-rpython_rlib_rmmap_py lang/pypy/pkg : PLIST Removed files: lang/pypy/patches: patch-ctypes_configure_cbuild_py Log message: Update lang/pypy to 6.0.0. Thanks to David Carlier for his alloc_noexec patch. Tested and OK'd by naddy@. Thanks Naddy also reports that this fixes the build with lld!
Re: update: telephony/baresip
On Sun, Oct 21, 2018 at 02:55:31PM +0200, Klemens Nanni wrote: > On Fri, Oct 19, 2018 at 02:13:08PM +0200, Sebastien Marie wrote: > > - telephony/baresip/baresip 0.5.10->0.5.11 > > > > ChangeLog: https://github.com/alfredh/baresip/blob/master/docs/ChangeLog > 0.5.11 depends on rem>=5.3 as per the changelog. OK kn if anyone wants > to commit this with the minimal version bumped in LIB_DEPENDS-main, > otherwise I'll do so next week unless I hear objections. yes, 0.5.11 depens on rem>=5.3, but as 0.5.10. I think porters has enough work to avoid such extra dependencies requirements if it serves to nothing. I doubt we want to support partially updated port tree package build. just my 2 cents. -- Sebastien Marie
Re: UPDATE: net/onionshare
fredl writes: > Hey, Hello, Both cli and -gui works on amd64. Thanks for the update. I think you need to drop the REVISION-gui (but I'm just a port-newb). Best, > attached is a update for net/onionshare. It updates onionshare to > version 1.3.1. (Changes: > https://github.com/micahflee/onionshare/releases/tag/v1.3.1) > > + license and readme.md files added. > > -- > fredl
Re: update: telephony/baresip
On Fri, Oct 19, 2018 at 02:13:08PM +0200, Sebastien Marie wrote: > - telephony/baresip/baresip 0.5.10->0.5.11 > > ChangeLog: https://github.com/alfredh/baresip/blob/master/docs/ChangeLog 0.5.11 depends on rem>=5.3 as per the changelog. OK kn if anyone wants to commit this with the minimal version bumped in LIB_DEPENDS-main, otherwise I'll do so next week unless I hear objections.
Re: Is sysutils/memtest86+ still useful?
On Sun, Oct 21, 2018 at 01:22:52PM +0200, Christian Weisgerber wrote: > Is the sysutils/memtest86+ port still useful? I doubt it. > Our version dates from 2011, but even the newest one listed on the > home page is from 2013. I tried to run memtest as built by our > port on my Thinkpad X230 (ca. 2012), but it just crashes. Does it > still run on any machines of recent vintage? You can download bootable images from their website as well. memtest-4.20p3 does not work through `boot memtest' on a X250 with stock firmware but just prints "entry point at 0x1000" and leaving the system in a halted state. On my machines MemTest86+ 5.01 is either built as coreboot payload or UEFI executable and works just fine. > The port does not build with ld.lld. I have a diff to fix that, > but I don't know if it produces a working executable. However, > that's pointless if even the one built with ld.bfd doesn't work. I consider our port useless and defunct so OK kn to remove it.
Re: UPDATE: devel/cmake (i386 build wanted)
On Sat, Oct 20, 2018 at 10:56:35PM +0200, Rafael Sadowski wrote: > Hi All. > > Please find below diff to update cmake to the latest stable version. The > last time it failed on i386 so it would be great if someone could build > it on i386 and report to me/ports@. Please ignore the wrong CVS tags, > the diff based on an old diff from sthen@. > > Rafael Sadowski Hi Rafael, Builds fine on 6.4-current i386 running in VMM VM on 6.4-current amd64. Tomorrow I will have access to real i386 laptop with Pentium 4. I could test there too. Have build libmusicbrainz and colobot packages that seem to use/require cmake. Jan
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/10/21 06:11:25 Modified files: sysutils/py-pynetbox: Makefile distinfo Log message: update to py-pynetbox-3.4.7
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/10/21 06:11:32 Modified files: sysutils/rclone: Makefile distinfo Log message: update to rclone-1.44
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/10/21 06:09:04 Modified files: textproc/py-pygfm: Makefile distinfo Log message: update to py-gfm-0.1.4
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/10/21 06:07:30 Modified files: x11/i3lock : Makefile distinfo Added files: x11/i3lock/patches: patch-Makefile_in patch-configure_ac Log message: update to i3lock-2.11.1
Re: Is sysutils/memtest86+ still useful?
On Sun, Oct 21, 2018 at 01:22:52PM +0200, Christian Weisgerber wrote: > Is the sysutils/memtest86+ port still useful? > > Our version dates from 2011, but even the newest one listed on the > home page is from 2013. I tried to run memtest as built by our > port on my Thinkpad X230 (ca. 2012), but it just crashes. Does it > still run on any machines of recent vintage? > > The port does not build with ld.lld. I have a diff to fix that, > but I don't know if it produces a working executable. However, > that's pointless if even the one built with ld.bfd doesn't work. > I normaly use a bootable USB stick that comes with a memtest utility and never used the port. -- :wq Claudio
Re: Is sysutils/memtest86+ still useful?
On Sun, Oct 21, 2018 at 01:22:52PM +0200, Christian Weisgerber wrote: > I tried to run memtest as built by our port on my Thinkpad X230 (ca. > 2012), but it just crashes. Does it still run on any machines of > recent vintage? On my x280, the screen goes dark, displays "Entry point 0x1" and stays there.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: kette...@cvs.openbsd.org2018/10/21 05:31:23 Modified files: sysutils/arm-trusted-firmware: Makefile distinfo Log message: update to ARM Trusted Firmware 2.0 Includes changes (already in 1.6) for the following advisory: Trusted Firmware A Security Advisory TFV 7 Trusted Firmware-A exposure to cache speculation vulnerability Variant 4 tested on rk3399 ok jsg@
Is sysutils/memtest86+ still useful?
Is the sysutils/memtest86+ port still useful? Our version dates from 2011, but even the newest one listed on the home page is from 2013. I tried to run memtest as built by our port on my Thinkpad X230 (ca. 2012), but it just crashes. Does it still run on any machines of recent vintage? The port does not build with ld.lld. I have a diff to fix that, but I don't know if it produces a working executable. However, that's pointless if even the one built with ld.bfd doesn't work. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: update sysutils/arm-trusted-firmware to 1.6
On Sun, Oct 21, 2018 at 01:08:47PM +0200, Mark Kettenis wrote: > > Date: Sun, 21 Oct 2018 11:24:01 +1100 > > From: Jonathan Gray > > > > On Sat, Oct 20, 2018 at 11:42:56PM +0200, Mark Kettenis wrote: > > > > Date: Sun, 23 Sep 2018 23:15:44 +1000 > > > > From: Jonathan Gray > > > > > > > > Update to 1.6 and build the newly added a64/h5 platform support that can > > > > hopefully replace the atf-allwinner port. > > > > > > > > Only compile tested for lack of hardware. > > > > > > FWIW, I've now also tested rk3399. So I think you should go ahead and > > > commit this. > > > > > > ok kettenis@ > > > > Thanks, committed. > > > > There is now a 2.0 release but that only seems to be concerned with removing > > deprecated apis. > > > > https://github.com/ARM-software/arm-trusted-firmware/blob/master/docs/change-log.rst#trusted-firmware-a-version-2-0 > > > > So we might as well wait till the next one, which should include more > > allwinner bits > > > > https://github.com/ARM-software/arm-trusted-firmware/pull/1633 > > > > 'This updates the Allwinner port to do some cleanups and minor > > improvements, also adding PMIC and proper SYSTEM_OFF and CPU_OFF support > > to bring it eventually on par (and beyond) the existing "legacy" ATF > > port,' > > Well, since they deprecated some interfaces in 2.0 and rk3399 is one > of the affected platforms I decided to test 2.0 anyway. The diff is > trivial, so we might as well go for it. > > ok? > sure, ok jsg@ > > > Index: sysutils/arm-trusted-firmware/Makefile > === > RCS file: /cvs/ports/sysutils/arm-trusted-firmware/Makefile,v > retrieving revision 1.6 > diff -u -p -r1.6 Makefile > --- sysutils/arm-trusted-firmware/Makefile21 Oct 2018 00:13:54 - > 1.6 > +++ sysutils/arm-trusted-firmware/Makefile21 Oct 2018 11:01:49 - > @@ -6,7 +6,7 @@ COMMENT= ARM Trusted Firmware > > GH_ACCOUNT= ARM-software > GH_PROJECT= arm-trusted-firmware > -GH_TAGNAME= v1.6 > +GH_TAGNAME= v2.0 > > CATEGORIES= sysutils > > Index: sysutils/arm-trusted-firmware/distinfo > === > RCS file: /cvs/ports/sysutils/arm-trusted-firmware/distinfo,v > retrieving revision 1.3 > diff -u -p -r1.3 distinfo > --- sysutils/arm-trusted-firmware/distinfo21 Oct 2018 00:13:54 - > 1.3 > +++ sysutils/arm-trusted-firmware/distinfo21 Oct 2018 11:01:49 - > @@ -1,2 +1,2 @@ > -SHA256 (arm-trusted-firmware-1.6.tar.gz) = > YhIDaPIZbT4SYpbIEW8zmVaOEAlgpRIuUgF9InZrcAk= > -SIZE (arm-trusted-firmware-1.6.tar.gz) = 3102529 > +SHA256 (arm-trusted-firmware-2.0.tar.gz) = > fWmaFoO7elkJ3je265G2442zLNb8WuSKCOsHGNZQSuQ= > +SIZE (arm-trusted-firmware-2.0.tar.gz) = 3058755 >
Re: update sysutils/arm-trusted-firmware to 1.6
> Date: Sun, 21 Oct 2018 11:24:01 +1100 > From: Jonathan Gray > > On Sat, Oct 20, 2018 at 11:42:56PM +0200, Mark Kettenis wrote: > > > Date: Sun, 23 Sep 2018 23:15:44 +1000 > > > From: Jonathan Gray > > > > > > Update to 1.6 and build the newly added a64/h5 platform support that can > > > hopefully replace the atf-allwinner port. > > > > > > Only compile tested for lack of hardware. > > > > FWIW, I've now also tested rk3399. So I think you should go ahead and > > commit this. > > > > ok kettenis@ > > Thanks, committed. > > There is now a 2.0 release but that only seems to be concerned with removing > deprecated apis. > > https://github.com/ARM-software/arm-trusted-firmware/blob/master/docs/change-log.rst#trusted-firmware-a-version-2-0 > > So we might as well wait till the next one, which should include more > allwinner bits > > https://github.com/ARM-software/arm-trusted-firmware/pull/1633 > > 'This updates the Allwinner port to do some cleanups and minor > improvements, also adding PMIC and proper SYSTEM_OFF and CPU_OFF support > to bring it eventually on par (and beyond) the existing "legacy" ATF > port,' Well, since they deprecated some interfaces in 2.0 and rk3399 is one of the affected platforms I decided to test 2.0 anyway. The diff is trivial, so we might as well go for it. ok? Index: sysutils/arm-trusted-firmware/Makefile === RCS file: /cvs/ports/sysutils/arm-trusted-firmware/Makefile,v retrieving revision 1.6 diff -u -p -r1.6 Makefile --- sysutils/arm-trusted-firmware/Makefile 21 Oct 2018 00:13:54 - 1.6 +++ sysutils/arm-trusted-firmware/Makefile 21 Oct 2018 11:01:49 - @@ -6,7 +6,7 @@ COMMENT=ARM Trusted Firmware GH_ACCOUNT=ARM-software GH_PROJECT=arm-trusted-firmware -GH_TAGNAME=v1.6 +GH_TAGNAME=v2.0 CATEGORIES=sysutils Index: sysutils/arm-trusted-firmware/distinfo === RCS file: /cvs/ports/sysutils/arm-trusted-firmware/distinfo,v retrieving revision 1.3 diff -u -p -r1.3 distinfo --- sysutils/arm-trusted-firmware/distinfo 21 Oct 2018 00:13:54 - 1.3 +++ sysutils/arm-trusted-firmware/distinfo 21 Oct 2018 11:01:49 - @@ -1,2 +1,2 @@ -SHA256 (arm-trusted-firmware-1.6.tar.gz) = YhIDaPIZbT4SYpbIEW8zmVaOEAlgpRIuUgF9InZrcAk= -SIZE (arm-trusted-firmware-1.6.tar.gz) = 3102529 +SHA256 (arm-trusted-firmware-2.0.tar.gz) = fWmaFoO7elkJ3je265G2442zLNb8WuSKCOsHGNZQSuQ= +SIZE (arm-trusted-firmware-2.0.tar.gz) = 3058755
Re: KDE Kate no longer works with kio-slaves
On Sun Oct 21, 2018 at 11:01:47AM +0200, Federico Giannici wrote: > After I upgraded from OpenBSD amd64 6.3 to 6.4 my KDE Kate program is no > longer able to open remote files (that is, to use kio-slaves). > > > When I try to open a remote (sftp:) file it says: KIO sftp support is no more part from base KIO. It's a part of x11/kde-applications/kio-extras which is not portet to OpenBSD yet. Sorry for the bad news you will have to wait for 6.5 or follow -current. Btw HTTPS, ftp still works with base KIO. Rafael Sadowski
KDE Kate no longer works with kio-slaves
After I upgraded from OpenBSD amd64 6.3 to 6.4 my KDE Kate program is no longer able to open remote files (that is, to use kio-slaves). When I try to open a remote (sftp:) file it says: kf5.kio.core: couldn't create slave: "klauncher said: Unknown protocol 'sftp'.\n" I was able to open an "http:" file supplying it via command line, but then it says: Non-native QFileDialog supports only local files All other KDE programs (Dolphin, Okular, Gwenview, etc.) are still able to correctly open "sftp:" urls, so the kio-slaves are correctly installed and working. I'd like to know if there is something bad in my system (maybe some missing package, or a conflict with older versions?) or it is a (very very important) limitation of the new Kate version. Thanks.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rsadow...@cvs.openbsd.org 2018/10/21 02:17:41 Modified files: multimedia/mkvtoolnix: Makefile distinfo multimedia/mkvtoolnix/pkg: PFRAG.no-no_x11 Log message: Update mkvtoolnix-28.0.0
Re: NEW: math/mlpack (and dependency math/armadillo)
Here's a revised version of the armadillo port, adding arpack and hdf5 as a dependency. And the missed COMPILER line, oops. Notes: - I definitely prefer to be explicit for CONFIGURE_STYLE - the library definitely requires linking with the gfortran library because some of lapack blas want symbols in there - with arpack added, no need for explicit depends on blas/lapack, as arpack needs them anyway. armadillo-9.100.5.tgz Description: armadillo-9.100.5.tgz