CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: rsadow...@cvs.openbsd.org 2018/05/24 23:45:47 Modified files: devel/qbs : Makefile distinfo Log message: Bugfix update qbs-1.11.1
multi-package and flavors
Hi, I'm making a port of devel/spyder that has the python3 FLAVOR. now, spyder and spyder3 have the same docs, so even when the applications files do not cause a conflict, the docs files do. So I just divided the port in 2 packages -main and -docs, but the idea is to create 3 at the end: - spyder - spyder3 - spyder-docs So no collisions and 0 problems, my question is, how do I set the multi-packages and the flavor without generating a 4th package. Example attached and also in openbsd-wip on github under devel/spyder. https://github.com/jasperla/openbsd-wip/blob/master/devel/spyder/Makefile Thanks for the help. Elias. Makefile Description: Binary data
New port - lv2 (LADSPA V2)
First time submitting a port, and pretty new to it all anyway. I have tested/built this on amd64 only. It is a pre-requisite to getting some other audio software ported over. So call it baby step #1. Ken lv2.tgz Description: application/tar-gz
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: bcal...@cvs.openbsd.org 2018/05/24 18:58:43 Modified files: www/liferea: Makefile distinfo www/liferea/patches: patch-src_ui_liferea_shell_c www/liferea/pkg: PLIST Log message: Update to liferea-1.12.3 ok aja@
Re: UPDATE: misc/redshift 1.11 => 1.12
On 05/24/18 09:43, Dumitru Mișu Moldovan wrote: Brian Callahanwrote: This is my "portroach rampage" contribution of the day, which is a new term I just learned. It is a mostly straightforward update of misc/redshift. I say mostly, because I noticed that while redshift was compiling with cc, it was linking with gcc -- it looks for your Objective-C compiler and uses that as the linker (and apparently defaults to gcc). But the override is trivial. It's nice to have the GTK+ systray widget bundled with this, but I kind of regret not having the option to install redshift without all the GTK+ 3.x and Python 3.x deps, like in FreeBSD, where the redshift package was pretty light. I know I'm not the only one to be bothered by this: https://www.tedunangst.com/flak/post/sct-set-color-temperature. :-] Maybe a flavor would be an option? There is sctd also, but I would still prefer a lighter flavor of redshift. Would a patch be welcome? I hope it's not an advanced topic, as I have pretty much zero experience with OpenBSD porting and packaging thus far… Sure. Send patches. You can even take over as MAINTAINER too. ~Brian
Re: update openldap to use modules
So it doesn't look like upstream is going to bite on removing the version numbers from the dynamically loadable modules, so here is an updated diff that includes a patch to do so and also installs the sample config file. It doesn't look like we can get rid of the aci flavor after all, if you try to build that as a module rather than bundled it fails and says that functionality does not support being a dynamically loaded module at this time. Index: Makefile === RCS file: /cvs/ports/databases/openldap/Makefile,v retrieving revision 1.161 diff -u -p -r1.161 Makefile --- Makefile12 Jan 2018 00:36:28 - 1.161 +++ Makefile24 May 2018 22:38:07 - @@ -9,7 +9,7 @@ DISTNAME = openldap-2.4.45 PKGNAME-main = ${DISTNAME:S/-/-client-/} PKGNAME-server = ${DISTNAME:S/-/-server-/} REVISION = 4 -REVISION-server = 5 +REVISION-server = 6 # overwrite -main pkgname/path to strip FLAVOR, aci only affects the server FULLPKGNAME-main = ${DISTNAME:S/-/-client-/}${FLAVOR_EXT:S/-aci//} @@ -32,7 +32,7 @@ MAINTAINER = Stuart Henderson
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ste...@cvs.openbsd.org 2018/05/24 15:23:50 Modified files: math/octave: Makefile distinfo math/octave/patches: patch-configure math/octave/pkg: PLIST Log message: update to octave 4.4.0
UPDATE: devel/py-twisted 17.9.0 -> 18.4.0
Dear all, Please find the update for devel/py-twisted attached. Changelog: https://github.com/twisted/twisted/blob/twisted-18.4.0/NEWS.rst -- With best regards, Pavel Korovin Index: Makefile === RCS file: /cvs/ports/devel/py-twisted/Makefile,v retrieving revision 1.7 diff -u -p -r1.7 Makefile --- Makefile6 Apr 2018 19:14:00 - 1.7 +++ Makefile24 May 2018 20:54:33 - @@ -2,7 +2,7 @@ COMMENT = asynchronous networking Python framework -MODPY_EGG_VERSION =17.9.0 +MODPY_EGG_VERSION =18.4.0 EXTRACT_SUFX = .tar.bz2 DISTNAME = Twisted-${MODPY_EGG_VERSION} @@ -45,7 +45,8 @@ RUN_DEPENDS +=devel/py-SOAPpy .endif SUBST_FILES = ${WRKSRC}/src/Twisted.egg-info/entry_points.txt \ - ${WRKSRC}/src/twisted/python/_setup.py + ${WRKSRC}/src/twisted/python/_setup.py \ + ${WRKSRC}/src/twisted/conch/test/test_ckeygen.py TEST_DEPENDS = ${FULLPKGNAME}:${BUILD_PKGPATH} TEST_ENV = LC_ALL=en_US.UTF-8 Index: distinfo === RCS file: /cvs/ports/devel/py-twisted/distinfo,v retrieving revision 1.2 diff -u -p -r1.2 distinfo --- distinfo6 Apr 2018 19:14:00 - 1.2 +++ distinfo24 May 2018 20:54:33 - @@ -1,2 +1,2 @@ -SHA256 (Twisted-17.9.0.tar.bz2) = DaGn411fyuN7yceXiXC1/rO8goIhVbhlTsY5JcBa91w= -SIZE (Twisted-17.9.0.tar.bz2) = 3019243 +SHA256 (Twisted-18.4.0.tar.bz2) = pMwWSngYWcdN5H8X8OhfS86KMyGp0IksAVyPgMQVitk= +SIZE (Twisted-18.4.0.tar.bz2) = 3037019 Index: patches/patch-src_twisted_conch_test_test_ckeygen_py === RCS file: patches/patch-src_twisted_conch_test_test_ckeygen_py diff -N patches/patch-src_twisted_conch_test_test_ckeygen_py --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_twisted_conch_test_test_ckeygen_py24 May 2018 20:54:33 - @@ -0,0 +1,26 @@ +$OpenBSD$ + +Index: src/twisted/conch/test/test_ckeygen.py +--- src/twisted/conch/test/test_ckeygen.py.orig src/twisted/conch/test/test_ckeygen.py +@@ -68,9 +68,9 @@ class KeyGenTests(TestCase): + def _testrun(self, keyType, keySize=None): + filename = self.mktemp() + if keySize is None: +-subprocess.call(['ckeygen', '-t', keyType, '-f', filename, '--no-passphrase']) ++subprocess.call(['ckeygen${MODPY_BIN_SUFFIX}', '-t', keyType, '-f', filename, '--no-passphrase']) + else: +-subprocess.call(['ckeygen', '-t', keyType, '-f', filename, '--no-passphrase', ++subprocess.call(['ckeygen${MODPY_BIN_SUFFIX}', '-t', keyType, '-f', filename, '--no-passphrase', + '-b', keySize]) + privKey = Key.fromFile(filename) + pubKey = Key.fromFile(filename + '.pub') +@@ -94,7 +94,7 @@ class KeyGenTests(TestCase): + def test_runBadKeytype(self): + filename = self.mktemp() + with self.assertRaises(subprocess.CalledProcessError): +-subprocess.check_call(['ckeygen', '-t', 'foo', '-f', filename]) ++subprocess.check_call(['ckeygen${MODPY_BIN_SUFFIX}', '-t', 'foo', '-f', filename]) + + + Index: patches/patch-src_twisted_internet_test_test_tcp_py === RCS file: patches/patch-src_twisted_internet_test_test_tcp_py diff -N patches/patch-src_twisted_internet_test_test_tcp_py --- /dev/null 1 Jan 1970 00:00:00 - +++ patches/patch-src_twisted_internet_test_test_tcp_py 24 May 2018 20:54:33 - @@ -0,0 +1,18 @@ +$OpenBSD$ + +Index: src/twisted/internet/test/test_tcp.py +--- src/twisted/internet/test/test_tcp.py.orig src/twisted/internet/test/test_tcp.py +@@ -2515,7 +2515,11 @@ class SimpleUtilityTests(TestCase): + accidentally block if it receives bad input.) + """ + err = self.assertRaises(socket.gaierror, _resolveIPv6, "::1", "http") +-self.assertEqual(err.args[0], socket.EAI_NONAME) ++if sys.platform.startswith(('openbsd')): ++# https://bugs.python.org/issue1282647 ++self.assertEqual(err.args[0], socket.EAI_SERVICE) ++else: ++self.assertEqual(err.args[0], socket.EAI_NONAME) + + if platform.isWindows(): + test_resolveNumericService.skip = ("The AI_NUMERICSERV flag is not " Index: patches/patch-src_twisted_python__setup_py === RCS file: /cvs/ports/devel/py-twisted/patches/patch-src_twisted_python__setup_py,v retrieving revision 1.1 diff -u -p -r1.1 patch-src_twisted_python__setup_py --- patches/patch-src_twisted_python__setup_py 6 Apr 2018 19:14:00 - 1.1 +++ patches/patch-src_twisted_python__setup_py 24 May 2018 20:54:33 - @@ -8,13 +8,14 @@ https://github.com/twisted/twisted/commi Index:
Re: new HOMEPAGE for net/onioncat
> > Missing bump :-) > > I though it was not needed in this case. Petter Hessler explained me it > was included in package metadata so it needed a bump REVISION. OK > > Index: Makefile > === > RCS file: /cvs/ports/net/onioncat/Makefile,v > retrieving revision 1.11 > diff -u -p -r1.11 Makefile > --- Makefile 2 Nov 2014 18:41:00 - 1.11 > +++ Makefile 24 May 2018 14:50:55 - > @@ -4,10 +4,10 @@ COMMENT = IP-Transparent Tor Hidden Serv > > V = 0.2.1 > DISTNAME = onioncat-$V > -REVISION = 0 > +REVISION = 1 >CATEGORIES = net > > -HOMEPAGE = http://www.abenteuerland.at/onioncat/ > +HOMEPAGE = https://www.cypherpunk.at/onioncat_trac/ > > MAINTAINER= Paul Irofti
py-wxPython extras
Hi, I was looking for Editra (text editor) and I found that there is not port of Editra by itself, It is contained within the py-wxPython port. https://github.com/openbsd/ports/blob/439514e8c82c269624be233d533b016d3a63a201/x11/py-wxPython/pkg/PLIST Shouldn't be better to remove the extras from py-wxPython and having those apart ? I mean, if you want wxPython maybe you do not want Editra. And looking for standalone applications inside other ports doesn't seem right, I might as well thought that there was no port of Editra. Cheers. Elias.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2018/05/24 11:06:17 Modified files: net/wireshark : Tag: OPENBSD_6_3 Makefile distinfo net/wireshark/patches: Tag: OPENBSD_6_3 patch-CMakeLists_txt Log message: update to wireshark-2.4.7
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2018/05/24 11:05:47 Modified files: net/wireshark : Makefile distinfo net/wireshark/patches: patch-CMakeLists_txt Added files: net/wireshark/patches: patch-wsutil_inet_addr_h Log message: update to wireshark-2.6.1
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/05/24 10:59:40 Modified files: sysutils/salt : Makefile sysutils/salt/files: pf.py vmctl.py Log message: address deprecation warnings for salt.utils.which
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: b...@cvs.openbsd.org2018/05/24 10:04:35 Modified files: security/lastpass-cli: Makefile distinfo Added files: security/lastpass-cli/patches: patch-config_c Log message: Update to lastpass-cli-1.3.1. This update fixes an issue with intermediate certificates resulting in a fatal error (https://github.com/lastpass/lastpass-cli/issues/409). Taken a patch from upstream to fix an issue resulting in a segmentation fault (https://github.com/lastpass/lastpass-cli/pull/411) While here change MAINTAINER mail address. OK sthen@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 09:54:09 Log message: Import gtk+4-3.93.0. Not hooked up yet because it's needs an updated libepoxy that matthieu@ should be able to commit soon. Some GNOME stuff (at least) will slowly start moving to gtk+4 in the near future so let's get this in early so we can start to play with it. Yes, this is obviously a development release, a stable gtk+4 is not officially out yet. Status: Vendor Tag: ajacoutot Release Tags: ajacoutot_20180524 N ports/x11/gtk+4/Makefile N ports/x11/gtk+4/distinfo N ports/x11/gtk+4/patches/patch-gtk_gtksettings_c N ports/x11/gtk+4/patches/patch-gtk_updateiconcache_c N ports/x11/gtk+4/patches/patch-modules_printbackends_meson_build N ports/x11/gtk+4/pkg/DESCR-cloudprint N ports/x11/gtk+4/pkg/DESCR-cups N ports/x11/gtk+4/pkg/PLIST-main N ports/x11/gtk+4/pkg/DESCR-main N ports/x11/gtk+4/pkg/PLIST-cloudprint N ports/x11/gtk+4/pkg/PLIST-cups N ports/x11/gtk+4/pkg/DESCR-guic N ports/x11/gtk+4/pkg/PLIST-guic N ports/x11/gtk+4/pkg/README-main No conflicts created by this import
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 09:53:37 Modified files: x11/gtk+3 : Makefile x11/gtk+3/pkg : README-main Log message: evince -> evince-previewer
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 09:38:00 Modified files: multimedia/gstreamer1/plugins-base: Makefile Log message: Add graphene support.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 09:37:13 Log message: Import graphene-1.8.0. Graphene contains optimizations for speeding up vector operations, those optimizations are optional, and used only if both Graphene was compiled with support for them and if the system you're running on has them. <...> ok robert@ Status: Vendor Tag: ajacoutot Release Tags: ajacoutot_20180524 N ports/graphics/graphene/Makefile N ports/graphics/graphene/distinfo N ports/graphics/graphene/pkg/DESCR N ports/graphics/graphene/pkg/PLIST No conflicts created by this import
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 09:37:40 Modified files: graphics : Makefile Log message: +graphene
Re: new HOMEPAGE for net/onioncat
Antoine Jacoutot writes: > On Thu, May 24, 2018 at 04:41:13PM +0200, Solene Rapenne wrote: >> The current onioncat ports homepage >> http://www.abenteuerland.at/onioncat/ redirects to a new one. >> >> Index: Makefile >> === >> RCS file: /cvs/ports/net/onioncat/Makefile,v >> retrieving revision 1.11 >> diff -u -p -r1.11 Makefile >> --- Makefile 2 Nov 2014 18:41:00 - 1.11 >> +++ Makefile 24 May 2018 14:39:34 - >> @@ -7,7 +7,7 @@ DISTNAME =onioncat-$V >> REVISION = 0 >> CATEGORIES = net >> >> -HOMEPAGE = http://www.abenteuerland.at/onioncat/ >> +HOMEPAGE = https://www.cypherpunk.at/onioncat_trac/ >> >> MAINTAINER= Paul Irofti> > Missing bump :-) I though it was not needed in this case. Petter Hessler explained me it was included in package metadata so it needed a bump REVISION. Index: Makefile === RCS file: /cvs/ports/net/onioncat/Makefile,v retrieving revision 1.11 diff -u -p -r1.11 Makefile --- Makefile 2 Nov 2014 18:41:00 - 1.11 +++ Makefile 24 May 2018 14:50:55 - @@ -4,10 +4,10 @@ COMMENT = IP-Transparent Tor Hidden Serv V = 0.2.1 DISTNAME = onioncat-$V -REVISION = 0 +REVISION = 1 CATEGORIES = net -HOMEPAGE = http://www.abenteuerland.at/onioncat/ +HOMEPAGE = https://www.cypherpunk.at/onioncat_trac/ MAINTAINER= Paul Irofti
Re: new HOMEPAGE for net/onioncat
On Thu, May 24, 2018 at 04:41:13PM +0200, Solene Rapenne wrote: > The current onioncat ports homepage > http://www.abenteuerland.at/onioncat/ redirects to a new one. > > Index: Makefile > === > RCS file: /cvs/ports/net/onioncat/Makefile,v > retrieving revision 1.11 > diff -u -p -r1.11 Makefile > --- Makefile 2 Nov 2014 18:41:00 - 1.11 > +++ Makefile 24 May 2018 14:39:34 - > @@ -7,7 +7,7 @@ DISTNAME =onioncat-$V > REVISION = 0 > CATEGORIES = net > > -HOMEPAGE = http://www.abenteuerland.at/onioncat/ > +HOMEPAGE = https://www.cypherpunk.at/onioncat_trac/ > > MAINTAINER= Paul IroftiMissing bump :-) -- Antoine
new HOMEPAGE for net/onioncat
The current onioncat ports homepage http://www.abenteuerland.at/onioncat/ redirects to a new one. Index: Makefile === RCS file: /cvs/ports/net/onioncat/Makefile,v retrieving revision 1.11 diff -u -p -r1.11 Makefile --- Makefile 2 Nov 2014 18:41:00 - 1.11 +++ Makefile 24 May 2018 14:39:34 - @@ -7,7 +7,7 @@ DISTNAME =onioncat-$V REVISION = 0 CATEGORIES = net -HOMEPAGE = http://www.abenteuerland.at/onioncat/ +HOMEPAGE = https://www.cypherpunk.at/onioncat_trac/ MAINTAINER= Paul Irofti
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: es...@cvs.openbsd.org 2018/05/24 08:35:20 Modified files: infrastructure/bin: update-plist Log message: so $d2 is not necessarily absolute, d2=dirname(d2) may loop on d2 = '.' infinite loop on a few ports (very few) also, be more specific on the exact prefix, depending whether the prefix string includes the trailing / in its expansion (MODPY_PYCACHE) or not.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 08:28:41 Modified files: net/wireshark : Makefile Log message: devel/mozjs17 -> devel/spidermonkey52
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 08:07:46 Modified files: x11/gtk+3 : Makefile Log message: Drop useless comment.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 08:07:23 Modified files: x11/gtk+3 : Makefile Log message: mozjs17 -> spidermonkey52
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: lan...@cvs.openbsd.org 2018/05/24 07:56:37 Modified files: x11/xfce4/xfce4-whiskermenu: Makefile distinfo x11/xfce4/xfce4-whiskermenu/pkg: PLIST Log message: Update to xfce4-whiskermenu 2.2.0.
Re: UPDATE: misc/redshift 1.11 => 1.12
Brian Callahanwrote: > > This is my "portroach rampage" contribution of the day, which is a > new term I just learned. It is a mostly straightforward update of > misc/redshift. I say mostly, because I noticed that while redshift > was compiling with cc, it was linking with gcc -- it looks for your > Objective-C compiler and uses that as the linker (and apparently > defaults to gcc). But the override is trivial. It's nice to have the GTK+ systray widget bundled with this, but I kind of regret not having the option to install redshift without all the GTK+ 3.x and Python 3.x deps, like in FreeBSD, where the redshift package was pretty light. I know I'm not the only one to be bothered by this: https://www.tedunangst.com/flak/post/sct-set-color-temperature. :-] Maybe a flavor would be an option? There is sctd also, but I would still prefer a lighter flavor of redshift. Would a patch be welcome? I hope it's not an advanced topic, as I have pretty much zero experience with OpenBSD porting and packaging thus far… pgp5xqJi1IyG8.pgp Description: Semnătură digitală OpenPGP
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: juan...@cvs.openbsd.org 2018/05/24 06:51:43 Modified files: mail/mozilla-thunderbird: Tag: OPENBSD_6_3 Makefile distinfo Added files: mail/mozilla-thunderbird/patches: Tag: OPENBSD_6_3 patch-mozilla_js_src_jit_ProcessExecutableMemory_h patch-mozilla_security_manager_ssl_nsNSSComponent_cpp Removed files: mail/mozilla-thunderbird/patches: Tag: OPENBSD_6_3 patch-mozilla_js_src_jit_ProcessExecutableMemory_cpp Log message: Update to mozilla-thunderbird 52.8.0. OK landry@ (MAINTAINER).
Re: www/mozilla-firefox: fix volume controls
On Tue, 15 May 2018 23:23:33 +0200, Alexandre Ratchovwrote: > Maybe you've noticed that somtimes sound volume in firefox doesn't > match the volume indicator until you touch it. > > This is because the firefox audio API has no volume getter and assumes > the initial volume is 1, while sndio saves volumes and allows volume > to be controlled externally, which makes firefox use a wrong > representation of the actual volume. > > The workaround is to do like alsa, pulseaudio and other backends: stop > using the native volume control and adjust the volume of the signal in > firefox. FTR, I've been running firefox since Friday with this diff and it fixes the problem without any regression spotted. Thanks! Daniel
Re: UPDATE: misc/redshift 1.11 => 1.12
On 2018/05/23 23:53, Brian Callahan wrote: > Hi ports -- > > This is my "portroach rampage" contribution of the day, which is a new term > I just learned. It is a mostly straightforward update of misc/redshift. I > say mostly, because I noticed that while redshift was compiling with cc, it > was linking with gcc -- it looks for your Objective-C compiler and uses that > as the linker (and apparently defaults to gcc). But the override is trivial. > > OK? Not sure about forcing OBJC=$CC... Seems better to restore the Makefile.in section to how it was in 1.11 for now and check with upstream? This was introduced when adding MacOS CoreLocation support. Index: src/Makefile.in --- src/Makefile.in.orig +++ src/Makefile.in @@ -661,7 +661,7 @@ windows/versioninfo.$(OBJEXT): windows/$(am__dirstamp) redshift$(EXEEXT): $(redshift_OBJECTS) $(redshift_DEPENDENCIES) $(EXTRA_redshift_DEPENDENCIES) @rm -f redshift$(EXEEXT) - $(AM_V_OBJCLD)$(OBJCLINK) $(redshift_OBJECTS) $(redshift_LDADD) $(LIBS) + $(AM_V_CCLD)$(LINK) $(redshift_OBJECTS) $(redshift_LDADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) Also: configure picks up geoclue2 if present at build time.
Re: PHP link sqlite to sqlite3 package
On 2018/05/24 12:49, Martijn van Duren wrote: > Hello ports@, > > Currently sqlite in PHP builds its own sqlite and links it internally. > This library is quite out of date (3.8.10.2 for 5.6 and 3.14.2 for 7.0). > > I would like to link PHP to our sqlite3 package, so that it can be kept > up to date. This is assumed to also be what happens with most Linux > packages[0]. That makes sense to me. > The motivation to split pdo_sqlite/sqlite3 into a separate package is > for two reasons: > 1) All other databases have their own package, so it's more consistent. > 2) Keeping the extensions in the SAPI would add another dependency on > the PHP-main package. > > Of course I can change the patch to just link to the sqlite package > if this is not enough motivation to split them up. > > If we split them up the packages that at least have sqlite calls in > them are, but might be in optional code-paths: > icinga/web2 > zabbix-agent > chive > nextcloud > owncloud > piwik It's definitely more than that - I know roundcube needs pdo_sqlite or another pdo_* - you'd need to look for things using pdo at all to find them, rather than just looking for sqlite. Splitting up database extensions with library deps does seem reasonable but I'm wondering maybe we should merge the $somedatabase and pdo_$somedatabase packages, I'm not sure there's much point in splitting them up, and that way the total number of subpackages doesn't increase.. www/chive is removed already btw.
PHP link sqlite to sqlite3 package
Hello ports@, Currently sqlite in PHP builds its own sqlite and links it internally. This library is quite out of date (3.8.10.2 for 5.6 and 3.14.2 for 7.0). I would like to link PHP to our sqlite3 package, so that it can be kept up to date. This is assumed to also be what happens with most Linux packages[0]. The motivation to split pdo_sqlite/sqlite3 into a separate package is for two reasons: 1) All other databases have their own package, so it's more consistent. 2) Keeping the extensions in the SAPI would add another dependency on the PHP-main package. Of course I can change the patch to just link to the sqlite package if this is not enough motivation to split them up. If we split them up the packages that at least have sqlite calls in them are, but might be in optional code-paths: icinga/web2 zabbix-agent chive nextcloud owncloud piwik I'm not 100% sure if I handled the PLIST-main correctly, but the sqlite3.h file can be removed anyway, since it's not used anymore. Also, I noticed a type-O in the pdo_pgsql description, so I changed it while here. Comments? martijn@ [0] https://marc.info/?l=php-internals=150031503332312=2 Index: Makefile.inc === RCS file: /cvs/ports/lang/php/Makefile.inc,v retrieving revision 1.104 diff -u -p -r1.104 Makefile.inc --- Makefile.inc23 May 2018 06:51:14 - 1.104 +++ Makefile.inc24 May 2018 10:44:31 - @@ -67,8 +67,6 @@ CONFIGURE_ARGS+= --enable-shared \ --with-pic \ --with-pear=${LOCALBASE}/share/php-${PV} \ --with-config-file-scan-dir=${SYSCONFDIR}/php-${PV} \ - --with-pdo-sqlite \ - --with-sqlite3 \ --program-suffix=-${PV} \ --with-readline @@ -237,11 +235,18 @@ WANTLIB-pdo_mysql=crypto m pthread ssl # pdo-pgsql PHP_EXTENSIONS+= -pdo_pgsql -COMMENT-pdo_pgsql= PDO database access extensions for php +COMMENT-pdo_pgsql= PDO pgsql database access extensions for php CONFIGURE_ARGS+=--with-pdo-pgsql=shared,${LOCALBASE} LIB_DEPENDS-pdo_pgsql= databases/postgresql WANTLIB-pdo_pgsql= pq>=2 pthread +# pdo-sqlite +PHP_EXTENSIONS+= -pdo_sqlite +COMMENT-pdo_sqlite=PDO sqlite3 database access extensions for php +CONFIGURE_ARGS+=--with-pdo-sqlite=shared,${LOCALBASE} +LIB_DEPENDS-pdo_sqlite=databases/sqlite3 +WANTLIB-pdo_sqlite=sqlite3 pthread + # pgsql PHP_EXTENSIONS+= -pgsql COMMENT-pgsql= pgsql database access extensions for php @@ -276,6 +281,13 @@ COMMENT-snmp= snmp protocol extensions CONFIGURE_ARGS+= --with-snmp=shared,${LOCALBASE} LIB_DEPENDS-snmp= net/net-snmp WANTLIB-snmp= crypto m netsnmp>=6.2 pthread + +# sqlite3 +PHP_EXTENSIONS+= -sqlite3 +COMMENT-sqlite3= sqlite3 database access extension for php +CONFIGURE_ARGS+= --with-sqlite3=shared,${LOCALBASE} +LIB_DEPENDS-sqlite3= databases/sqlite3 +WANTLIB-sqlite3= sqlite3 pthread # pdo-dblib PHP_EXTENSIONS+= -pdo_dblib Index: 5.6/Makefile === RCS file: /cvs/ports/lang/php/5.6/Makefile,v retrieving revision 1.64 diff -u -p -r1.64 Makefile --- 5.6/Makefile23 May 2018 06:51:14 - 1.64 +++ 5.6/Makefile24 May 2018 10:44:31 - @@ -5,7 +5,7 @@ BROKEN-alpha= pcre_jit_compile.c:65:2: e PV=5.6 V= ${PV}.35 REVISION= 1 -REVISION-main= 2 +REVISION-main= 3 MASTER_SITES0= https://download.suhosin.org/ Index: 5.6/pkg/PLIST-main === RCS file: /cvs/ports/lang/php/5.6/pkg/PLIST-main,v retrieving revision 1.11 diff -u -p -r1.11 PLIST-main --- 5.6/pkg/PLIST-main 23 May 2018 06:51:14 - 1.11 +++ 5.6/pkg/PLIST-main 24 May 2018 10:44:31 - @@ -5,9 +5,7 @@ @conflict php-fpm->=5.6,<5.7 @conflict php5-core->=5.2 @conflict php5-core-fastcgi->=5.2 -@conflict php5-pdo_sqlite->=5.2 @pkgpath lang/php/5.4 -@pkgpath lang/php/5.4,-pdo_sqlite @pkgpath lang/php/5.6,-fpm @pkgpath lang/php/5.6,-main,ap2 @extraunexec rm -f ${SYSCONFDIR}/php-${PV}/php-${PV}.sample/* @@ -281,9 +279,6 @@ share/php-${PV}/include/ext/spl/spl_func share/php-${PV}/include/ext/spl/spl_heap.h share/php-${PV}/include/ext/spl/spl_iterators.h share/php-${PV}/include/ext/spl/spl_observer.h -share/php-${PV}/include/ext/sqlite3/ -share/php-${PV}/include/ext/sqlite3/libsqlite/ -share/php-${PV}/include/ext/sqlite3/libsqlite/sqlite3.h share/php-${PV}/include/ext/standard/ share/php-${PV}/include/ext/standard/base64.h share/php-${PV}/include/ext/standard/basic_functions.h Index: 5.6/pkg/PLIST-pdo_sqlite === RCS file: 5.6/pkg/PLIST-pdo_sqlite diff -N 5.6/pkg/PLIST-pdo_sqlite --- /dev/null 1 Jan
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2018/05/24 04:25:28 Modified files: textproc/solr : Tag: OPENBSD_6_3 Makefile distinfo textproc/solr/patches: Tag: OPENBSD_6_3 patch-bin_solr textproc/solr/pkg: Tag: OPENBSD_6_3 PLIST Log message: update to solr-7.3.1, CVE-2018-8010: XXE vulnerability due to Apache Solr configset upload
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2018/05/24 04:24:37 Modified files: textproc/solr : Makefile distinfo textproc/solr/pkg: PLIST Log message: update to solr-7.3.1, if you aren't restricting to local-only traffic then you probably want this .. CVE-2018-8010: XXE vulnerability due to Apache Solr configset upload
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: st...@cvs.openbsd.org 2018/05/24 03:33:42 Modified files: graphics/ImageMagick: Makefile distinfo graphics/ImageMagick/pkg: PLIST Log message: update to ImageMagick-6.9.9-47, seems to fix crashes seen in geo/postgis build and probably others. noting here just for the sake of it, my test case reduced to: convert -size 200x200 xc:none -draw "path 'M 110,190 M 70,40'" tmp0.png
Re: UPDATE: misc/redshift 1.11 => 1.12
On Wed, May 23, 2018 at 11:53:23PM -0400, Brian Callahan wrote: > This is my "portroach rampage" contribution of the day, which is a new term > I just learned. It is a mostly straightforward update of misc/redshift. I > say mostly, because I noticed that while redshift was compiling with cc, it > was linking with gcc -- it looks for your Objective-C compiler and uses that > as the linker (and apparently defaults to gcc). But the override is trivial. Looks good to me, works as expected on my X250. I noticed this at the end of `fake', though: gtk-update-icon-cache -f -t /usr/local/share/icons/hicolor || :; gtk-update-icon-cache: Failed to open file /usr/local/share/icons/hicolor/.icon-theme.cache : Permission denied Glancing at the configure script and invoked compiler commands during `build', OBJCFLAGS seems unset. Did you take a closer look at these just to make sure they don't interfere with CFLAGS, DEBUG and the like?
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/05/24 01:43:09 Modified files: devel : Makefile Log message: +py-dtopt,python3
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: jas...@cvs.openbsd.org 2018/05/24 01:42:18 Modified files: devel/py-dtopt : Makefile devel/py-dtopt/pkg: PLIST Log message: add python3 flavor
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 01:27:45 Modified files: www/owncloud : Makefile Log message: http -> https prodded by jca@
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 01:24:06 Modified files: sysutils/amazon-ssm-agent: Makefile distinfo Log message: Update to amazon-ssm-agent-2.2.607.0.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 01:21:28 Modified files: x11/gnome/libdazzle: Makefile distinfo Log message: Update to libdazzle-3.28.2.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 00:41:17 Modified files: devel/quirks : Makefile devel/quirks/files: Quirks.pm Log message: Register git-bz removal.
CVS: cvs.openbsd.org: ports
CVSROOT:/cvs Module name:ports Changes by: ajacou...@cvs.openbsd.org 2018/05/24 00:40:29 Modified files: devel : Makefile Removed files: devel/git-bz : Makefile distinfo devel/git-bz/pkg: DESCR PLIST Log message: Remove git-bz. It's unmaintained and lost most of its interest after GNOME moved to GitLab. ok jasper@