From glen at delfi.ee Wed Apr 1 01:37:43 2015 From: glen at delfi.ee (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Wed, 01 Apr 2015 02:37:43 +0300 Subject: [packages/php-pecl-http] - R += pecl-propro and pecl-raphf In-Reply-To: <20150331153416.GA30151@adamg.eu> References: <66ed567bf73502d3c2f1fff0254087f0dc592a8a_refs_heads_master@pld-linux.org> <20150331140426.GA22593@adamg.eu> <551AB0C1.40708@pld-linux.org> <20150331153416.GA30151@adamg.eu> Message-ID: <551B2FC7.1070105@delfi.ee> On 31.03.2015 18:34, Adam Golebiowski wrote: > On Tue, Mar 31, 2015 at 05:35:45PM +0300, Elan Ruusam?e wrote: >> On 31.03.2015 17:04, Adam Golebiowski wrote: >>> [adamg at adamg ~]$ php -r '' >>> PHP Warning: PHP Startup: Unable to load dynamic library '/usr/lib/php56/http.so' - /usr/lib/php56/http.so: undefined symbol: php_persistent_handle_abandon in Unknown on line 0 >>> [adamg at adamg ~]$ sudo mv /etc/php56/conf.d/{,z_}httpd.ini >>> [adamg at adamg ~]$ php -r '' >>> [adamg at adamg ~]$ >>> >>> Not only we need these two, but we need them to load before http.so. >>> >>> Thinking about adopting apache-style prefixing with numbers: >>> - 00-29 - core php modules >>> - 30-99 - extra (pecl / non-pecl) >> i've tried to postpone introducing numbering to ini files as long as >> possible. because the numbering sucks(!). > but it's easy and solves the problem. creates problems elsewhere, need to migrate configs properly, and i hate when config files rename themselves, as i keep them in vcs. >> maybe rename your ini file so that LC_ALL=C locale would sort it last: >> >> like: >> http.ini -> http.ini (no changes) >> propro.ini -> http_propro.ini >> raphf.ini -> http_raphf.ini > until we add some `boo' extension that depends on raphf, and that's just > another version of prefixing. but you need to handle only these two packages. not all 199 php-related packages, and don't even know when you need to reorder again because you guessed initial number wrong and *all* dependencies need to be shifted by some number. and the schema is "_.ini which is quite clear after what to order. > >> ideally php-core should be smart itself (somehow) to be able to load >> modules in different order based on dependencies. >> possible solution would be to open shared library with lazy loading to >> get dependency structure and then do the real open (with RTDL_NOW) after >> dependency sort. >> this is just theory, haven't tried. > nice, but a bit overcomplicated. > and we would end up maintaining yet another patch. and what prevents submitting it upstream? -- glen From adamg at pld-linux.org Wed Apr 1 07:54:21 2015 From: adamg at pld-linux.org (Adam Golebiowski) Date: Wed, 1 Apr 2015 07:54:21 +0200 Subject: [packages/php-pecl-http] - R += pecl-propro and pecl-raphf In-Reply-To: <551B2FC7.1070105@delfi.ee> References: <66ed567bf73502d3c2f1fff0254087f0dc592a8a_refs_heads_master@pld-linux.org> <20150331140426.GA22593@adamg.eu> <551AB0C1.40708@pld-linux.org> <20150331153416.GA30151@adamg.eu> <551B2FC7.1070105@delfi.ee> Message-ID: <20150401055421.GB30151@adamg.eu> On Wed, Apr 01, 2015 at 02:37:43AM +0300, Elan Ruusam?e wrote: > On 31.03.2015 18:34, Adam Golebiowski wrote: > > On Tue, Mar 31, 2015 at 05:35:45PM +0300, Elan Ruusam?e wrote: > >> On 31.03.2015 17:04, Adam Golebiowski wrote: > >>> [adamg at adamg ~]$ php -r '' > >>> PHP Warning: PHP Startup: Unable to load dynamic library '/usr/lib/php56/http.so' - /usr/lib/php56/http.so: undefined symbol: php_persistent_handle_abandon in Unknown on line 0 > >>> [adamg at adamg ~]$ sudo mv /etc/php56/conf.d/{,z_}httpd.ini > >>> [adamg at adamg ~]$ php -r '' > >>> [adamg at adamg ~]$ > >>> > >>> Not only we need these two, but we need them to load before http.so. > >>> > >>> Thinking about adopting apache-style prefixing with numbers: > >>> - 00-29 - core php modules > >>> - 30-99 - extra (pecl / non-pecl) > >> i've tried to postpone introducing numbering to ini files as long as > >> possible. because the numbering sucks(!). > > but it's easy and solves the problem. > > creates problems elsewhere, need to migrate configs properly, and i hate > when config files rename themselves, as i keep them in vcs. migration - not that big pita, most of the files vcs - point taken, but we do rename files every now and then. > >> maybe rename your ini file so that LC_ALL=C locale would sort it last: > >> > >> like: > >> http.ini -> http.ini (no changes) > >> propro.ini -> http_propro.ini > >> raphf.ini -> http_raphf.ini > > until we add some `boo' extension that depends on raphf, and that's just > > another version of prefixing. > > but you need to handle only these two packages. not all 199 php-related > packages, and don't even know when you need to reorder again because you > guessed initial number wrong and *all* dependencies need to be shifted > by some number. Not if we choose initial number to be in the middle. > and the schema is "_.ini which is > quite clear after what to order. Until we add a second package that depends on where it starts to get funny. And in this case, it won't help. [adamg at adamg ~]$ echo -e "http.ini\nhttp_raphf.ini\nhttp_propro.ini" | LC_ALL=C sort http.ini http_propro.ini http_raphf.ini [adamg at adamg ~]$ > >> ideally php-core should be smart itself (somehow) to be able to load > >> modules in different order based on dependencies. > >> possible solution would be to open shared library with lazy loading to > >> get dependency structure and then do the real open (with RTDL_NOW) after > >> dependency sort. > >> this is just theory, haven't tried. > > nice, but a bit overcomplicated. > > and we would end up maintaining yet another patch. > and what prevents submitting it upstream? Nothing, but I am not sure it would be accepted as it would make modules loading path inconsistent between platforms (not really sure you can do the lazy loading trick on Windows). That said, I am not willing to code it (feel free), but I do take the vcs argument, so unless someone makes the patch, I would go with a different approach: - for now, use ugly way solve the problem for this one case (http.ini -> z_http.ini). - wait for next major (7.0.0 - Dec 2015) / minor (5.7.0 - Aug 2015) PHP release before implementing numbering approach That will make yours and other people's vcs happy. -- adamg From glen at pld-linux.org Thu Apr 2 09:53:21 2015 From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Thu, 02 Apr 2015 10:53:21 +0300 Subject: [packages/php-pecl-http] - R += pecl-propro and pecl-raphf In-Reply-To: <20150401055421.GB30151@adamg.eu> References: <66ed567bf73502d3c2f1fff0254087f0dc592a8a_refs_heads_master@pld-linux.org> <20150331140426.GA22593@adamg.eu> <551AB0C1.40708@pld-linux.org> <20150331153416.GA30151@adamg.eu> <551B2FC7.1070105@delfi.ee> <20150401055421.GB30151@adamg.eu> Message-ID: <551CF571.8070404@pld-linux.org> On 01.04.2015 08:54, Adam Golebiowski wrote: >>>> > >>maybe rename your ini file so that LC_ALL=C locale would sort it last: >>>> > >> >>>> > >>like: >>>> > >>http.ini -> http.ini (no changes) >>>> > >>propro.ini -> http_propro.ini >>>> > >>raphf.ini -> http_raphf.ini >>> > >until we add some `boo' extension that depends on raphf, and that's just >>> > >another version of prefixing. two more ideas: 1. create modules-load.d dir, which would use the numbering (or whatever scheme) it would contain only "extension=modname.mod" lines the actual module specific configs stay in current paths 2. create script which creates directory containing ordered symlinks automatically, symlinks pointing to current conf.d similar to mods-available, mods-enabled scheme used in debian/ubuntu the order would be somehow automatically detected, maybe extra file from package saying what needs to be loaded first. ps: we have SAPINAME.d besides conf.d as well. maybe it is no longer neccessary as loading cli only module probably does not crash webserver-sapi modules anymore. -- glen From glen at pld-linux.org Sun Apr 12 11:08:39 2015 From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Sun, 12 Apr 2015 12:08:39 +0300 Subject: monitoring plugins Message-ID: <552A3617.8050301@pld-linux.org> 11:06:39 glen> https://github.com/monitoring-plugins/monitoring-plugins 11:06:46 glen> https://github.com/nagios-plugins/nagios-plugins 11:07:01 glen> i wonder, should create new package? or use just different tarball in nagios-plugins.spec -- glen From glen at pld-linux.org Tue Apr 21 06:47:06 2015 From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Tue, 21 Apr 2015 07:47:06 +0300 Subject: Fwd: [packages/atlas-upload-cli: 3/3] fix issue when invoked via builder script In-Reply-To: References: Message-ID: <5535D64A.8040801@pld-linux.org> since when our build macros to not reset env to be clean? here's snip of the build log: Executing(%build): /bin/sh -e /home/users/glen/tmp/rpm-tmp.91559 TMP=/home/users/glen/tmp SSH_AUTH_SOCK=/tmp/ssh-dpsZYbTNrr/agent.251855 HOME=/home/users/glen PATH=/bin:/usr/bin:/usr/sbin:/sbin:/usr/X11R6/bin TMPDIR=/home/users/glen/tmp + umask 022 + cd /home/users/glen/rpm/packages/BUILD.i686-linux + cd atlas-upload-cli-0.2.0 + env + grep GIT GIT_DIR=/home/users/glen/rpm/packages/atlas-upload-cli/.git GIT_WORK_TREE=/home/users/glen/rpm/packages/atlas-upload-cli + pwd + export 'GOPATH=/home/users/glen/rpm/packages/BUILD.i686-linux/atlas-upload-cli-0.2.0/vendor' -------- Forwarded Message -------- Subject: [packages/atlas-upload-cli: 3/3] fix issue when invoked via builder script Date: Tue, 21 Apr 2015 06:43:42 +0200 From: glen Reply-To: pld-devel-en at lists.pld-linux.org, pld-devel-pl at lists.pld-linux.org To: pld-cvs-commit at lists.pld-linux.org commit ba85c2197a01f3ec94a9bd3cf3b87f893553720a Author: Elan Ruusam?e Date: Tue Apr 21 07:43:12 2015 +0300 fix issue when invoked via builder script atlas-upload-cli.spec | 3 +++ 1 file changed, 3 insertions(+) --- diff --git a/atlas-upload-cli.spec b/atlas-upload-cli.spec index 6859e4b..304c675 100644 --- a/atlas-upload-cli.spec +++ b/atlas-upload-cli.spec @@ -38,6 +38,9 @@ deploys. #package github.com/mitchellh/ioprogress: cannot download, $GOPATH not set. For more details see: go help gopath %build +# these interfere with go download -- the git vars point to .spec repo +unset GIT_DIR GIT_WORK_TREE + export GOPATH=$(pwd)/vendor # make -j1 because https://github.com/hashicorp/atlas-upload-cli/pull/10 %{__make} -j1 ================================================================ ---- gitweb: http://git.pld-linux.org/gitweb.cgi/packages/atlas-upload-cli.git/commitdiff/ba85c2197a01f3ec94a9bd3cf3b87f893553720a _______________________________________________ pld-cvs-commit mailing list pld-cvs-commit at lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit From baggins at pld-linux.org Tue Apr 21 07:41:44 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Tue, 21 Apr 2015 07:41:44 +0200 Subject: [packages/gdm] Update to 3.16.1.1 In-Reply-To: <787341bb0d552aea6a9944ec80f3fd73a5d7d71e_refs_heads_master@pld-linux.org> References: <787341bb0d552aea6a9944ec80f3fd73a5d7d71e_refs_heads_master@pld-linux.org> Message-ID: <20150421054143.GA3493@home.lan> Does it work without systemd and without weston / non-root X? On Mon, 20 Apr 2015, megabajt wrote: > commit 787341bb0d552aea6a9944ec80f3fd73a5d7d71e > Author: Marcin Banasiak > Date: Mon Apr 20 22:22:25 2015 +0200 > > Update to 3.16.1.1 > > gdm.spec | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > --- > diff --git a/gdm.spec b/gdm.spec > index d2b213d..1e60635 100644 > --- a/gdm.spec > +++ b/gdm.spec > @@ -16,13 +16,13 @@ Summary(pt_BR.UTF-8): Gerenciador de Entrada do GNOME > Summary(ru.UTF-8): ?????????? ???????? GNOME > Summary(uk.UTF-8): ?????????? ???????? GNOME > Name: gdm > -Version: 3.16.0 > +Version: 3.16.1.1 > Release: 1 > Epoch: 2 > License: GPL/LGPL > Group: X11/Applications > Source0: http://ftp.gnome.org/pub/GNOME/sources/gdm/3.16/%{name}-%{version}.tar.xz > -# Source0-md5: aacdff89e695cb66c8d4a6ff0a00e293 > +# Source0-md5: 0b315af5fe0ee66055edb6f301510820 > Source1: %{name}.pamd > Source2: %{name}.init > Source3: %{name}-pld-logo.png > ================================================================ > > ---- gitweb: > > http://git.pld-linux.org/gitweb.cgi/packages/gdm.git/commitdiff/787341bb0d552aea6a9944ec80f3fd73a5d7d71e > > _______________________________________________ > pld-cvs-commit mailing list > pld-cvs-commit at lists.pld-linux.org > http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From megabajt at pld-linux.org Tue Apr 21 18:54:26 2015 From: megabajt at pld-linux.org (Marcin Banasiak) Date: Tue, 21 Apr 2015 18:54:26 +0200 Subject: [packages/gdm] Update to 3.16.1.1 In-Reply-To: <20150421054143.GA3493@home.lan> References: <787341bb0d552aea6a9944ec80f3fd73a5d7d71e_refs_heads_master@pld-linux.org> <20150421054143.GA3493@home.lan> Message-ID: 2015-04-21 7:41 GMT+02:00 Jan R?korajski : > Does it work without systemd and without weston / non-root X? It should. Wayland is used by default, but there is a fallback to Xorg. You can also disable wayland in /etc/gdm/custom.conf. That's theory. I've tried to make it work with both wayland and xorg, but none of them works for me :( I'm afraid we should remove this package from th-test. -- Marcin Banasiak From baggins at pld-linux.org Tue Apr 21 19:07:20 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Tue, 21 Apr 2015 19:07:20 +0200 Subject: [packages/gdm] Update to 3.16.1.1 In-Reply-To: References: <787341bb0d552aea6a9944ec80f3fd73a5d7d71e_refs_heads_master@pld-linux.org> <20150421054143.GA3493@home.lan> Message-ID: <20150421170718.GA3450@home.lan> On Tue, 21 Apr 2015, Marcin Banasiak wrote: > 2015-04-21 7:41 GMT+02:00 Jan R?korajski : > > Does it work without systemd and without weston / non-root X? > > It should. Wayland is used by default, but there is a fallback to > Xorg. You can also disable wayland in /etc/gdm/custom.conf. That's > theory. > > I've tried to make it work with both wayland and xorg, but none of > them works for me :( I was afraid so, I tried to run 3.16.0, failed and thought maybe they did fix something in update. > I'm afraid we should remove this package from th-test. Removed. -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From jajcus at jajcus.net Wed Apr 22 19:38:27 2015 From: jajcus at jajcus.net (Jacek Konieczny) Date: Wed, 22 Apr 2015 19:38:27 +0200 Subject: python In-Reply-To: <5537C13C.20104@pld-linux.org> References: <5537C13C.20104@pld-linux.org> Message-ID: <5537DC93.7060804@jajcus.net> On 2015-04-22 17:41, Elan Ruusam?e wrote: > http://git.pld-linux.org/?p=packages/python-setuptools.git;a=commitdiff;h=64c5aa5e97487c2879a621605724e4e64f85a933 > > blah, i do not understand what should i use in new specs > python-setuptools or python-distribute? setuptools > i thought setuptools is old and distribute is new? That was true at some time. But that had changed again long time ago. So its complicated. In short: Python has provided 'distutils' in its standard library for long time, but that was quite limited, so setuptools appeared. For some it was not enough either, so it was forked and 'distribute' appeared. Soon, 'distribute' become the standard. In the meantime another project 'distutils2' was announced, which was to supersed both setuptools and distribute, but that didn't really happen. Instead, 'distribute' was merged back to 'setuptools' and 'setuptools' is what the world uses now. Now 'distribute' is only a dummy package providing egg-info data for other packages tha might require it. http://stackoverflow.com/questions/6344076/differences-between-distribute-distutils-setuptools-and-distutils2/14753678#14753678 > could you update template-specs/python.spec accordingly so that would be > source of truth Already done. I updated the template when I was updating the setuptools and distribute packages. Greets, Jacek From wrobell at pld-linux.org Wed Apr 22 19:47:42 2015 From: wrobell at pld-linux.org (Artur Wroblewski) Date: Wed, 22 Apr 2015 18:47:42 +0100 Subject: python In-Reply-To: <5537DC93.7060804@jajcus.net> References: <5537C13C.20104@pld-linux.org> <5537DC93.7060804@jajcus.net> Message-ID: On Wed, Apr 22, 2015 at 6:38 PM, Jacek Konieczny wrote: > On 2015-04-22 17:41, Elan Ruusam?e wrote: >> http://git.pld-linux.org/?p=packages/python-setuptools.git;a=commitdiff;h=64c5aa5e97487c2879a621605724e4e64f85a933 >> >> blah, i do not understand what should i use in new specs >> python-setuptools or python-distribute? > > setuptools > >> i thought setuptools is old and distribute is new? > > That was true at some time. But that had changed again long time ago. > > So its complicated. In short: > > Python has provided 'distutils' in its standard library for long time, > but that was quite limited, so setuptools appeared. For some it was not > enough either, so it was forked and 'distribute' appeared. Soon, > 'distribute' become the standard. > > In the meantime another project 'distutils2' was announced, which was to > supersed both setuptools and distribute, but that didn't really happen. > Instead, 'distribute' was merged back to 'setuptools' and 'setuptools' > is what the world uses now. Also, pip is the blessed way of installing the modules now, see https://www.python.org/dev/peps/pep-0453/#abstract Some notes about setuptools https://www.python.org/dev/peps/pep-0453/#automatic-installation-of-setuptools [...] Regards, w From jajcus at jajcus.net Wed Apr 22 20:48:51 2015 From: jajcus at jajcus.net (Jacek Konieczny) Date: Wed, 22 Apr 2015 20:48:51 +0200 Subject: python In-Reply-To: References: <5537C13C.20104@pld-linux.org> <5537DC93.7060804@jajcus.net> Message-ID: <5537ED13.7080209@jajcus.net> On 2015-04-22 19:47, Artur Wroblewski wrote: > On Wed, Apr 22, 2015 at 6:38 PM, Jacek Konieczny wrote: >> On 2015-04-22 17:41, Elan Ruusam?e wrote: >> In the meantime another project 'distutils2' was announced, which was to >> supersed both setuptools and distribute, but that didn't really happen. >> Instead, 'distribute' was merged back to 'setuptools' and 'setuptools' >> is what the world uses now. > > Also, pip is the blessed way of installing the modules now, see > > https://www.python.org/dev/peps/pep-0453/#abstract Yeah? blessed by developers, who don't care about system consistency etc. Installing packages with pip doesn't go well with providing consistent Linux distribution packages. We could probably minimize the damages pip can do by making it somehow install 'system-wide' packages to /usr/local instead of /usr? but that would probably require changes in our Python packages and can cause new compatibility problems. Currently our pip installs packages into /usr/share/site-packages, overwirtting whatever is already there (from PLD RPM packages) making a real mess of the system. Additionally our pip is stipped of the private copies of some libraries it uses (requests, urllib2), which are sometimes used by other python packages. Yes, Python developers went the Java way ? ?instead of creating stable interfaces, let's just redistribute random version of all required libraries with our software. Or ask to use virtualenv for everything. Twenty copies of ten different versions of 'requests'? What could go wrong? Like, if a security problem was found? Jacek From mateusz-lists at ant.gliwice.pl Thu Apr 23 10:04:21 2015 From: mateusz-lists at ant.gliwice.pl (Mateusz Korniak) Date: Thu, 23 Apr 2015 10:04:21 +0200 Subject: python In-Reply-To: <5537ED13.7080209@jajcus.net> References: <5537C13C.20104@pld-linux.org> <5537ED13.7080209@jajcus.net> Message-ID: <2680997.tgCvMUgdvJ@matkor-toshiba> On Wednesday 22 April 2015 20:48:51 Jacek Konieczny wrote: > > Also, pip is the blessed way of installing the modules now, see > > > > https://www.python.org/dev/peps/pep-0453/#abstract > > Yeah? blessed by developers, who don't care about system consistency > etc. (...) > We could probably minimize the damages pip can do (...) PIP can be easily used inside any Python virtual environment when really needed, so IMHO, there is no need to propagate PIP mess system wide, and we should system-wide installs be from rpms only by default. Not only many versions of same packages is problem with PIP, but we also lose track of library dependencies in binary packages. -- Mateusz Korniak "(...) mam brata - powa?ny, domator, liczykrupa, hipokryta, pobo?ni?, kr?tko m?wi?c - podpora spo?ecze?stwa." Nikos Kazantzakis - "Grek Zorba" From jajcus at jajcus.net Thu Apr 23 10:25:00 2015 From: jajcus at jajcus.net (Jacek Konieczny) Date: Thu, 23 Apr 2015 10:25:00 +0200 Subject: python In-Reply-To: <2680997.tgCvMUgdvJ@matkor-toshiba> References: <5537C13C.20104@pld-linux.org> <5537ED13.7080209@jajcus.net> <2680997.tgCvMUgdvJ@matkor-toshiba> Message-ID: <5538AC5C.4070001@jajcus.net> On 2015-04-23 10:04, Mateusz Korniak wrote: > On Wednesday 22 April 2015 20:48:51 Jacek Konieczny wrote: >>> Also, pip is the blessed way of installing the modules now, see >>> >>> https://www.python.org/dev/peps/pep-0453/#abstract >> >> Yeah? blessed by developers, who don't care about system consistency >> etc. (...) >> We could probably minimize the damages pip can do (...) > > PIP can be easily used inside any Python virtual environment when really > needed, so IMHO, there is no need to propagate PIP mess system wide, and we > should system-wide installs be from rpms only by default. > > Not only many versions of same packages is problem with PIP, but we also lose > track of library dependencies in binary packages. But the PEP you linked above suggests that 'pip' and 'ensurepip' should always be available system-wide and capable of upgrading pip itself. Even if we agree using those system wide is not a good idea, we should make it work in a way, that minimizes the damages. Greets, Jacek From glen at pld-linux.org Thu Apr 23 11:07:23 2015 From: glen at pld-linux.org (=?ISO-8859-1?Q?Elan_Ruusam=E4e?=) Date: Thu, 23 Apr 2015 12:07:23 +0300 Subject: rpm -e in deleted directory Message-ID: <5538B64B.5010207@pld-linux.org> this particular dir was removed by rpm itself # poldek -e hwdata-0.275-1.noarch mark hwdata-0.275-1.noarch Processing dependencies... hwdata-0.275-1.noarch marks pciutils-3.3.1-2.x86_64 (req hwdata >= 0.243-2) There are 2 packages to remove (1 marked by dependencies): R hwdata-0.275-1.noarch D pciutils-3.3.1-2.x86_64 This operation will free 5.6MB of disk space. Proceed? [N/y] y Running rpm --erase --root /... rpmdb: BDB2570 no absolute path for the current directory: No such file or directory rpmdb: BDB2571 log file auto-remove: No such file or directory but i recall it behaving the same when invoking initial erase from dir that has already been removed -- glen From glen at pld-linux.org Thu Apr 23 14:54:23 2015 From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Thu, 23 Apr 2015 15:54:23 +0300 Subject: Fwd: [packages/nagios] fix $HOME env var In-Reply-To: <72682420b4037546e22f06ed3a32242cac375ef0_refs_heads_master@pld-linux.org> References: <72682420b4037546e22f06ed3a32242cac375ef0_refs_heads_master@pld-linux.org> Message-ID: <5538EB7F.6070707@pld-linux.org> imho we should set nagios user $HOME to be /var/lib/nagios currently it's /usr/lib/nagios on i686 and /usr/lib64/nagios on x86_64 it's first inconsistent (can't share configs) and secondly some plugins require putting files to $HOME, and writing to /usr/lib/nagios by sysadmin/apps is wrong. -------- Original Message -------- Subject: [packages/nagios] fix $HOME env var Date: Thu, 23 Apr 2015 14:51:41 +0200 From: glen Reply-To: pld-devel-en at lists.pld-linux.org, pld-devel-pl at lists.pld-linux.org To: pld-cvs-commit at lists.pld-linux.org commit 72682420b4037546e22f06ed3a32242cac375ef0 Author: Elan Ruusam?e Date: Thu Apr 23 15:51:26 2015 +0300 fix $HOME env var nagios.init | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- diff --git a/nagios.init b/nagios.init index cf6348f..b23587e 100755 --- a/nagios.init +++ b/nagios.init @@ -27,6 +27,10 @@ fi cfg_file=/etc/nagios/nagios.cfg +# value to use for $HOME +# as pld initscript resets HOME=/tmp, plugins may inherit bad value +nagios_home=/usr/lib/nagios + # check for precache precached_object_file=$(awk -F= '/^precached_object_file/{print $2}' $cfg_file) @@ -83,7 +87,7 @@ start() { # we're safe to use -x as we did verify config prior startup # precached object file also is created in configtest. - daemon /usr/sbin/nagios ${precached_object_file:+-u} -x -d $cfg_file + daemon env -i PATH="$PATH" HOME="$nagios_home" /usr/sbin/nagios ${precached_object_file:+-u} -x -d $cfg_file RETVAL=$? if [ $RETVAL -eq 0 ]; then touch /var/lock/subsys/nagios ================================================================ ---- gitweb: http://git.pld-linux.org/gitweb.cgi/packages/nagios.git/commitdiff/72682420b4037546e22f06ed3a32242cac375ef0 _______________________________________________ pld-cvs-commit mailing list pld-cvs-commit at lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit From baggins at pld-linux.org Sat Apr 25 21:30:01 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Sat, 25 Apr 2015 21:30:01 +0200 Subject: [packages/libass] try to disable asm on x32 In-Reply-To: References: <5e58a9c8df07ba4897742d9388e895b92855cb31_refs_heads_master@pld-linux.org> Message-ID: <20150425193001.GA3423@home.lan> If it uses nasm try passing proper elf format settings '-f elfx32' On Sat, 25 Apr 2015, atler wrote: > commit c42901fed838c359ad23900dd23f87e81b372b83 > Author: Jan Palus > Date: Sat Apr 25 21:13:54 2015 +0200 > > try to disable asm on x32 > > libass.spec | 3 +++ > 1 file changed, 3 insertions(+) > --- > diff --git a/libass.spec b/libass.spec > index c657ae8..9d5f844 100644 > --- a/libass.spec > +++ b/libass.spec > @@ -78,6 +78,9 @@ Statyczna biblioteka LibASS. > %{__autoheader} > %{__automake} > %configure \ > +%ifarch x32 > + --disable-asm \ > +%endif > --disable-silent-rules > > %{__make} > ================================================================ > > ---- gitweb: > > http://git.pld-linux.org/gitweb.cgi/packages/libass.git/commitdiff/c42901fed838c359ad23900dd23f87e81b372b83 > > _______________________________________________ > pld-cvs-commit mailing list > pld-cvs-commit at lists.pld-linux.org > http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From baggins at pld-linux.org Mon Apr 27 18:11:43 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Mon, 27 Apr 2015 18:11:43 +0200 Subject: [packages/xfmedia] - rel.13 In-Reply-To: <19b64aa531d9a7b6d4afa65b558ed192deee08f5_refs_heads_master@pld-linux.org> References: <4c09c10c6db092b3240c728a243f8adbc909cb61_refs_heads_master@pld-linux.org> <19b64aa531d9a7b6d4afa65b558ed192deee08f5_refs_heads_master@pld-linux.org> Message-ID: <20150427161142.GA3589@home.lan> It's dead and obsoleted by parole. On Mon, 27 Apr 2015, blues wrote: > commit 19b64aa531d9a7b6d4afa65b558ed192deee08f5 > Author: Pawe? Go?aszewski > Date: Mon Apr 27 13:49:41 2015 +0200 > > - rel.13 > > xfmedia.spec | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > --- > diff --git a/xfmedia.spec b/xfmedia.spec > index 2ecfe4b..0be257d 100644 > --- a/xfmedia.spec > +++ b/xfmedia.spec > @@ -6,7 +6,7 @@ Summary: Xfmedia - lightweight media player based on the xine engine > Summary(pl.UTF-8): Xfmedia - lekki odtwarzacz multimedialny oparty na silniku xine > Name: xfmedia > Version: 0.9.2 > -Release: 12 > +Release: 13 > License: GPL v2 > Group: X11/Applications/Multimedia > Source0: http://spuriousinterrupt.org/projects/xfmedia/files/%{name}-%{version}.tar.bz2 > ================================================================ > > ---- gitweb: > > http://git.pld-linux.org/gitweb.cgi/packages/xfmedia.git/commitdiff/19b64aa531d9a7b6d4afa65b558ed192deee08f5 > > _______________________________________________ > pld-cvs-commit mailing list > pld-cvs-commit at lists.pld-linux.org > http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From baggins at pld-linux.org Mon Apr 27 18:15:02 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Mon, 27 Apr 2015 18:15:02 +0200 Subject: [packages/xfce4-battery-plugin] - rel.6 In-Reply-To: <7b16a94ad7a4c1555058b7681e5cede7cdbd5f5e_refs_heads_master@pld-linux.org> References: <1228585b32881e427ee49cb1ff920d814f02e829_refs_heads_master@pld-linux.org> <7b16a94ad7a4c1555058b7681e5cede7cdbd5f5e_refs_heads_master@pld-linux.org> Message-ID: <20150427161502.GB3589@home.lan> It's been merged into xfce4-power-manager, I will removed it from th-test again On Mon, 27 Apr 2015, blues wrote: > commit 7b16a94ad7a4c1555058b7681e5cede7cdbd5f5e > Author: Pawe? Go?aszewski > Date: Mon Apr 27 13:53:26 2015 +0200 > > - rel.6 > > xfce4-battery-plugin.spec | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > --- > diff --git a/xfce4-battery-plugin.spec b/xfce4-battery-plugin.spec > index 7fe822b..74fd78c 100644 > --- a/xfce4-battery-plugin.spec > +++ b/xfce4-battery-plugin.spec > @@ -2,7 +2,7 @@ Summary: Battery monitor panel plugin for Xfce > Summary(pl.UTF-8): Monitor zu?ycia baterii dla panelu Xfce > Name: xfce4-battery-plugin > Version: 1.0.5 > -Release: 5 > +Release: 6 > License: GPL v2 > Group: X11/Applications > Source0: http://archive.xfce.org/src/panel-plugins/xfce4-battery-plugin/1.0/%{name}-%{version}.tar.bz2 > ================================================================ > > ---- gitweb: > > http://git.pld-linux.org/gitweb.cgi/packages/xfce4-battery-plugin.git/commitdiff/7b16a94ad7a4c1555058b7681e5cede7cdbd5f5e > > _______________________________________________ > pld-cvs-commit mailing list > pld-cvs-commit at lists.pld-linux.org > http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From baggins at pld-linux.org Tue Apr 28 20:05:41 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Tue, 28 Apr 2015 20:05:41 +0200 Subject: [packages/pwdutils] Revert "- updated gettext BR" In-Reply-To: <7dd6707732bd39e8126426a2ef984ed6727774e1_refs_heads_master@pld-linux.org> References: <7dd6707732bd39e8126426a2ef984ed6727774e1_refs_heads_master@pld-linux.org> Message-ID: <20150428180541.GA3412@home.lan> Why did you revert it? The whole -devel package is not needed to regenerate .po files. On Tue, 28 Apr 2015, glen wrote: > commit 7dd6707732bd39e8126426a2ef984ed6727774e1 > Author: Elan Ruusam?e > Date: Tue Apr 28 20:06:51 2015 +0300 > > Revert "- updated gettext BR" > > This reverts commit 0bc92006af89196f5949282fe02f058648940100. > > pwdutils.spec | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > --- > diff --git a/pwdutils.spec b/pwdutils.spec > index e5c56b5..1a17ba0 100644 > --- a/pwdutils.spec > +++ b/pwdutils.spec > @@ -41,7 +41,7 @@ BuildRequires: autoconf > BuildRequires: automake >= 1:1.9 > %{?with_bioapi:BuildRequires: bioapi-devel} > BuildRequires: gcc >= 5:3.2 > -BuildRequires: gettext-tools > +BuildRequires: gettext-devel > %{?with_gnutls:BuildRequires: gnutls-devel >= 1.0.0} > BuildRequires: libnscd-devel > %{?with_selinux:BuildRequires: libselinux-devel} > ================================================================ > > ---- gitweb: > > http://git.pld-linux.org/gitweb.cgi/packages/pwdutils.git/commitdiff/476e0c813c7626146165cee8787116c1ec986f27 > > _______________________________________________ > pld-cvs-commit mailing list > pld-cvs-commit at lists.pld-linux.org > http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From baggins at pld-linux.org Wed Apr 29 07:53:06 2015 From: baggins at pld-linux.org (Jan =?utf-8?Q?R=C4=99korajski?=) Date: Wed, 29 Apr 2015 07:53:06 +0200 Subject: [packages/pwdutils] Revert "- updated gettext BR" In-Reply-To: <5540662C.1090209@pld-linux.org> References: <7dd6707732bd39e8126426a2ef984ed6727774e1_refs_heads_master@pld-linux.org> <20150428180541.GA3412@home.lan> <5540662C.1090209@pld-linux.org> Message-ID: <20150429055305.GA3421@home.lan> On Wed, 29 Apr 2015, Elan Ruusam?e wrote: > because you blindly changed these > and makes simplier to build ac version > less conditions, etc What kind of argument is this? All packages on HEAD have up-to-date BRs - for other packages on HEAD, not some unneeded, historical ones. Ac place is on ac-branch. Revert this revert, and don't remove mailing list from To:. > > On 28.04.2015 21:05, Jan R?korajski wrote: > > Why did you revert it? The whole -devel package is not needed to > > regenerate .po files. > > > > On Tue, 28 Apr 2015, glen wrote: > > > >> commit 7dd6707732bd39e8126426a2ef984ed6727774e1 > >> Author: Elan Ruusam?e > >> Date: Tue Apr 28 20:06:51 2015 +0300 > >> > >> Revert "- updated gettext BR" > >> > >> This reverts commit 0bc92006af89196f5949282fe02f058648940100. > >> > >> pwdutils.spec | 2 +- > >> 1 file changed, 1 insertion(+), 1 deletion(-) > >> --- > >> diff --git a/pwdutils.spec b/pwdutils.spec > >> index e5c56b5..1a17ba0 100644 > >> --- a/pwdutils.spec > >> +++ b/pwdutils.spec > >> @@ -41,7 +41,7 @@ BuildRequires: autoconf > >> BuildRequires: automake >= 1:1.9 > >> %{?with_bioapi:BuildRequires: bioapi-devel} > >> BuildRequires: gcc >= 5:3.2 > >> -BuildRequires: gettext-tools > >> +BuildRequires: gettext-devel > >> %{?with_gnutls:BuildRequires: gnutls-devel >= 1.0.0} > >> BuildRequires: libnscd-devel > >> %{?with_selinux:BuildRequires: libselinux-devel} > > > > -- > glen > -- Jan R?korajski | PLD/Linux SysAdm | bagginspld-linux.org | http://www.pld-linux.org/ From glen at pld-linux.org Wed Apr 29 08:02:01 2015 From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=) Date: Wed, 29 Apr 2015 09:02:01 +0300 Subject: [packages/pwdutils] Revert "- updated gettext BR" In-Reply-To: <20150429055305.GA3421@home.lan> References: <7dd6707732bd39e8126426a2ef984ed6727774e1_refs_heads_master@pld-linux.org> <20150428180541.GA3412@home.lan> <5540662C.1090209@pld-linux.org> <20150429055305.GA3421@home.lan> Message-ID: <554073D9.9060306@pld-linux.org> On 29.04.2015 08:53, Jan R?korajski wrote: > and don't remove mailing list from To:. don't add me to to:/cc: when posting to list, my mails go to different folder that get sent privately. -- glen From jajcus at jajcus.net Thu Apr 30 16:08:55 2015 From: jajcus at jajcus.net (Jacek Konieczny) Date: Thu, 30 Apr 2015 16:08:55 +0200 Subject: CEPH Message-ID: <55423777.5020007@jajcus.net> Hi, Does anyone actually use the CEPH packages in PLD? The spec file seems actively maintained, but the init script seems unusable. Currently that is an over-complicated upstream startup script not updated for PLD (except disabling it through the chkconfig header). It doesn't work and it seems it cannot reliably work in PLD, for multiple reasons. I think that should be replaced with some sane and PLD-specific init scripts and systemd units. But for this some decisions need to be made (e.g. what user to run that as) which may break whatever configuration anyone uses. I would also like to update the 'ceph-deploy' package (currently not in PLD) to support PLD and that will also require more changes that just adding poldek support ? currently it abuses '/sbin/service' to pass arguments to the init script 'start' action. This won't work reliably under systemd. Jacek