From sparky at pld-linux.org  Tue Feb  1 19:30:14 2011
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Tue, 1 Feb 2011 19:30:14 +0100
Subject: rc-scripts-user
Message-ID: <20110201183014.GA4429@pld-linux.org>


rc-scripts-user / userscripts allows normal users to schedule and
execute tasks on system startup and shutdown. Please test and comment.
Once it gets on any production system it will be hard to change
interface without breaking user configuration.

The interface is:

* All scripts in ~/.config/init.d/ (hardcoded path, no xdg support)
 - any better ideas ?

* all *~ and *.bak files are ignored
 - ignore something else ?

* limited number of scripts per user
 - default 20 should be plenty

* executed in alphabetic order
 - maybe invert the order on shutdown ?

* Scripts must accept "start" and "stop" arguments.
 - I don't see any use for restart or reload, do you ?

* Very limited time to perform the task before shutdown
 - We don't want users to lock the machine.


Comments ?

If none, those things will stay as they are.


Best regards,
	Przemys?aw Iskra.

-- 
 ____    sparky  --  Przemyslaw  ................ LANG...Pl,Ca,Es,En
/____) ___  ___  _ _ || Iskra   : WWW . http://ppcrcd.pld-linux.org/
\____\| -_)'___| ||^'||//\\//   : WWW2 ............ http://rsget.pl/
(____/||   (_-_|_||  ||\\ ||    : eMail ..... <sparky at pld-linux.org>


From kornet at camk.edu.pl  Tue Feb  1 19:51:42 2011
From: kornet at camk.edu.pl (Kacper Kornet)
Date: Tue, 1 Feb 2011 19:51:42 +0100
Subject: %cmake macro
Message-ID: <20110201185142.GA27352@camk.edu.pl>

I think that current %cmake macro is broken. For example soprano.spec is build
with options: -O2 ... -O3, which is effectively -O3. As far as I understand the
preferred option would be -O2, 

In cmake the options are build form two variables as following as:

CMAKE_C_FLAGS CMAKE_C_FLAGS_<Buildtype>. 

while %cmake defines CMAKE_C_FLAGS and uses
Builtype = RELEASE. 

So I think there are the following possibilities:


1. Define CMAKE_C_FLAGS_RELEASE in %cmake

Flaws:
A package can define that variable internally in such a way that  it has
precedence.  For example packages from kde4 do it.


2. Change the build type we use to None. Then only CMAKE_C_FLAGS will be used

Flaws:
Same as above. But I don't know if any package redefines that variable
unconditionally.

Also if a package uses some special options in release flags, i.e. to avoid
debug output, we would need to include them in out definition of CMAKE_C_FLAGS
for this package.


3. Use or own build type like: cmake
-DCMAKE_C_FLAGS_PLD=""${CFLAGS:-%{rpmcflags}}"

Flaws:
If a package uses some special options in release flags, i.e. to avoid debug
output, we would need to put them into our flags. 


I can implement the third solution. But first I am waiting for any objections
or better ideas.

-- 
  Kacper Kornet


From blues at pld-linux.org  Tue Feb  1 19:58:55 2011
From: blues at pld-linux.org (Pawel Golaszewski)
Date: Tue, 1 Feb 2011 19:58:55 +0100 (CET)
Subject: rc-scripts-user
In-Reply-To: <20110201183014.GA4429@pld-linux.org>
References: <20110201183014.GA4429@pld-linux.org>
Message-ID: <alpine.LNX.2.00.1102011954500.28788@fgnss.gnfx.tqn.cy>

On Tue, 1 Feb 2011, Przemyslaw Iskra wrote:
> rc-scripts-user / userscripts allows normal users to schedule and 
> execute tasks on system startup and shutdown. Please test and comment. 
> Once it gets on any production system it will be hard to change 
> interface without breaking user configuration.
> 
> The interface is:
> 
> * All scripts in ~/.config/init.d/ (hardcoded path, no xdg support)
>  - any better ideas ?

~/.config/init/

> * all *~ and *.bak files are ignored
>  - ignore something else ?

.rpmnew, .rpmsave...

There is few more temporary filetypes.

> * executed in alphabetic order
>  - maybe invert the order on shutdown ?

I see no use.

> * Scripts must accept "start" and "stop" arguments.
>  - I don't see any use for restart or reload, do you ?

status should be, if state is noticed anywhere.

-- 
pozdr.  Pawe? Go?aszewski          jid:blues<at>jabber<dot>gda<dot>pl
--------------------------------------------------------------------------
If you think of MS-DOS as mono, and Windows as stereo, then Linux is Dolby
Pro-Logic Surround Sound with Bass Boost and all the music is free.

From kornet at camk.edu.pl  Tue Feb  1 20:08:25 2011
From: kornet at camk.edu.pl (Kacper Kornet)
Date: Tue, 1 Feb 2011 20:08:25 +0100
Subject: rc-scripts-user
In-Reply-To: <20110201183014.GA4429@pld-linux.org>
References: <20110201183014.GA4429@pld-linux.org>
Message-ID: <20110201190825.GB27352@camk.edu.pl>

On Tue, Feb 01, 2011 at 07:30:14PM +0100, Przemyslaw Iskra wrote:
> rc-scripts-user / userscripts allows normal users to schedule and
> execute tasks on system startup and shutdown. 

Out of my own curiosity, what do you need it for?

-- 
  Kacper Kornet


From pawelz at pld-linux.org  Tue Feb  1 21:21:30 2011
From: pawelz at pld-linux.org (=?ISO-8859-2?Q?Pawe=B3_Zuzelski?=)
Date: Tue, 1 Feb 2011 21:21:30 +0100
Subject: rc-scripts-user
In-Reply-To: <20110201183014.GA4429@pld-linux.org>
References: <20110201183014.GA4429@pld-linux.org>
Message-ID: <AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>

On Tue, Feb 1, 2011 at 7:30 PM, Przemyslaw Iskra <sparky at pld-linux.org> wrote:
> rc-scripts-user / userscripts allows normal users to schedule and
> execute tasks on system startup and shutdown. Please test and comment.
> Once it gets on any production system it will be hard to change
> interface without breaking user configuration.

Why is it better than @reboot flag in user's crontab?

-- 
Regards,
Pawe?


From sparky at pld-linux.org  Tue Feb  1 21:27:23 2011
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Tue, 1 Feb 2011 21:27:23 +0100
Subject: rc-scripts-user
In-Reply-To: <AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>
References: <20110201183014.GA4429@pld-linux.org>
	<AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>
Message-ID: <20110201202723.GB7391@pld-linux.org>

On Tue, Feb 01, 2011 at 09:21:30PM +0100, Pawe? Zuzelski wrote:
> On Tue, Feb 1, 2011 at 7:30 PM, Przemyslaw Iskra <sparky at pld-linux.org> wrote:
> > rc-scripts-user / userscripts allows normal users to schedule and
> > execute tasks on system startup and shutdown. Please test and comment.
> > Once it gets on any production system it will be hard to change
> > interface without breaking user configuration.
> 
> Why is it better than @reboot flag in user's crontab?

Does that work on system shutdown as well ? If not, how do I stop any of
my daemons on shutdown ?


Best regards,
	Przemys?aw Iskra.

-- 
 ____    sparky  --  Przemyslaw  ................ LANG...Pl,Ca,Es,En
/____) ___  ___  _ _ || Iskra   : WWW . http://ppcrcd.pld-linux.org/
\____\| -_)'___| ||^'||//\\//   : WWW2 ............ http://rsget.pl/
(____/||   (_-_|_||  ||\\ ||    : eMail ..... <sparky at pld-linux.org>


From patrys at pld-linux.org  Tue Feb  1 21:36:02 2011
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Tue, 1 Feb 2011 21:36:02 +0100
Subject: rc-scripts-user
In-Reply-To: <20110201202723.GB7391@pld-linux.org>
References: <20110201183014.GA4429@pld-linux.org>
	<AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>
	<20110201202723.GB7391@pld-linux.org>
Message-ID: <AANLkTim6TBayHEfmfo+KC7_ve1CXC4Co2cobka5rneZ2@mail.gmail.com>

On Tue, Feb 1, 2011 at 9:27 PM, Przemyslaw Iskra <sparky at pld-linux.org> wrote:
> Does that work on system shutdown as well ? If not, how do I stop any of
> my daemons on shutdown ?

It's really seldom that a daemon manages to survive a shutdown ;)

But really, do you need to do something special on shutdown that the
usual TERM/KILL combo won't handle?

-- 
Patryk Zawadzki


From sparky at pld-linux.org  Tue Feb  1 21:43:37 2011
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Tue, 1 Feb 2011 21:43:37 +0100
Subject: rc-scripts-user
In-Reply-To: <AANLkTim6TBayHEfmfo+KC7_ve1CXC4Co2cobka5rneZ2@mail.gmail.com>
References: <20110201183014.GA4429@pld-linux.org>
	<AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>
	<20110201202723.GB7391@pld-linux.org>
	<AANLkTim6TBayHEfmfo+KC7_ve1CXC4Co2cobka5rneZ2@mail.gmail.com>
Message-ID: <20110201204337.GC7391@pld-linux.org>

On Tue, Feb 01, 2011 at 09:36:02PM +0100, Patryk Zawadzki wrote:
> On Tue, Feb 1, 2011 at 9:27 PM, Przemyslaw Iskra <sparky at pld-linux.org> wrote:
> > Does that work on system shutdown as well ? If not, how do I stop any of
> > my daemons on shutdown ?
> 
> It's really seldom that a daemon manages to survive a shutdown ;)
> 
> But really, do you need to do something special on shutdown that the
> usual TERM/KILL combo won't handle?

I'd like it do be killed when I still have working internet connection.

/quit Ragequit !

-- 
 ____    sparky  --  Przemyslaw  ................ LANG...Pl,Ca,Es,En
/____) ___  ___  _ _ || Iskra   : WWW . http://ppcrcd.pld-linux.org/
\____\| -_)'___| ||^'||//\\//   : WWW2 ............ http://rsget.pl/
(____/||   (_-_|_||  ||\\ ||    : eMail ..... <sparky at pld-linux.org>


From pawelz at pld-linux.org  Tue Feb  1 22:09:39 2011
From: pawelz at pld-linux.org (=?ISO-8859-2?Q?Pawe=B3_Zuzelski?=)
Date: Tue, 1 Feb 2011 22:09:39 +0100
Subject: rc-scripts-user
In-Reply-To: <20110201202723.GB7391@pld-linux.org>
References: <20110201183014.GA4429@pld-linux.org>
	<AANLkTi=MNLamo=OY+pTvt3N7Vuwi3mucMQTbSbUjYFuX@mail.gmail.com>
	<20110201202723.GB7391@pld-linux.org>
Message-ID: <AANLkTi=JEkPv2z-59QAt5qyHASwZ2Pb9hxYy27a9NdB3@mail.gmail.com>

On Tue, Feb 1, 2011 at 9:27 PM, Przemyslaw Iskra <sparky at pld-linux.org> wrote:
>> Why is it better than @reboot flag in user's crontab?
>
> Does that work on system shutdown as well ? If not, how do I stop any of
> my daemons on shutdown ?

AFAIK no. Good point.

-- 
Pawe?


From glen at pld-linux.org  Wed Feb  2 08:55:56 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Wed, 02 Feb 2011 09:55:56 +0200
Subject: rc-scripts-user
In-Reply-To: <alpine.LNX.2.00.1102011954500.28788@fgnss.gnfx.tqn.cy>
References: <20110201183014.GA4429@pld-linux.org>
	<alpine.LNX.2.00.1102011954500.28788@fgnss.gnfx.tqn.cy>
Message-ID: <4D490E0C.4050304@pld-linux.org>

On 01.02.2011 20:58, Pawel Golaszewski wrote:
> On Tue, 1 Feb 2011, Przemyslaw Iskra wrote:
>> rc-scripts-user / userscripts allows normal users to schedule and
>> execute tasks on system startup and shutdown. Please test and comment.
>> Once it gets on any production system it will be hard to change
>> interface without breaking user configuration.
>>
>> The interface is:
>>
>> * All scripts in ~/.config/init.d/ (hardcoded path, no xdg support)
>>   - any better ideas ?
> ~/.config/init/
symmetric to /etc/init then ~/.config/init would be scripts for upstart, 
so i'd say ~/.config/init.d is fine

xdg itself is simple, but loading user configuration to it, is 
difficult, so i'd say leave hardcoded .config, this can be changed later 
if somebody really interested :)
>> * all *~ and *.bak files are ignored
>>   - ignore something else ?
> .rpmnew, .rpmsave...
>
> There is few more temporary filetypes.
i suggest to process only extensions *.sh and *.init, whitelisting 
always works better than blacklisting.

-- 
glen



From wojciech at blaszkowski.com  Thu Feb  3 11:58:00 2011
From: wojciech at blaszkowski.com (Wojciech =?utf-8?q?B=C5=82aszkowski?=)
Date: Thu, 3 Feb 2011 11:58:00 +0100
Subject: INFO Th: new mysql 5.5
In-Reply-To: <201101311036.58317.arekm@maven.pl>
References: <201101311036.58317.arekm@maven.pl>
Message-ID: <201102031158.01071@wojtosz>

Dnia Monday 31 January 2011, Arkadiusz Miskiewicz napisa?:
> Hi,
> 
> New mysql 5.5 (.8 at this moment) is going to be included into Th. Be
> warned  (as upgrading to it requires proper procedure described on
> mysql.com documentation pages).

Some of us :) are still using PHP 5.2.X, and that makes the problem with MySQL 
5.5, like:

poldek:/all-avail> upgrade mysql-* --test
Przetwarzanie zale?no?ci...
mysql-libs-5.1.52-4.x86_64 zostanie zast?piony przez mysql-libs-5.5.8-2.x86_64
mysql-client-5.1.52-4.x86_64 zostanie zast?piony przez mysql-
client-5.5.8-2.x86_64
mysql-5.1.52-4.x86_64 zostanie zast?piony przez mysql-5.5.8-2.x86_64
b??d: mysql-5.5.8-2.x86_64: nie znaleziono wymaganego libaio.so.1()(64bit)
b??d: mysql-5.5.8-2.x86_64: nie znaleziono wymaganego libaio.so.1(LIBAIO_0.1)
(64bit)
b??d: mysql-5.5.8-2.x86_64: nie znaleziono wymaganego libaio.so.1(LIBAIO_0.4)
(64bit)
b??d: libmysqlclient_r.so.16()(64bit) is required by installed php-pdo-
mysql-5.2.17-1.x86_64
b??d: libmysqlclient_r.so.16(libmysqlclient_16)(64bit) is required by 
installed php-pdo-mysql-5.2.17-1.x86_64
b??d: libmysqlclient_r.so.16()(64bit) is required by installed php-
mysqli-5.2.17-1.x86_64
b??d: libmysqlclient_r.so.16(libmysqlclient_16)(64bit) is required by 
installed php-mysqli-5.2.17-1.x86_64
b??d: libmysqlclient_r.so.16()(64bit) is required by installed php-
mysql-5.2.17-1.x86_64
b??d: libmysqlclient_r.so.16(libmysqlclient_16)(64bit) is required by 
installed php-mysql-5.2.17-1.x86_64
mysql-charsets-5.1.52-4.x86_64 zostanie zast?piony przez mysql-
charsets-5.5.8-2.x86_64
There are 4 packages to install, 4 to remove:
I mysql-5.5.8-2.x86_64, mysql-charsets-5.5.8-2.x86_64, mysql-
client-5.5.8-2.x86_64, mysql-libs-5.5.8-2.x86_64
R mysql-5.1.52-4.x86_64, mysql-charsets-5.1.52-4.x86_64, mysql-
client-5.1.52-4.x86_64, mysql-libs-5.1.52-4.x86_64
This operation will use 27.2MB of disk space.
Potrzeba pobra? 7.1MB archiw?w (7.1MB do pobrania).

Can somebody rebuild PHP 5.2.17 according to above ?

-- 
Pozdrawiam, Best regards, Mit freundlichen Gr??en,

Wojciech B?aszkowski
www.blaszkowski.com
GSM: +48 600 197 207
JID: wojtosz at jabber.biz.pl


From marcin.rybak at gmail.com  Thu Feb  3 12:11:10 2011
From: marcin.rybak at gmail.com (Marcin Rybak)
Date: Thu, 3 Feb 2011 12:11:10 +0100
Subject: Future direction of official PLD website (Was: A couple questions
	about the website)
In-Reply-To: <AANLkTim=EkyytgOYVNmrv56-OScZ_Pyc3Jb=nd0RnCzk@mail.gmail.com>
References: <AANLkTim=EkyytgOYVNmrv56-OScZ_Pyc3Jb=nd0RnCzk@mail.gmail.com>
Message-ID: <AANLkTinU1HzUt6iTnf=a3ACFrjQGMJTAtCBhhrtEheOS@mail.gmail.com>

2010/11/20 Caleb Maclennan <caleb at pld-linux.org>

> 2010/11/11 Elan Ruusam?e <glen at pld-linux.org>:
>  > i'd finish the data migration and replace the moinmoun install
>

just to continue the idea of migration, someone done a nice migration script
http://www.dokuwiki.org/tips:moinmoin2doku
---
Marcin Rybak
http://marcinrybak.com


From qboosh at pld-linux.org  Thu Feb  3 18:10:29 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 3 Feb 2011 18:10:29 +0100
Subject: %cmake macro
In-Reply-To: <20110201185142.GA27352@camk.edu.pl>
References: <20110201185142.GA27352@camk.edu.pl>
Message-ID: <20110203171029.GB5518@stranger.qboosh.pl>

On Tue, Feb 01, 2011 at 07:51:42PM +0100, Kacper Kornet wrote:
> I think that current %cmake macro is broken. For example soprano.spec is build
> with options: -O2 ... -O3, which is effectively -O3. As far as I understand the
> preferred option would be -O2, 
> 
> In cmake the options are build form two variables as following as:
> 
> CMAKE_C_FLAGS CMAKE_C_FLAGS_<Buildtype>. 
> 
> while %cmake defines CMAKE_C_FLAGS and uses
> Builtype = RELEASE. 
> 
> So I think there are the following possibilities:
> 
> 
> 1. Define CMAKE_C_FLAGS_RELEASE in %cmake
> 
> Flaws:
> A package can define that variable internally in such a way that  it has
> precedence.  For example packages from kde4 do it.
> 
> 
> 2. Change the build type we use to None. Then only CMAKE_C_FLAGS will be used
> 
> Flaws:
> Same as above. But I don't know if any package redefines that variable
> unconditionally.

Some packages set build type to some value (mostly Release) if it's not
specified/None.

> Also if a package uses some special options in release flags, i.e. to avoid
> debug output, we would need to include them in out definition of CMAKE_C_FLAGS
> for this package.
> 
> 
> 3. Use or own build type like: cmake
> -DCMAKE_C_FLAGS_PLD=""${CFLAGS:-%{rpmcflags}}"

+ "-DNDEBUG" I think

> Flaws:
> If a package uses some special options in release flags, i.e. to avoid debug
> output, we would need to put them into our flags. 
> 
> 
> I can implement the third solution. But first I am waiting for any objections
> or better ideas.

Needs some testing probably.

BTW, I mentioned this cmake issue on -devel-? list some months ago.


-- 
Jakub Bogusz    http://qboosh.pl/


From kornet at camk.edu.pl  Thu Feb  3 18:55:48 2011
From: kornet at camk.edu.pl (Kacper Kornet)
Date: Thu, 3 Feb 2011 18:55:48 +0100
Subject: %cmake macro
In-Reply-To: <20110203171029.GB5518@stranger.qboosh.pl>
References: <20110201185142.GA27352@camk.edu.pl>
	<20110203171029.GB5518@stranger.qboosh.pl>
Message-ID: <20110203175548.GA24885@camk.edu.pl>

On Thu, Feb 03, 2011 at 06:10:29PM +0100, Jakub Bogusz wrote:
> On Tue, Feb 01, 2011 at 07:51:42PM +0100, Kacper Kornet wrote:
> > 2. Change the build type we use to None. Then only CMAKE_C_FLAGS will be used

> > Flaws:
> > Same as above. But I don't know if any package redefines that variable
> > unconditionally.

> Some packages set build type to some value (mostly Release) if it's not
> specified/None.

I have only seen packages with condition:

if(NOT CMAKE_BUILD_TYPE)
  set(CMAKE_BUILD_TYPE "Release")
endif()

> > 3. Use or own build type like: cmake
> > -DCMAKE_C_FLAGS_PLD=""${CFLAGS:-%{rpmcflags}}"

> + "-DNDEBUG" I think

Why not put it to %optflags?

Another problem is where to put -DNO_QT_DEBUG for builds with Qt.

> > Flaws:
> > If a package uses some special options in release flags, i.e. to avoid debug
> > output, we would need to put them into our flags. 


> > I can implement the third solution. But first I am waiting for any objections
> > or better ideas.

> Needs some testing probably.

So far I have found, that at least in one package there is as build type
specified in .spec file.

> BTW, I mentioned this cmake issue on -devel-? list some months ago.

I have checked in the archive. It seems that you have similar idea
earlier than me.

-- 
  Kacper Kornet


From qboosh at pld-linux.org  Fri Feb  4 19:20:37 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 4 Feb 2011 19:20:37 +0100
Subject: packages: allegro/allegro.spec - new cmake makro
In-Reply-To: <packages_allegro_allegro.spec.1.87@pld-linux.org>
References: <packages_allegro_allegro.spec.1.86@pld-linux.org>
	<packages_allegro_allegro.spec.1.87@pld-linux.org>
Message-ID: <20110204182037.GD17235@stranger.qboosh.pl>

On Fri, Feb 04, 2011 at 12:13:07AM +0100, draenog wrote:
> +BuildRequires:	rpmbuild(macros) >= 1.577

>  %cmake .. \
> -	-DCMAKE_BUILD_TYPE=%{!?debug:Release}%{?debug:Debug} \
> -	-DCMAKE_C_FLAGS_RELEASE="-DNDEBUG" \
> -	-DCMAKE_INSTALL_PREFIX=%{_prefix} \
> -	-DCMAKE_VERBOSE_MAKEFILE=1 \
> -	-DMANDIR=%{_mandir} \
>  	-DINFODIR=%{_infodir} \
>  	-DPLATFORM_LIBS=-ldl \
>  	-DWANT_LINUX_CONSOLE=1 \
>  	%{!?with_vga:-DWANT_LINUX_VGA=off} \
> -	%{!?with_svga:-DWANT_LINUX_SVGALIB=off} \
> -%if "%{_lib}" == "lib64"
> -	-DLIB_SUFFIX=64
> -%endif
> +	%{!?with_svga:-DWANT_LINUX_SVGALIB=off}

Omitting CMAKE_C_FLAGS_RELEASE requires macros >= 1.604


-- 
Jakub Bogusz    http://qboosh.pl/


From qboosh at pld-linux.org  Fri Feb  4 19:22:58 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 4 Feb 2011 19:22:58 +0100
Subject: %cmake macro
In-Reply-To: <20110203175548.GA24885@camk.edu.pl>
References: <20110201185142.GA27352@camk.edu.pl>
	<20110203171029.GB5518@stranger.qboosh.pl>
	<20110203175548.GA24885@camk.edu.pl>
Message-ID: <20110204182258.GC5518@stranger.qboosh.pl>

On Thu, Feb 03, 2011 at 06:55:48PM +0100, Kacper Kornet wrote:
> On Thu, Feb 03, 2011 at 06:10:29PM +0100, Jakub Bogusz wrote:
> > On Tue, Feb 01, 2011 at 07:51:42PM +0100, Kacper Kornet wrote:
[...
> > > 3. Use or own build type like: cmake
> > > -DCMAKE_C_FLAGS_PLD=""${CFLAGS:-%{rpmcflags}}"
> 
> > + "-DNDEBUG" I think
> 
> Why not put it to %optflags?

It's rather cmake-specific (i.e. specific to packages which use
cmake as buildsystem).
Autotools or other build systems generally don't set it.


-- 
Jakub Bogusz    http://qboosh.pl/


From kornet at camk.edu.pl  Fri Feb  4 19:45:38 2011
From: kornet at camk.edu.pl (Kacper Kornet)
Date: Fri, 4 Feb 2011 19:45:38 +0100
Subject: %cmake macro
In-Reply-To: <20110204182258.GC5518@stranger.qboosh.pl>
References: <20110201185142.GA27352@camk.edu.pl>
	<20110203171029.GB5518@stranger.qboosh.pl>
	<20110203175548.GA24885@camk.edu.pl>
	<20110204182258.GC5518@stranger.qboosh.pl>
Message-ID: <20110204184538.GA5590@camk.edu.pl>

On Fri, Feb 04, 2011 at 07:22:58PM +0100, Jakub Bogusz wrote:
> On Thu, Feb 03, 2011 at 06:55:48PM +0100, Kacper Kornet wrote:
> > On Thu, Feb 03, 2011 at 06:10:29PM +0100, Jakub Bogusz wrote:
> > > On Tue, Feb 01, 2011 at 07:51:42PM +0100, Kacper Kornet wrote:
> [...
> > > > 3. Use or own build type like: cmake
> > > > -DCMAKE_C_FLAGS_PLD=""${CFLAGS:-%{rpmcflags}}"

> > > + "-DNDEBUG" I think

> > Why not put it to %optflags?

> It's rather cmake-specific (i.e. specific to packages which use
> cmake as buildsystem).
> Autotools or other build systems generally don't set it.

I don't think it is specific only for packages that use cmake. Function
assert depends on it (see man 2 assert). For example audacious uses
assert, but not cmake. But some cmake based packages add additional meaning to DNDEBUG.

So the question is whether assert should have no effect only in cmake
based packages or in all programs.
-- 
  Kacper Kornet


From sparky at pld-linux.org  Fri Feb  4 19:59:26 2011
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Fri, 4 Feb 2011 19:59:26 +0100
Subject: %cmake macro
In-Reply-To: <20110204184538.GA5590@camk.edu.pl>
References: <20110201185142.GA27352@camk.edu.pl>
	<20110203171029.GB5518@stranger.qboosh.pl>
	<20110203175548.GA24885@camk.edu.pl>
	<20110204182258.GC5518@stranger.qboosh.pl>
	<20110204184538.GA5590@camk.edu.pl>
Message-ID: <20110204185926.GA14432@pld-linux.org>

On Fri, Feb 04, 2011 at 07:45:38PM +0100, Kacper Kornet wrote:

> So the question is whether assert should have no effect only in cmake
> based packages or in all programs.

Problem is that not everyone knows how to use assert() makro, so if
a package has not been prepared by the developers to work with -DNDEBUG
flag it may break.


-- 
 ____    sparky  --  Przemyslaw  ................ LANG...Pl,Ca,Es,En
/____) ___  ___  _ _ || Iskra   : WWW . http://ppcrcd.pld-linux.org/
\____\| -_)'___| ||^'||//\\//   : WWW2 ............ http://rsget.pl/
(____/||   (_-_|_||  ||\\ ||    : eMail ..... <sparky at pld-linux.org>


From kornet at camk.edu.pl  Fri Feb  4 20:19:10 2011
From: kornet at camk.edu.pl (Kacper Kornet)
Date: Fri, 4 Feb 2011 20:19:10 +0100
Subject: %cmake macro
In-Reply-To: <20110204185926.GA14432@pld-linux.org>
References: <20110201185142.GA27352@camk.edu.pl>
	<20110203171029.GB5518@stranger.qboosh.pl>
	<20110203175548.GA24885@camk.edu.pl>
	<20110204182258.GC5518@stranger.qboosh.pl>
	<20110204184538.GA5590@camk.edu.pl>
	<20110204185926.GA14432@pld-linux.org>
Message-ID: <20110204191910.GB5590@camk.edu.pl>

On Fri, Feb 04, 2011 at 07:59:26PM +0100, Przemyslaw Iskra wrote:
> On Fri, Feb 04, 2011 at 07:45:38PM +0100, Kacper Kornet wrote:

> > So the question is whether assert should have no effect only in cmake
> > based packages or in all programs.

> Problem is that not everyone knows how to use assert() makro, so if
> a package has not been prepared by the developers to work with -DNDEBUG
> flag it may break.

OK. So I will add DNDEBUG only to cmake makro.
-- 
  Kacper Kornet


From qboosh at pld-linux.org  Sat Feb  5 13:14:11 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sat, 5 Feb 2011 13:14:11 +0100
Subject: packages: glib2/glib2.spec - BR: fam-devel -> gamin-devel
In-Reply-To: <packages_glib2_glib2.spec.1.230@pld-linux.org>
References: <packages_glib2_glib2.spec.1.229@pld-linux.org>
	<packages_glib2_glib2.spec.1.230@pld-linux.org>
Message-ID: <20110205121411.GE17235@stranger.qboosh.pl>

On Sat, Feb 05, 2011 at 11:29:23AM +0100, areq wrote:
> Author: areq                         Date: Sat Feb  5 10:29:23 2011 GMT
> Module: packages                      Tag: HEAD
> ---- Log message:
> - BR: fam-devel -> gamin-devel

It links to libfam not libgamin, so I'd say fam-devel is a proper
virtual name here.


-- 
Jakub Bogusz    http://qboosh.pl/


From udvzsolt at gmail.com  Wed Feb  9 21:36:58 2011
From: udvzsolt at gmail.com (Zsolt Udvari)
Date: Wed, 9 Feb 2011 21:36:58 +0100
Subject: MySQL start
Message-ID: <AANLkTi=2UNMn1zixQDHrOD71aQe+sy6f0-AYRsbBt1ab@mail.gmail.com>

Hi all!

# service mysql restart
Stopping MySQL /var/lib/mysql
service...................................................................................................................[
DONE ]
Starting MySQL /var/lib/mysql
service...................................................................................................................[
BUSY ]
110209 21:34:40 [Note] Flashcache bypass: disabled
110209 21:34:40 [Note] Flashcache setup error is : ioctl failed


Waiting for MySQL to
start..............................................................................................................................[
DONE ]


What's wrong? All file is the shipped by mysql package.

Zsolt


From arekm at maven.pl  Wed Feb  9 22:30:02 2011
From: arekm at maven.pl (Arkadiusz Miskiewicz)
Date: Wed, 9 Feb 2011 22:30:02 +0100
Subject: MySQL start
In-Reply-To: <AANLkTi=2UNMn1zixQDHrOD71aQe+sy6f0-AYRsbBt1ab@mail.gmail.com>
References: <AANLkTi=2UNMn1zixQDHrOD71aQe+sy6f0-AYRsbBt1ab@mail.gmail.com>
Message-ID: <201102092230.02220.arekm@maven.pl>

On Wednesday 09 of February 2011, Zsolt Udvari wrote:
> Hi all!
> 
> # service mysql restart
> Stopping MySQL /var/lib/mysql
> service....................................................................
> ...............................................[ DONE ]
> Starting MySQL /var/lib/mysql
> service....................................................................
> ...............................................[ BUSY ]
> 110209 21:34:40 [Note] Flashcache bypass: disabled
> 110209 21:34:40 [Note] Flashcache setup error is : ioctl failed
> 
> 
> Waiting for MySQL to
> start......................................................................
> ........................................................[ DONE ]
> 
> 
> What's wrong? All file is the shipped by mysql package.

Nothing is wrong. mysqld is too loud. flashcache is for ssd cache.

We need to figure out how to silence this. Likely by disabling flashcache by 
default via config.

> Zsolt

-- 
Arkadiusz Mi?kiewicz        PLD/Linux Team
arekm / maven.pl            http://ftp.pld-linux.org/


From glen at pld-linux.org  Fri Feb 11 12:01:26 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Fri, 11 Feb 2011 13:01:26 +0200
Subject: packages: perl-version/perl-version.spec - up to 0.88 - build
	using	Makefil...
In-Reply-To: <packages_perl_version_perl_version.spec.1.24@pld-linux.org>
References: <packages_perl_version_perl_version.spec.1.23@pld-linux.org>
	<packages_perl_version_perl_version.spec.1.24@pld-linux.org>
Message-ID: <4D551706.9000404@pld-linux.org>

On 11.02.2011 12:54, lisu wrote:
> - build using Makefile.pl instead of Build.pl
> - more verbose man files
and why is that?

so far i've understood that Makefile.PL is deprecated method (hacks to 
wrap over make(1) where Build.PL is more native (pure Perl and more 
cross platform meaning less hacks)

-- 
glen



From lisu87 at gmail.com  Fri Feb 11 12:11:36 2011
From: lisu87 at gmail.com (=?UTF-8?B?TWljaGHFgiBMaXNvd3NraQ==?=)
Date: Fri, 11 Feb 2011 12:11:36 +0100
Subject: packages: perl-version/perl-version.spec - up to 0.88 - build
	using	Makefil...
In-Reply-To: <4D551706.9000404@pld-linux.org>
References: <packages_perl_version_perl_version.spec.1.23@pld-linux.org>
	<packages_perl_version_perl_version.spec.1.24@pld-linux.org>
	<4D551706.9000404@pld-linux.org>
Message-ID: <4D551968.3080405@gmail.com>

W dniu 11.02.2011 12:01, Elan Ruusam?e pisze:
> On 11.02.2011 12:54, lisu wrote:
>> - build using Makefile.pl instead of Build.pl
>> - more verbose man files
> and why is that?
>
> so far i've understood that Makefile.PL is deprecated method (hacks to 
> wrap over make(1) where Build.PL is more native (pure Perl and more 
> cross platform meaning less hacks)
>
There is no Build.PL script anymore in some perl sources. So I'm very 
sorry, but I was not able to use Build.PL.


From glen at pld-linux.org  Fri Feb 11 12:58:18 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Fri, 11 Feb 2011 13:58:18 +0200
Subject: packages: perl-version/perl-version.spec - up to 0.88 - build
	using	Makefil...
In-Reply-To: <4D551968.3080405@gmail.com>
References: <packages_perl_version_perl_version.spec.1.23@pld-linux.org>	<packages_perl_version_perl_version.spec.1.24@pld-linux.org>	<4D551706.9000404@pld-linux.org>
	<4D551968.3080405@gmail.com>
Message-ID: <4D55245A.7000102@pld-linux.org>

On 11.02.2011 13:11, Micha? Lisowski wrote:
> W dniu 11.02.2011 12:01, Elan Ruusam?e pisze:
>> On 11.02.2011 12:54, lisu wrote:
>>> - build using Makefile.pl instead of Build.pl
>>> - more verbose man files
>> and why is that?
>>
>> so far i've understood that Makefile.PL is deprecated method (hacks 
>> to wrap over make(1) where Build.PL is more native (pure Perl and 
>> more cross platform meaning less hacks)
>>
> There is no Build.PL script anymore in some perl sources. So I'm very 
> sorry, but I was not able to use Build.PL.

then you commit message should had been less misleading :)

"build using Makefile.PL as Build.PL is gone"

would had been better :P

-- 
glen



From qboosh at pld-linux.org  Fri Feb 11 15:51:17 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 11 Feb 2011 15:51:17 +0100
Subject: packages: perl-version/perl-version.spec - up to 0.88 - build
	using Makefil...
In-Reply-To: <4D551706.9000404@pld-linux.org>
References: <packages_perl_version_perl_version.spec.1.23@pld-linux.org>
	<packages_perl_version_perl_version.spec.1.24@pld-linux.org>
	<4D551706.9000404@pld-linux.org>
Message-ID: <20110211145117.GA908@stranger.qboosh.pl>

On Fri, Feb 11, 2011 at 01:01:26PM +0200, Elan Ruusam?e wrote:
> On 11.02.2011 12:54, lisu wrote:
> > - build using Makefile.pl instead of Build.pl
> > - more verbose man files
> and why is that?
> 
> so far i've understood that Makefile.PL is deprecated method (hacks to 
> wrap over make(1) where Build.PL is more native (pure Perl and more 
> cross platform meaning less hacks)

These are two most commonly used build systems for Perl modules:
1) ExtUtils::MakeMaker (which uses Makefile.PL + make)
2) Module::Build (which uses Build.PL, in this case Makefile.PL is just
  a wrapper)

In second case you're right that calling Build.PL directly is preferred,
but in the first case there is no alternative.


-- 
Jakub Bogusz    http://qboosh.pl/


From lisu87 at gmail.com  Tue Feb 15 09:59:23 2011
From: lisu87 at gmail.com (=?UTF-8?B?TWljaGHFgiBMaXNvd3NraQ==?=)
Date: Tue, 15 Feb 2011 09:59:23 +0100
Subject: wormux -> warmux
Message-ID: <4D5A406B.40008@gmail.com>

cp please


From lisu87 at gmail.com  Tue Feb 15 10:26:53 2011
From: lisu87 at gmail.com (=?UTF-8?B?TWljaGHFgiBMaXNvd3NraQ==?=)
Date: Tue, 15 Feb 2011 10:26:53 +0100
Subject: wormux-bonusmaps -> warmux-bonusmaps
Message-ID: <4D5A46DD.5090005@gmail.com>

cp please.


From glen at pld-linux.org  Tue Feb 15 11:17:14 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Tue, 15 Feb 2011 12:17:14 +0200
Subject: wormux -> warmux
In-Reply-To: <4D5A406B.40008@gmail.com>
References: <4D5A406B.40008@gmail.com>
Message-ID: <4D5A52AA.70406@pld-linux.org>

On 15.02.2011 10:59, Micha? Lisowski wrote:
> cp please

Re: wormux -> warmux
Re: wormux-bonusmaps -> warmux-bonusmaps

copied dirs, renamed .spec and patches inside

-- 
glen



From lisu87 at gmail.com  Tue Feb 15 11:30:32 2011
From: lisu87 at gmail.com (=?UTF-8?B?TWljaGHFgiBMaXNvd3NraQ==?=)
Date: Tue, 15 Feb 2011 11:30:32 +0100
Subject: wormux -> warmux
In-Reply-To: <4D5A52AA.70406@pld-linux.org>
References: <4D5A406B.40008@gmail.com> <4D5A52AA.70406@pld-linux.org>
Message-ID: <4D5A55C8.5000603@gmail.com>

W dniu 15.02.2011 11:17, Elan Ruusam?e pisze:
> On 15.02.2011 10:59, Micha? Lisowski wrote:
>> cp please
>
> Re: wormux -> warmux
> Re: wormux-bonusmaps -> warmux-bonusmaps
>
> copied dirs, renamed .spec and patches inside
>
Thank you


From megabajt at pld-linux.org  Tue Feb 15 16:15:37 2011
From: megabajt at pld-linux.org (Marcin Banasiak)
Date: Tue, 15 Feb 2011 16:15:37 +0100
Subject: cvs mv: Thunar-vfs -> thunar-vfs
Message-ID: <AANLkTikQNsqUhj0GzLOThQ+tizbCC9Gsw7MOpj8HTA7w@mail.gmail.com>

Please rename Thunar-vfs to thunar-vfs. Thanks.

-- 
Marcin Banasiak


From glen at pld-linux.org  Tue Feb 15 19:12:53 2011
From: glen at pld-linux.org (=?ISO-8859-1?Q?Elan_Ruusam=E4e?=)
Date: Tue, 15 Feb 2011 20:12:53 +0200
Subject: cvs mv: Thunar-vfs -> thunar-vfs
In-Reply-To: <AANLkTikQNsqUhj0GzLOThQ+tizbCC9Gsw7MOpj8HTA7w@mail.gmail.com>
References: <AANLkTikQNsqUhj0GzLOThQ+tizbCC9Gsw7MOpj8HTA7w@mail.gmail.com>
Message-ID: <4D5AC225.7070706@pld-linux.org>

On 15.02.2011 17:15, Marcin Banasiak wrote:
> Please rename Thunar-vfs to thunar-vfs. Thanks.
done

-- 
glen



From glen at pld-linux.org  Sat Feb 19 16:55:20 2011
From: glen at pld-linux.org (=?ISO-8859-1?Q?Elan_Ruusam=E4e?=)
Date: Sat, 19 Feb 2011 17:55:20 +0200
Subject: package %post service restart
Message-ID: <4D5FE7E8.30805@pld-linux.org>

hi

i have an idea to replace current '%service foo restart' expansion with
'/sbin/service foo restart' from rpm-build-macros

goal is to support restart action from upstart, as it does not really
confirm with /var/lock/subsys/foo lockfiles detecting is service
"running" or "up"

additionally it opens ways to support other init systems, like systemd

simpliest it would be to expand %service as /sbin/service and replace
"restart" action with "try-restart" (perhaps in macro not in .spec), so
the behaviour is the same, if service is running, it is restarted. this
has the downside that message printed "run \"/sbin/service foo start\"
to start foo" is not presented to end user.

this could be workarounded perhaps by adding new custom action (or
option) that emulates this message printing. good side is that it maybe
localized if the gettext messages get translated.

also for migaration time would fallback to current behaviour if it is
found that initscript doesn't support "try-restart" action

altho i haven't tought out other versions, i'd like still hear some
expressions. as the currently described one sounds like way to do it.
i'll start implementing it meanwhile

-- 
glen


From qboosh at pld-linux.org  Sat Feb 19 21:40:15 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sat, 19 Feb 2011 21:40:15 +0100
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <packages_pam_pam.spec.1.310@pld-linux.org>
References: <packages_pam_pam.spec.1.309@pld-linux.org>
	<packages_pam_pam.spec.1.310@pld-linux.org>
Message-ID: <20110219204015.GA17732@stranger.qboosh.pl>

On Fri, Feb 18, 2011 at 04:57:18PM +0100, glen wrote:
> Author: glen                         Date: Fri Feb 18 15:57:18 2011 GMT
> Module: packages                      Tag: HEAD
> ---- Log message:
> - lang tag trigrams (old change)

Are these files used/selected depending on user locale?


> ================================================================
> Index: packages/pam/pam.spec
> diff -u packages/pam/pam.spec:1.309 packages/pam/pam.spec:1.310
> --- packages/pam/pam.spec:1.309	Mon Jan 24 02:55:59 2011
> +++ packages/pam/pam.spec	Fri Feb 18 16:57:12 2011
> @@ -417,8 +417,16 @@
>  %attr(755,root,root) %config(noreplace) %verify(not md5 mtime size) /etc/security/namespace.init
>  %config(noreplace) %verify(not md5 mtime size) /etc/security/pam_env.conf
>  %config(noreplace) %verify(not md5 mtime size) /etc/security/time.conf
> -%config(noreplace) %verify(not md5 mtime size) /etc/security/trigram*
> -%config /etc/security/console.perms.d/50-default.perms
> +%config(noreplace) %verify(not md5 mtime size) /etc/security/trigram
> +%config(noreplace) %verify(not md5 mtime size) /etc/security/trigram.en
> +%config(noreplace) %verify(not md5 mtime size) %lang(de) /etc/security/trigram.de
> +%config(noreplace) %verify(not md5 mtime size) %lang(es) /etc/security/trigram.es
> +%config(noreplace) %verify(not md5 mtime size) %lang(fi) /etc/security/trigram.fi
> +%config(noreplace) %verify(not md5 mtime size) %lang(it) /etc/security/trigram.it
> +%config(noreplace) %verify(not md5 mtime size) %lang(ja) /etc/security/trigram.ja
> +%config(noreplace) %verify(not md5 mtime size) %lang(nb) /etc/security/trigram.no
> +%config(noreplace) %verify(not md5 mtime size) %lang(pl) /etc/security/trigram.pl
> +%config(noreplace) %verify(not md5 mtime size) /etc/security/console.perms.d/50-default.perms
>  %attr(600,root,root) %config(noreplace) %verify(not md5 mtime size) /etc/security/opasswd
>  %attr(755,root,root) %{_bindir}/pam_pwgen
>  %attr(755,root,root) %{_sbindir}/mkhomedir_helper
> @@ -539,6 +547,9 @@
>  All persons listed below can be reached at <cvs_login>@pld-linux.org
>  
>  $Log$
> +Revision 1.310  2011/02/18 15:57:12  glen
> +- lang tag trigrams (old change)
> +
>  Revision 1.309  2011/01/24 01:55:59  sparky
>  - BR: glibc-misc
>  
> ================================================================
> 
> ---- CVS-web:
>     http://cvs.pld-linux.org/cgi-bin/cvsweb.cgi/packages/pam/pam.spec?r1=1.309&r2=1.310&f=u
> 
> _______________________________________________
> pld-cvs-commit mailing list
> pld-cvs-commit at lists.pld-linux.org
> http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit

-- 
Jakub Bogusz    http://qboosh.pl/


From qboosh at pld-linux.org  Sun Feb 20 08:41:19 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sun, 20 Feb 2011 08:41:19 +0100
Subject: ur_PK locale [Re: packages:
	xfce4-notes-plugin/xfce4-notes-plugin.spec - merged from DEVEL]
In-Reply-To: <packages_xfce4_notes_plugin_xfce4_notes_plugin.spec.1.19@pld-linux.org>
References: <packages_xfce4_notes_plugin_xfce4_notes_plugin.spec.1.18@pld-linux.org>
	<packages_xfce4_notes_plugin_xfce4_notes_plugin.spec.1.19@pld-linux.org>
Message-ID: <20110220074118.GB17732@stranger.qboosh.pl>

On Wed, Feb 16, 2011 at 11:02:03PM +0100, megabajt wrote:
> Author: megabajt                     Date: Wed Feb 16 22:02:03 2011 GMT
> Module: packages                      Tag: HEAD
> ---- Log message:
> - merged from DEVEL

> -rm -r $RPM_BUILD_ROOT%{_datadir}/locale/ur_PK
> +%{__rm} -r $RPM_BUILD_ROOT%{_datadir}/locale/ur_PK

ur_PK locale exists, but its directory is /usr/share/locale/ur
(see glibc.spec and supported locales)

So use mv instead of rm.


-- 
Jakub Bogusz    http://qboosh.pl/


From qboosh at pld-linux.org  Sun Feb 20 08:54:03 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sun, 20 Feb 2011 08:54:03 +0100
Subject: ur_PK locale [Re: packages:
	xfce4-notes-plugin/xfce4-notes-plugin.spec - merged from DEVEL]
In-Reply-To: <20110220074118.GB17732@stranger.qboosh.pl>
References: <packages_xfce4_notes_plugin_xfce4_notes_plugin.spec.1.18@pld-linux.org>
	<packages_xfce4_notes_plugin_xfce4_notes_plugin.spec.1.19@pld-linux.org>
	<20110220074118.GB17732@stranger.qboosh.pl>
Message-ID: <20110220075403.GB17727@stranger.qboosh.pl>

On Sun, Feb 20, 2011 at 08:41:19AM +0100, Jakub Bogusz wrote:
> On Wed, Feb 16, 2011 at 11:02:03PM +0100, megabajt wrote:
> > Author: megabajt                     Date: Wed Feb 16 22:02:03 2011 GMT
> > Module: packages                      Tag: HEAD
> > ---- Log message:
> > - merged from DEVEL
> 
> > -rm -r $RPM_BUILD_ROOT%{_datadir}/locale/ur_PK
> > +%{__rm} -r $RPM_BUILD_ROOT%{_datadir}/locale/ur_PK
> 
> ur_PK locale exists, but its directory is /usr/share/locale/ur
> (see glibc.spec and supported locales)
> 
> So use mv instead of rm.

And similarly, tl_PH is supported - its dir is "tl" (in HEAD glibc).

Are there any more locales killed not already documented in glibc.spec?

If there appears some new locale with no directory, please consult
glibc.spec and eventually update it. Or ask on the list if you are not
sure.


-- 
Jakub Bogusz    http://qboosh.pl/


From baggins at sith.mimuw.edu.pl  Sun Feb 20 09:57:43 2011
From: baggins at sith.mimuw.edu.pl (Jan =?utf-8?Q?R=C4=99korajski?=)
Date: Sun, 20 Feb 2011 09:57:43 +0100
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <20110219204015.GA17732@stranger.qboosh.pl>
References: <packages_pam_pam.spec.1.309@pld-linux.org>
	<packages_pam_pam.spec.1.310@pld-linux.org>
	<20110219204015.GA17732@stranger.qboosh.pl>
Message-ID: <20110220085743.GA28814@sith.mimuw.edu.pl>

On Sat, 19 Feb 2011, Jakub Bogusz wrote:

> On Fri, Feb 18, 2011 at 04:57:18PM +0100, glen wrote:
> > Author: glen                         Date: Fri Feb 18 15:57:18 2011 GMT
> > Module: packages                      Tag: HEAD
> > ---- Log message:
> > - lang tag trigrams (old change)
> 
> Are these files used/selected depending on user locale?

No, and glen - please revert it, what if someone wants to generate a
foreign language sounding password, but don't want to install all locale stuff?

-- 
Jan R?korajski            |  ALL SUSPECTS ARE GUILTY. PERIOD!
baggins<at>mimuw.edu.pl   |  OTHERWISE THEY WOULDN'T BE SUSPECTS, WOULD THEY?
BOFH, MANIAC              |                   -- TROOPS by Kevin Rubio


From arekm at maven.pl  Mon Feb 21 13:22:57 2011
From: arekm at maven.pl (Arkadiusz Miskiewicz)
Date: Mon, 21 Feb 2011 13:22:57 +0100
Subject: INFO Th: new mysql 5.5 [upgrade info]
In-Reply-To: <201101311036.58317.arekm@maven.pl>
References: <201101311036.58317.arekm@maven.pl>
Message-ID: <201102211322.57784.arekm@maven.pl>

On Monday 31 of January 2011, Arkadiusz Miskiewicz wrote:
> Hi,
> 
> New mysql 5.5 (.8 at this moment) is going to be included into Th. Be
> warned (as upgrading to it requires proper procedure described on
> mysql.com documentation pages).

I'm running 5.5 on few hosts (including one with 30GB database) and there is 
no problem.

Remember that if you want to upgrade without dump & restore you have to 
upgrade from series to next series for example:

4.1 -> 5.0; run mysql_upgrade script
5.0 -> 5.1; run mysql_upgrade script
5.1 -> 5.5; run mysql_upgrade script

You cannot upgrade directly from 5.0 to 5.5 for example.

5.5 will land in main this week.

-- 
Arkadiusz Mi?kiewicz        PLD/Linux Team
arekm / maven.pl            http://ftp.pld-linux.org/


From wojciech at blaszkowski.com  Mon Feb 21 13:59:09 2011
From: wojciech at blaszkowski.com (Wojciech =?iso-8859-2?q?B=B3aszkowski?=)
Date: Mon, 21 Feb 2011 13:59:09 +0100
Subject: INFO Th: new mysql 5.5 [upgrade info]
In-Reply-To: <201102211322.57784.arekm@maven.pl>
References: <201101311036.58317.arekm@maven.pl>
	<201102211322.57784.arekm@maven.pl>
Message-ID: <201102211359.09519@wojtosz>

Dnia Monday 21 of February 2011, Arkadiusz Miskiewicz napisa?:
> > New mysql 5.5 (.8 at this moment) is going to be included into Th. Be
> > warned (as upgrading to it requires proper procedure described on
> > mysql.com documentation pages).
> 
> I'm running 5.5 on few hosts (including one with 30GB database) and there
> is  no problem.
> 
> Remember that if you want to upgrade without dump & restore you have to 
> upgrade from series to next series for example:
> 
> 4.1 -> 5.0; run mysql_upgrade script
> 5.0 -> 5.1; run mysql_upgrade script
> 5.1 -> 5.5; run mysql_upgrade script
> 
> You cannot upgrade directly from 5.0 to 5.5 for example.
> 
> 5.5 will land in main this week.

Again - no chance to see PHP 5.2 working with MySQ 5.5 on FTP obsolete 
directory ?

-- 
Pozdrawiam, Best regards, Mit freundlichen Gr??en,

Wojciech B?aszkowski
www.blaszkowski.com
GSM: +48 600 197 207
JID: wojtosz at jabber.biz.pl


From glen at pld-linux.org  Wed Feb 23 07:17:31 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Wed, 23 Feb 2011 08:17:31 +0200
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <20110220085743.GA28814@sith.mimuw.edu.pl>
References: <packages_pam_pam.spec.1.309@pld-linux.org>	<packages_pam_pam.spec.1.310@pld-linux.org>	<20110219204015.GA17732@stranger.qboosh.pl>
	<20110220085743.GA28814@sith.mimuw.edu.pl>
Message-ID: <4D64A67B.8040409@pld-linux.org>

On 20/02/11 10:57, Jan R?korajski wrote:
> On Sat, 19 Feb 2011, Jakub Bogusz wrote:
> 
>> On Fri, Feb 18, 2011 at 04:57:18PM +0100, glen wrote:
>>> Author: glen                         Date: Fri Feb 18 15:57:18 2011 GMT
>>> Module: packages                      Tag: HEAD
>>> ---- Log message:
>>> - lang tag trigrams (old change)
>>
>> Are these files used/selected depending on user locale?
> 
> No, and glen - please revert it, what if someone wants to generate a
> foreign language sounding password, but don't want to install all locale stuff?

i too want time to time enable some %_install_langs to particular
package only, mostly web apps. i don't care russian manual pages in
system, but i care if webmail app can allow user choosing russian language.

here rather it be rpm that had possibility to "stick" some
%_install_lang preference by package.

so should this pam package be an exception then?

-- 
glen


From baggins at sith.mimuw.edu.pl  Wed Feb 23 11:01:31 2011
From: baggins at sith.mimuw.edu.pl (Jan =?utf-8?Q?R=C4=99korajski?=)
Date: Wed, 23 Feb 2011 11:01:31 +0100
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <4D64A67B.8040409@pld-linux.org>
References: <packages_pam_pam.spec.1.309@pld-linux.org>
	<packages_pam_pam.spec.1.310@pld-linux.org>
	<20110219204015.GA17732@stranger.qboosh.pl>
	<20110220085743.GA28814@sith.mimuw.edu.pl>
	<4D64A67B.8040409@pld-linux.org>
Message-ID: <20110223100131.GF12774@sith.mimuw.edu.pl>

On Wed, 23 Feb 2011, Elan Ruusam?e wrote:

> On 20/02/11 10:57, Jan R?korajski wrote:
> > On Sat, 19 Feb 2011, Jakub Bogusz wrote:
> > 
> >> On Fri, Feb 18, 2011 at 04:57:18PM +0100, glen wrote:
> >>> Author: glen                         Date: Fri Feb 18 15:57:18 2011 GMT
> >>> Module: packages                      Tag: HEAD
> >>> ---- Log message:
> >>> - lang tag trigrams (old change)
> >>
> >> Are these files used/selected depending on user locale?
> > 
> > No, and glen - please revert it, what if someone wants to generate a
> > foreign language sounding password, but don't want to install all locale stuff?
> 
> i too want time to time enable some %_install_langs to particular
> package only, mostly web apps. i don't care russian manual pages in
> system, but i care if webmail app can allow user choosing russian language.
> 
> here rather it be rpm that had possibility to "stick" some
> %_install_lang preference by package.
> 
> so should this pam package be an exception then?

Yes, as, IMO, should web apps you mentioned. It's better to waste some
space than constantly fool around with %_install_langs.

-- 
Jan R?korajski            |  ALL SUSPECTS ARE GUILTY. PERIOD!
baggins<at>mimuw.edu.pl   |  OTHERWISE THEY WOULDN'T BE SUSPECTS, WOULD THEY?
BOFH, MANIAC              |                   -- TROOPS by Kevin Rubio


From gotar at polanet.pl  Wed Feb 23 12:25:38 2011
From: gotar at polanet.pl (Tomasz Pala)
Date: Wed, 23 Feb 2011 12:25:38 +0100
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <4D64A67B.8040409@pld-linux.org>
References: <packages_pam_pam.spec.1.309@pld-linux.org>
	<packages_pam_pam.spec.1.310@pld-linux.org>
	<20110219204015.GA17732@stranger.qboosh.pl>
	<20110220085743.GA28814@sith.mimuw.edu.pl>
	<4D64A67B.8040409@pld-linux.org>
Message-ID: <20110223112538.GA7488@polanet.pl>

On Wed, Feb 23, 2011 at 08:17:31 +0200, Elan Ruusam?e wrote:

>>> Are these files used/selected depending on user locale?
>> 
>> No, and glen - please revert it, what if someone wants to generate a
>> foreign language sounding password, but don't want to install all locale stuff?
> 
> i too want time to time enable some %_install_langs to particular
> package only, mostly web apps. i don't care russian manual pages in
> system, but i care if webmail app can allow user choosing russian language.

Sure. Which webapp has such files tagged? Out of the ones I've got
installed I see wordpress and squirrelmail (but maybe old versions) only
- it should be removed.

> so should this pam package be an exception then?

The same exception as phpMyAdmin, phpPgAdmin, coppermine-gallery,
mediawiki etc.

-- 
Tomasz Pala <gotar at pld-linux.org>


From glen at pld-linux.org  Wed Feb 23 16:50:01 2011
From: glen at pld-linux.org (=?UTF-8?B?RWxhbiBSdXVzYW3DpGU=?=)
Date: Wed, 23 Feb 2011 17:50:01 +0200
Subject: packages: pam/pam.spec - lang tag trigrams (old change)
In-Reply-To: <20110223100131.GF12774@sith.mimuw.edu.pl>
References: <packages_pam_pam.spec.1.309@pld-linux.org>	<packages_pam_pam.spec.1.310@pld-linux.org>	<20110219204015.GA17732@stranger.qboosh.pl>	<20110220085743.GA28814@sith.mimuw.edu.pl>	<4D64A67B.8040409@pld-linux.org>
	<20110223100131.GF12774@sith.mimuw.edu.pl>
Message-ID: <4D652CA9.4040908@pld-linux.org>

On 23.02.2011 12:01, Jan R?korajski wrote:
> On Wed, 23 Feb 2011, Elan Ruusam?e wrote:
>
>> On 20/02/11 10:57, Jan R?korajski wrote:
>>> On Sat, 19 Feb 2011, Jakub Bogusz wrote:
>>>
>>>> On Fri, Feb 18, 2011 at 04:57:18PM +0100, glen wrote:
>>>>> Author: glen                         Date: Fri Feb 18 15:57:18 2011 GMT
>>>>> Module: packages                      Tag: HEAD
>>>>> ---- Log message:
>>>>> - lang tag trigrams (old change)
>>>> Are these files used/selected depending on user locale?
>>> No, and glen - please revert it, what if someone wants to generate a
>>> foreign language sounding password, but don't want to install all locale stuff?
>> i too want time to time enable some %_install_langs to particular
>> package only, mostly web apps. i don't care russian manual pages in
>> system, but i care if webmail app can allow user choosing russian language.
>>
>> here rather it be rpm that had possibility to "stick" some
>> %_install_lang preference by package.
>>
>> so should this pam package be an exception then?
> Yes, as, IMO, should web apps you mentioned. It's better to waste some
> space than constantly fool around with %_install_langs.
nah, i just stick with extra noise, and rm -rf /usr/share/man if disk 
space is an issue.

-- 
glen



From lisu87 at gmail.com  Wed Feb 23 18:25:05 2011
From: lisu87 at gmail.com (=?UTF-8?B?TWljaGHFgiBMaXNvd3NraQ==?=)
Date: Wed, 23 Feb 2011 18:25:05 +0100
Subject: packages: mixxx/mixxx.spec - skins-core subpackage - remove patch0
	- remove...
In-Reply-To: <packages_mixxx_mixxx.spec.1.18@pld-linux.org>
References: <packages_mixxx_mixxx.spec.1.17@pld-linux.org>
	<packages_mixxx_mixxx.spec.1.18@pld-linux.org>
Message-ID: <4D6542F1.3000506@gmail.com>

W dniu 23.02.2011 13:06, uzsolt pisze:
> Author: uzsolt                       Date: Wed Feb 23 12:06:11 2011 GMT
> Module: packages                      Tag: HEAD
> ---- Log message:
> - skins-core subpackage
> - remove patch0
Ok, and now remove patch file from cvs.


From udvzsolt at gmail.com  Wed Feb 23 21:19:36 2011
From: udvzsolt at gmail.com (Zsolt Udvari)
Date: Wed, 23 Feb 2011 21:19:36 +0100
Subject: packages: mixxx/mixxx.spec - skins-core subpackage - remove
	patch0 - remove...
In-Reply-To: <4D6542F1.3000506@gmail.com>
References: <packages_mixxx_mixxx.spec.1.17@pld-linux.org>
	<packages_mixxx_mixxx.spec.1.18@pld-linux.org>
	<4D6542F1.3000506@gmail.com>
Message-ID: <AANLkTimSD5nv66RZyKL13wFiBWSXdgtBCgqy5FOBx02v@mail.gmail.com>

Done.

2011/2/23 Micha? Lisowski <lisu87 at gmail.com>

> W dniu 23.02.2011 13:06, uzsolt pisze:
>
>> Author: uzsolt                       Date: Wed Feb 23 12:06:11 2011 GMT
>> Module: packages                      Tag: HEAD
>> ---- Log message:
>> - skins-core subpackage
>> - remove patch0
>>
> Ok, and now remove patch file from cvs.
> _______________________________________________
> pld-devel-en mailing list
> pld-devel-en at lists.pld-linux.org
> http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
>


From qboosh at pld-linux.org  Thu Feb 24 07:43:47 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 24 Feb 2011 07:43:47 +0100
Subject: packages: OGLFT/OGLFT.spec - include Qt support, remove .la,
	rel. 2
In-Reply-To: <packages_OGLFT_OGLFT.spec.1.2@pld-linux.org>
References: <packages_OGLFT_OGLFT.spec.1.1@pld-linux.org>
	<packages_OGLFT_OGLFT.spec.1.2@pld-linux.org>
Message-ID: <20110224064347.GA4157@mail>

On Wed, Feb 23, 2011 at 11:38:44PM +0100, gotar wrote:
> Author: gotar                        Date: Wed Feb 23 22:38:44 2011 GMT
> Module: packages                      Tag: HEAD
> ---- Log message:
> - include Qt support, remove .la, rel. 2

> @@ -69,7 +75,6 @@
>  %doc doc/html
>  %{_includedir}/oglft
>  %attr(755,root,root) %{_libdir}/libOGLFT.so
> -%{_libdir}/libOGLFT.la
>  
>  %files static
>  %defattr(644,root,root,755)

Without pkgconfig file (or similar) it makes -static useless.
There is no way to know its dependencies (e.g. optional qt).


-- 
Jakub Bogusz    http://qboosh.pl/


From gotar at polanet.pl  Fri Feb 25 00:48:31 2011
From: gotar at polanet.pl (Tomasz Pala)
Date: Fri, 25 Feb 2011 00:48:31 +0100
Subject: packages: OGLFT/OGLFT.spec - include Qt support, remove .la,
	rel. 2
In-Reply-To: <20110224064347.GA4157@mail>
References: <packages_OGLFT_OGLFT.spec.1.1@pld-linux.org>
	<packages_OGLFT_OGLFT.spec.1.2@pld-linux.org>
	<20110224064347.GA4157@mail>
Message-ID: <20110224234831.GA24734@polanet.pl>

On Thu, Feb 24, 2011 at 07:43:47 +0100, Jakub Bogusz wrote:

>> -%{_libdir}/libOGLFT.la
> 
> Without pkgconfig file (or similar) it makes -static useless.
> There is no way to know its dependencies (e.g. optional qt).

I can't compile it properly anyway... at this point x86_64 build tried
to link against /usr/lib/libstdc++.so:
http://buildlogs.pld-linux.org/index.php?dist=th&arch=x86_64&ok=0&name=OGLFT&id=46a4776a-6cea-4bb5-a723-efaa66f7388f&action=tail
and without autocrap regeneration there are missing symbols on all archs:
http://buildlogs.pld-linux.org/index.php?dist=th&arch=i686&ok=0&name=OGLFT&id=3258ac9a-13ed-4321-914d-2fc84f813009&action=tail

-- 
Tomasz Pala <gotar at pld-linux.org>


From qboosh at pld-linux.org  Fri Feb 25 06:46:58 2011
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 25 Feb 2011 06:46:58 +0100
Subject: packages: OGLFT/OGLFT.spec - include Qt support, remove .la,
	rel. 2
In-Reply-To: <20110224234831.GA24734@polanet.pl>
References: <packages_OGLFT_OGLFT.spec.1.1@pld-linux.org>
	<packages_OGLFT_OGLFT.spec.1.2@pld-linux.org>
	<20110224064347.GA4157@mail> <20110224234831.GA24734@polanet.pl>
Message-ID: <20110225054658.GA7667@mail>

On Fri, Feb 25, 2011 at 12:48:31AM +0100, Tomasz Pala wrote:
> On Thu, Feb 24, 2011 at 07:43:47 +0100, Jakub Bogusz wrote:
> 
> >> -%{_libdir}/libOGLFT.la
> > 
> > Without pkgconfig file (or similar) it makes -static useless.
> > There is no way to know its dependencies (e.g. optional qt).
> 
> I can't compile it properly anyway... at this point x86_64 build tried
> to link against /usr/lib/libstdc++.so:
> http://buildlogs.pld-linux.org/index.php?dist=th&arch=x86_64&ok=0&name=OGLFT&id=46a4776a-6cea-4bb5-a723-efaa66f7388f&action=tail

You need to trace where "-L/usr/lib" in libtool --mode=link comes from.
Maybe qt setup ("-L$QTDIR/lib" is common bug).

> and without autocrap regeneration there are missing symbols on all archs:
> http://buildlogs.pld-linux.org/index.php?dist=th&arch=i686&ok=0&name=OGLFT&id=3258ac9a-13ed-4321-914d-2fc84f813009&action=tail

Looks like too old libtool (no CXX tag support?)


-- 
Jakub Bogusz    http://qboosh.pl/


From gotar at polanet.pl  Fri Feb 25 11:07:24 2011
From: gotar at polanet.pl (Tomasz Pala)
Date: Fri, 25 Feb 2011 11:07:24 +0100
Subject: packages: OGLFT/OGLFT.spec - include Qt support, remove .la,
	rel. 2
In-Reply-To: <20110225054658.GA7667@mail>
References: <packages_OGLFT_OGLFT.spec.1.1@pld-linux.org>
	<packages_OGLFT_OGLFT.spec.1.2@pld-linux.org>
	<20110224064347.GA4157@mail> <20110224234831.GA24734@polanet.pl>
	<20110225054658.GA7667@mail>
Message-ID: <20110225100724.GA28272@polanet.pl>

On Fri, Feb 25, 2011 at 06:46:58 +0100, Jakub Bogusz wrote:

>> I can't compile it properly anyway... at this point x86_64 build tried
>> to link against /usr/lib/libstdc++.so:
>> http://buildlogs.pld-linux.org/index.php?dist=th&arch=x86_64&ok=0&name=OGLFT&id=46a4776a-6cea-4bb5-a723-efaa66f7388f&action=tail
> 
> You need to trace where "-L/usr/lib" in libtool --mode=link comes from.
> Maybe qt setup ("-L$QTDIR/lib" is common bug).

Probably, because rel. 1 without Qt support has been build.


As for .la I can repackage it in -static (after reading previous
discussion), because it's useless and wrong in devel and _all of them_ are
nondeterministic anyway (but it's less probable that most of the -static
would be installed on builders), if you want to revive this discussion I
got no response for the latter:

http://www.mail-archive.com/pld-devel-en at lists.pld-linux.org/msg05937.html
http://www.mail-archive.com/pld-devel-en at lists.pld-linux.org/msg05951.html

Either BR: *-static for static builds _or better_, do some .la postprocessing to
make them more deterministic (like s/-l(.*)/\1.la/ and disabling
autorequires for other .la files). I prefer the second solution unless
it creates other problems (for example: what if we don't have
appropriate .la file, because library is pc-enabled? would it be enough
to ship stub .la file instead?).

-- 
Tomasz Pala <gotar at pld-linux.org>