From dchmelik at gmail.com Wed Mar 1 04:43:39 2017 From: dchmelik at gmail.com (David Melik) Date: Tue, 28 Feb 2017 20:43:39 -0800 Subject: [Slackbuilds-users] switching a SlackBuild to pip? (Python) Giving up interruptingcow, mpmath SlackBuilds Message-ID: 'How's hacking,' slackbuilds-users, I maintain Mathics SlackBuild, which had update problems with libraries, and the Mathics maintainers aren't always that informative, so I just found a last October update works. However, it uses pip, which downloads source code I already downloaded. If I update my SlackBuild, how should I use pip? I'm not sure Mathics SlackBuild is even needed anymore, as anyone can download, read source code documentation for installation that says do 'pip install mathics'... but I'll maintain this for convenience, if you think some wouldn't go to that extent or know that's possible or how. I give up interruptingcow, mpmath SlackBuilds (Mathics dependencies) since pip installs those for Mathics. If you want to maintain one of those, email me directly or just get permission from SlackBuilds.org administrators and probably have someone inform me. I'm on the slackbuilds-users list but only read the Gmane newsgroup every so often, and I'm on freenode IRC ##slackware, #slackbulids, #freeslack sometimes. Sincerely, David http://davidchmelik.com/ (homesite, updated first, OFFLINE for maintenance) http://www.cwu.edu/~melikd/ (most stable homepage) From baildon.research at googlemail.com Wed Mar 1 12:07:11 2017 From: baildon.research at googlemail.com (David Spencer) Date: Wed, 1 Mar 2017 12:07:11 +0000 Subject: [Slackbuilds-users] switching a SlackBuild to pip? (Python) Giving up interruptingcow, mpmath SlackBuilds In-Reply-To: References: Message-ID: > I maintain Mathics SlackBuild, which had update problems with > libraries, and the Mathics maintainers aren't always that informative, > so I just found a last October update works. However, it uses pip, > which downloads source code I already downloaded. If I update my > SlackBuild, how should I use pip? *Don't* use pip. The upstream instructions say "The easiest way to install Mathics is with pip", but the existing setup.py recipe still works with mathics-1.0. We don't like downloads inside SlackBuilds, for multiple reasons, the only exception being node if it's unavoidable. Also, see below. > I'm not sure Mathics SlackBuild is even needed anymore, as anyone can > download, read source code documentation for installation that says do > 'pip install mathics'... but I'll maintain this for convenience, if > you think some wouldn't go to that extent or know that's possible or > how. The whole point of distro packages instead of pip is that you get packages that can be handled the same as everything else on your system, or indeed scalably and repeatably on a whole cluster of systems, configuration managed, regularly upgraded etc. pip, npm, cpan, gem, cabal, etc, all have their place (single user box where the user is both competent and lazy), but they are not capable of substituting for proper distro packages. Hence SlackBuilds for everything, and also npm2tgz, cpan2tgz and gem2tgz. (There isn't a pip2tgz -- well, there is, but it does something else -- but even if there was, it would be a dead end, because the future is probably wheels instead of pip.) afaict your existing SlackBuild is easy to upgrade for mathics-1.0: just change the tar xvf (the download tarball is now named .tgz), and the patch isn't needed. > I give up interruptingcow, mpmath SlackBuilds (Mathics dependencies) > since pip installs those for Mathics. You might want to reconsider that? Thanks -D. From thedoogster at gmail.com Wed Mar 1 15:39:11 2017 From: thedoogster at gmail.com (Doogster) Date: Wed, 1 Mar 2017 07:39:11 -0800 Subject: [Slackbuilds-users] Doom Retro 2.4.1 released when 2.4.0 still in pending Message-ID: Doom Retro 2.4.1 was released yesterday. 2.4.0 is still in pending and I can't delete it. Could you please remove doomretro.tar.bz2 from pending? From kingbeowulf at gmail.com Wed Mar 1 23:58:17 2017 From: kingbeowulf at gmail.com (King Beowulf) Date: Wed, 1 Mar 2017 15:58:17 -0800 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) Message-ID: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> Hello 1st time compiling LibreOffice and not using the pre-compiled bins. So far so good with jdk and hunspell. just one glitch: In the past I installed (both install to /opt/) office/libreoffice office/libreoffice-helppack Now, when I access the help function, I get the online help (browser fires up and ends up with a usually an unwritten or empty help page) and not the offline help. The off-line help files appear to be missing or not configured even though they seem to exist in the 2nd 'src' file - at least the fetch log indicates a "libreoffice-help-5.3.0.3.tar.xz" was downloaded. 1. Does this source install provide offline help? 2. If not, how do I get offline help? cp from office/libreoffice-helppack? Thanks Ed -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From atelszewski at gmail.com Thu Mar 2 00:07:12 2017 From: atelszewski at gmail.com (Andrzej Telszewski) Date: Thu, 2 Mar 2017 01:07:12 +0100 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) In-Reply-To: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> References: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> Message-ID: <0ee70b8f-966f-cb30-e193-c6de7efb4802@gmail.com> On 02/03/17 00:58, King Beowulf wrote: > Now, when I access the help function, I get the online help (browser > fires up and ends up with a usually an unwritten or empty help page) and > not the offline help. Wow, everything there sends to website. WTF? -- Best regards, Andrzej Telszewski From kingbeowulf at gmail.com Thu Mar 2 02:16:40 2017 From: kingbeowulf at gmail.com (King Beowulf) Date: Wed, 1 Mar 2017 18:16:40 -0800 Subject: [Slackbuilds-users] BIG: nvidia-driver-275/39 Message-ID: <1e73bfd1-9562-06a9-c07e-dbaa1c85dcaf@gmail.com> Seems I goofed some symlinks. If you use nvidia-smi and get: "NVIDIA-SMI couldn't find libnvidia-ml.so library in your system. Please make sure that the NVIDIA Display Driver is properly installed and present in your system. Please also try adding directory that contains libnvidia-ml.so to your system PATH." Do this: cd /usr/lib[64] ln -s libnvidia-ml.so.375.39 libnvidia-ml.so.1 I'll fix this in the next iteration. -Ed ======================= diff --git a/system/nvidia-driver/nvidia-driver.SlackBuild b/system/nvidia-driver/nvidia-driver.SlackBuild index a5d9257..809f72c 100755 --- a/system/nvidia-driver/nvidia-driver.SlackBuild +++ b/system/nvidia-driver/nvidia-driver.SlackBuild @@ -165,6 +165,7 @@ cd $PKG/usr/lib$LIBDIRSUFFIX ln -sf libGLESv1_CM_nvidia.so.$VERSION libGLESv1_CM_nvidia.so.1 ln -sf libGLESv2_nvidia.so.$VERSION libGLESv2_nvidia.so.2 ln -sf libnvidia-tls.so.$VERSION libnvidia-tls.so + ln -sf libnvidia-ml.so.$VERSION libnvidia-ml.so.1 ln -sf vdpau/libvdpau_nvidia.so.$VERSION libvdpau_nvidia.so cd - cd $PKG/usr/lib$LIBDIRSUFFIX/tls @@ -225,6 +226,7 @@ if [ "$ARCH" = "x86_64" -a "${COMPAT32}" = "yes" ]; then ln -sf libGLESv1_CM_nvidia.so.$VERSION libGLESv1_CM_nvidia.so.1 ln -sf libGLESv2_nvidia.so.$VERSION libGLESv2_nvidia.so.2 ln -sf libnvidia-tls.so.$VERSION libnvidia-tls.so + ln -sf libnvidia-ml.so.$VERSION libnvidia-ml.so.1 ln -sf vdpau/libvdpau_nvidia.so.$VERSION libvdpau_nvidia.so cd - cd $PKG/usr/lib/tls ============================ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From willysr at slackbuilds.org Thu Mar 2 05:56:17 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Thu, 2 Mar 2017 12:56:17 +0700 Subject: [Slackbuilds-users] Doom Retro 2.4.1 released when 2.4.0 still in pending In-Reply-To: References: Message-ID: > Doom Retro 2.4.1 was released yesterday. 2.4.0 is still in pending and > I can't delete it. > > Could you please remove doomretro.tar.bz2 from pending? updated in my branch to 2.4.1 -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From d.zlatanidis at gmail.com Thu Mar 2 11:25:56 2017 From: d.zlatanidis at gmail.com (Dimitris Zlatanidis) Date: Thu, 2 Mar 2017 13:25:56 +0200 Subject: [Slackbuilds-users] task Message-ID: Hi, task slackbuild seems download source url has changed. I get checksum error https://www.slackbuilds.org/repository/14.2/office/task/ DONWLOAD="http://www.taskwarrior.org/download/task-2.5.1.tar.gz" MD5SUM="cb07f7933ac6f3ad19606215829ada16" Replaced with: DOWNLOAD="https://taskwarrior.org/download/task-2.5.1.tar.gz" MD5SUM="bcd984a00d6d1eb6b40faf567419f784" Thanks -- Dimitris Zlatanidis From chris.willing at iinet.net.au Thu Mar 2 11:42:56 2017 From: chris.willing at iinet.net.au (Christoph Willing) Date: Thu, 2 Mar 2017 21:42:56 +1000 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) In-Reply-To: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> References: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> Message-ID: <459298f9-41b5-8def-9d6d-378635d9fc33@iinet.net.au> On 02/03/17 09:58, King Beowulf wrote: > Hello > > 1st time compiling LibreOffice and not using the pre-compiled bins. So > far so good with jdk and hunspell. just one glitch: > > In the past I installed (both install to /opt/) > office/libreoffice > office/libreoffice-helppack > > Now, when I access the help function, I get the online help (browser > fires up and ends up with a usually an unwritten or empty help page) and > not the offline help. > Sorry about that. An updated SlackBuild is on its way but in the meantime you could fix it locally by changing: --with-help=common \ configure option to: --with-help \ Thanks for pointing out the problem. chris -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: OpenPGP digital signature URL: From chris.willing at iinet.net.au Thu Mar 2 12:22:09 2017 From: chris.willing at iinet.net.au (Christoph Willing) Date: Thu, 2 Mar 2017 22:22:09 +1000 Subject: [Slackbuilds-users] openjdk was: Re: Updates - 20170225.1 In-Reply-To: <291706d3-8f27-db7b-2d7f-f76a84825acc@googlemail.com> References: <6f713cb3-6546-c9ad-9612-32963f5f5a55@slackbuilds.org> <7153f517-b926-3a7e-ad8c-51fae47539da@iinet.net.au> <39182867-607c-87e6-59a9-d52b98703467@iinet.net.au> <130f67f8-f710-e98a-6f20-7529416c8933@googlemail.com> <5de3fbc8-b8fd-04d2-c34d-2bb4d46094f9@googlemail.com> <58dd001e-2e63-134e-e57a-7dd25dc0b49f@gmail.com> <291706d3-8f27-db7b-2d7f-f76a84825acc@googlemail.com> Message-ID: <7c1ca4e0-5e1a-bb9c-0d77-c4982829c5d7@iinet.net.au> I just had reason to update the SlackBuild for another matter (missing local help reader) and took the opportunity to update the Java capabilities at the same time. I had some sympathy with the statement that: "some people will prefer openjdk because of licensing. People who care about the "Libre" in the LibreOffice name might care about this." so I've changed the default to openjdk with alternatives selected by setting a JAVA environment variable to the name of relevant SBo SlackBuild; either openjdk, openjdk8, jdk or no (don't build with Java support). chris From atelszewski at gmail.com Thu Mar 2 13:43:46 2017 From: atelszewski at gmail.com (Andrzej Telszewski) Date: Thu, 2 Mar 2017 14:43:46 +0100 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) In-Reply-To: <459298f9-41b5-8def-9d6d-378635d9fc33@iinet.net.au> References: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> <459298f9-41b5-8def-9d6d-378635d9fc33@iinet.net.au> Message-ID: On 02/03/17 12:42, Christoph Willing wrote: > On 02/03/17 09:58, King Beowulf wrote: >> Hello >> >> 1st time compiling LibreOffice and not using the pre-compiled bins. So >> far so good with jdk and hunspell. just one glitch: >> >> In the past I installed (both install to /opt/) >> office/libreoffice >> office/libreoffice-helppack >> >> Now, when I access the help function, I get the online help (browser >> fires up and ends up with a usually an unwritten or empty help page) and >> not the offline help. >> > Sorry about that. An updated SlackBuild is on its way but in the > meantime you could fix it locally by changing: > --with-help=common \ > configure option to: > --with-help \ > > Thanks for pointing out the problem. > > chris > > That is nice. Just to make sure, my WTF in the previous message wasn't about you (SlackBuild provider), but about the general tendency of making everything web based. -- Best regards, Andrzej Telszewski From willysr at slackbuilds.org Thu Mar 2 16:02:08 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Thu, 2 Mar 2017 23:02:08 +0700 Subject: [Slackbuilds-users] task In-Reply-To: References: Message-ID: > task slackbuild seems download source url has changed. I get checksum error > > https://www.slackbuilds.org/repository/14.2/office/task/ > > > DONWLOAD="http://www.taskwarrior.org/download/task-2.5.1.tar.gz" > > MD5SUM="cb07f7933ac6f3ad19606215829ada16" > > Replaced with: > > DOWNLOAD="https://taskwarrior.org/download/task-2.5.1.tar.gz" Fixed in my branch Thanks -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From ts at websafe.pl Thu Mar 2 18:17:23 2017 From: ts at websafe.pl (Thomas Szteliga) Date: Thu, 2 Mar 2017 19:17:23 +0100 Subject: [Slackbuilds-users] osinfo-db-tools: wrong MD5SUM / project moved to gitlab Message-ID: <08e47639-c29e-e168-dc8d-cface08f79e6@websafe.pl> +============================================================================== | MD5SUM check for osinfo-db-tools-1.1.0.tar.gz [ FAILED ] +============================================================================== | Expected: 5b346d7e361a7f510aa62da923cbdd8b | Found: a5619251bd9f91b5063d8a0d8a618f71 +============================================================================== New location: https://gitlab.com/libosinfo/libosinfo Infrastructure/Fedorahosted-retirement: https://fedoraproject.org/wiki/Infrastructure/Fedorahosted-retirement -- Thomas Szteliga From matteo.bernardini at gmail.com Thu Mar 2 19:54:39 2017 From: matteo.bernardini at gmail.com (Matteo Bernardini) Date: Thu, 2 Mar 2017 20:54:39 +0100 Subject: [Slackbuilds-users] osinfo-db-tools: wrong MD5SUM / project moved to gitlab In-Reply-To: <08e47639-c29e-e168-dc8d-cface08f79e6@websafe.pl> References: <08e47639-c29e-e168-dc8d-cface08f79e6@websafe.pl> Message-ID: 2017-03-02 19:17 GMT+01:00 Thomas Szteliga : > +============================================================================== > | MD5SUM check for osinfo-db-tools-1.1.0.tar.gz [ FAILED ] > +============================================================================== > | Expected: 5b346d7e361a7f510aa62da923cbdd8b > | Found: a5619251bd9f91b5063d8a0d8a618f71 > +============================================================================== > > > New location: https://gitlab.com/libosinfo/libosinfo thanks Thomas for the report! It seems libosinfo and friends are now hosted on pagure.org https://libosinfo.org/download/ > Infrastructure/Fedorahosted-retirement: > https://fedoraproject.org/wiki/Infrastructure/Fedorahosted-retirement this most probably affects also these other scripts desktop/gtk-nodoka-engine libraries/libpwquality python/python-augeas python/python-suds python/python3-augeas system/cronie system/microcode_ctl system/targetcli-fb system/virt-viewer I'll have a look. Matteo From chris.willing at iinet.net.au Thu Mar 2 20:43:09 2017 From: chris.willing at iinet.net.au (Christoph Willing) Date: Fri, 3 Mar 2017 06:43:09 +1000 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) In-Reply-To: References: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> <459298f9-41b5-8def-9d6d-378635d9fc33@iinet.net.au> Message-ID: On 02/03/17 23:43, Andrzej Telszewski wrote: > On 02/03/17 12:42, Christoph Willing wrote: >> On 02/03/17 09:58, King Beowulf wrote: >>> Hello >>> >>> 1st time compiling LibreOffice and not using the pre-compiled bins. So >>> far so good with jdk and hunspell. just one glitch: >>> >>> In the past I installed (both install to /opt/) >>> office/libreoffice >>> office/libreoffice-helppack >>> >>> Now, when I access the help function, I get the online help (browser >>> fires up and ends up with a usually an unwritten or empty help page) and >>> not the offline help. >>> >> Sorry about that. An updated SlackBuild is on its way but in the >> meantime you could fix it locally by changing: >> --with-help=common \ >> configure option to: >> --with-help \ >> >> Thanks for pointing out the problem. >> >> chris >> >> > > That is nice. > > Just to make sure, my WTF in the previous message wasn't about you > (SlackBuild provider), but about the general tendency of making > everything web based. > No worries. Actually I had noticed the help problem a long time ago and intended to sort it out sometime but, of course, forgot all about it till now; it probably deserves a WTF anyway :) chris From kingbeowulf at gmail.com Thu Mar 2 21:34:45 2017 From: kingbeowulf at gmail.com (King Beowulf) Date: Thu, 2 Mar 2017 13:34:45 -0800 Subject: [Slackbuilds-users] LibreOffice-5.3.0.3 (source install) In-Reply-To: References: <1d1a36b3-8d2d-d4ec-8dd6-54efaba5c066@gmail.com> <459298f9-41b5-8def-9d6d-378635d9fc33@iinet.net.au> Message-ID: <4d4d57bb-9a7d-3b34-6919-9eafdc9fb06b@gmail.com> On 03/02/2017 12:43 PM, Christoph Willing wrote: > On 02/03/17 23:43, Andrzej Telszewski wrote: >> On 02/03/17 12:42, Christoph Willing wrote: >>> On 02/03/17 09:58, King Beowulf wrote: >>>> Now, when I access the help function, I get the online help (browser >>>> fires up and ends up with a usually an unwritten or empty help page) and >>>> not the offline help. >>>> >>> Sorry about that. An updated SlackBuild is on its way but in the >>> meantime you could fix it locally by changing: >>> --with-help=common \ >>> configure option to: >>> --with-help \ >>> >>> Thanks for pointing out the problem. >>> >>> chris >> >> That is nice. >> >> Just to make sure, my WTF in the previous message wasn't about you >> (SlackBuild provider), but about the general tendency of making >> everything web based. >> > No worries. Actually I had noticed the help problem a long time ago and > intended to sort it out sometime but, of course, forgot all about it > till now; it probably deserves a WTF anyway :) > > chris Heh, I have all sorts of 'wtf' moments lately tying to get nvidia-driver to behave! -Ed -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From dchmelik at gmail.com Fri Mar 3 02:31:07 2017 From: dchmelik at gmail.com (David Melik) Date: Thu, 2 Mar 2017 18:31:07 -0800 Subject: [Slackbuilds-users] updating Mathics SlackBuild, giving up interruptingcow SlackBuild Message-ID: >*Don't* use pip. The upstream instructions say "The easiest way to install Mathics is with pip", but the existing setup.py recipe still works with mathics-1.0. Okay. >afaict your existing SlackBuild is easy to upgrade for mathics-1.0: just change the tar xvf (the download tarball is now named .tgz), and the patch isn't needed. Yes, except for dependencies, for which it may need a new patch. Also, the last version of python-django I used (from source, changing SlackBuild version number) caused errors and it seems that SlackBuild hasn't been updated since sometime in Slackware 14.1. >>I give up interruptingcow, mpmath SlackBuilds (Mathics dependencies) since pip installs those for Mathics. > >You might want to reconsider that? Okay, I'll keep mpmath. The other is no longer necessary for Mathics. From duncan_roe at acslink.net.au Wed Mar 1 22:06:59 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Thu, 2 Mar 2017 09:06:59 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails Message-ID: <20170301220658.GA24715@dimstar.local.net> This looks like an upstream bug. The gdal SB initially fails as below: > running build_ext > building 'osgeo._gdal' extension > creating build/temp.linux-x86_64-2.7 > creating build/temp.linux-x86_64-2.7/extensions > /bin/sh /tmp/SBo/gdal-2.1.3/libtool --mode=compile --tag=CC gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -o build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o > libtool: compile: gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -fPIC -DPIC -o build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o > cc1plus: warning: command line option '-Wstrict-prototypes' is valid for C/ObjC but not for C++ > In file included from /usr/include/python2.7/Python.h:133:0, > from extensions/gdal_wrap.cpp:155: > extensions/gdal_wrap.cpp: In function 'PyObject* _wrap_MajorObject_SetMetadata__SWIG_0(PyObject*, PyObject*)': > /usr/include/python2.7/abstract.h:1354:62: warning: deprecated conversion from string constant to 'char*' [-Wwrite-strings] > #define PyMapping_Items(O) PyObject_CallMethod(O,"items",NULL) > ^ > extensions/gdal_wrap.cpp:9115:31: note: in expansion of macro 'PyMapping_Items' > PyObject *item_list = PyMapping_Items( obj1 ); > ^ > g++ -pthread -shared build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o -L../../.libs -L../../ -L/usr/lib64 -L/tmp/SBo/gdal-2.1.3/lib -lpython2.7 -lgdal -o build/lib.linux-x86_64-2.7/osgeo/_gdal.so > g++: error: build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o: No such file or directory > error: command 'g++' failed with exit status 1 > GNUmakefile:74: recipe for target 'build' failed > make[2]: *** [build] Error 1 > make[2]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig/python' > GNUmakefile:30: recipe for target 'build' failed > make[1]: *** [build] Error 2 > make[1]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig' > GNUmakefile:104: recipe for target 'swig-modules' failed > make: *** [swig-modules] Error 2 On manually re-running make, it runs to completion. However make install now fails identically. The error above happens because gdal_wrap.o is in build/temp.linux-x86_64-2.7/extensions/.libs (within /tmp/SBo/gdal-2.1.3/swig/python). build/temp.linux-x86_64-2.7/extensions contains gdal_wrap.lo: > + cat gdal_wrap.lo > # build/temp.linux-x86_64-2.7/extensions/gdal_wrap.lo - a libtool object file > # Generated by libtool (GNU libtool) 2.4 > # > # Please DO NOT delete this file! > # It is necessary for linking the library. > > # Name of the PIC object. > pic_object='.libs/gdal_wrap.o' > > # Name of the non-PIC object > non_pic_object=none That looks as if it ought to work I guess, but it doesn't. I salvaged the build by: > 08:40:26# cd build/temp.linux-x86_64-2.7/extensions > 08:47:54# mv .libs/gdal_wrap.o . > 08:48:22# rmdir .libs > 08:48:30# ln -s . .libs > 08:50:46# cd /tmp/SBo/gdal-2.1.3 > 08:50:57# make install DESTDIR=$PKG then running the rest of the SB commands by hand. (I had run some of the previous commands already, e.g. to set $PKG). Do I need a newer version of ld or what? Cheers ... Duncan. From willysr at slackbuilds.org Fri Mar 3 05:56:39 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Fri, 3 Mar 2017 12:56:39 +0700 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170301220658.GA24715@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> Message-ID: > The gdal SB initially fails as below: > >> running build_ext >> building 'osgeo._gdal' extension >> creating build/temp.linux-x86_64-2.7 >> creating build/temp.linux-x86_64-2.7/extensions >> /bin/sh /tmp/SBo/gdal-2.1.3/libtool --mode=compile --tag=CC gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -o build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o >> libtool: compile: gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -fPIC -DPIC -o build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o >> cc1plus: warning: command line option '-Wstrict-prototypes' is valid for C/ObjC but not for C++ >> In file included from /usr/include/python2.7/Python.h:133:0, >> from extensions/gdal_wrap.cpp:155: >> extensions/gdal_wrap.cpp: In function 'PyObject* _wrap_MajorObject_SetMetadata__SWIG_0(PyObject*, PyObject*)': >> /usr/include/python2.7/abstract.h:1354:62: warning: deprecated conversion from string constant to 'char*' [-Wwrite-strings] >> #define PyMapping_Items(O) PyObject_CallMethod(O,"items",NULL) >> ^ >> extensions/gdal_wrap.cpp:9115:31: note: in expansion of macro 'PyMapping_Items' >> PyObject *item_list = PyMapping_Items( obj1 ); >> ^ >> g++ -pthread -shared build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o -L../../.libs -L../../ -L/usr/lib64 -L/tmp/SBo/gdal-2.1.3/lib -lpython2.7 -lgdal -o build/lib.linux-x86_64-2.7/osgeo/_gdal.so >> g++: error: build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o: No such file or directory >> error: command 'g++' failed with exit status 1 >> GNUmakefile:74: recipe for target 'build' failed >> make[2]: *** [build] Error 1 >> make[2]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig/python' >> GNUmakefile:30: recipe for target 'build' failed >> make[1]: *** [build] Error 2 >> make[1]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig' >> GNUmakefile:104: recipe for target 'swig-modules' failed >> make: *** [swig-modules] Error 2 > > On manually re-running make, it runs to completion. > > However make install now fails identically. > > The error above happens because gdal_wrap.o is in > build/temp.linux-x86_64-2.7/extensions/.libs (within > /tmp/SBo/gdal-2.1.3/swig/python). > > build/temp.linux-x86_64-2.7/extensions contains gdal_wrap.lo: > >> + cat gdal_wrap.lo >> # build/temp.linux-x86_64-2.7/extensions/gdal_wrap.lo - a libtool object file >> # Generated by libtool (GNU libtool) 2.4 >> # >> # Please DO NOT delete this file! >> # It is necessary for linking the library. >> >> # Name of the PIC object. >> pic_object='.libs/gdal_wrap.o' >> >> # Name of the non-PIC object >> non_pic_object=none > > That looks as if it ought to work I guess, but it doesn't. > > I salvaged the build by: > >> 08:40:26# cd build/temp.linux-x86_64-2.7/extensions >> 08:47:54# mv .libs/gdal_wrap.o . >> 08:48:22# rmdir .libs >> 08:48:30# ln -s . .libs >> 08:50:46# cd /tmp/SBo/gdal-2.1.3 >> 08:50:57# make install DESTDIR=$PKG > > then running the rest of the SB commands by hand. (I had run some of the > previous commands already, e.g. to set $PKG). > > Do I need a newer version of ld or what? works fine here proj: MD5SUM check for proj-4.9.3.tar.gz ... OK Building package proj-4.9.3-x86_64-1_SBo.tgz ... OK Installing package proj-4.9.3-x86_64-1_SBo.tgz ... OK geos: MD5SUM check for geos-3.6.1.tar.bz2 ... OK Building package geos-3.6.1-x86_64-1_SBo.tgz ... OK Installing package geos-3.6.1-x86_64-1_SBo.tgz ... OK gdal: MD5SUM check for gdal-2.1.3.tar.xz ... OK Building package gdal-2.1.3-x86_64-1_SBo.tgz ... OK Installing package gdal-2.1.3-x86_64-1_SBo.tgz ... OK -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From yalhcru at gmail.com Fri Mar 3 07:35:52 2017 From: yalhcru at gmail.com (B Watson) Date: Fri, 3 Mar 2017 02:35:52 -0500 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170301220658.GA24715@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> Message-ID: On 3/1/17, Duncan Roe wrote: > This looks like an upstream bug. > > The gdal SB initially fails as below: Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything other than -j1). That's a snap diagnosis, I don't know what gdal even is. From yalhcru at gmail.com Fri Mar 3 08:38:21 2017 From: yalhcru at gmail.com (B Watson) Date: Fri, 3 Mar 2017 03:38:21 -0500 Subject: [Slackbuilds-users] rlog source MIA Message-ID: --2017-03-03 03:28:28-- http://rlog.googlecode.com/files/rlog-1.4.tar.gz Resolving rlog.googlecode.com (rlog.googlecode.com)... 2607:f8b0:4002:c06::52, 74.125.196.82 Connecting to rlog.googlecode.com (rlog.googlecode.com)|2607:f8b0:4002:c06::52|:80... connected. HTTP request sent, awaiting response... 404 Not Found 2017-03-03 03:28:28 ERROR 404: Not Found. There's a copy on sbosrcarch: http://slackware.uk/sbosrcarch//by-md5/c/2/c29f74e0f50d66b20312d049b683ff82/rlog-1.4.tar.gz There's also a copy here: http://distcache.freebsd.org/ports-distfiles/rlog-1.4.tar.gz It would probably be better to use the freebsd.org link in the .info file, that way the sbosrcarch mirror has something to be a mirror of (if that makes sense). From duncan_roe at acslink.net.au Fri Mar 3 08:44:33 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Fri, 3 Mar 2017 19:44:33 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> Message-ID: <20170303084433.GA1333@dimstar.local.net> On Fri, Mar 03, 2017 at 12:56:39PM +0700, Willy Sudiarto Raharjo wrote: > > The gdal SB initially fails as below: > > > >> running build_ext > >> building 'osgeo._gdal' extension > >> creating build/temp.linux-x86_64-2.7 > >> creating build/temp.linux-x86_64-2.7/extensions > >> /bin/sh /tmp/SBo/gdal-2.1.3/libtool --mode=compile --tag=CC gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -o build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o > >> libtool: compile: gcc -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I/usr/lib64/python2.7/site-packages/numpy/core/include -I/tmp/SBo/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -fPIC -DPIC -o build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o > >> cc1plus: warning: command line option '-Wstrict-prototypes' is valid for C/ObjC but not for C++ > >> In file included from /usr/include/python2.7/Python.h:133:0, > >> from extensions/gdal_wrap.cpp:155: > >> extensions/gdal_wrap.cpp: In function 'PyObject* _wrap_MajorObject_SetMetadata__SWIG_0(PyObject*, PyObject*)': > >> /usr/include/python2.7/abstract.h:1354:62: warning: deprecated conversion from string constant to 'char*' [-Wwrite-strings] > >> #define PyMapping_Items(O) PyObject_CallMethod(O,"items",NULL) > >> ^ > >> extensions/gdal_wrap.cpp:9115:31: note: in expansion of macro 'PyMapping_Items' > >> PyObject *item_list = PyMapping_Items( obj1 ); > >> ^ > >> g++ -pthread -shared build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o -L../../.libs -L../../ -L/usr/lib64 -L/tmp/SBo/gdal-2.1.3/lib -lpython2.7 -lgdal -o build/lib.linux-x86_64-2.7/osgeo/_gdal.so > >> g++: error: build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o: No such file or directory > >> error: command 'g++' failed with exit status 1 > >> GNUmakefile:74: recipe for target 'build' failed > >> make[2]: *** [build] Error 1 > >> make[2]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig/python' > >> GNUmakefile:30: recipe for target 'build' failed > >> make[1]: *** [build] Error 2 > >> make[1]: Leaving directory '/tmp/SBo/gdal-2.1.3/swig' > >> GNUmakefile:104: recipe for target 'swig-modules' failed > >> make: *** [swig-modules] Error 2 > > > > On manually re-running make, it runs to completion. > > > > However make install now fails identically. > > > > The error above happens because gdal_wrap.o is in > > build/temp.linux-x86_64-2.7/extensions/.libs (within > > /tmp/SBo/gdal-2.1.3/swig/python). > > > > build/temp.linux-x86_64-2.7/extensions contains gdal_wrap.lo: > > > >> + cat gdal_wrap.lo > >> # build/temp.linux-x86_64-2.7/extensions/gdal_wrap.lo - a libtool object file > >> # Generated by libtool (GNU libtool) 2.4 > >> # > >> # Please DO NOT delete this file! > >> # It is necessary for linking the library. > >> > >> # Name of the PIC object. > >> pic_object='.libs/gdal_wrap.o' > >> > >> # Name of the non-PIC object > >> non_pic_object=none > > > > That looks as if it ought to work I guess, but it doesn't. > > > > I salvaged the build by: > > > >> 08:40:26# cd build/temp.linux-x86_64-2.7/extensions > >> 08:47:54# mv .libs/gdal_wrap.o . > >> 08:48:22# rmdir .libs > >> 08:48:30# ln -s . .libs > >> 08:50:46# cd /tmp/SBo/gdal-2.1.3 > >> 08:50:57# make install DESTDIR=$PKG > > > > then running the rest of the SB commands by hand. (I had run some of the > > previous commands already, e.g. to set $PKG). > > > > Do I need a newer version of ld or what? > > works fine here > > proj: > MD5SUM check for proj-4.9.3.tar.gz ... OK > Building package proj-4.9.3-x86_64-1_SBo.tgz ... OK > Installing package proj-4.9.3-x86_64-1_SBo.tgz ... OK > > geos: > MD5SUM check for geos-3.6.1.tar.bz2 ... OK > Building package geos-3.6.1-x86_64-1_SBo.tgz ... OK > Installing package geos-3.6.1-x86_64-1_SBo.tgz ... OK > > gdal: > MD5SUM check for gdal-2.1.3.tar.xz ... OK > Building package gdal-2.1.3-x86_64-1_SBo.tgz ... OK > Installing package gdal-2.1.3-x86_64-1_SBo.tgz ... OK > > > -- > Willy Sudiarto Raharjo > I also had libwebp installed, but none of the other optional dependencies. I tried removepkg libwebp then rebuild, but it still failed for me. Did you have any other optional deps installed? Cheers ... Duncan. From willysr at slackbuilds.org Fri Mar 3 10:17:35 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Fri, 3 Mar 2017 17:17:35 +0700 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170303084433.GA1333@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170303084433.GA1333@dimstar.local.net> Message-ID: > I also had libwebp installed, but none of the other optional dependencies. > > I tried removepkg libwebp then rebuild, but it still failed for me. > > Did you have any other optional deps installed? No, i always build it on a clean VM -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From baildon.research at googlemail.com Fri Mar 3 12:19:09 2017 From: baildon.research at googlemail.com (David Spencer) Date: Fri, 3 Mar 2017 12:19:09 +0000 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> Message-ID: > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > other than -j1). That's a snap diagnosis, I don't know what gdal even is. gdal already has 'make -j1' hard-coded, because it definitely fails with multiple parallel make jobs. For some reason that I don't understand, on Duncan's system gcc is being invoked through libtool. The compilation explicitly has '-o build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that file does not exist at the time the next command 'g++ -pthread -shared build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. So this really does look like a multiple parallel make jobs failure. Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? Are you on -current? Do you have CC and/or CXX defined in the environment? could you put your config.log on pastebin or somewhere similar? Here's the corresponding part of a clean build on my box. running build_ext building 'osgeo._gdal' extension creating build/temp.linux-x86_64-2.7 creating build/temp.linux-x86_64-2.7/extensions gcc -pthread -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I../../port -I../../gcore -I../../alg -I../../ogr/ -I../../ogr/ogrsf_frmts -I../../gnm -I../../apps -I/usr/include/python2.7 -I. -I/tmp/review/gdal-2.1.3/include -c extensions/gdal_wrap.cpp -o build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o cc1plus: warning: command line option ?-Wstrict-prototypes? is valid for C/ObjC but not for C++ In file included from /usr/include/python2.7/Python.h:133:0, from extensions/gdal_wrap.cpp:155: extensions/gdal_wrap.cpp: In function ?PyObject* _wrap_MajorObject_SetMetadata__SWIG_0(PyObject*, PyObject*)?: /usr/include/python2.7/abstract.h:1354:62: warning: deprecated conversion from string constant to ?char*? [-Wwrite-strings] #define PyMapping_Items(O) PyObject_CallMethod(O,"items",NULL) ^ extensions/gdal_wrap.cpp:9115:31: note: in expansion of macro ?PyMapping_Items? PyObject *item_list = PyMapping_Items( obj1 ); ^ g++ -pthread -shared build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o -L../../.libs -L../../ -L/usr/lib64 -L/tmp/review/gdal-2.1.3/lib -lpython2.7 -lgdal -o build/lib.linux-x86_64-2.7/osgeo/_gdal.so building 'osgeo._gdalconst' extension [...] I see that Duncan has numpy, whereas I don't. Clearly I should document numpy as a requirement for the Python bindings in README, or maybe make it a hard dep... A bit perplexed -D. From fellype at gmail.com Fri Mar 3 14:15:52 2017 From: fellype at gmail.com (Fellype do Nascimento) Date: Fri, 3 Mar 2017 11:15:52 -0300 Subject: [Slackbuilds-users] extra-cmake-modules.SlackBuild - Approved version does not gzip manpages Message-ID: <792580aa-eaba-324f-c66a-05ce2bc327e7@gmail.com> Dear SBo admins, I verified that in the approved version of the extra-cmake-modules.SlackBuild the piece of code that compress the manpages was removed from the script. Then, the manpages generated in the build process will not be gzipped, and this is not the standard. In the attached script there is a more simple piece of code (just a line) that compress the generated manpages: # Compress manpages gzip -9 $PKG/usr/man/man7/*.7 Please, include this in the extra-cmake-modules.SlackBuild before the next update. If you think it is more appropriate to submit the script again via submission form, please let me know and I will do so. Best regards, Fellype -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: extra-cmake-modules.SlackBuild Type: all/slackware-build-script Size: 3066 bytes Desc: not available URL: From willysr at slackbuilds.org Fri Mar 3 14:22:24 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Fri, 3 Mar 2017 21:22:24 +0700 Subject: [Slackbuilds-users] extra-cmake-modules.SlackBuild - Approved version does not gzip manpages In-Reply-To: <792580aa-eaba-324f-c66a-05ce2bc327e7@gmail.com> References: <792580aa-eaba-324f-c66a-05ce2bc327e7@gmail.com> Message-ID: <3756f495-c23e-d93a-d2bb-d441cf223fb8@slackbuilds.org> > I verified that in the approved version of the > extra-cmake-modules.SlackBuild the piece of code that compress the > manpages was removed from the script. Then, the manpages generated in > the build process will not be gzipped, and this is not the standard. > In the attached script there is a more simple piece of code (just a > line) that compress the generated manpages: > > # Compress manpages > gzip -9 $PKG/usr/man/man7/*.7 > > Please, include this in the extra-cmake-modules.SlackBuild before the > next update. Done -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From duncan_roe at acslink.net.au Fri Mar 3 08:47:05 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Fri, 3 Mar 2017 19:47:05 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> Message-ID: <20170303084705.GB1333@dimstar.local.net> On Fri, Mar 03, 2017 at 02:35:52AM -0500, B Watson wrote: > On 3/1/17, Duncan Roe wrote: > > This looks like an upstream bug. > > > > The gdal SB initially fails as below: > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > other than -j1). That's a snap diagnosis, I don't know what gdal even is. > _______________________________________________ Thanks for the guess, but gdal.SlackBuild has explicitly make -j1 Cheers .. Duncan. From rshepard at appl-ecosys.com Fri Mar 3 17:41:30 2017 From: rshepard at appl-ecosys.com (Rich Shepard) Date: Fri, 3 Mar 2017 09:41:30 -0800 (PST) Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 Message-ID: Recently I upgraded Python3 from the 3.5 series to the 3.6 series, yet /usr/lib/ still has a python3.5/site-packages/ tree rather than a python3.6/site-packages/ tree. I tried installing an application requiring PyQt5 and it could not find it: Checking dependencies... Python Version 3.6.0 -> OK PyQt not found -> Sorry, PyQt 5.4 or higher is required and must be built for Python 3 Installed here are: python3-3.6.0-i586-3_SBo python3-PyQt5-5.6-i486-1_SBo python3-sip-4.18-i486-1_SBo Please point me to the resolution of this problem. Rich From matteo.bernardini at gmail.com Fri Mar 3 17:54:29 2017 From: matteo.bernardini at gmail.com (Matteo Bernardini) Date: Fri, 3 Mar 2017 18:54:29 +0100 Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 In-Reply-To: References: Message-ID: 2017-03-03 18:41 GMT+01:00 Rich Shepard : > Recently I upgraded Python3 from the 3.5 series to the 3.6 series, yet > /usr/lib/ still has a python3.5/site-packages/ tree rather than a > python3.6/site-packages/ tree. > > I tried installing an application requiring PyQt5 and it could not find > it: > > Checking dependencies... > Python Version 3.6.0 -> OK > PyQt not found -> Sorry, PyQt 5.4 or higher is required and must be built > for Python 3 > > Installed here are: > > python3-3.6.0-i586-3_SBo > python3-PyQt5-5.6-i486-1_SBo > python3-sip-4.18-i486-1_SBo > > Please point me to the resolution of this problem. hi Rich, have you rebuilt all of yours python3-* (including python3-PyQt5 and python3-sip) packages after upgrading to the new python3? if you had built them before upgrading it they will probably be still in /usr/lib/python3.5/site-packages instead of /usr/lib/python3.6/site-packages... Matteo From rshepard at appl-ecosys.com Fri Mar 3 18:02:09 2017 From: rshepard at appl-ecosys.com (Rich Shepard) Date: Fri, 3 Mar 2017 10:02:09 -0800 (PST) Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 In-Reply-To: References: Message-ID: On Fri, 3 Mar 2017, Matteo Bernardini wrote: > have you rebuilt all of yours python3-* (including python3-PyQt5 and > python3-sip) packages after upgrading to the new python3? Matteo, Oops! No. > if you had built them before upgrading it they will probably be still in > /usr/lib/python3.5/site-packages instead of > /usr/lib/python3.6/site-packages... Got it. Thanks for the insight, Rich From rshepard at appl-ecosys.com Fri Mar 3 18:30:42 2017 From: rshepard at appl-ecosys.com (Rich Shepard) Date: Fri, 3 Mar 2017 10:30:42 -0800 (PST) Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 In-Reply-To: References: Message-ID: On Fri, 3 Mar 2017, Matteo Bernardini wrote: > have you rebuilt all of yours python3-* (including python3-PyQt5 and > python3-sip) packages after upgrading to the new python3? Matteo, When I tried to re-build python3-PyQt5 the configuration failed: configure.py: error: '/usr/include/python3.6' is not a directory so I looked in /usr/include. I found a bunch of directories: python2.7/ python3.3m/ python3.5/ python3.6m/ Since 2.7 and 3.3m are old I can remove them. But, why is the directory python3.6m/ rather than python3.6/? Do I rebuild and re-install python3-3.6 to fix this? Thanks, Rich From ekinakoglu at gmail.com Fri Mar 3 19:03:52 2017 From: ekinakoglu at gmail.com (Ekin Akoglu) Date: Fri, 3 Mar 2017 22:03:52 +0300 Subject: [Slackbuilds-users] New submissions Message-ID: Dear Willy, I see that you have approved my submissions of arpack, g3data and lapack95 for all of which I am the new maintainer. However, my submission for "arpack" is seen as "lapack" in the approved submissions page and in cgit (https://slackbuilds.org/cgit/slackbuilds/commit/?id=080e60704). Further, my submission of "lapack95" is not visible at all. Could there be a mistake? Thank you for your attention. Best, Ekin From matteo.bernardini at gmail.com Fri Mar 3 20:01:07 2017 From: matteo.bernardini at gmail.com (Matteo Bernardini) Date: Fri, 3 Mar 2017 21:01:07 +0100 Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 In-Reply-To: References: Message-ID: 2017-03-03 19:30 GMT+01:00 Rich Shepard : > On Fri, 3 Mar 2017, Matteo Bernardini wrote: > >> have you rebuilt all of yours python3-* (including python3-PyQt5 and >> python3-sip) packages after upgrading to the new python3? > > > Matteo, > > When I tried to re-build python3-PyQt5 the configuration failed: > > configure.py: error: '/usr/include/python3.6' is not a directory > > so I looked in /usr/include. I found a bunch of directories: > > python2.7/ > python3.3m/ > python3.5/ > python3.6m/ > > Since 2.7 and 3.3m are old I can remove them. But, why is the directory > python3.6m/ rather than python3.6/? Do I rebuild and re-install python3-3.6 > to fix this? you shouldn't remove /usr/include/python-2.7 as it cointains file belonging to Slackware's stock python, that you should leave installed (as other packages in the system depends on it). /usr/include/python3.3m and /usr/include/python3.5 are probably leftovers of old versions of python3 installed in your system: you could have a look into them to have a hint on whatever needs to be rebuilt. /usr/include/python3.6m containes pyhton-3.6.0 headers and should be fine where it is. your issue is because you are trying to rebuild python3-PyQt5 before having rebuilt also its dependency python3-sip: this last will install a file /usr/include/python3.6/sip.h that is exactly what the python3-PyQt5 build system is looking for. Matteo From rshepard at appl-ecosys.com Fri Mar 3 20:10:09 2017 From: rshepard at appl-ecosys.com (Rich Shepard) Date: Fri, 3 Mar 2017 12:10:09 -0800 (PST) Subject: [Slackbuilds-users] Upgrading python3-3.5 to python3-3.6 In-Reply-To: References: Message-ID: On Fri, 3 Mar 2017, Matteo Bernardini wrote: > you shouldn't remove /usr/include/python-2.7 as it cointains file > belonging to Slackware's stock python, that you should leave installed > (as other packages in the system depends on it). Matteo, I realized this and did not remove it. > your issue is because you are trying to rebuild python3-PyQt5 before > having rebuilt also its dependency python3-sip: this last will install a > file /usr/include/python3.6/sip.h that is exactly what the python3-PyQt5 > build system is looking for. Mea culpa! I forgot the installation sequence and was working my way through my SBo/Python directories. Thanks and carpe weekend, Rich From willysr at slackbuilds.org Sat Mar 4 00:24:19 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sat, 4 Mar 2017 07:24:19 +0700 Subject: [Slackbuilds-users] rlog source MIA In-Reply-To: References: Message-ID: <6736e865-0c69-9a03-c154-d652dfc0d356@slackbuilds.org> > --2017-03-03 03:28:28-- http://rlog.googlecode.com/files/rlog-1.4.tar.gz > Resolving rlog.googlecode.com (rlog.googlecode.com)... > 2607:f8b0:4002:c06::52, 74.125.196.82 > Connecting to rlog.googlecode.com > (rlog.googlecode.com)|2607:f8b0:4002:c06::52|:80... connected. > HTTP request sent, awaiting response... 404 Not Found > 2017-03-03 03:28:28 ERROR 404: Not Found. > > There's a copy on sbosrcarch: > > http://slackware.uk/sbosrcarch//by-md5/c/2/c29f74e0f50d66b20312d049b683ff82/rlog-1.4.tar.gz > > There's also a copy here: > > http://distcache.freebsd.org/ports-distfiles/rlog-1.4.tar.gz > > It would probably be better to use the freebsd.org link in the .info > file, that way the sbosrcarch mirror has something to be a mirror of > (if that makes sense). Fixed in my branch thanks -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From willysr at slackbuilds.org Sat Mar 4 01:33:54 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sat, 4 Mar 2017 08:33:54 +0700 Subject: [Slackbuilds-users] Updates - 20170304.1 Message-ID: <5152d1a5-e3ed-ea19-9d55-38d4878dd292@slackbuilds.org> First batch of update in March bullet-pending branch is now merged into master and 14.2 since glPortal has a working tree upstream and it's now included in this batch. FlightGear also released their first update in 2017. Good news for tor users, tor 0.2.9.x is now marked LTS series and it will be supported until January 2020. There are some scripts being dropped: - bar -> lemonbar - electron-prebuild -> electron - thermal_daemon -> thermald As you may noticed, there are several commits involving changing GitHub URLs. The new URLs will results in consistent source filename downloaded both via wget or browser. If your scripts uses sources from GitHub, it's probably worth to check and update them accordingly. Sat Mar 4 00:29:15 UTC 2017 academic/arpack: new maintainer. academic/ent: Fix license URL. academic/g3data: new maintainer. academic/link-grammar: Update DEP. academic/mendeleydesktop: Updated for version 1.17.8. accessibility/espeak: Fix license URL. accessibility/unclutter-xfixes: Use long-format github URL. accessibility/wgetpaste: Fix license URL. accessibility/xdotool: Minor script cleanup. audio/calf: Use long-format github URL. audio/foo-yc20: Fix license URL. audio/lingot: Fix license URL. audio/lsmi: Fix license URL. audio/lv2file: Use long-format github URL. audio/mididings: Fix license URL. audio/mpd: Updated for version 0.20.5. audio/vocoder: Fix license URL. desktop/adwaita-qt: Updated for version 0.98. desktop/bar: Removed (replaced with lemonbar). desktop/bspwm: Updated for version 0.9.2 + new maintainer. desktop/cbatticon: Added (battery applet). desktop/gtk-nodoka-engine: Fixed download link. desktop/j4-dmenu-desktop: Use long-format github URL. desktop/lemonbar: Added (lightweight bar). desktop/sxhkd: Updated for version 0.5.7 + new maintainer. desktop/wmcliphist: Updated for version 2.1. desktop/xdgmenumaker: Use long-format github URL. desktop/xdo: Added (X Utility). desktop/xtitle: Updated for version 0.4.3 + new maintainer. development/Arduino: Update DEP. development/apache-log4cxx: Added (A logging framework for C++). development/atom: Updated for version 1.14.3. development/electron-prebuild: Removed (replaced with electron). development/extra-cmake-modules: Updated for version 5.31.0. development/git-lfs: Updated for version 2.0.0. development/inform: Fix license URL. development/kicad: Updated for version 4.0.5. development/mutagen: Updated for version 1.37. development/rstudio-desktop: Update DEP. development/urjtag: Added (Universal JTAG Library). games/FlightGear-data: Updated for version 2017.1.1. games/FlightGear: Updated for version 2017.1.1. games/bastet: Use long-format github URL. games/chocolate-doom: Allow building 3.0.0-beta1. games/commandergenius: Updated for version 1.9.8.7beta. games/doomretro: Updated for version 2.4.1. games/dungeon: Fix license URL. games/freeblocks: Use long-format github URL. games/freedoom: Updated for version 0.11.1. games/glPortal: Add patch to build with bullet 2.86. games/hatari: Fix license URL. games/javacpc: Updated for version 2.9.5d. games/mame: Use long-format github URL. games/opendune: Use long-format github URL. games/rott: Fix license URL. games/snes9x: Use long-format github URL. games/supermariowar: Use long-format github URL. games/vbam: Use long-format github URL. games/wordwarvi: Use long-format github URL, add .desktop. games/zoom: Fix license URL. games/zork: Fix license URL. games/ztools: Fix license URL. gis/Fiona: Updated for version 1.7.4. gis/GMT: Updated for version 5.3.2. gis/josm: Fixed SlackBuild. gis/qgis: Updated for version 2.18.4. gis/saga-gis: Updated for version 4.0.0. graphics/psftools: Fix license URL. graphics/qcad: Updated for version 3.16.5.0 haskell/haskell-cryptonite: Updated for version 0.22. libraries/SimGear: Updated for version 2017.1.1. libraries/bullet: Updated for version 2.86.1 libraries/dropbox-python: Updated for version 7.2.1. libraries/libicns: Fix license URL. libraries/libmseed: Updated for version 2.19.2. libraries/libosinfo: Fixed download link. libraries/lvtk: Use long-format github URL. libraries/polyclipping: Updated for version 6.4.2. libraries/rlog: Update DOWNLOAD url. libraries/zita-convolver: Fix license URL. misc/heyu: Fix license URL. misc/sutils: Updated for version 0.2 + new maintainer. multimedia/glyr: Use long-format github URL. multimedia/l-smash: Use long-format github URL. multimedia/mtpfs: Fix license URL. multimedia/obs-studio: Updated for version 18.0.0 multimedia/plexmediaserver: Updated for version 1.4.2.3400. multimedia/spotify: Updated source. multimedia/vitunes-mplayer: Fix license URL. multimedia/x265: Add also Main-10 and Main-12 color profiles. multimedia/xawtv: Added (TV and webcam tools). multimedia/youtube-viewer: Updated for version 3.2.5. network/authbind: Fix license URL. network/bmon: Use long-format github URL. network/emailrelay: Fix doinst.sh. network/irssi_otr: Use long-format github URL. network/opera: Updated for version 43.0.2442.1144. network/psad: Updated for version 2.4.4. network/sslscan: Updated for version 1.11.8_rbsec + new maintainer. network/suricata: Compress manpage. network/tor: Updated for version 0.2.9.10. network/verm: Fixed slack-desc. network/weechat: Updated for new maintainer and version 1.7. network/yturl: Use long-format github URL. office/LibreOffice: Full local help. New JDK options office/MasterPDFEditor: Updated for version 4.0.60. office/ProjectLibre: Updated for version 1.7.0. office/calibre: Updated for version 2.80.0 office/keepassxc: Updated for version 2.1.3. office/task: Update DOWNLOAD url. office/zathura-cb: Fix license URL. perl/perl-Capture-Tiny: Updated for version 0.46. perl/perl-Date-Manip: Updated for version 6.58. perl/perl-IPTables-ChainMgr: Added (manipulate iptables policies). perl/perl-IPTables-Parse: Added (parse iptables/ip6tables policies). perl/perl-Mojolicious: Updated for version 7.27. python/obspy: Updated for version 1.0.3. python/ply: Updated for version 3.10. python/python-augeas: Fixed download link. python/python-autobahn: Updated for version 0.17.2. python/python-requests: Updated for version 2.13.0. python/python-sh: Updated for version 1.12.10 python/python-suds: Fixed download link. python/python-txaio: Updated for version 2.6.1. python/python3-augeas: Fixed download link. ruby/rubygem-gosu: Updated for version 0.11.2. system/Iosevka: Updated for version 1.11.1. system/aha: Updated for version 0.4.10.5. system/cronie: Fixed download link. system/firetools: Updated for version 0.9.46. system/fsv: Use long-format github URL. system/isomd5sum: Use long-format github URL. system/iucode_tool: Updated for version 2.1.2. system/lbzip2: Use long-format github URL. system/lz4: Use long-format github URL. system/microcode_ctl: Fixed download link. system/nvidia-driver: Fix path. system/openrc-services: Updated for version 20170218. system/openrc: Fixed SlackBuild. system/osinfo-db-tools: Fixed download link. system/osinfo-db: Fixed download link. system/reptyr: Use long-format github URL. system/sysdig: Updated for version 0.15.0 system/targetcli-fb: Fixed download links. system/thermal_daemon: Removed (replaced with thermald). system/thermald: Added (Linux Thermal Daemon). system/udiskie: Updated for version 1.6.1. system/vagrant: Updated for version 1.9.2. system/vcp: Fix license URL. system/virt-viewer: Fixed download link. system/xarchiver: Updated for version 0.5.4.8. +--------------------------+ -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From jebrhansen+SBo at gmail.com Sat Mar 4 02:18:47 2017 From: jebrhansen+SBo at gmail.com (Jeremy Hansen) Date: Fri, 3 Mar 2017 21:18:47 -0500 Subject: [Slackbuilds-users] discord submission Message-ID: I am looking at submitting a SlackBuild for the chat program discord. Unfortunately, there is already a discord SlackBuild on SBo for a completely different program under the audio category (Brainwave entrainment for Linux - binaural beats, pulse beats, phase beats). Any suggestions on what name I should use in place? Or is it possible to rename the other SlackBuild since I feel that the majority of people looking for discord would want the chat program rather than the audio one? Jeremy -------------- next part -------------- An HTML attachment was scrubbed... URL: From n0nb at n0nb.us Sat Mar 4 02:45:34 2017 From: n0nb at n0nb.us (Nate Bargmann) Date: Fri, 3 Mar 2017 20:45:34 -0600 Subject: [Slackbuilds-users] Proper installation directory for TCL files? Message-ID: <20170304024534.GC9349@n0nb.us> Hi. I am working through an existing Slackbuild script that has not seen an update for about four years. I am the upstream maintainer for the library the build script is for so I am quite familiar with its build procedure. I have not yet contacted the listed maintainer about taking over the package as of yet. The library in question uses Swig to generate language bindings for Perl, Python, TCL, and Lua. So far I have the Perl and Python file placements in the package matching other files on my Slackware64 14.2 build system and by advice found in the autotools-template.SlackBuild file. Right now the TCL files are being placed in /usr/lib[64]/tcl by the library's build system, yet I find tcl8, tcl8.6, and tclx8.4 under /usr/lib[64]. Which of these directories should the generated files be placed into? Also, there is a generated pkgIndex.tcl file that would appear to conflict with any of the above full installation directories. As noted, I am working on a freshly installed Slackware64 14.2 system (VM, actually) that has been updated to the latest 14.2 packages. Thanks! - Nate -- "The optimist proclaims that we live in the best of all possible worlds. The pessimist fears this is true." Ham radio, Linux, bikes, and more: http://www.n0nb.us From andrew at stevens.is Sat Mar 4 02:52:38 2017 From: andrew at stevens.is (Andrew) Date: Fri, 03 Mar 2017 21:52:38 -0500 Subject: [Slackbuilds-users] discord submission In-Reply-To: References: Message-ID: As an idea have you considered calling it ``discordapp" as that is what the website domain is? Just my two cents. Really look forward to this, thanks Jeremy! Warm Regards, Andrew Stevens andrew at stevens.is -------- Original Message -------- Subject: [Slackbuilds-users] discord submission Local Time: March 4, 2017 3:18 PM UTC Time: March 4, 2017 2:18 AM From: jebrhansen+SBo at gmail.com To: slackbuilds-users I am looking at submitting a SlackBuild for the chat program discord. Unfortunately, there is already a discord SlackBuild on SBo for a completely different program under the audio category (Brainwave entrainment for Linux - binaural beats, pulse beats, phase beats). Any suggestions on what name I should use in place? Or is it possible to rename the other SlackBuild since I feel that the majority of people looking for discord would want the chat program rather than the audio one? Jeremy -------------- next part -------------- An HTML attachment was scrubbed... URL: From ryanpcmcquen at member.fsf.org Sat Mar 4 03:39:31 2017 From: ryanpcmcquen at member.fsf.org (Ryan P.C. McQuen) Date: Fri, 03 Mar 2017 19:39:31 -0800 Subject: [Slackbuilds-users] discord submission In-Reply-To: References: Message-ID: <15a9766a899.100ad469824310.7649678974579160116@member.fsf.org> >Any suggestions on what name I should use in place?discord_chat -Ryan [ryanpcmcquen.org] -------------- next part -------------- An HTML attachment was scrubbed... URL: From Hedgehog87 at hotmail.it Sat Mar 4 10:04:53 2017 From: Hedgehog87 at hotmail.it (Cristiano Urban) Date: Sat, 4 Mar 2017 10:04:53 +0000 Subject: [Slackbuilds-users] Atom doesn't build Message-ID: Hi, I've tried to update atom via sbotools (# sboupgrade atom) for three consecutive times and it fails to build. Here below I attach a .txt with some info. Does anyone have the same issue? Thanks to all in advance, Cristiano. ___________________________________________ Site/blog @ https://crish4cks.net/ Registered user #488179 @ http://linuxcounter.net Slackware, because it works! =) -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: atom.txt URL: From atelszewski at gmail.com Sat Mar 4 10:21:00 2017 From: atelszewski at gmail.com (Andrzej Telszewski) Date: Sat, 4 Mar 2017 11:21:00 +0100 Subject: [Slackbuilds-users] LibreOffice.SlackBuild: java selection code Message-ID: <5399a5ee-2c45-07d8-30de-81fb22e7865a@gmail.com> Hi, In the latest LibreOffice.SlackBuild you have: JAVA=${JAVA:-openjdk} if [ "$JAVA" = "openjdk" ]; then . /etc/profile.d/openjdk.sh use_java="--with-java=java" elif [ "$JAVA" = "openjdk8" ]; then . /etc/profile.d/openjdk8.sh use_java="--with-java=java" elif [ "$JAVA" = "jdk" ]; then . /etc/profile.d/jdk.sh use_java="--with-java=java" elif [ "$JAVA" = "no" ]; then use_java="--with-java=no" fi Isn't it the task of the shell to actually source the correct file? -- Best regards, Andrzej Telszewski From duncan_roe at acslink.net.au Sat Mar 4 10:48:36 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Sat, 4 Mar 2017 21:48:36 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> Message-ID: <20170304104836.GA2826@dimstar.local.net> On Fri, Mar 03, 2017 at 12:19:09PM +0000, David Spencer wrote: > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > > other than -j1). That's a snap diagnosis, I don't know what gdal even is. > > gdal already has 'make -j1' hard-coded, because it definitely fails > with multiple parallel make jobs. > > For some reason that I don't understand, on Duncan's system gcc is > being invoked through libtool. > The compilation explicitly has '-o > build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that > file does not exist at the time the next command 'g++ -pthread -shared > build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. > > So this really does look like a multiple parallel make jobs failure. > Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? > Are you on -current? > Do you have CC and/or CXX defined in the environment? > could you put your config.log on pastebin or somewhere similar? > Hi David, On 14.2 unpatched env has CC=gcc Paste is http://pastebin.com/v4gXJbDr Cheers ... Duncan. From andreas.guldstrand at gmail.com Sat Mar 4 10:49:07 2017 From: andreas.guldstrand at gmail.com (Andreas Guldstrand) Date: Sat, 4 Mar 2017 11:49:07 +0100 Subject: [Slackbuilds-users] Proper installation directory for TCL files? In-Reply-To: <20170304024534.GC9349@n0nb.us> References: <20170304024534.GC9349@n0nb.us> Message-ID: On 4 March 2017 at 03:45, Nate Bargmann wrote: > Right now the TCL files are being placed in /usr/lib[64]/tcl by the > library's build system, yet I find tcl8, tcl8.6, and tclx8.4 under > /usr/lib[64]. Which of these directories should the generated files be > placed into? Also, there is a generated pkgIndex.tcl file that would > appear to conflict with any of the above full installation directories. Hi, while I don't exactly know things about TCL, the pkgIndex.tcl file sounds similar to the perllocal.pod that is usually generated by perl installation tools - that file is simply removed in the template for perl slackbuilds, and I'd assume that would be the right thing to do for this file as well. But as I said, I don't really know much about TCL, so take my words with a couple grains of salt. Regards / Andreas From andreas.guldstrand at gmail.com Sat Mar 4 11:02:09 2017 From: andreas.guldstrand at gmail.com (Andreas Guldstrand) Date: Sat, 4 Mar 2017 12:02:09 +0100 Subject: [Slackbuilds-users] LibreOffice.SlackBuild: java selection code In-Reply-To: <5399a5ee-2c45-07d8-30de-81fb22e7865a@gmail.com> References: <5399a5ee-2c45-07d8-30de-81fb22e7865a@gmail.com> Message-ID: On 4 March 2017 at 11:21, Andrzej Telszewski wrote: > Isn't it the task of the shell to actually source the correct file? That would force the user to log out and log in again between installing openjdk/openjdk8/jdk and LibreOffice (or have them source it manually of course, but that wouldn't be "the shell" doing it anyway) From duncan_roe at acslink.net.au Sat Mar 4 11:06:09 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Sat, 4 Mar 2017 22:06:09 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170304104836.GA2826@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> Message-ID: <20170304110608.GB2826@dimstar.local.net> On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: > On Fri, Mar 03, 2017 at 12:19:09PM +0000, David Spencer wrote: > > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > > > other than -j1). That's a snap diagnosis, I don't know what gdal even is. > > > > gdal already has 'make -j1' hard-coded, because it definitely fails > > with multiple parallel make jobs. > > > > For some reason that I don't understand, on Duncan's system gcc is > > being invoked through libtool. > > The compilation explicitly has '-o > > build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that > > file does not exist at the time the next command 'g++ -pthread -shared > > build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. > > > > So this really does look like a multiple parallel make jobs failure. > > Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? > > Are you on -current? > > Do you have CC and/or CXX defined in the environment? > > could you put your config.log on pastebin or somewhere similar? > > > Hi David, > > On 14.2 unpatched > > env has CC=gcc > > Paste is http://pastebin.com/v4gXJbDr > And, I did not override -j1. This pocessor does have 2 cores though. Cheers ... Duncan. From duncan_roe at acslink.net.au Sat Mar 4 11:13:14 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Sat, 4 Mar 2017 22:13:14 +1100 Subject: [Slackbuilds-users] SOLVED: gdal SlackBuild fails In-Reply-To: <20170304110608.GB2826@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> Message-ID: <20170304111314.GC2826@dimstar.local.net> On Sat, Mar 04, 2017 at 10:06:09PM +1100, Duncan Roe wrote: > On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: > > On Fri, Mar 03, 2017 at 12:19:09PM +0000, David Spencer wrote: > > > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > > > > other than -j1). That's a snap diagnosis, I don't know what gdal even is. > > > > > > gdal already has 'make -j1' hard-coded, because it definitely fails > > > with multiple parallel make jobs. > > > > > > For some reason that I don't understand, on Duncan's system gcc is > > > being invoked through libtool. > > > The compilation explicitly has '-o > > > build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that > > > file does not exist at the time the next command 'g++ -pthread -shared > > > build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. > > > > > > So this really does look like a multiple parallel make jobs failure. > > > Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? > > > Are you on -current? > > > Do you have CC and/or CXX defined in the environment? > > > could you put your config.log on pastebin or somewhere similar? > > > > > Hi David, > > > > On 14.2 unpatched > > > > env has CC=gcc > > > > Paste is http://pastebin.com/v4gXJbDr > > > And, I did not override -j1. This pocessor does have 2 cores though. > I removed CC=gcc from the environment and the SB ran to completion. It's been in my .bashrc for years, will take it out. Cheers ... Duncan. From matteo.bernardini at gmail.com Sat Mar 4 11:15:29 2017 From: matteo.bernardini at gmail.com (Matteo Bernardini) Date: Sat, 4 Mar 2017 12:15:29 +0100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170304110608.GB2826@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> Message-ID: 2017-03-04 12:06 GMT+01:00 Duncan Roe : > On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: >> On Fri, Mar 03, 2017 at 12:19:09PM +0000, David Spencer wrote: >> > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything >> > > other than -j1). That's a snap diagnosis, I don't know what gdal even is. >> > >> > gdal already has 'make -j1' hard-coded, because it definitely fails >> > with multiple parallel make jobs. >> > >> > For some reason that I don't understand, on Duncan's system gcc is >> > being invoked through libtool. >> > The compilation explicitly has '-o >> > build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that >> > file does not exist at the time the next command 'g++ -pthread -shared >> > build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. >> > >> > So this really does look like a multiple parallel make jobs failure. >> > Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? >> > Are you on -current? >> > Do you have CC and/or CXX defined in the environment? >> > could you put your config.log on pastebin or somewhere similar? >> > >> Hi David, >> >> On 14.2 unpatched >> >> env has CC=gcc >> >> Paste is http://pastebin.com/v4gXJbDr >> > And, I did not override -j1. This pocessor does have 2 cores though. Duncan, sorry, before anyone else asks: the paste above looks done from your unprivilege used environment. it could be there are no differences but maybe it's better if you post the one you got from running the SlackBuild (you can take it directly from the /tm/SBo/gdal-* directory) from the root environment (in which you get by running "su -l" from your user shell). Matteo From baildon.research at googlemail.com Sat Mar 4 11:50:30 2017 From: baildon.research at googlemail.com (David Spencer) Date: Sat, 4 Mar 2017 11:50:30 +0000 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170304110608.GB2826@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> Message-ID: >> env has CC=gcc aaaAAAAhhhhh!!! Add this to the ./configure arguments --without-libtool \ and/or try it again without CC in the environment. (There have been reports on the gdal mailing list where setting CC/CXX in the environment causes gdal's configure to use libtool, leading to python-related build failures. Both steps in that chain of causality are bizarre and unwelcome; let's hope the workround works.) Thanks -D. From baildon.research at googlemail.com Sat Mar 4 11:58:33 2017 From: baildon.research at googlemail.com (David Spencer) Date: Sat, 4 Mar 2017 11:58:33 +0000 Subject: [Slackbuilds-users] SOLVED: gdal SlackBuild fails In-Reply-To: <20170304111314.GC2826@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> <20170304111314.GC2826@dimstar.local.net> Message-ID: > I removed CC=gcc from the environment and the SB ran to completion. Sorry for the noise everybody: I missed your solution before I replied just now :( Anyway I hope the additional info was useful. As I implied previously, gdal's configure is quite fragile in multiple ways. I'll add --without-libtool to the SlackBuild, because explicitly setting a default in an environment variable shouldn't break stuff. Thanks Duncan, we got there in the end :) -D. From n0nb at n0nb.us Sat Mar 4 12:36:03 2017 From: n0nb at n0nb.us (Nate Bargmann) Date: Sat, 4 Mar 2017 06:36:03 -0600 Subject: [Slackbuilds-users] Proper installation directory for TCL files? In-Reply-To: References: <20170304024534.GC9349@n0nb.us> Message-ID: <20170304123603.GD9349@n0nb.us> * On 2017 04 Mar 04:50 -0600, Andreas Guldstrand wrote: > On 4 March 2017 at 03:45, Nate Bargmann wrote: > > Right now the TCL files are being placed in /usr/lib[64]/tcl by the > > library's build system, yet I find tcl8, tcl8.6, and tclx8.4 under > > /usr/lib[64]. Which of these directories should the generated files be > > placed into? Also, there is a generated pkgIndex.tcl file that would > > appear to conflict with any of the above full installation directories. > > Hi, while I don't exactly know things about TCL, the pkgIndex.tcl file > sounds similar to the perllocal.pod that is usually generated by perl > installation tools - that file is simply removed in the template for > perl slackbuilds, and I'd assume that would be the right thing to do > for this file as well. But as I said, I don't really know much about > TCL, so take my words with a couple grains of salt. Hi Andreas. My post was incomplete. Here is the tree for the TCL files as installed by the package's Makefile: $ tree ./usr/lib64/tcl/ ./usr/lib64/tcl/ `-- Library |-- librarytcl-99.99.so |-- librarytcl.la `-- pkgIndex.tcl As can seen, the installed files will be in their own directory, so any existing pkgIndex.tcl would not be overwritten. My question now is even a bit more simple, and perhaps by looking things over again, I've found an acceptable answer. I see the sqlite package is installed under /usr/lib[64]/tcl8.6/sqlite3 which contains libtclsqlite3.so and pkgIndex.tcl. It appears as though pkgIndex.tcl is necessary to map the .so file for the tcl interpreter, so I shall proceed by having the Slackbuild script rename 'tcl' to 'tcl8.6', remove librarytcl.la, and continue testing. Thanks. - Nate -- "The optimist proclaims that we live in the best of all possible worlds. The pessimist fears this is true." Ham radio, Linux, bikes, and more: http://www.n0nb.us From andreas.guldstrand at gmail.com Sat Mar 4 12:41:01 2017 From: andreas.guldstrand at gmail.com (Andreas Guldstrand) Date: Sat, 4 Mar 2017 13:41:01 +0100 Subject: [Slackbuilds-users] Atom doesn't build In-Reply-To: References: Message-ID: On 4 March 2017 at 11:04, Cristiano Urban wrote: > I've tried to update atom via sbotools (# sboupgrade atom) for three > consecutive times and it fails to build. > > Here below I attach a .txt with some info. > > Does anyone have the same issue? I had the same issue when trying to install it too (not upgrade because I didn't have it before). But I noticed that the prerequisite, nodejs, had something it installed into /etc/profile.d, so I assumed I needed that loaded before installing atom, so I logged out, and logged in again using a proper root shell and tried again, and then it worked. Regards / Andreas From nikos.giotis at gmail.com Sat Mar 4 13:34:21 2017 From: nikos.giotis at gmail.com (Nikos Giotis) Date: Sat, 4 Mar 2017 15:34:21 +0200 Subject: [Slackbuilds-users] Submitting packages with cyclic dependencies via website Message-ID: <51360219-7c50-bec6-1711-ba1cbded6dde@gmail.com> Hi, how are we supposed to submit new packages with cyclic dependencies via the website? This time, I am trying to submit two python packages fixtures and testtools. I tried to be imaginative and removed testtools from REQUIRES, then submitted testtools and then tried to submit again fixtures with the testtools included in the REQUIRES, but "a submission by the name 'fixtures' is already pending approval." :) How should I proceed so that we can have these two packages added properly? Also, Nikos From nikos.giotis at gmail.com Sat Mar 4 13:41:03 2017 From: nikos.giotis at gmail.com (Nikos Giotis) Date: Sat, 4 Mar 2017 15:41:03 +0200 Subject: [Slackbuilds-users] Question about git workflow Message-ID: Hi, is there a guide or any sort of documentation about how to submit patches to slackbuilds.org via git? Nikos From duncan_roe at acslink.net.au Sat Mar 4 14:02:27 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Sun, 5 Mar 2017 01:02:27 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> Message-ID: <20170304140227.GA2739@dimstar.local.net> On Sat, Mar 04, 2017 at 12:15:29PM +0100, Matteo Bernardini wrote: > 2017-03-04 12:06 GMT+01:00 Duncan Roe : > > On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: > >> On Fri, Mar 03, 2017 at 12:19:09PM +0000, David Spencer wrote: > >> > > Looks like the kind of issue caused by using MAKEFLAGS=-j3 (or anything > >> > > other than -j1). That's a snap diagnosis, I don't know what gdal even is. > >> > > >> > gdal already has 'make -j1' hard-coded, because it definitely fails > >> > with multiple parallel make jobs. > >> > > >> > For some reason that I don't understand, on Duncan's system gcc is > >> > being invoked through libtool. > >> > The compilation explicitly has '-o > >> > build/temp.linux-x86_64-2.7/extensions/.libs/gdal_wrap.o', but that > >> > file does not exist at the time the next command 'g++ -pthread -shared > >> > build/temp.linux-x86_64-2.7/extensions/gdal_wrap.o' is invoked. > >> > > >> > So this really does look like a multiple parallel make jobs failure. > >> > Duncan, for avoidance of doubt you haven't removed -j1 or overridden it somehow? > >> > Are you on -current? > >> > Do you have CC and/or CXX defined in the environment? > >> > could you put your config.log on pastebin or somewhere similar? > >> > > >> Hi David, > >> > >> On 14.2 unpatched > >> > >> env has CC=gcc > >> > >> Paste is http://pastebin.com/v4gXJbDr > >> > > And, I did not override -j1. This pocessor does have 2 cores though. > > Duncan, sorry, before anyone else asks: the paste above looks done > from your unprivilege used environment. > it could be there are no differences but maybe it's better if you post > the one you got from running the SlackBuild (you can take it directly > from the /tm/SBo/gdal-* directory) from the root environment (in which > you get by running "su -l" from your user shell). > > Matteo Hi Matteo, I ran the build from a logged-in-as-root environment, as evidenced by: > 22:00:11# echo $0 > -bash I just do su but then immediately start a fresh xterm (with a pink background, so I know it's root); then exit from the su. What looks unprivileged in the paste? Cheers ... Duncan. From matteo.bernardini at gmail.com Sat Mar 4 14:11:36 2017 From: matteo.bernardini at gmail.com (Matteo Bernardini) Date: Sat, 4 Mar 2017 15:11:36 +0100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: <20170304140227.GA2739@dimstar.local.net> References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> <20170304140227.GA2739@dimstar.local.net> Message-ID: 2017-03-04 15:02 GMT+01:00 Duncan Roe : > On Sat, Mar 04, 2017 at 12:15:29PM +0100, Matteo Bernardini wrote: >> 2017-03-04 12:06 GMT+01:00 Duncan Roe : >> > On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: >> >> Paste is http://pastebin.com/v4gXJbDr >> >> >> > And, I did not override -j1. This pocessor does have 2 cores though. >> >> Duncan, sorry, before anyone else asks: the paste above looks done >> from your unprivilege used environment. >> it could be there are no differences but maybe it's better if you post >> the one you got from running the SlackBuild (you can take it directly >> from the /tm/SBo/gdal-* directory) from the root environment (in which >> you get by running "su -l" from your user shell). >> >> Matteo > > Hi Matteo, > > I ran the build from a logged-in-as-root environment, as evidenced by: > >> 22:00:11# echo $0 >> -bash > > I just do su but then immediately start a fresh xterm (with a pink background, > so I know it's root); then exit from the su. > > What looks unprivileged in the paste? in the paste you have these lines PATH: /home/dunc/bin64 PATH: /home/dunc/bin PATH: /home/dunc/command_line_tools/bin PATH: /usr/local/sbin PATH: /usr/sbin PATH: /sbin PATH: /usr/local/bin PATH: /usr/bin PATH: /bin PATH: /usr/games PATH: /usr/lib/java/bin PATH: /usr/lib64/kde4/libexec PATH: /usr/lib64/qt/bin PATH: /usr/share/texmf/bin these mean that you have executed just "su" and not "su -" (or "su -l") so your root environment, not having being launched as a true login shell, was polluted by user variables (see point 3 of the howto https://slackbuilds.org/howto/). you shouldn't run SBo's build scripts executing just "su". Matteo From duncan_roe at acslink.net.au Sat Mar 4 15:51:55 2017 From: duncan_roe at acslink.net.au (Duncan Roe) Date: Sun, 5 Mar 2017 02:51:55 +1100 Subject: [Slackbuilds-users] gdal SlackBuild fails In-Reply-To: References: <20170301220658.GA24715@dimstar.local.net> <20170304104836.GA2826@dimstar.local.net> <20170304110608.GB2826@dimstar.local.net> <20170304140227.GA2739@dimstar.local.net> Message-ID: <20170304155155.GB2739@dimstar.local.net> On Sat, Mar 04, 2017 at 03:11:36PM +0100, Matteo Bernardini wrote: > 2017-03-04 15:02 GMT+01:00 Duncan Roe : > > On Sat, Mar 04, 2017 at 12:15:29PM +0100, Matteo Bernardini wrote: > >> 2017-03-04 12:06 GMT+01:00 Duncan Roe : > >> > On Sat, Mar 04, 2017 at 09:48:36PM +1100, Duncan Roe wrote: > >> >> Paste is http://pastebin.com/v4gXJbDr > >> >> > >> > And, I did not override -j1. This pocessor does have 2 cores though. > >> > >> Duncan, sorry, before anyone else asks: the paste above looks done > >> from your unprivilege used environment. > >> it could be there are no differences but maybe it's better if you post > >> the one you got from running the SlackBuild (you can take it directly > >> from the /tm/SBo/gdal-* directory) from the root environment (in which > >> you get by running "su -l" from your user shell). > >> > >> Matteo > > > > Hi Matteo, > > > > I ran the build from a logged-in-as-root environment, as evidenced by: > > > >> 22:00:11# echo $0 > >> -bash > > > > I just do su but then immediately start a fresh xterm (with a pink background, > > so I know it's root); then exit from the su. > > > > What looks unprivileged in the paste? > > in the paste you have these lines > > PATH: /home/dunc/bin64 > PATH: /home/dunc/bin > PATH: /home/dunc/command_line_tools/bin > PATH: /usr/local/sbin > PATH: /usr/sbin > PATH: /sbin > PATH: /usr/local/bin > PATH: /usr/bin > PATH: /bin > PATH: /usr/games > PATH: /usr/lib/java/bin > PATH: /usr/lib64/kde4/libexec > PATH: /usr/lib64/qt/bin > PATH: /usr/share/texmf/bin > > these mean that you have executed just "su" and not "su -" (or "su > -l") so your root environment, not having being launched as a true > login shell, was polluted by user variables (see point 3 of the howto > https://slackbuilds.org/howto/). > you shouldn't run SBo's build scripts executing just "su". > > Matteo No, they don't mean that at all: > 00:47:13# ls -l ~/.bashrc > lrwxrwxrwx 1 root root 18 May 5 2014 /root/.bashrc -> /home/dunc/.bashrc Cheers ... Duncan. From atelszewski at gmail.com Sat Mar 4 17:09:43 2017 From: atelszewski at gmail.com (Andrzej Telszewski) Date: Sat, 4 Mar 2017 18:09:43 +0100 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: Message-ID: On 04/03/17 14:41, Nikos Giotis wrote: > is there a guide or any sort of documentation about how to submit > patches to slackbuilds.org via git? I think it's not supported. -- Best regards, Andrzej Telszewski From Hedgehog87 at hotmail.it Sat Mar 4 17:20:14 2017 From: Hedgehog87 at hotmail.it (Cristiano Urban) Date: Sat, 4 Mar 2017 17:20:14 +0000 Subject: [Slackbuilds-users] Atom doesn't build In-Reply-To: References: , Message-ID: I've just turned on my laptop and tried to upgrade atom using su - instead of su. Now it works. Andreas, thank you for the tip! ___________________________________________ Site/blog @ https://crish4cks.net/ Registered user #488179 @ http://linuxcounter.net Slackware, because it works! =) ________________________________ Da: SlackBuilds-users per conto di Andreas Guldstrand Inviato: sabato 4 marzo 2017 13:41:01 A: SlackBuilds.org Users List Oggetto: Re: [Slackbuilds-users] Atom doesn't build On 4 March 2017 at 11:04, Cristiano Urban wrote: > I've tried to update atom via sbotools (# sboupgrade atom) for three > consecutive times and it fails to build. > > Here below I attach a .txt with some info. > > Does anyone have the same issue? I had the same issue when trying to install it too (not upgrade because I didn't have it before). But I noticed that the prerequisite, nodejs, had something it installed into /etc/profile.d, so I assumed I needed that loaded before installing atom, so I logged out, and logged in again using a proper root shell and tried again, and then it worked. Regards / Andreas _______________________________________________ SlackBuilds-users mailing list SlackBuilds-users at slackbuilds.org http://lists.slackbuilds.org/mailman/listinfo/slackbuilds-users Archives - https://lists.slackbuilds.org/pipermail/slackbuilds-users/ FAQ - https://slackbuilds.org/faq/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From willysr at slackbuilds.org Sat Mar 4 18:38:38 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 01:38:38 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: Message-ID: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> >> is there a guide or any sort of documentation about how to submit >> patches to slackbuilds.org via git? > > I think it's not supported. Not officially supported, but i personally have accepted PR from public github repository for some people for their own scripts (not updating others nor new submission). -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From willysr at slackbuilds.org Sun Mar 5 02:33:32 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 09:33:32 +0700 Subject: [Slackbuilds-users] Submitting packages with cyclic dependencies via website In-Reply-To: <51360219-7c50-bec6-1711-ba1cbded6dde@gmail.com> References: <51360219-7c50-bec6-1711-ba1cbded6dde@gmail.com> Message-ID: <582360df-e8ae-eb4e-44f7-3fc0eb49ac67@slackbuilds.org> > how are we supposed to submit new packages with cyclic dependencies > via the website? > > This time, I am trying to submit two python packages fixtures and testtools. > I tried to be imaginative and removed testtools from REQUIRES, then > submitted testtools and then tried to submit again fixtures with the > testtools included in the REQUIRES, but > "a submission by the name 'fixtures' is already pending approval." :) > > How should I proceed so that we can have these two packages added properly? Unless you want to resubmit everything, i have removed fixture from testtools in my branch and after public update, i will remove fixture from pending (it can't be deleted now since it's listed as dep for testtools). cyclic dependency should be listed in the README. I think opencv and ffmpeg also had the same problem -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From nikos.giotis at gmail.com Sun Mar 5 03:07:08 2017 From: nikos.giotis at gmail.com (Nikos Giotis) Date: Sun, 5 Mar 2017 05:07:08 +0200 Subject: [Slackbuilds-users] Submitting packages with cyclic dependencies via website In-Reply-To: <582360df-e8ae-eb4e-44f7-3fc0eb49ac67@slackbuilds.org> References: <51360219-7c50-bec6-1711-ba1cbded6dde@gmail.com> <582360df-e8ae-eb4e-44f7-3fc0eb49ac67@slackbuilds.org> Message-ID: On 03/05/2017 04:33 AM, Willy Sudiarto Raharjo wrote: >> how are we supposed to submit new packages with cyclic dependencies >> via the website? >> >> This time, I am trying to submit two python packages fixtures and testtools. >> I tried to be imaginative and removed testtools from REQUIRES, then >> submitted testtools and then tried to submit again fixtures with the >> testtools included in the REQUIRES, but >> "a submission by the name 'fixtures' is already pending approval." :) >> >> How should I proceed so that we can have these two packages added properly? > > Unless you want to resubmit everything, i have removed fixture from > testtools in my branch and after public update, i will remove fixture > from pending (it can't be deleted now since it's listed as dep for > testtools). > > cyclic dependency should be listed in the README. > I think opencv and ffmpeg also had the same problem I will wait for the next public update, then resubmit fixture and testtools with the appropriate REQUIRES string and with fixed README files. -- Nikos From willysr at slackbuilds.org Sun Mar 5 04:50:10 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 11:50:10 +0700 Subject: [Slackbuilds-users] Updates - 20170305.1 Message-ID: Hi folks, Sorry for the unscheduled updates but the xen update seems quite important. We also had so many new python submission for Flask applications. Happy Weekend :) Sun Mar 5 04:30:32 UTC 2017 academic/lapack95: New maintainer. desktop/gtk-theme-Greybird: Updated for version 3.22.12. games/FlightGear-data: Updated for version 2017.1.2. games/FlightGear: Updated for version 2017.1.2. games/chessx: Updated for version 1.4.4. libraries/SimGear: Updated for version 2017.1.2. libraries/html5lib: New maintainer libraries/libseccomp: Updated for version 2.3.2. misc/grc: Updated for version 1.10.1. misc/subsurface: Updated for version 4.6.3 network/krb5: Updated for version 1.15.1. network/sslscan: Fix DOWNLOAD url. network/weechat: Fix path on x86_64 perl/perl-Math-BigInt: Updated for version 1.999810. perl/perl-Sidef: Updated for version 2.36. python/Flask-Babel: Added (i18n/l10n support to Flask). python/Flask-Gravatar: Added (Flask extension for Gravatar service). python/Flask-HTMLmin: Added (Minimize rendered templates html). python/Flask-Mail: Added (Flask extension for sending email). python/Flask-Principal: Added (Identity management for flask). python/Flask-Security: Added (Simple security for Flask apps). python/Flask-WTF: Added (Simple integration of Flask and WTForms). python/PyStemmer: Added (Snowball stemming algorithms). python/WTForms: Added (web forms validation). python/alabaster: Added (Sphinx theme). python/babel: Update DEP. python/extras: Added (Python bindings for augeas). python/htmlmin: Added (An HTML Minifier). python/imagesize: Added (python library for images). python/linecache2: Added (Backports of the linecache python module). python/pyrsistent: Added (data structures). python/python-augeas: Update DOWNLOAD url. python/python-mimeparse: Added (module for parsing mime-type). python/snowballstemmer: Added (stemmer algorithms for python). python/speaklater: Added (Python module). python/testtools: Added (Python unit testing framework extension). python/traceback2: Addded (Backports of the traceback module). system/letsencrypt: Updated for version 0.12.0. system/vinterm: Fix icon/desktop installation. system/xen: XSA 207-210 update. +--------------------------+ -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From lists at osh.id.au Sun Mar 5 07:10:42 2017 From: lists at osh.id.au (David O'Shaughnessy) Date: Sun, 5 Mar 2017 17:10:42 +1000 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> Message-ID: <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> On 03/05/2017 04:38 AM, Willy Sudiarto Raharjo wrote: >>> is there a guide or any sort of documentation about how to submit >>> patches to slackbuilds.org via git? >> >> I think it's not supported. > > Not officially supported, but i personally have accepted PR from public > github repository for some people for their own scripts (not updating > others nor new submission). I've always wondered about this. Wouldn't it be easier for the maintenance team to deal with _some_ form of version control, rather than updated archives delivered through slackbuilds.org? -- Dave From yalhcru at gmail.com Sun Mar 5 07:20:56 2017 From: yalhcru at gmail.com (B Watson) Date: Sun, 5 Mar 2017 02:20:56 -0500 Subject: [Slackbuilds-users] gnulib is evil Message-ID: Heads up to maintainers. If your build uses gnulib, there's a good chance it's modifying the timestamp of /usr/lib when it runs. I don't mean "if your build has gnulib as a dependency", because the normal way to use gnulib is to bundle it with the source code. The top-level directory will generally have a gnulib/ directory (along with src/ and whatever else is in the tarball). What it's doing is checking $PREFIX/lib (along with some other directories) to see whether it's able to create a file whose name is longer than 14 characters. man-db was doing this, and I had to patch the configure script to stop it. This appears to be severe brain damage to me: - The /usr/lib test will always fail if you're compiling as a normal user - Nobody still runs a 1989 version of SCO or SVr3 or whatever - ./configure and make should never write to system directories (only 'make install' should do that) Probably worth looking at how I patched it, if you want to do something similar for your own build. Patch is attached to this email (also in my user/urchlay/updates git branch). -------------- next part -------------- A non-text attachment was scrubbed... Name: dont_write_in_usr.diff.gz Type: application/x-gzip Size: 853 bytes Desc: not available URL: From willysr at slackbuilds.org Sun Mar 5 08:10:17 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 15:10:17 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> Message-ID: <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> > I've always wondered about this. Wouldn't it be easier for the > maintenance team to deal with _some_ form of version control, rather > than updated archives delivered through slackbuilds.org? We do have and we used git see the web-based interface here: https://slackbuilds.org/cgit/slackbuilds -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From jebrhansen+SBo at gmail.com Sun Mar 5 14:38:39 2017 From: jebrhansen+SBo at gmail.com (Jeremy Hansen) Date: Sun, 05 Mar 2017 14:38:39 +0000 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> Message-ID: On Sun, Mar 5, 2017, 3:10 AM Willy Sudiarto Raharjo wrote: > > I've always wondered about this. Wouldn't it be easier for the > > maintenance team to deal with _some_ form of version control, rather > > than updated archives delivered through slackbuilds.org? > > We do have and we used git > > see the web-based interface here: > https://slackbuilds.org/cgit/slackbuilds Do you prefer one method over another? Jeremy > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From atelszewski at gmail.com Sun Mar 5 16:20:38 2017 From: atelszewski at gmail.com (Andrzej Telszewski) Date: Sun, 5 Mar 2017 17:20:38 +0100 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> Message-ID: On 05/03/17 15:38, Jeremy Hansen wrote: > > > On Sun, Mar 5, 2017, 3:10 AM Willy Sudiarto Raharjo > > wrote: > > > I've always wondered about this. Wouldn't it be easier for the > > maintenance team to deal with _some_ form of version control, rather > > than updated archives delivered through slackbuilds.org > ? > > We do have and we used git > > see the web-based interface here: > https://slackbuilds.org/cgit/slackbuilds > > > Do you prefer one method over another? > > Jeremy > For the moment SBo prefers tarballs, well, for the moment that's the only option. There is no git for ordinary users and maintainers as of now. -- Best regards, Andrzej Telszewski From willysr at slackbuilds.org Sun Mar 5 16:22:20 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 23:22:20 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> Message-ID: <2e8f2a80-c6cf-c3ff-e108-d2cf3057109f@slackbuilds.org> > Do you prefer one method over another? We only support git. cgit is just web interface so can see other's work -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From willysr at slackbuilds.org Sun Mar 5 16:24:34 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Sun, 5 Mar 2017 23:24:34 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> Message-ID: <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> > For the moment SBo prefers tarballs, well, for the moment that's the > only option. > > There is no git for ordinary users and maintainers as of now. That's not true we do allow "some" maintainers to have their own branches in our GIT repository. See https://slackbuilds.org/cgit/slackbuilds/ -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From harald.achitz at gmail.com Sun Mar 5 17:42:54 2017 From: harald.achitz at gmail.com (Harald Achitz) Date: Sun, 5 Mar 2017 18:42:54 +0100 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> Message-ID: I am pretty sure that this, more cvs/svn orientated branching strategy is not optimal. There is no need for users to have branches, they should have forks. There they can branch, and create for these branches pull requests. Of course you can create also branches per user, but this only pollutes the repo with unnecessary stuff. You could take pull requests from every trust able person fore each package, the maintainer models is total obsolete that only scales bad, distributes work bad, and is a bottleneck by design. Honestly, if, for example, Niki would make a pull request for some package update because he needs it in MELD, I would pull it in if he is the maintainer or not. Maybe with a 'next release branch' for some time to give early adopter chance to do additional testing, which his more a useful branching strategy than per 'trusted user'. And this is more the open source, or general how it works, model of today, but of course just IMHO and I do not want to convince anybody of anything, as usual this are just my 2 cents ;-) Regards Harald 2017-03-05 17:24 GMT+01:00 Willy Sudiarto Raharjo : > > For the moment SBo prefers tarballs, well, for the moment that's the > > only option. > > > > There is no git for ordinary users and maintainers as of now. > > That's not true > we do allow "some" maintainers to have their own branches in our GIT > repository. See https://slackbuilds.org/cgit/slackbuilds/ > > > > -- > Willy Sudiarto Raharjo > > > _______________________________________________ > SlackBuilds-users mailing list > SlackBuilds-users at slackbuilds.org > http://lists.slackbuilds.org/mailman/listinfo/slackbuilds-users > Archives - https://lists.slackbuilds.org/pipermail/slackbuilds-users/ > FAQ - https://slackbuilds.org/faq/ > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From baildon.research at googlemail.com Sun Mar 5 19:37:00 2017 From: baildon.research at googlemail.com (David Spencer) Date: Sun, 5 Mar 2017 19:37:00 +0000 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> Message-ID: > There is no need for users to have branches, they should have forks. > There they can branch, and create for these branches pull requests. We do have the vague intention of using Gitlab like that, but it is an ongoing research project, because first we need an answer to the question below: > You could take pull requests from every trust able person fore each package, > the maintainer models is total obsolete that only scales bad, distributes > work bad, and is a bottleneck by design. Who, or what, is going to do the required testing? Cheers -D. "Never ask a question, unless you already know the answer" ;-) From edps.mundognu at gmail.com Sun Mar 5 20:11:15 2017 From: edps.mundognu at gmail.com (Edinaldo) Date: Sun, 05 Mar 2017 17:11:15 -0300 Subject: [Slackbuilds-users] gecko-mediaplayer Message-ID: Here's i found a 404 error when trying downloading gecko-mediaplayer. So... please change download URL to "sboarch": http://slackware.uk/sbosrcarch/by-md5/d/1/d13303664fcbb7f3b1273be4b1aa1b1e/gecko-mediaplayer-1.0.8.tar.gz or to: https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/gecko-mediaplayer/gecko-mediaplayer-1.0.8.tar.gz Cheers, Edinaldo. From kingbeowulf at gmail.com Sun Mar 5 20:18:40 2017 From: kingbeowulf at gmail.com (Edward Koenig) Date: Sun, 5 Mar 2017 12:18:40 -0800 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> Message-ID: <2CB9E608-3F67-47B7-8945-922D0CA934FA@gmail.com> >> You could take pull requests from every trust able person fore each package, >> the maintainer models is total obsolete that only scales bad, distributes >> work bad, and is a bottleneck by design. > > Who, or what, is going to do the required testing? Well, I agree with David. I think the maintainer model serves SBo quite well as opposed to a more free-for-all F/OSS approach. I expect that most maintainers USE most of the scripts (I do). I also gladly accept patches and suggestions to be included in my next update. ( I might be a bit slow....) This allows deeper testing in many cases. Automated tools can only help to a certain extent. As we all know, source documentation is not a favorite task for many projects (and programmers!). To go with a "typical" repo development methodology could seriously impact quality. Does SBo really want to set up a build bot? Is a script tested as "ok" if it runs to completion? Compiles? Often, bugs or missing dependencies aren't evident until you actually install and run the created package. Overall, I think the SBo model creates a somewhat better repository quality than those "other guys". -Ed From willysr at slackbuilds.org Sun Mar 5 23:45:24 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Mon, 6 Mar 2017 06:45:24 +0700 Subject: [Slackbuilds-users] gecko-mediaplayer In-Reply-To: References: Message-ID: > Here's i found a 404 error when trying downloading gecko-mediaplayer. > > So... please change download URL to "sboarch": > > http://slackware.uk/sbosrcarch/by-md5/d/1/d13303664fcbb7f3b1273be4b1aa1b1e/gecko-mediaplayer-1.0.8.tar.gz > > > or to: > > https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/gecko-mediaplayer/gecko-mediaplayer-1.0.8.tar.gz Fixed thanks -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From willysr at slackbuilds.org Mon Mar 6 00:05:52 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Mon, 6 Mar 2017 07:05:52 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> Message-ID: > I am pretty sure that this, more cvs/svn orientated branching strategy is > not optimal. > > There is no need for users to have branches, they should have forks. > There they can branch, and create for these branches pull requests. > > Of course you can create also branches per user, but this only pollutes the > repo with unnecessary stuff. this branch strategy works like forks, but instead you are given a privileges to write your changes to our GIT repository directly instead of doing it on your own public space. Also, you no longer need to send a pull requests, since by using cgit, we can see other's changes and notify them if something needs to be fixed before next public update. Once a week, i simply run a script to merge them all to master and then remove all those branches and maintainers can rebase their branches to the latest master again. It's all very efficient in GIT, but maybe not in CVS/SVN. > You could take pull requests from every trust able person fore each > package, the maintainer models is total obsolete that only scales bad, > distributes work bad, and is a bottleneck by design. That's what we do :) we gave "some" maintainers privileges to push it directly to our GIT repository. We will be adding more maintainers from time to time. There's no written requirements, but these two should be a good start: - active maintainer - understand GIT basics - having public GIT repository is an added value :) Please note that maintainers in SBo differs from maintainers in Linux Kernel project where they took all changes in their subsystem from other people as well. In SBo, maintainers are responsible for their own scripts. That's why they write it on their own branch. > Honestly, if, for example, Niki would make a pull request for some package > update because he needs it in MELD, I would pull it in if he is the > maintainer or not. > Maybe with a 'next release branch' for some time to give early adopter > chance to do additional testing, which his more a useful branching strategy > than per 'trusted user'. since all the changes can be seen in the cgit, users who are willing to try can just clone the SBo repository and switch to anyone's branch and test whether it works or not. -- Willy Sudiarto Raharjo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From klaatu at SDF.ORG Sun Mar 5 19:49:30 2017 From: klaatu at SDF.ORG (klaatu la terible) Date: Sun, 5 Mar 2017 19:49:30 +0000 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <1c77f136-3ac2-6d0e-62d2-696470406a02@slackbuilds.org> Message-ID: <20170305194930.v45xub6ojs4vcgrw@SDF.ORG> On Sun, Mar 05, 2017 at 11:24:34PM +0700, Willy Sudiarto Raharjo wrote: > > For the moment SBo prefers tarballs, well, for the moment that's the > > only option. > > > > There is no git for ordinary users and maintainers as of now. > > That's not true > we do allow "some" maintainers to have their own branches in our GIT > repository. See https://slackbuilds.org/cgit/slackbuilds/ > > -- > Willy Sudiarto Raharjo > I can't help but ask, what's the deciding factor on who gets a branch? I'd love to be able to update my builds with git. From lists at osh.id.au Mon Mar 6 03:28:17 2017 From: lists at osh.id.au (David O'Shaughnessy) Date: Mon, 6 Mar 2017 13:28:17 +1000 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: <2e8f2a80-c6cf-c3ff-e108-d2cf3057109f@slackbuilds.org> References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <2e8f2a80-c6cf-c3ff-e108-d2cf3057109f@slackbuilds.org> Message-ID: On 03/06/2017 02:22 AM, Willy Sudiarto Raharjo wrote: >> Do you prefer one method over another? > > We only support git. cgit is just web interface so can see other's work I upload my SlackBuilds to GitHub (https://github.com/drgibbon/slackbuilds), is there any way for that to be incorporated into the workflow? I find making a tar archive/uploading to the website to be quite unwieldy when I want to update a script (although I'm doing quite low maintenance stuff, but still). -- Dave From willysr at slackbuilds.org Mon Mar 6 06:36:21 2017 From: willysr at slackbuilds.org (Willy Sudiarto Raharjo) Date: Mon, 6 Mar 2017 13:36:21 +0700 Subject: [Slackbuilds-users] Question about git workflow In-Reply-To: References: <3a96c626-cfb4-e5b0-6b2f-8c1edf7d56d2@slackbuilds.org> <0d06b15e-e320-d893-4486-cf9c4bef631f@osh.id.au> <614a9f92-2e35-09d3-7107-25e8bc23c22c@slackbuilds.org> <2e8f2a80-c6cf-c3ff-e108-d2cf3057109f@slackbuilds.org> Message-ID: <691e9e67-06c8-8646-7de3-43825365c842@slackbuilds.org> >> We only support git. cgit is just web interface so can see other's work > > I upload my SlackBuilds to GitHub > (https://github.com/drgibbon/slackbuilds), is there any way for that to > be incorporated into the workflow? I find making a tar archive/uploading > to the website to be quite unwieldy when I want to update a script > (although I'm doing quite low maintenance stuff, but still). it's better to fork from https://github.com/SlackBuildsOrg/slackbuilds/ and then create your own branch and start making changes Some notes: - always rebase from latest master branch - Only 1 commit per script - Use the following commit message template for version bump "/