From maciekp at japhy.fnord.org  Wed Nov  1 00:13:29 2006
From: maciekp at japhy.fnord.org (Maciek Pasternacki)
Date: Wed, 01 Nov 2006 00:13:29 +0100
Subject: [SPECS] Diff to bzr.spec (latest stable version).
In-Reply-To: <b7eadbe40610301404j333de521q61c9dfc0288ec302@mail.gmail.com>
	(arkh's message of "Mon\, 30 Oct 2006 23\:04\:58 +0100")
References: <b7eadbe40610301404j333de521q61c9dfc0288ec302@mail.gmail.com>
Message-ID: <m3ac3c3wpy.fsf@japhy.fnord.org>

On Pungenday, The Aftermath 11, 3172 YOLD, arkh wrote:

> - diff to bzr.spec - latest stable version
> - new spec for bzrtools (I'm not sure if polish translation is good -
> can anybody check if it's OK? ;)

Commited (bzrtools.spec from the next post).  Slightly modified
translation -- changed English `patch'y' to Polish equivalent `?at'.

Thanks,

-- 
__    Maciek Pasternacki <maciekp at japhy.fnord.org> [ http://japhy.fnord.org/ ]
`| _   |_\  /   {    Ibi nullus timet mortem
,|{-}|}| }\/                         sed pro Baccho mittunt sortem.   }
\/   |____/                                           ( Carmina Burana )  -><-


From public.arkh at gmail.com  Wed Nov  1 00:35:20 2006
From: public.arkh at gmail.com (arkh)
Date: Wed, 1 Nov 2006 00:35:20 +0100
Subject: [SPECS] Diff to bzr.spec (latest stable version).
In-Reply-To: <m3ac3c3wpy.fsf@japhy.fnord.org>
References: <b7eadbe40610301404j333de521q61c9dfc0288ec302@mail.gmail.com>
	<m3ac3c3wpy.fsf@japhy.fnord.org>
Message-ID: <b7eadbe40610311535j5697b3des9cbcd7990e7692e6@mail.gmail.com>

First update to bzrtools.spec (newer version) :)

Could you add AC-branch tag to bzr.spec? - it's a stable version.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bzrtools.spec.diff
Type: text/x-diff
Size: 738 bytes
Desc: not available
URL: </mailman/pipermail/pld-devel-en/attachments/20061101/6a8d2ec4/attachment.bin>

From maciekp at japhy.fnord.org  Wed Nov  1 01:04:20 2006
From: maciekp at japhy.fnord.org (Maciek Pasternacki)
Date: Wed, 01 Nov 2006 01:04:20 +0100
Subject: [SPECS] Diff to bzr.spec (latest stable version).
In-Reply-To: <b7eadbe40610311535j5697b3des9cbcd7990e7692e6@mail.gmail.com>
	(arkh's message of "Wed\, 1 Nov 2006 00\:35\:20 +0100")
References: <b7eadbe40610301404j333de521q61c9dfc0288ec302@mail.gmail.com>
	<m3ac3c3wpy.fsf@japhy.fnord.org>
	<b7eadbe40610311535j5697b3des9cbcd7990e7692e6@mail.gmail.com>
Message-ID: <m364e03ud7.fsf@japhy.fnord.org>

On Setting Orange, The Aftermath 13, 3172 YOLD, arkh wrote:

> First update to bzrtools.spec (newer version) :)
>
> Could you add AC-branch tag to bzr.spec? - it's a stable version.

I don't use Ac, so I can't test if it works on Ac.  It's not just
matter of bzr version, but also of different environment on Th and
Ac.  But it doesn't matter since glen already commited and sent to Ac
builders :)

-- 
__    Maciek Pasternacki <maciekp at japhy.fnord.org> [ http://japhy.fnord.org/ ]
`| _   |_\  / { ...mamy dzie?, mamy ka?d? noc, mamy swoje sny i swoje drogi,
,|{-}|}| }\/      tak, jak ty rozgl?damy si?, chcemy zna? ostatni?
\/   |____/      wersj? prawdy... }                       ( L. Janerka )  -><-


From maciekp at japhy.fnord.org  Wed Nov  1 01:08:54 2006
From: maciekp at japhy.fnord.org (Maciek Pasternacki)
Date: Wed, 01 Nov 2006 01:08:54 +0100
Subject: apache vs autoconf 2.60
In-Reply-To: <20061019140534.GB8896@jajo.axeos-gliwice> (Jacek Konieczny's
	message of "Thu\, 19 Oct 2006 16\:05\:34 +0200")
References: <20061013194051.GA9923@fngna.oyu>
	<200610161220.32877.glen@delfi.ee> <20061016183426.GA15746@fngna.oyu>
	<20061019140534.GB8896@jajo.axeos-gliwice>
Message-ID: <m31woo3u5l.fsf@japhy.fnord.org>

On Boomtime, Bureaucracy 73, 3172 YOLD, Jacek Konieczny wrote:

> On Mon, Oct 16, 2006 at 08:34:26PM +0200, Jakub Bogusz wrote:
>> With autoconf 2.60 layout overrides configure arguments. And that's the
>> main problem.
>
> It seems apr doesn't compile because that. Any ideas how to solve that
> system-wide? autoconf fix?

One-liner fix to apxs.  Apxs didn't read libdir variable from
config_variables.mk and libexecdir is referring to libdir now (and
didn't with earlier autoconf).  Merged into apache-apxs.patch, apache
up to rel 7.

-- 
__    Maciek Pasternacki <maciekp at japhy.fnord.org> [ http://japhy.fnord.org/ ]
`| _   |_\  / { ...I will wear you white feather, I will carry
,|{-}|}| }\/       your white flag, I will swear I have no nation,
\/   |____/        but I'm proud to own my heart... }           ( Fish )  -><-


From blues at pld-linux.org  Wed Nov  1 12:19:35 2006
From: blues at pld-linux.org (=?ISO-8859-2?Q?Pawe=B3_Go=B3aszewski?=)
Date: Wed, 1 Nov 2006 12:19:35 +0100 (CET)
Subject: SPECS: webalizer.spec - -base instead of -cron as suggested by
	blu...
In-Reply-To: <20061031135238.GA9937@gruby.cs.net.pl>
References: <SPECS_webalizer.spec.1.84@pld-linux.org>
	<SPECS_webalizer.spec.1.85@pld-linux.org>
	<20061026072626.GB17497@gruby.cs.net.pl>
	<Pine.LNX.4.64N.0610261330240.21931@wby.gnfx.tqn.cy>
	<20061031135238.GA9937@gruby.cs.net.pl>
Message-ID: <Pine.LNX.4.64N.0611011212200.20851@wby.gnfx.tqn.cy>

On Tue, 31 Oct 2006, Jakub Bogusz wrote:
> > After all - change proposed by glen would be great PITA too - after 
> > upgrade from package before split you'll get not working cron. Even 
> > after installing cron package (glen gave no triggers and that's why 
> > I've started talking to him about that :) ).
> > 
> > Besides - it sounds logical for me:
> > * -base (or basic) with pure application without any additions
> > * main package as full-set.
> #define "full-set"?

In this case: everything what was present in package until now.
In general: standard solution from distribution. If someone doesn't like 
it - there is base (or basic) package for those who want to prepare 
everything on their way.

> > IMvHO separating very basic package-set to -base package should be 
> > general rule. But - it's only IMO :D
> It depends what we define as "base".

look above.

-- 
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 qboosh at pld-linux.org  Wed Nov  1 12:53:03 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 1 Nov 2006 12:53:03 +0100
Subject: SPECS: webalizer.spec - -base instead of -cron as suggested by
	blu...
In-Reply-To: <Pine.LNX.4.64N.0611011212200.20851@wby.gnfx.tqn.cy>
References: <SPECS_webalizer.spec.1.84@pld-linux.org>
	<SPECS_webalizer.spec.1.85@pld-linux.org>
	<20061026072626.GB17497@gruby.cs.net.pl>
	<Pine.LNX.4.64N.0610261330240.21931@wby.gnfx.tqn.cy>
	<20061031135238.GA9937@gruby.cs.net.pl>
	<Pine.LNX.4.64N.0611011212200.20851@wby.gnfx.tqn.cy>
Message-ID: <20061101115303.GA30382@fngna.oyu>

On Wed, Nov 01, 2006 at 12:19:35PM +0100, Pawe? Go?aszewski wrote:
> On Tue, 31 Oct 2006, Jakub Bogusz wrote:
> > > After all - change proposed by glen would be great PITA too - after 
> > > upgrade from package before split you'll get not working cron. Even 
> > > after installing cron package (glen gave no triggers and that's why 
> > > I've started talking to him about that :) ).
> > > 
> > > Besides - it sounds logical for me:
> > > * -base (or basic) with pure application without any additions
> > > * main package as full-set.
> > #define "full-set"?
> 
> In this case: everything what was present in package until now.
> In general: standard solution from distribution. If someone doesn't like 
> it - there is base (or basic) package for those who want to prepare 
> everything on their way.

OK (for already existing packages).


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


From glen at delfi.ee  Wed Nov  1 13:04:39 2006
From: glen at delfi.ee (Elan =?iso-8859-2?q?Ruusam=E4e?=)
Date: Wed, 1 Nov 2006 14:04:39 +0200
Subject: SPECS: webalizer.spec - -base instead of -cron as suggested by
	blu...
In-Reply-To: <20061101115303.GA30382@fngna.oyu>
References: <SPECS_webalizer.spec.1.84@pld-linux.org>
	<Pine.LNX.4.64N.0611011212200.20851@wby.gnfx.tqn.cy>
	<20061101115303.GA30382@fngna.oyu>
Message-ID: <200611011404.40040.glen@delfi.ee>

On Wednesday 01 November 2006 13:53, Jakub Bogusz wrote:
> > > > Besides - it sounds logical for me:
> > > > * -base (or basic) with pure application without any additions
> > > > * main package as full-set.
there also exists schema -progs in PLD (libxml2, libxsl, putty, gd), where are 
only programs for application (or library).

> > >
> > > #define "full-set"?
> >
> > In this case: everything what was present in package until now.
> > In general: standard solution from distribution. If someone doesn't like
> > it - there is base (or basic) package for those who want to prepare
> > everything on their way.
>
> OK (for already existing packages).
will it stay this way (-base) or goes with the  -cron?

-- 
glen


From qboosh at pld-linux.org  Wed Nov  1 14:47:54 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 1 Nov 2006 14:47:54 +0100
Subject: SPECS: webalizer.spec - -base instead of -cron as suggested by
	blu...
In-Reply-To: <200611011404.40040.glen@delfi.ee>
References: <SPECS_webalizer.spec.1.84@pld-linux.org>
	<Pine.LNX.4.64N.0611011212200.20851@wby.gnfx.tqn.cy>
	<20061101115303.GA30382@fngna.oyu>
	<200611011404.40040.glen@delfi.ee>
Message-ID: <20061101134754.GA2106@fngna.oyu>

On Wed, Nov 01, 2006 at 02:04:39PM +0200, Elan Ruusam?e wrote:
> On Wednesday 01 November 2006 13:53, Jakub Bogusz wrote:
> > > > > Besides - it sounds logical for me:
> > > > > * -base (or basic) with pure application without any additions
> > > > > * main package as full-set.
> there also exists schema -progs in PLD (libxml2, libxsl, putty, gd), where are 
> only programs for application (or library).

It depends what is the main purpose of package.
If sth is mainly library, the main package should contain library and
some optional programs can be placed in subpackage (-progs, sometimes
-utils or -tools; I don't even remember which one was chosen as
"standard" last time...). "libxml-libs" would look somewhat funny ;)

> > > > #define "full-set"?
> > >
> > > In this case: everything what was present in package until now.
> > > In general: standard solution from distribution. If someone doesn't like
> > > it - there is base (or basic) package for those who want to prepare
> > > everything on their way.
> >
> > OK (for already existing packages).
> will it stay this way (-base) or goes with the  -cron?

As previous ftp versions of webalizer contained cron job in main
package, let it stay with -base.

For new packages, if cron job is optional addon (not essential part of
functionality), I would place it in -cron subpackage.


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


From zawadaa at gmail.com  Wed Nov  1 21:53:39 2006
From: zawadaa at gmail.com (Andrzej Zawadzki)
Date: Wed, 01 Nov 2006 21:53:39 +0100
Subject: [AC] mc, readline-5.2-1 and _
Message-ID: <45490953.6080607@gmail.com>

Hello!

After upgrade to readline-5.2-1, mc works bad when dir has "_" in name.
For example: dir = a_a
After <enter> on dir I've got:
Warning: Cannot change to /home/users/zawada/a_a.

Recompile mc with new readline doesn't help.

What's wrong?

-- 
Andrzej Zawadzki


From qboosh at pld-linux.org  Wed Nov  1 21:58:26 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 1 Nov 2006 21:58:26 +0100
Subject: [AC] mc, readline-5.2-1 and _
In-Reply-To: <45490953.6080607@gmail.com>
References: <45490953.6080607@gmail.com>
Message-ID: <20061101205826.GA15122@fngna.oyu>

On Wed, Nov 01, 2006 at 09:53:39PM +0100, Andrzej Zawadzki wrote:
> Hello!
> 
> After upgrade to readline-5.2-1, mc works bad when dir has "_" in name.
> For example: dir = a_a
> After <enter> on dir I've got:
> Warning: Cannot change to /home/users/zawada/a_a.
> 
> Recompile mc with new readline doesn't help.
> 
> What's wrong?

It's not readline, but bash 3.2; see comment on top of bash.spec.

$ echo -e '\137'
\137
$ echo -e '\0137'
_


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


From wrobell at pld-linux.org  Thu Nov  2 00:13:33 2006
From: wrobell at pld-linux.org (wrobell)
Date: Wed, 1 Nov 2006 23:13:33 +0000
Subject: vim mimetype
In-Reply-To: <200610270156.18882.glen@delfi.ee>
References: <200610251925.20596.glen@delfi.ee>
	<200610261155.38197.glen@delfi.ee> <20061026224341.GC26414@borg>
	<200610270156.18882.glen@delfi.ee>
Message-ID: <20061101231333.GF1844@borg>

On Fri, Oct 27, 2006 at 01:56:18AM +0300, Elan Ruusam?e wrote:
> On Friday 27 October 2006 01:43, wrobell wrote:
> > On Thu, Oct 26, 2006 at 11:55:37AM +0300, Elan Ruusam?e wrote:
> > > On Thursday 26 October 2006 02:30, wrobell wrote:
> > > > > anyway back to topic, so it's kde bug it's not opening terminal (that
> > > > > desktop file works ok if started from menu)?
> > > >
> > > > it seems. in gnome menu it works perfectly
> > >
> > > so, does it work also without mimetype=text/plain in vim.desktop?
> >
> > as i remember - no. but i can be wrong.
> could you test? 
> it should be hard as invoking:
> 
> $ sudo vim /usr/share/applications/vim.desktop

seems to work

    wrobell <wrobell at pld-linux.org>


From wrobell at pld-linux.org  Thu Nov  2 00:21:39 2006
From: wrobell at pld-linux.org (wrobell)
Date: Wed, 1 Nov 2006 23:21:39 +0000
Subject: SPECS (DEVEL): mozilla-firefox.spec - added more configure
	options...
In-Reply-To: <SPECS_mozilla_firefox.spec.1.95.2.33@pld-linux.org>
References: <SPECS_mozilla_firefox.spec.1.95.2.32@pld-linux.org>
	<SPECS_mozilla_firefox.spec.1.95.2.33@pld-linux.org>
Message-ID: <20061101232139.GG1844@borg>

On Wed, Nov 01, 2006 at 01:48:26AM +0100, glen wrote:
> Author: glen                         Date: Wed Nov  1 00:48:26 2006 GMT
> Module: SPECS                         Tag: DEVEL
> ---- Log message:
> - added more configure options, reviewed them
> - install section cleanup and skip tgz step
> - firefox-chrome+xpcom-generate cleanup and comments
> - postun dropped (should be done as ghost files instead)
> 
> ---- Files affected:
> SPECS:
>    mozilla-firefox.spec (1.95.2.32 -> 1.95.2.33) 
> 
> ---- Diffs:
> 
> ================================================================
> Index: SPECS/mozilla-firefox.spec
> diff -u SPECS/mozilla-firefox.spec:1.95.2.32 SPECS/mozilla-firefox.spec:1.95.2.33
> --- SPECS/mozilla-firefox.spec:1.95.2.32	Thu Oct 26 11:26:09 2006
> +++ SPECS/mozilla-firefox.spec	Wed Nov  1 01:48:20 2006
> @@ -8,13 +8,17 @@
[...]
> +BuildRequires:	jdk

will it build on ppc? gcj is enough? does it introduce java runtime
dependency?

[...]

    wrobell <wrobell at pld-linux.org>


From glen at delfi.ee  Thu Nov  2 00:27:28 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 2 Nov 2006 02:27:28 +0300
Subject: SPECS (DEVEL): mozilla-firefox.spec - added more configure
	options...
In-Reply-To: <20061101232139.GG1844@borg>
References: <SPECS_mozilla_firefox.spec.1.95.2.32@pld-linux.org>
	<SPECS_mozilla_firefox.spec.1.95.2.33@pld-linux.org>
	<20061101232139.GG1844@borg>
Message-ID: <200611020127.28404.glen@delfi.ee>

On Thursday 02 November 2006 01:21, wrobell wrote:
> > +++ SPECS/mozilla-firefox.spec????????Wed Nov ?1 01:48:20 2006
> > @@ -8,13 +8,17 @@
>
> [...]
>
> > +BuildRequires:???????jdk
>
> will it build on ppc? gcj is enough? 
no idea, testing yet.
it needs jni.h and javac, java as i looked compile progress

> does it introduce java runtime 
> dependency?
nope.

-- 
glen


From arekm at pld-linux.org  Thu Nov  2 13:42:46 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Thu, 2 Nov 2006 13:42:46 +0100
Subject: SOURCES (DEVEL): mysql.init - using previously saved sql dumps to
	...
In-Reply-To: <SOURCES_mysql.init.1.105.2.1@pld-linux.org>
References: <SOURCES_mysql.init.1.105@pld-linux.org>
	<SOURCES_mysql.init.1.105.2.1@pld-linux.org>
Message-ID: <200611021342.46834.arekm@pld-linux.org>

On Thursday 02 November 2006 13:13, glen wrote:
> Author: glen                         Date: Thu Nov  2 12:13:49 2006 GMT
> Module: SOURCES                       Tag: DEVEL
> ---- Log message:
> - using previously saved sql dumps to init mysql database

What about InnoDB tables here?

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


From qboosh at pld-linux.org  Thu Nov  2 14:37:58 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 2 Nov 2006 14:37:58 +0100
Subject: SPECS: nss.spec - added nss-config (subst based on suse spec) -
	ad...
In-Reply-To: <SPECS_nss.spec.1.69@pld-linux.org>
References: <SPECS_nss.spec.1.68@pld-linux.org>
	<SPECS_nss.spec.1.69@pld-linux.org>
Message-ID: <20061102133758.GD4017@gruby.cs.net.pl>

On Thu, Nov 02, 2006 at 12:57:24AM +0100, glen wrote:
> Author: glen                         Date: Wed Nov  1 23:57:24 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - added nss-config (subst based on suse spec)

What uses it?
A general trend in many projects is to drop *-config script in favour
of *.pc files.


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


From qboosh at pld-linux.org  Thu Nov  2 14:56:04 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 2 Nov 2006 14:56:04 +0100
Subject: SPECS: nss.spec - added nss-config (subst based on suse spec) -
	ad...
In-Reply-To: <20061102133758.GD4017@gruby.cs.net.pl>
References: <SPECS_nss.spec.1.68@pld-linux.org>
	<SPECS_nss.spec.1.69@pld-linux.org>
	<20061102133758.GD4017@gruby.cs.net.pl>
Message-ID: <20061102135604.GE4017@gruby.cs.net.pl>

On Thu, Nov 02, 2006 at 02:37:58PM +0100, Jakub Bogusz wrote:
> On Thu, Nov 02, 2006 at 12:57:24AM +0100, glen wrote:
> > Author: glen                         Date: Wed Nov  1 23:57:24 2006 GMT
> > Module: SPECS                         Tag: HEAD
> > ---- Log message:
> > - added nss-config (subst based on suse spec)
> 
> What uses it?

Aaah. I see, mozilla.org (firefox) itself, without providing it anywhere.
Great :/


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


From twittner at o2.pl  Thu Nov  2 21:24:36 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Thu, 2 Nov 2006 21:24:36 +0100
Subject: [Ac] kaddressbook - sigsegv while adding IM
Message-ID: <200611022124.36577.twittner@o2.pl>

Hi,

kaddressbook segfaults while adding IM address.
$ qf `which kaddressbook`
kdepim-kaddressbook-9:3.5.4-2 athlon.ac.pld-linux.org

I've tested kaddressbook from Debian/unstable:
$ dpkg-query -W kaddressbook
kaddressbook    4:3.5.5.dfsg.1-1
and it doesn't crashes in the same situation.
BTW: I saw that kde*.spec on AC-branch have been upgraded by glen to 3.5.5 
version (but they have "dotted" releases).

-- 
Tomasz Wittner


From glen at delfi.ee  Thu Nov  2 21:45:43 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 2 Nov 2006 23:45:43 +0300
Subject: mozilla*
Message-ID: <200611022245.43813.glen@delfi.ee>

$ ac-provides libxpconnect.so
mozilla-firefox-1.5.0.7-2
mozilla-thunderbird-1.5.0.7-1
seamonkey-1.0.5-1

this is no good, is it?

is there way to exclude the provides for whole dir 
like /usr/lib/mozilla-thunderbird/components/* ?

altho those mozilla packages don't require it via rpm deps:
$ ac-requires  libxpconnect.so
$

-- 
glen


From qboosh at pld-linux.org  Thu Nov  2 22:10:11 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 2 Nov 2006 22:10:11 +0100
Subject: mozilla*
In-Reply-To: <200611022245.43813.glen@delfi.ee>
References: <200611022245.43813.glen@delfi.ee>
Message-ID: <20061102211011.GA29818@fngna.oyu>

On Thu, Nov 02, 2006 at 11:45:43PM +0300, Elan Ruusam?e wrote:
> $ ac-provides libxpconnect.so
> mozilla-firefox-1.5.0.7-2
> mozilla-thunderbird-1.5.0.7-1
> seamonkey-1.0.5-1
> 
> this is no good, is it?
> 
> is there way to exclude the provides for whole dir 
> like /usr/lib/mozilla-thunderbird/components/* ?

Something like
%define		_noautoprovfiles	%{_libdir}/mozilla.*

I'm not sure about quoting, please test it.
IIRC it's regexp, not shellglob.


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


From patrys at pld-linux.org  Thu Nov  2 22:27:00 2006
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Thu, 02 Nov 2006 22:27:00 +0100
Subject: rpm 4.4.7 + poldek
Message-ID: <1162502820.8503.14.camel@localhost>

As you might have already guessed, Jeff dropped all the deprecated stuff
from this release of rpm. Now it's there in Th. Don't upgrade.

poldek makes use of most of the deprecated functions that are now gone
(Jeff says some of them were marked as deprecated for as long as 6 whole
years).

DEVEL branch contains my attempt at making poldek usable again. There is
a poldek-deprecated.patch that fixes the building process but
unfortunately, other things are not that easy to fix:

[patrys at meaw SOURCES]$ LC_ALL=C poldek
Loading [pndir]th...
Loading [pndir]th...
Loading [pndir]th-test...
Loading [pndir]th-test...
13041 packages read
Removed 11 duplicate packages from available set
error: db4 error(146746632) from db->set_cachesize: Unknown error 146746632
error: db4 error(22) from db->close: Invalid argument
error: cannot open Packages index using db3 - Unknown error 146746632 (146746632)
error: //var/lib/rpm: open rpm database failed
error: Load installed packages failed

While I am completely unfamiliar with the above, can anyone try and fix
it?

Pawel Gajda (poldek's author) is not answering emails sent to
mis at pld-linux.org

-- 
Patryk Zawadzki <patrys at pld-linux.org>
PLD Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: To jest cz??? listu	podpisana cyfrowo
URL: </mailman/pipermail/pld-devel-en/attachments/20061102/6c93489f/attachment.sig>

From glen at delfi.ee  Thu Nov  2 23:40:03 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Fri, 3 Nov 2006 01:40:03 +0300
Subject: php 5.2
Message-ID: <200611030040.03319.glen@delfi.ee>

hi

in php 5.2 pcre extension needs to be compiled statically as libphp_common.so 
uses symbols from it.

./.libs/libphp_common.so: undefined reference to 
`pcre_get_compiled_regex_cache'
./.libs/libphp_common.so: undefined reference to `php_pcre_replace_impl'
./.libs/libphp_common.so: undefined reference to `pcre_get_compiled_regex'
./.libs/libphp_common.so: undefined reference to `php_pcre_split_impl'
./.libs/libphp_common.so: undefined reference to `php_pcre_match_impl'
collect2: ld returned 1 exit status

but adding Provides:  %{name}-pcre = %{epoch}:%{version}-%{release} will not 
make poldek happy, and upgrade will break.

error: php-common = 4:5.1.6-1 is required by php-pcre-5.1.6-1
error: php-common = 4:5.1.6-1 is required by php-pcre-5.1.6-1

$ rpm -qp ../RPMS/php-common-5.2.0-0.1.i686.rpm --provides|grep pcre
php-pcre = 4:5.2.0-0.1

probably php-common should obsolete php-pcre? like:
Obsoletes:  php-pcre < 4:5.2.0

but how to do that so that it won't obsolete php4-pcre which has:
$ q php4-pcre --provides
php-pcre = 3:4.4.4-5

bump epoch+=2 so that php4 has epoch=5 and php5 has epoch=6?

-- 
glen


From glen at delfi.ee  Fri Nov  3 01:21:41 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Fri, 3 Nov 2006 03:21:41 +0300
Subject: glib2 G_GNUC_UNUSED
Message-ID: <200611030221.41731.glen@delfi.ee>

is glib2 defining the macro incompatible for gcc in Ac, or there's some other cause?

$ g++ -c t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags gtk+-2.0`
t.c:5: error: syntax error before `__attribute__'

$ cat t.c
#include <gtk/gtk.h>
#include <glib.h>

static void open_url_cb( GtkWidget *pButton G_GNUC_UNUSED, gpointer data )
{
}

$ g++ -E t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags gtk+-2.0`>t.C
$ grep open_url_cb t.C
static void open_url_cb( GtkWidget *pButton __attribute__((__unused__)), gpointer data )

$ q gcc glib2 gtk+2
gcc-5:3.3.6-4
glib2-1:2.10.3-1
gtk+2-2:2.8.20-1
$

http://developer.gnome.org/doc/API/2.0/glib/glib-Miscellaneous-Macros.html#G-GNUC-UNUSED:CAPS

-- 
glen


From glen at delfi.ee  Fri Nov  3 13:49:04 2006
From: glen at delfi.ee (Elan =?iso-8859-2?q?Ruusam=E4e?=)
Date: Fri, 3 Nov 2006 14:49:04 +0200
Subject: SOURCES (DEVEL): mysql.init - using previously saved sql dumps to
	...
In-Reply-To: <200611021342.46834.arekm@pld-linux.org>
References: <SOURCES_mysql.init.1.105@pld-linux.org>
	<SOURCES_mysql.init.1.105.2.1@pld-linux.org>
	<200611021342.46834.arekm@pld-linux.org>
Message-ID: <200611031449.04221.glen@delfi.ee>

On Thursday 02 November 2006 14:42, Arkadiusz Miskiewicz wrote:
> On Thursday 02 November 2006 13:13, glen wrote:
> > Author: glen                         Date: Thu Nov  2 12:13:49 2006 GMT
> > Module: SOURCES                       Tag: DEVEL
> > ---- Log message:
> > - using previously saved sql dumps to init mysql database
>
> What about InnoDB tables here?
restored previous behaviour. 
didn't test :)

-- 
glen


From arekm at pld-linux.org  Fri Nov  3 14:54:38 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Fri, 3 Nov 2006 14:54:38 +0100
Subject: SOURCES (DEVEL): mysql.init - using previously saved sql dumps to
	...
In-Reply-To: <200611031449.04221.glen@delfi.ee>
References: <SOURCES_mysql.init.1.105@pld-linux.org>
	<200611021342.46834.arekm@pld-linux.org>
	<200611031449.04221.glen@delfi.ee>
Message-ID: <200611031454.38221.arekm@pld-linux.org>

On Friday 03 November 2006 13:49, Elan Ruusam?e wrote:
> On Thursday 02 November 2006 14:42, Arkadiusz Miskiewicz wrote:
> > On Thursday 02 November 2006 13:13, glen wrote:
> > > Author: glen                         Date: Thu Nov  2 12:13:49 2006 GMT
> > > Module: SOURCES                       Tag: DEVEL
> > > ---- Log message:
> > > - using previously saved sql dumps to init mysql database
> >
> > What about InnoDB tables here?
>
> restored previous behaviour.
> didn't test :)

Not really ;-) Some (core mysql) tables can not be InnoDB type... 

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


From qboosh at pld-linux.org  Fri Nov  3 18:58:45 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 3 Nov 2006 18:58:45 +0100
Subject: libxslt broken? (Re: ERRORS: libxcb.spec OK: xcb-proto.spec)
In-Reply-To: <pld-builder.1162574912.395139.20066@th-686>
References: <a1a8ef16-2909-4093-83a1-30d080b974d6@pld.src.builder>
	<pld-builder.1162574912.395139.20066@th-686>
Message-ID: <20061103175845.GA18685@gruby.cs.net.pl>

On Fri, Nov 03, 2006 at 05:28:32PM +0000, PLD th-i686 builder wrote:
[...]
> /usr/bin/xsltproc --stringparam mode source \
> 	            --stringparam base-path /usr/share/xcb/ \
> 	            --stringparam extension-path /usr/share/xcb/ \
> 	            -o xproto.c ./c-client.xsl xproto.xml 
> make[1]: *** [xproto.c] Segmentation fault
> make[1]: Leaving directory `/home/users/builder/rpm/BUILD/libxcb-0.9.93/src'
> make: *** [all-recursive] Error 1
> error: Bad exit status from /var/tmp/rpm-tmp.80655 (%build)

It crashes inside libxslt.
With older libxml2/libxslt it builds.

#0  0x40204b77 in free () from /lib/tls/libc.so.6
#1  0x400594ec in xsltFreeDocumentKeys () from /usr/lib/libxslt.so.1
#2  0x4005796e in xsltReleaseRVT () from /usr/lib/libxslt.so.1
#3  0x40068d41 in xsltChoose () from /usr/lib/libxslt.so.1
#4  0x400694fd in xsltProcessOneNode () from /usr/lib/libxslt.so.1
#5  0x4006a18d in xsltApplyTemplates () from /usr/lib/libxslt.so.1
#6  0x40067798 in xsltCopyTextString () from /usr/lib/libxslt.so.1
#7  0x40068d0e in xsltChoose () from /usr/lib/libxslt.so.1
#8  0x400694fd in xsltProcessOneNode () from /usr/lib/libxslt.so.1
#9  0x400699b6 in xsltProcessOneNode () from /usr/lib/libxslt.so.1
#10 0x4006a18d in xsltApplyTemplates () from /usr/lib/libxslt.so.1
#11 0x40067798 in xsltCopyTextString () from /usr/lib/libxslt.so.1
#12 0x400583d2 in xsltCreateRVT () from /usr/lib/libxslt.so.1
#13 0x400585bf in xsltCreateRVT () from /usr/lib/libxslt.so.1
#14 0x40058b2c in xsltXPathVariableLookup () from /usr/lib/libxslt.so.1
#15 0x400fee9e in xmlXPathVariableLookup () from /usr/lib/libxml2.so.2
#16 0x4010017f in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#17 0x400ff1e8 in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#18 0x400ffb68 in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#19 0x400ffbd9 in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#20 0x400ff1e8 in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#21 0x4010114e in xmlXPathFunctionLookup () from /usr/lib/libxml2.so.2
#22 0x40105916 in xmlXPathEval () from /usr/lib/libxml2.so.2
#23 0x40105b12 in xmlXPathCompiledEval () from /usr/lib/libxml2.so.2
#24 0x400581ab in xsltCreateRVT () from /usr/lib/libxslt.so.1
#25 0x400cdb3f in xmlBuildQName () from /usr/lib/libxml2.so.2
#26 0x400cdbcb in xmlHashScanFull () from /usr/lib/libxml2.so.2
#27 0x400cdc1a in xmlHashScan () from /usr/lib/libxml2.so.2
#28 0x4005764c in xsltEvalGlobalVariables () from /usr/lib/libxslt.so.1
#29 0x4006d690 in xsltNewTransformContext () from /usr/lib/libxslt.so.1
#30 0x4006de53 in xsltRunStylesheetUser () from /usr/lib/libxslt.so.1
#31 0x080498a3 in ?? ()
#32 0x080587f0 in ?? ()
#33 0x08083c48 in ?? ()
#34 0x0804caa0 in xmlFree ()
#35 0x0804d2d8 in ?? ()
#36 0x00000000 in ?? ()

Sorry, no debuginfo here.


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


From twittner at o2.pl  Fri Nov  3 23:37:05 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Fri, 3 Nov 2006 23:37:05 +0100
Subject: smtpdaemon and sendmail wrapper
In-Reply-To: <200610312241.31555.glen@delfi.ee>
References: <200610312241.31555.glen@delfi.ee>
Message-ID: <200611032337.05800.twittner@o2.pl>

On Tue 31. October 2006 21:41, Elan Ruusam?e wrote:
> imho they are different things
> 1. smtpdaemon means something that listens port 25
> 2. sendmail wrapper is package that provides /usr/lib/sendmail
> 
> are there any program that really needs smtpdaemon in this context?
> 
> i'd propose to use /usr/lib/sendmail Provides/Requires pairs when program 
> needs /usr/lib/sendmail binary (and drop smtpdaemon if none actually needs 
> it).
> 
IMO, for simplicity and clarity: `P/O: smtpdaemon' should be renamed to 
`P/O: smtpd'
and what acts as smtp client (has sendmail(1) command) should have
`P/O: smtp'.

-- 
Tomasz Wittner


From twittner at o2.pl  Fri Nov  3 23:53:02 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Fri, 3 Nov 2006 23:53:02 +0100
Subject: smtpdaemon and sendmail wrapper
In-Reply-To: <200610312241.31555.glen@delfi.ee>
References: <200610312241.31555.glen@delfi.ee>
Message-ID: <200611032353.02898.twittner@o2.pl>

On Tue 31. October 2006 21:41, Elan Ruusam?e wrote:
> imho they are different things
> 1. smtpdaemon means something that listens port 25
> 2. sendmail wrapper is package that provides /usr/lib/sendmail
> 
> are there any program that really needs smtpdaemon in this context?
> 
> i'd propose to use /usr/lib/sendmail Provides/Requires pairs when program 
> needs /usr/lib/sendmail binary (and drop smtpdaemon if none actually needs 
> it).
P/O: smtpdaemon (imo better P/O: smtpd) is needed by MTAs for obsoleting each 
other.(?)
> 

-- 
Tomasz Wittner


From pluto at agmk.net  Sat Nov  4 19:20:00 2006
From: pluto at agmk.net (=?utf-8?q?Pawe=C5=82_Sikora?=)
Date: Sat, 4 Nov 2006 19:20:00 +0100
Subject: libxslt broken? (Re: ERRORS: libxcb.spec OK: xcb-proto.spec)
In-Reply-To: <20061103175845.GA18685@gruby.cs.net.pl>
References: <a1a8ef16-2909-4093-83a1-30d080b974d6@pld.src.builder>
	<pld-builder.1162574912.395139.20066@th-686>
	<20061103175845.GA18685@gruby.cs.net.pl>
Message-ID: <200611041920.00827.pluto@agmk.net>

On Friday 03 November 2006 18:58, Jakub Bogusz wrote:
> On Fri, Nov 03, 2006 at 05:28:32PM +0000, PLD th-i686 builder wrote:
> [...]
>
> > /usr/bin/xsltproc --stringparam mode source \
> > 	            --stringparam base-path /usr/share/xcb/ \
> > 	            --stringparam extension-path /usr/share/xcb/ \
> > 	            -o xproto.c ./c-client.xsl xproto.xml
> > make[1]: *** [xproto.c] Segmentation fault
> > make[1]: Leaving directory
> > `/home/users/builder/rpm/BUILD/libxcb-0.9.93/src' make: ***
> > [all-recursive] Error 1
> > error: Bad exit status from /var/tmp/rpm-tmp.80655 (%build)
>
> It crashes inside libxslt.

(gdb) bt
#0  0x00002b71716a42a5 in free () from /lib64/libc.so.6
#1  0x00002b7170eb96ab in xsltFreeKeyTable (keyt=0x74fa40) at keys.c:154
#2  0x00002b7170eb973f in xsltFreeKeyTableList (keyt=0x0) at keys.c:177
#3  0x00002b7170eba9da in xsltFreeDocumentKeys (idoc=0x8045b0) at keys.c:852

Breakpoint 1, xsltFreeKeyTableList (keyt=0x74fa40) at keys.c:174
174         while (keyt != NULL) {

(gdb) p *keyt
$2 = {
  next = 0x0,
  name = 0xb1 <Address 0xb1 out of bounds>,
  nameURI = 0x8045b0 "",
  keys = 0x7510f0
}

one item in list with crappy data.
freeing keyt->name crashes.


From twittner at o2.pl  Sun Nov  5 09:27:43 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Sun, 5 Nov 2006 09:27:43 +0100
Subject: [Ac] kaddressbook - sigsegv while adding IM
In-Reply-To: <200611022124.36577.twittner@o2.pl>
References: <200611022124.36577.twittner@o2.pl>
Message-ID: <200611050927.44435.twittner@o2.pl>

On Thu 2. November 2006 21:24, Tomasz Wittner wrote:
> Hi,
> 
> kaddressbook segfaults while adding IM address.
> $ qf `which kaddressbook`
> kdepim-kaddressbook-9:3.5.4-2 athlon.ac.pld-linux.org
> 
> I've tested kaddressbook from Debian/unstable:
> $ dpkg-query -W kaddressbook
> kaddressbook    4:3.5.5.dfsg.1-1
> and it doesn't crashes in the same situation.
> BTW: I saw that kde*.spec on AC-branch have been upgraded by glen to 3.5.5 
> version (but they have "dotted" releases).
> 
Upgrading kde to 3.5.5 didn't help:
[KCrash handler]
#5  0xb79b85d6 in KPluginInfo::name () from /usr/lib/libkutils.so.1
#6  0xb7f5520a in IMAddressWidget::slotProtocolChanged ()
   from /usr/lib/libkaddressbook.so.0
#7  0xb7f54d2b in IMAddressWidget::init () from /usr/lib/libkaddressbook.so.0
#8  0xb7f54806 in IMAddressWidget::IMAddressWidget ()
   from /usr/lib/libkaddressbook.so.0
#9  0xb7f52362 in IMEditorWidget::slotAdd () from /usr/lib/libkaddressbook.so.0
#10 0xb7f53760 in IMEditorWidget::qt_invoke ()
[...]

$ qf `which kaddressbook`
kdepim-kaddressbook-9:3.5.5-1 athlon.ac.pld-linux.org

-- 
Tomasz Wittner


From twittner at o2.pl  Sun Nov  5 10:23:19 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Sun, 5 Nov 2006 10:23:19 +0100
Subject: [Th] There are more than one package which provide "/usr/lib/debug"
	...
Message-ID: <200611051023.20805.twittner@o2.pl>

From poldek:

There are more than one package which provide "/usr/lib/debug":
a) filesystem-debuginfo-3.0-12
b) galago-daemon-0.5.0-2
c) tcl-tls-1.5.0-2

Please remember that /usr/lib{,64} contains debug directory which should go 
into -debuginfo subpackage thus sth like:
%files
%{_libdir}/*
is wrong.

I've already corrected galago-daemon.spec and tcl-tls.spec in cvs.

-- 
Tomasz Wittner


From jajcus at jajcus.net  Mon Nov  6 14:21:00 2006
From: jajcus at jajcus.net (Jacek Konieczny)
Date: Mon, 6 Nov 2006 14:21:00 +0100
Subject: smtpdaemon and sendmail wrapper
In-Reply-To: <200611032337.05800.twittner@o2.pl>
References: <200610312241.31555.glen@delfi.ee>
	<200611032337.05800.twittner@o2.pl>
Message-ID: <20061106132100.GG30585@jajo.axeos-gliwice>

On Fri, Nov 03, 2006 at 11:37:05PM +0100, Tomasz Wittner wrote:
> and what acts as smtp client (has sendmail(1) command) should have
> `P/O: smtp'.

Do you suggest that "sendmail" command is the same as smtp client? It is not.
There could be a sendmail command implementation doing no SMTP and SMTP
client which do not provide sendmail command (there are a lot of those).

Greets,
        Jacek


From glen at delfi.ee  Mon Nov  6 14:40:28 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Mon, 6 Nov 2006 15:40:28 +0200
Subject: smtpdaemon and sendmail wrapper
In-Reply-To: <20061106132100.GG30585@jajo.axeos-gliwice>
References: <200610312241.31555.glen@delfi.ee>
	<200611032337.05800.twittner@o2.pl>
	<20061106132100.GG30585@jajo.axeos-gliwice>
Message-ID: <200611061540.29049.glen@delfi.ee>

On Monday 06 November 2006 15:21, Jacek Konieczny wrote:
> On Fri, Nov 03, 2006 at 11:37:05PM +0100, Tomasz Wittner wrote:
> > and what acts as smtp client (has sendmail(1) command) should have
> > `P/O: smtp'.
>
> Do you suggest that "sendmail" command is the same as smtp client? It is
> not. There could be a sendmail command implementation doing no SMTP and
> SMTP client which do not provide sendmail command (there are a lot of
> those).
those (if compatible with sendmail commandline args) could 
provide /usr/lib/sendmail symlink

-- 
glen


From jajcus at jajcus.net  Mon Nov  6 15:32:39 2006
From: jajcus at jajcus.net (Jacek Konieczny)
Date: Mon, 6 Nov 2006 15:32:39 +0100
Subject: glibc compilation problem
Message-ID: <20061106143238.GH30585@jajo.axeos-gliwice>

Hello,

I have encountered a big problem with glibc compilation on
PLD-Th-derived system:

glibc 2.4 build (from glibc.spec rev. 1.648) fails with:

i686-pld-linux-gcc   -nostdlib -nostartfiles -r -o /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/librtld.map.o '-Wl,-(' /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a -lgcc '-Wl,-)' -Wl,-Map,/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/librtld.mapT
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(chk_fail.os): In function `*__GI___chk_fail':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/debug/chk_fail.c:28: multiple definition of `__GI___chk_fail'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf/dl-minimal.c:354: first defined here
/usr/bin/ld: Warning: size of symbol `__GI___chk_fail' changed from 27 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os to 64 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(chk_fail.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(chk_fail.os): In function `*__GI___chk_fail':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/debug/chk_fail.c:28: multiple definition of `__chk_fail'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf/dl-minimal.c:354: first defined here
/usr/bin/ld: Warning: size of symbol `__chk_fail' changed from 27 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os to 64 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(chk_fail.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(init-first.os):(.data+0x0): multiple definition of `__libc_multiple_libcs'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf/rtld.c:825: first defined here
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(_itoa.os): In function `_itoa':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/stdio-common/_itoa.c:210: multiple definition of `_itoa'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf/dl-minimal.c:295: first defined here
/usr/bin/ld: Warning: size of symbol `_itoa' changed from 192 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os to 612 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(_itoa.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(sbrk.os): In function `*__GI___sbrk':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/misc/sbrk.c:34: multiple definition of `__sbrk'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf/../misc/sbrk.c:34: first defined here
/usr/bin/ld: Warning: size of symbol `__sbrk' changed from 113 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/dl-allobjs.os to 108 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a(sbrk.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/libc_pic.a: could not read symbols: Bad value
collect2: ld returned 1 exit status
make[2]: *** [/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/builddir/elf/librtld.map] Error 1
make[2]: Leaving directory `/home/users/jacek/axeos-rpm/BUILD/glibc-2.4/elf'
make[1]: *** [elf/subdir_lib] Error 2
make[1]: Leaving directory `/home/users/jacek/axeos-rpm/BUILD/glibc-2.4'
make: *** [all] Error 2
error: Bad exit status from /var/tmp/rpm-tmp.73973 (%build)

binutils-2.17.50.0.6-1.aos1.i686
gcc-4.2.0-0.20060916r116994.3.aos2.i686

(.aos* releases don't differ much from PLD Th specs -- only release change in these cases).

Any ideas what may be wrong? Where to look for solution? 

I managed to build with glibc when bootstraping my build environment (I used
older binutils/gcc versions then (about month ago), probably from PLD Th), but
I need to rebuild it now.

Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?

Greets,
        Jacek


From jajcus at jajcus.net  Mon Nov  6 21:17:12 2006
From: jajcus at jajcus.net (Jacek Konieczny)
Date: Mon, 6 Nov 2006 21:17:12 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106143238.GH30585@jajo.axeos-gliwice>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
Message-ID: <20061106201712.GB410@nic.nigdzie>

On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?

I have tried. The same problem:

: /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a
i686-pld-linux-gcc   -nostdlib -nostartfiles -r -o /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/librtld.map.o '-Wl,-(' /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a -lgcc '-Wl,-)' -Wl,-Map,/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/librtld.mapT
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(chk_fail.os): In function `*__GI___chk_fail':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/debug/chk_fail.c:28: multiple definition of `__GI___chk_fail'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/elf/dl-minimal.c:354: first defined here
/usr/bin/ld: Warning: size of symbol `__GI___chk_fail' changed from 27 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os to 64 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(chk_fail.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(chk_fail.os): In function `*__GI___chk_fail':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/debug/chk_fail.c:28: multiple definition of `__chk_fail'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/elf/dl-minimal.c:354: first defined here
/usr/bin/ld: Warning: size of symbol `__chk_fail' changed from 27 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os to 64 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(chk_fail.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(init-first.os):(.data+0x0): multiple definition of `__libc_multiple_libcs'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/elf/rtld.c:827: first defined here
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(_itoa.os): In function `_itoa':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/stdio-common/_itoa.c:210: multiple definition of `_itoa'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/elf/dl-minimal.c:295: first defined here
/usr/bin/ld: Warning: size of symbol `_itoa' changed from 192 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os to 612 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(_itoa.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(sbrk.os): In function `*__GI___sbrk':
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/misc/sbrk.c:34: multiple definition of `__sbrk'
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os:/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/elf/../misc/sbrk.c:34: first defined here
/usr/bin/ld: Warning: size of symbol `__sbrk' changed from 113 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/dl-allobjs.os to 108 in /home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a(sbrk.os)
/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/libc_pic.a: could not read symbols: Bad value
collect2: ld returned 1 exit status
make[2]: *** [/home/users/jacek/axeos-rpm/BUILD/glibc-2.5/builddir/elf/librtld.map] Error 1

Greets,
        Jacek


From qboosh at pld-linux.org  Mon Nov  6 22:06:14 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Mon, 6 Nov 2006 22:06:14 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106201712.GB410@nic.nigdzie>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<20061106201712.GB410@nic.nigdzie>
Message-ID: <20061106210614.GA26183@fngna.oyu>

On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> 
> I have tried. The same problem:

Maybe too new gcc?
glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
gcc-4.1.0-1.athlon.


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


From pluto at agmk.net  Mon Nov  6 22:08:02 2006
From: pluto at agmk.net (=?utf-8?q?Pawe=C5=82_Sikora?=)
Date: Mon, 6 Nov 2006 22:08:02 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106210614.GA26183@fngna.oyu>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<20061106201712.GB410@nic.nigdzie>
	<20061106210614.GA26183@fngna.oyu>
Message-ID: <200611062208.02983.pluto@agmk.net>

On Monday 06 November 2006 22:06, Jakub Bogusz wrote:
> On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> > On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> >
> > I have tried. The same problem:
>
> Maybe too new gcc?
> glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
> gcc-4.1.0-1.athlon.

works for me for gcc-4.2.0-0.20061030r118195.3.


From qboosh at pld-linux.org  Mon Nov  6 22:10:01 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Mon, 6 Nov 2006 22:10:01 +0100
Subject: glibc compilation problem
In-Reply-To: <200611062208.02983.pluto@agmk.net>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<20061106201712.GB410@nic.nigdzie>
	<20061106210614.GA26183@fngna.oyu>
	<200611062208.02983.pluto@agmk.net>
Message-ID: <20061106211001.GB26183@fngna.oyu>

On Mon, Nov 06, 2006 at 10:08:02PM +0100, Pawe? Sikora wrote:
> On Monday 06 November 2006 22:06, Jakub Bogusz wrote:
> > On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> > > On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > > > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> > >
> > > I have tried. The same problem:
> >
> > Maybe too new gcc?
> > glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
> > gcc-4.1.0-1.athlon.
> 
> works for me for gcc-4.2.0-0.20061030r118195.3.

Which architectures?


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


From sparky at pld-linux.org  Mon Nov  6 22:11:24 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Mon, 6 Nov 2006 22:11:24 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106211001.GB26183@fngna.oyu>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<20061106201712.GB410@nic.nigdzie>
	<20061106210614.GA26183@fngna.oyu>
	<200611062208.02983.pluto@agmk.net>
	<20061106211001.GB26183@fngna.oyu>
Message-ID: <20061106211124.GA1757@pld-linux.org>

On Mon, Nov 06, 2006 at 10:10:01PM +0100, Jakub Bogusz wrote:
> On Mon, Nov 06, 2006 at 10:08:02PM +0100, Pawe? Sikora wrote:
> > On Monday 06 November 2006 22:06, Jakub Bogusz wrote:
> > > On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> > > > On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > > > > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> > > >
> > > > I have tried. The same problem:
> > >
> > > Maybe too new gcc?
> > > glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
> > > gcc-4.1.0-1.athlon.
> > 
> > works for me for gcc-4.2.0-0.20061030r118195.3.
> 
> Which architectures?

x86_64 probably


works for me, on ppc:

[sparky at forge ~]$ rpm -q glibc gcc binutils
glibc-2.5-0.2
gcc-4.1.2-0.20060928r117266.2
binutils-2.17.50.0.2-2


-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From zawadaa at gmail.com  Tue Nov  7 00:04:55 2006
From: zawadaa at gmail.com (Andrzej Zawadzki)
Date: Tue, 07 Nov 2006 00:04:55 +0100
Subject: fusebase specs names
Message-ID: <454FBF97.8070603@gmail.com>

Hello!

I'd like to ask what is a proper name (maybe is some convection) for
fusebase specs?
For example we have:
name - .spec
cvsfs - cvsfs-fuse.spec
fusesmb - fusesmb.spec

I "made" spec for curlftpfs and I don't know how to name this.

ps. it's works pretty good with FTP over SSL!

-- 
Andrzej Zawadzki


From pluto at agmk.net  Tue Nov  7 06:27:21 2006
From: pluto at agmk.net (=?iso-8859-2?q?Pawe=B3_Sikora?=)
Date: Tue, 7 Nov 2006 06:27:21 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106211001.GB26183@fngna.oyu>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<200611062208.02983.pluto@agmk.net>
	<20061106211001.GB26183@fngna.oyu>
Message-ID: <200611070627.21493.pluto@agmk.net>

On Monday 06 November 2006 22:10, Jakub Bogusz wrote:
> On Mon, Nov 06, 2006 at 10:08:02PM +0100, Pawe? Sikora wrote:
> > On Monday 06 November 2006 22:06, Jakub Bogusz wrote:
> > > On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> > > > On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > > > > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> > > >
> > > > I have tried. The same problem:
> > >
> > > Maybe too new gcc?
> > > glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
> > > gcc-4.1.0-1.athlon.
> >
> > works for me for gcc-4.2.0-0.20061030r118195.3.
>
> Which architectures?

th-x86_64.


From jajcus at jajcus.net  Tue Nov  7 08:29:43 2006
From: jajcus at jajcus.net (Jacek Konieczny)
Date: Tue, 7 Nov 2006 08:29:43 +0100
Subject: glibc compilation problem
In-Reply-To: <200611062208.02983.pluto@agmk.net>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
	<20061106201712.GB410@nic.nigdzie>
	<20061106210614.GA26183@fngna.oyu>
	<200611062208.02983.pluto@agmk.net>
Message-ID: <20061107072942.GA8058@jajo.axeos-gliwice>

On Mon, Nov 06, 2006 at 10:08:02PM +0100, Pawe? Sikora wrote:
> On Monday 06 November 2006 22:06, Jakub Bogusz wrote:
> > On Mon, Nov 06, 2006 at 09:17:12PM +0100, Jacek Konieczny wrote:
> > > On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> > > > Or maybe I should try with glibc 2.5? Is it, as in PLD CVS, usable?
> > >
> > > I have tried. The same problem:
> >
> > Maybe too new gcc?
> > glibc 2.5 builds for me with binutils-2.17.50.0.6-1.athlon and
> > gcc-4.1.0-1.athlon.
> 
> works for me for gcc-4.2.0-0.20061030r118195.3.

And for me not. :-(  The same problem.

Probably I have something broken in my build environment, but I have no
idea where to look...

Architecture is i686 and i586.

Greets,
        Jacek


From sparky at pld-linux.org  Tue Nov  7 18:22:19 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Tue, 7 Nov 2006 18:22:19 +0100
Subject: RFC: Macro for building modules
Message-ID: <20061107172219.GA19010@pld-linux.org>

Right now we are copying code for building kernel 2.6 modules to each
spec which provides them. It is a little hard to maintain and to add
some functionality. I have created macro for building modules based on
that code, a little expanded (with cross-compilation support). Now in
spec file thete will only have to be some thing like:

%if %{with kernel}
cd module/place
%build_kernel_modules module_name1 module_name2
cd ../..
%endif


If some additional options for make modules are required they may be
passed like: -1 option1 -2 option2 -3 option3; up to 9 options are
supported.

Exemple when some additional option in make modules is required:

%build_kernel_modules -1 T=$TMPDIR mol


There may be 2 options passed to make prepare scripts: -p o1 -P o2.
Module patching is supported via shell script called pld-module.patch.

Exemple with additional patching:

%if %{with kernel}
cp -f pwc-if.c pwc-if.c.orig

cat > pld-module.patch <<'EOF'
set -e -x
if grep -q "CONFIG_PREEMPT_RT" o/.config; then
	sed 's/SPIN_LOCK_UNLOCKED/SPIN_LOCK_UNLOCKED(pdev->ptrlock)/' \
		pwc-if.c.orig > pwc-if.c
else
	cat pwc-if.c.orig > pwc-if.c
fi
EOF

%build_kernel_modules pwc
%endif


Now, take a look at the code and propose your additions! Number and use
of backshashes is scary, but it has to be this way as %define is expanded
before anything.


-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org
-------------- next part --------------

%build_kernel_modules(1:2:3:4:5:6:7:8:9:p:P:)					\
%{!?1:%{error:build_kernel_modules: Required module name missing}}		\
										\
%define Opts 									\\\\\\\
%if "%{_target_base_arch}" != "%{_arch}"					\\\
	%if "%{_arch}" == "x86_64" && "%{_target_base_arch}" == "i386"		\\\
	CC="%{__cc}" CPP="%{__cpp}" ARCH=%{_target_base_arch}			\\\
	%else									\\\
	ARCH=%{_target_base_arch} CROSS_COMPILE=%{_target_cpu}-pld-linux-	\\\
	%endif									\\\
%else										\\\
	CC="%{__cc}" CPP="%{__cpp}"						\\\
%endif										\
%define	MakeOpts HOSTCC="%{__cc}" %{?with_verbose:V=1} %{Opts}			\
										\
for cfg in %{?with_dist_kernel:%{?with_smp:smp} up}%{!?with_dist_kernel:nondist}; do \
	if [ ! -r "%{_kernelsrcdir}/config-$cfg" ]; then			\
		exit 1								\
	fi									\
	install -d o/include/linux						\
	ln -sf %{_kernelsrcdir}/config-$cfg o/.config				\
	ln -sf %{_kernelsrcdir}/Module.symvers-$cfg o/Module.symvers		\
	ln -sf %{_kernelsrcdir}/include/linux/autoconf-$cfg.h o/include/linux/autoconf.h \
										\
	[ ! -e pld-module.patch ] || %__spec_build_shell pld-module.patch	\
										\
	%if %{with dist_kernel}							\
		%{__make} -j1 -C %{_kernelsrcdir} prepare scripts		\\\
			O=$PWD/o %{-p*} %{-P*}					\\\
			%{MakeOpts}						\
	%else									\
		install -d o/include/config					\
		touch o/include/config/MARKER					\
		ln -sf %{_kernelsrcdir}/scripts o/scripts			\
	%endif									\
										\
	%{__make} -C %{_kernelsrcdir} modules					\\\
		M=$PWD O=$PWD/o							\\\
		%{-1*} %{-2*} %{-3*} %{-4*} %{-5*} %{-6*} %{-7*} %{-8*} %{-9*}	\\\
		%{MakeOpts}							\
										\
	for MODULE in %{*}; do							\
		mv ${MODULE}{,-$cfg}.ko						\
	done									\
done										\
%{nil}

From glen at delfi.ee  Tue Nov  7 19:40:12 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Tue, 7 Nov 2006 20:40:12 +0200
Subject: RFC: Macro for building modules
In-Reply-To: <20061107172219.GA19010@pld-linux.org>
References: <20061107172219.GA19010@pld-linux.org>
Message-ID: <200611072040.12404.glen@delfi.ee>

On Tuesday 07 November 2006 19:22, sparky at pld-linux.org wrote:
> Right now we are copying code for building kernel 2.6 modules to each
> spec which provides them. It is a little hard to maintain and to add
> some functionality. I have created macro for building modules based on
> that code, a little expanded (with cross-compilation support). Now in
> spec file thete will only have to be some thing like:
>
> %if %{with kernel}
> cd module/place
> %build_kernel_modules module_name1 module_name2
> cd ../..
> %endif
could you post the sample specfile and initial macros?

> If some additional options for make modules are required they may be
> passed like: -1 option1 -2 option2 -3 option3; up to 9 options are
> supported.
i belive this "limitation" can be avoided by using %* macro for parameters and 
say -m modulename1,modulename2 for passing module name(s)


> Exemple when some additional option in make modules is required:
>
> %build_kernel_modules -1 T=$TMPDIR mol
>
>
> There may be 2 options passed to make prepare scripts: -p o1 -P o2.
> Module patching is supported via shell script called pld-module.patch.
>
> Exemple with additional patching:
>
> %if %{with kernel}
> cp -f pwc-if.c pwc-if.c.orig
>
> cat > pld-module.patch <<'EOF'
perhaps name it with .sh extension?


-- 
glen


From sparky at pld-linux.org  Tue Nov  7 20:46:16 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Tue, 7 Nov 2006 20:46:16 +0100
Subject: RFC: Macro for building modules
In-Reply-To: <200611072040.12404.glen@delfi.ee>
References: <20061107172219.GA19010@pld-linux.org>
	<200611072040.12404.glen@delfi.ee>
Message-ID: <20061107194616.GA1546@pld-linux.org>

On Tue, Nov 07, 2006 at 08:40:12PM +0200, Elan Ruusam?e wrote:
> On Tuesday 07 November 2006 19:22, sparky at pld-linux.org wrote:
> > Right now we are copying code for building kernel 2.6 modules to each
> > spec which provides them. It is a little hard to maintain and to add
> > some functionality. I have created macro for building modules based on
> > that code, a little expanded (with cross-compilation support). Now in
> > spec file thete will only have to be some thing like:
> >
> > %if %{with kernel}
> > cd module/place
> > %build_kernel_modules module_name1 module_name2
> > cd ../..
> > %endif
> could you post the sample specfile and initial macros?

spec for ieee80211, and initial macros, you mean SOURCES/rpm.macros ?
for now I'm testing it in my ~/.rpmmacros

> > If some additional options for make modules are required they may be
> > passed like: -1 option1 -2 option2 -3 option3; up to 9 options are
> > supported.
> i belive this "limitation" can be avoided by using %* macro for parameters and 
> say -m modulename1,modulename2 for passing module name(s)

good point, changed

> > Exemple when some additional option in make modules is required:
> >
> > %build_kernel_modules -1 T=$TMPDIR mola

now it becomes:
%build_kernel_modules T=$TMPDIR -m mol

> > There may be 2 options passed to make prepare scripts: -p o1 -P o2.
> > Module patching is supported via shell script called pld-module.patch.
> >
> > Exemple with additional patching:
> >
> > %if %{with kernel}
> > cp -f pwc-if.c pwc-if.c.orig
> >
> > cat > pld-module.patch <<'EOF'
> perhaps name it with .sh extension?

I knew pld-module.patch won't be the best name just after it came to my
mind, changed.


What I don't understand is why %{error: blabla} does not stop building,
it's just displayed at the beginning.

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org
-------------- next part --------------
#
%build_kernel_modules(p:P:m:)							\
%{!?-m:%{error:%{0}: Required module name/list missing}}			\
										\
%define Opts 									\\\\\\\
%if "%{_target_base_arch}" != "%{_arch}"					\\\
	%if "%{_arch}" == "x86_64" && "%{_target_base_arch}" == "i386"		\\\
	CC="%{__cc}" CPP="%{__cpp}" ARCH=%{_target_base_arch}			\\\
	%else									\\\
	ARCH=%{_target_base_arch} CROSS_COMPILE=%{_target_cpu}-pld-linux-	\\\
	%endif									\\\
%else										\\\
	CC="%{__cc}" CPP="%{__cpp}"						\\\
%endif										\
%define	MakeOpts HOSTCC="%{__cc}" SYSSRC=%{_kernelsrcdir} SYSOUT=$PWD/o		\\\\\\\
		O=$PWD/o %{?with_verbose:V=1} %{Opts}				\
										\
for cfg in %{?with_dist_kernel:%{?with_smp:smp} up}%{!?with_dist_kernel:nondist}; do \
	if [ ! -r "%{_kernelsrcdir}/config-$cfg" ]; then			\
		exit 1								\
	fi									\
	install -d o/include/linux						\
	ln -sf %{_kernelsrcdir}/config-$cfg o/.config				\
	ln -sf %{_kernelsrcdir}/Module.symvers-$cfg o/Module.symvers		\
	ln -sf %{_kernelsrcdir}/include/linux/autoconf-$cfg.h o/include/linux/autoconf.h \
										\
	[ ! -e pld-module.sh ] || %__spec_build_shell pld-module.sh		\
										\
	%if %{with dist_kernel}							\
		%{__make} -j1 -C %{_kernelsrcdir} prepare scripts		\\\
			%{-p*} %{-P*}						\\\
			%{MakeOpts}						\
	%else									\
		install -d o/include/config					\
		touch o/include/config/MARKER					\
		ln -sf %{_kernelsrcdir}/scripts o/scripts			\
	%endif									\
										\
	%{__make} -C %{_kernelsrcdir} clean					\\\
		RCS_FIND_IGNORE="-name '*.ko' -o"				\\\
		M=$PWD %{MakeOpts}						\
	%{__make} -C %{_kernelsrcdir} modules					\\\
		%{*}								\\\
		M=$PWD %{MakeOpts}						\
										\
	for MODULE in {%{-m*},}; do						\
		[ -z "${MODULE}" ] || mv ${MODULE}{,-$cfg}.ko			\
	done									\
done										\
%{nil}
-------------- next part --------------
# $Revision: 1.28 $, $Date: 2006/09/06 13:48:33 $
#
# Conditional build:
%bcond_without	dist_kernel	# allow non-distribution kernel
%bcond_without	smp		# don't build SMP module
%bcond_with	verbose		# verbose build (V=1)
#
%ifarch sparc
%undefine	with_smp
%endif
#
%define		modname	ieee80211
%define		_rel	1
Summary:	Linux kernel module for the ieee80211 networking stack
Summary(de):	Linux Kernel Treiber f?r den ieee80211 Netz Stapel
Summary(pl):	Modu? j?dra Linuksa do stosu sieciowego ieee80211
Name:		kernel%{_alt_kernel}-net-%{modname}
Version:	1.2.15
Release:	%{_rel}@%{_kernel_ver_str}
License:	GPL v2
Group:		Base/Kernel
Source0:	http://dl.sourceforge.net/ieee80211/%{modname}-%{version}.tgz
# Source0-md5:	499d5272fd1326ae65ebef80d9726e4d
URL:		http://ieee80211.sourceforge.net/
%{?with_dist_kernel:BuildRequires:	kernel%{_alt_kernel}-module-build >= 3:2.6.14}
BuildRequires:	rpmbuild(macros) >= 1.308
BuildRequires:	sed >= 4.0
%{?with_dist_kernel:%requires_releq_kernel_up}
Requires(post,postun):	/sbin/depmod
Requires:	module-init-tools >= 3.2.2-2
BuildRoot:	%{tmpdir}/%{name}-%{version}-root-%(id -u -n)

%description
Linux kernel module for the ieee80211 networking stack.

%description -l de
Linux Kernel Modul f?r den ieee80211 Netz Stapel.

%description -l pl
Modu? j?dra Linuksa do stosu sieciowego ieee80211.

%package -n kernel%{_alt_kernel}-smp-net-%{modname}
Summary:	Linux SMP kernel module for the ieee80211 networking stack
Summary(de):	Linux SMP Kernel Modul f?r den ieee80211 Netz Stapel
Summary(pl):	Modu? j?dra Linuksa SMP do stosu sieciowego ieee80211
Release:	%{_rel}@%{_kernel_ver_str}
Group:		Base/Kernel
%{?with_dist_kernel:%requires_releq_kernel_smp}
Requires(post,postun):	/sbin/depmod
Requires:	module-init-tools >= 3.2.2-2

%description -n kernel%{_alt_kernel}-smp-net-%{modname}
Linux SMP kernel module for the ieee80211 networking stack.

%description -n kernel%{_alt_kernel}-smp-net-%{modname} -l de
Linux SMP Kernel Modul f?r den ieee80211 Netz Stapel.

%description -n kernel%{_alt_kernel}-smp-net-%{modname} -l pl
Modu? j?dra Linuksa SMP do stosu sieciowego ieee80211.

%package -n %{modname}-devel
Summary:	Development header files for the ieee80211 networking stack
Summary(de):	Development Header Dateien f?r den ieee80211 Netz Stapel
Summary(pl):	Pliki nag??wkowe do stosu sieciowego ieee80211
Release:	%{_rel}@%{_kernel_ver_str}
%{?with_dist_kernel:%requires_releq kernel-module-build}
Group:		Development/Libraries

%description -n %{modname}-devel
Development header files for the ieee80211 networking stack.

%description -n %{modname}-devel -l de
Development Header Dateien f?r den ieee80211 Netz Stapel.

%description -n %{modname}-devel -l pl
Pliki nag??wkowe do stosu sieciowego ieee80211.

%prep
%setup -q -n %{modname}-%{version}
%{__sed} -i 's:<net/ieee80211.h>:"net/ieee80211.h":g' *.c

%build
%build_kernel_modules -m ieee80211,ieee80211_crypt{,_wep,_ccmp,_tkip}

%install
rm -rf $RPM_BUILD_ROOT
install -d $RPM_BUILD_ROOT/lib/modules/%{_kernel_ver}{,smp}/misc \
	$RPM_BUILD_ROOT%{_sysconfdir}/modprobe.d/%{_kernel_ver}{,smp} \
	$RPM_BUILD_ROOT%{_kernelsrcdir}/include/net

for MOD in ieee80211 ieee80211_crypt ieee80211_crypt_wep \
		ieee80211_crypt_ccmp ieee80211_crypt_tkip; do
	install $MOD-%{?with_dist_kernel:up}%{!?with_dist_kernel:nondist}.ko \
		$RPM_BUILD_ROOT/lib/modules/%{_kernel_ver}/misc/${MOD}_current.ko
	echo "alias $MOD ${MOD}_current" \
		>> $RPM_BUILD_ROOT%{_sysconfdir}/modprobe.d/%{_kernel_ver}/ieee80211.conf
done

%if %{with smp} && %{with dist_kernel}
for MOD in ieee80211 ieee80211_crypt ieee80211_crypt_wep \
		ieee80211_crypt_ccmp ieee80211_crypt_tkip; do
	install $MOD-smp.ko \
		$RPM_BUILD_ROOT/lib/modules/%{_kernel_ver}smp/misc/${MOD}_current.ko
	echo "alias $MOD ${MOD}_current" \
		>> $RPM_BUILD_ROOT%{_sysconfdir}/modprobe.d/%{_kernel_ver}smp/ieee80211.conf
done
%endif

install net/* \
	$RPM_BUILD_ROOT%{_kernelsrcdir}/include/net

%clean
rm -rf $RPM_BUILD_ROOT

%post	-n kernel%{_alt_kernel}-net-%{modname}
%depmod %{_kernel_ver}

%postun	-n kernel%{_alt_kernel}-net-%{modname}
%depmod %{_kernel_ver}

%post	-n kernel%{_alt_kernel}-smp-net-%{modname}
%depmod %{_kernel_ver}smp

%postun	-n kernel%{_alt_kernel}-smp-net-%{modname}
%depmod %{_kernel_ver}smp

%files -n kernel%{_alt_kernel}-net-%{modname}
%defattr(644,root,root,755)
/lib/modules/%{_kernel_ver}/misc/ieee80211*.ko*
%{_sysconfdir}/modprobe.d/%{_kernel_ver}/ieee80211.conf

%if %{with smp} && %{with dist_kernel}
%files -n kernel%{_alt_kernel}-smp-net-%{modname}
%defattr(644,root,root,755)
/lib/modules/%{_kernel_ver}smp/misc/ieee80211*.ko*
%{_sysconfdir}/modprobe.d/%{_kernel_ver}smp/ieee80211.conf
%endif

%files -n %{modname}-devel
%defattr(644,root,root,755)
# should go to a versioned directory
%{_kernelsrcdir}/include/net/*

%define date	%(echo `LC_ALL="C" date +"%a %b %d %Y"`)
%changelog
* %{date} PLD Team <feedback at pld-linux.org>
All persons listed below can be reached at <cvs_login>@pld-linux.org

$Log: kernel-net-ieee80211.spec,v $
Revision 1.28  2006/09/06 13:48:33  sparky
- mass atack: -j1 for make scripts

Revision 1.27  2006/08/31 21:53:38  glen
- remove excess escaping

Revision 1.26  2006/08/31 14:02:41  glen
- make -j1 for scripts

Revision 1.25  2006/08/21 10:02:25  shadzik
- 1.2.15

Revision 1.24  2006/06/27 21:55:06  lmasko
- alternative kernel support

Revision 1.23  2006/06/12 20:13:53  shadzik
- up to 1.1.14

Revision 1.22  2006/05/09 10:14:08  qboosh
- typo

Revision 1.21  2006/05/01 00:15:22  shadzik
- added german description
- some cleanups
- STBR

Revision 1.20  2006/04/16 21:49:21  lmasko
- up to 1.1.13

Revision 1.19  2006/04/12 12:15:40  glen
- add kernel epoch, adapterized

Revision 1.18  2006/03/12 14:21:10  glen
- module-init-tools with /etc/modprobe.d dir

Revision 1.17  2006/03/12 12:31:50  qboosh
- more updates

Revision 1.16  2006/03/12 11:02:56  blues
- adapt to new building system
  by ?ukasz Ma?ko masko at ipipan dot waw dot pl

Revision 1.15  2006/03/11 12:44:32  blues
- do the aliases theRightWat(TM)

Revision 1.14  2006/03/11 12:11:48  blues
- modules are not in conflict with modules from kernel. release 2.

Revision 1.13  2006/03/06 23:17:20  djurban
- update to 1.1.12

Revision 1.12  2006/02/08 23:02:58  zawadaa
- up to 1.1.11

Revision 1.11  2006/01/22 20:19:13  blues
- 1.0.9 - builds, seems to be working (no chance to test it :(

Revision 1.10  2006/01/15 23:34:35  blues
- don't overwrite kernel module - put it into misc (rel.2)

Revision 1.9  2006/01/12 20:26:23  arekm
- sparc/ppc fixes

Revision 1.8  2006/01/09 23:52:38  zawadaa
- up to 1.1.8

Revision 1.7  2006/01/07 19:36:31  zawadaa
- up to 1.1.7

Revision 1.6  2005/10/25 15:43:46  zawadaa
- up to 1.1.6, test...

Revision 1.5  2005/09/16 13:06:35  jpc
- make -devel depend on kernel-module-build version

Revision 1.4  2005/09/15 18:25:05  sparky
- prepared for kernel >= 2.6.13: link Module.symvers

Revision 1.3  2005/08/29 23:34:26  jpc
- sort R:
- move the ieee80211 includes to %%{_kernelsrcdir}

Revision 1.2  2005/07/21 18:51:33  qboosh
- pl, typos

Revision 1.1  2005/07/20 15:23:16  jpc
- new spec based on ipw2200.spec

From glen at delfi.ee  Tue Nov  7 21:52:51 2006
From: glen at delfi.ee (Elan =?iso-8859-15?q?Ruusam=E4e?=)
Date: Tue, 7 Nov 2006 23:52:51 +0300
Subject: RFC: Macro for building modules
In-Reply-To: <20061107194616.GA1546@pld-linux.org>
References: <20061107172219.GA19010@pld-linux.org>
	<200611072040.12404.glen@delfi.ee>
	<20061107194616.GA1546@pld-linux.org>
Message-ID: <200611072252.51610.glen@delfi.ee>

On Tuesday 07 November 2006 21:46, sparky at pld-linux.org wrote:
> > > cat > pld-module.patch <<'EOF'
> >
> > perhaps name it with .sh extension?
>
> I knew pld-module.patch won't be the best name just after it came to my
> mind, changed.
just another idea, could use optional <<EOF for "patching section"

%build_kernel_module -m foo <<EOF
some scripting around here
EOF

something similar that %banner does.

hint: 'cat somescript.sh | sh' invokes 'somescript.sh'

> What I don't understand is why %{error: blabla} does not stop building,
> it's just displayed at the beginning.

i once asked the same from jbj, he replied me with a patch. i didn't apply it 
that time. but the similar patch is taken (attached) from openpkg.

oh, and don't use %Opts, it could conflict with spec macros. 
(hint: %__build_kernel_module_opts)

-- 
glen
-------------- next part --------------
A non-text attachment was scrubbed...
Name: rpm-errormacro.patch
Type: text/x-diff
Size: 612 bytes
Desc: not available
URL: </mailman/pipermail/pld-devel-en/attachments/20061107/55f5bcd9/attachment.bin>

From qboosh at pld-linux.org  Tue Nov  7 21:53:30 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 7 Nov 2006 21:53:30 +0100
Subject: php 5.2
In-Reply-To: <200611030040.03319.glen@delfi.ee>
References: <200611030040.03319.glen@delfi.ee>
Message-ID: <20061107205330.GA29880@fngna.oyu>

On Fri, Nov 03, 2006 at 01:40:03AM +0300, Elan Ruusam?e wrote:
> hi
> 
> in php 5.2 pcre extension needs to be compiled statically as libphp_common.so 
> uses symbols from it.
> 
> ./.libs/libphp_common.so: undefined reference to 
> `pcre_get_compiled_regex_cache'
> ./.libs/libphp_common.so: undefined reference to `php_pcre_replace_impl'
> ./.libs/libphp_common.so: undefined reference to `pcre_get_compiled_regex'
> ./.libs/libphp_common.so: undefined reference to `php_pcre_split_impl'
> ./.libs/libphp_common.so: undefined reference to `php_pcre_match_impl'
> collect2: ld returned 1 exit status
> 
> but adding Provides:  %{name}-pcre = %{epoch}:%{version}-%{release} will not 
> make poldek happy, and upgrade will break.

It requires Obsoletes to work.

> error: php-common = 4:5.1.6-1 is required by php-pcre-5.1.6-1
> error: php-common = 4:5.1.6-1 is required by php-pcre-5.1.6-1
> 
> $ rpm -qp ../RPMS/php-common-5.2.0-0.1.i686.rpm --provides|grep pcre
> php-pcre = 4:5.2.0-0.1
> 
> probably php-common should obsolete php-pcre? like:
> Obsoletes:  php-pcre < 4:5.2.0
> 
> but how to do that so that it won't obsolete php4-pcre which has:
> $ q php4-pcre --provides
> php-pcre = 3:4.4.4-5
> 
> bump epoch+=2 so that php4 has epoch=5 and php5 has epoch=6?

No, very ugly.

Maybe let's drop "php-module" Provides in php4 and use (versioned?)
"php(module)" Provides in both versions (and update deps in other
packages)?
Using virtual Provides with other _real_ package name brings many
troubles and should be avoided.


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


From qboosh at pld-linux.org  Tue Nov  7 22:05:07 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 7 Nov 2006 22:05:07 +0100
Subject: SPECS: mc.spec - i believe -bash-3.2.patch breaks elder bash?
In-Reply-To: <SPECS_mc.spec.1.245@pld-linux.org>
References: <SPECS_mc.spec.1.244@pld-linux.org>
	<SPECS_mc.spec.1.245@pld-linux.org>
Message-ID: <20061107210506.GA7905@gruby.cs.net.pl>

On Tue, Nov 07, 2006 at 09:54:28PM +0100, glen wrote:
> Author: glen                         Date: Tue Nov  7 20:54:28 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - i believe -bash-3.2.patch breaks elder bash?

bash < 2.05b, not < 3.2
See bash's NEWS:

[...]
This is a terse description of the new features added to bash-2.05b since
the release of bash-2.05a.  As always, the manual page (doc/bash.1) is
the place to look for complete descriptions.
[...]
ee. The `echo' builtin now accepts \0xxx (zero to three octal digits following
    the `0') in addition to \xxx (one to three octal digits) for SUSv3/XPG6/
    POSIX.1-2001 compliance.


BTW: could sb check these standards if they allow to omit leading '0'?
If yes, there is serious regression. I think it breaks not only mc.


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


From sparky at pld-linux.org  Tue Nov  7 23:49:54 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Tue, 7 Nov 2006 23:49:54 +0100
Subject: RFC: Macro for building modules
In-Reply-To: <200611072252.51610.glen@delfi.ee>
References: <20061107172219.GA19010@pld-linux.org>
	<200611072040.12404.glen@delfi.ee>
	<20061107194616.GA1546@pld-linux.org>
	<200611072252.51610.glen@delfi.ee>
Message-ID: <20061107224954.GA13060@pld-linux.org>

On Tue, Nov 07, 2006 at 11:52:51PM +0300, Elan Ruusam?e wrote:

> just another idea, could use optional <<EOF for "patching section"
> 
> %build_kernel_module -m foo <<EOF
> some scripting around here
> EOF

ok, got something

> oh, and don't use %Opts, it could conflict with spec macros. 
> (hint: %__build_kernel_module_opts)

it isn't any problem, macros defined inside other macro are visible only
internally

the only problem with internal macros is they are expanded before anything,
that's why %if is placed inside Opts definition and not viceversa

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org
-------------- next part --------------
#
%build_kernel_modules(p:P:m:)							\
%{!?-m:%{error:%{0}: Required module name/list missing} exit 1}			\
										\
%define Opts 									\\\\\\\
%if "%{_target_base_arch}" != "%{_arch}"					\\\
	%if "%{_arch}" == "x86_64" && "%{_target_base_arch}" == "i386"		\\\
	CC="%{__cc}" CPP="%{__cpp}" ARCH=%{_target_base_arch}			\\\
	%else									\\\
	ARCH=%{_target_base_arch} CROSS_COMPILE=%{_target_cpu}-pld-linux-	\\\
	%endif									\\\
%else										\\\
	CC="%{__cc}" CPP="%{__cpp}"						\\\
%endif										\
%define	MakeOpts HOSTCC="%{__cc}" SYSSRC=%{_kernelsrcdir} SYSOUT=$PWD/o		\\\\\\\
		O=$PWD/o %{?with_verbose:V=1} %{Opts}				\
compile() {									\
	L="<"; [[ '%{*}' != *$L$L* ]] || PATCH_SH="set -x -e;$(cat)"		\
	set -e -x								\
										\
for cfg in %{?with_dist_kernel:%{?with_smp:smp} up}%{!?with_dist_kernel:nondist}; do \
	if [ ! -r "%{_kernelsrcdir}/config-$cfg" ]; then			\
		exit 1								\
	fi									\
	install -d o/include/linux						\
	ln -sf %{_kernelsrcdir}/config-$cfg o/.config				\
	ln -sf %{_kernelsrcdir}/Module.symvers-$cfg o/Module.symvers		\
	ln -sf %{_kernelsrcdir}/include/linux/autoconf-$cfg.h o/include/linux/autoconf.h \
										\
	set +x									\
	[ -z "$PATCH_SH" ] || echo "$PATCH_SH" | %__spec_build_shell		\
	set -x									\
										\
	%if %{with dist_kernel}							\
		%{__make} -j1 -C %{_kernelsrcdir} prepare scripts		\\\
			%{-p*} %{-P*}						\\\
			%{MakeOpts}						\
	%else									\
		install -d o/include/config					\
		touch o/include/config/MARKER					\
		ln -sf %{_kernelsrcdir}/scripts o/scripts			\
	%endif									\
										\
	%{__make} -C %{_kernelsrcdir} clean					\\\
		RCS_FIND_IGNORE="-name '*.ko' -o"				\\\
		M=$PWD %{MakeOpts}						\
	%{__make} -C %{_kernelsrcdir} modules					\\\
		${1+"$@"}							\\\
		M=$PWD %{MakeOpts}						\
										\
	for MODULE in {%{-m*},}; do						\
		[ -z "${MODULE}" ] || mv ${MODULE}{,-$cfg}.ko			\
	done									\
done										\
}										\
compile %{*}									\
%{nil}

From glen at delfi.ee  Wed Nov  8 00:06:38 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 8 Nov 2006 02:06:38 +0300
Subject: php 5.2
In-Reply-To: <20061107205330.GA29880@fngna.oyu>
References: <200611030040.03319.glen@delfi.ee>
	<20061107205330.GA29880@fngna.oyu>
Message-ID: <200611080106.38176.glen@delfi.ee>

On Tuesday 07 November 2006 22:53, Jakub Bogusz wrote:
> > but how to do that so that it won't obsolete php4-pcre which has:
> > $ q php4-pcre --provides
> > php-pcre = 3:4.4.4-5
> >
> > bump epoch+=2 so that php4 has epoch=5 and php5 has epoch=6?
>
> No, very ugly.
agree.

> Maybe let's drop "php-module" Provides in php4 and use (versioned?)
> "php(module)" Provides in both versions (and update deps in other
> packages)?
> Using virtual Provides with other _real_ package name brings many
> troubles and should be avoided.

versions are needed, some packages work with php4, some with php5.
but that could be solved by macthing php-common >= 4:4.4.3 or the like...
so the versions in virtuals aren't necessarily needed...

haven't seen yet where there could be need to specify:
php-common >= 4:4.1.0
php-pcre >= 4:4.4.0

but if to go with versions in virtuals, then it's going to be like :
php(pcre) = 5.2.0?

ie no epoches around those virtuals?

-- 
glen


From qboosh at pld-linux.org  Wed Nov  8 00:47:04 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 00:47:04 +0100
Subject: php 5.2
In-Reply-To: <200611080106.38176.glen@delfi.ee>
References: <200611030040.03319.glen@delfi.ee>
	<20061107205330.GA29880@fngna.oyu>
	<200611080106.38176.glen@delfi.ee>
Message-ID: <20061107234704.GC29880@fngna.oyu>

On Wed, Nov 08, 2006 at 02:06:38AM +0300, Elan Ruusam?e wrote:
[...]
> but if to go with versions in virtuals, then it's going to be like :
> php(pcre) = 5.2.0?
> 
> ie no epoches around those virtuals?

Sure, no epochs. That's the next advantage of using virtual name
different than any real package name.


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


From jajcus at jajcus.net  Wed Nov  8 08:37:17 2006
From: jajcus at jajcus.net (Jacek Konieczny)
Date: Wed, 8 Nov 2006 08:37:17 +0100
Subject: glibc compilation problem
In-Reply-To: <20061106143238.GH30585@jajo.axeos-gliwice>
References: <20061106143238.GH30585@jajo.axeos-gliwice>
Message-ID: <20061108073717.GA17125@jajo.axeos-gliwice>

On Mon, Nov 06, 2006 at 03:32:39PM +0100, Jacek Konieczny wrote:
> Hello,
> 
> I have encountered a big problem with glibc compilation on
> PLD-Th-derived system:

[...]

Problem solved.

It was '-feliminate-dwarf2-dups' still used in CFLAGS by my RPM.

Greets,
        Jacek


From arekm at pld-linux.org  Wed Nov  8 10:37:33 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Wed, 8 Nov 2006 10:37:33 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
Message-ID: <200611081037.33841.arekm@pld-linux.org>


We use tons of brace expansions (like man{1,5}) in our specs. Unfortunately 
there is a problem since POSIX/SUSv3 shell specification doesn't say a word 
about {} expansion. What now - should {} in spec files be forbidden?

I was thinking about switching from buggy and unmaintained pdksh to some other 
shell for Th than pdksh.

I've tested zsh as /bin/sh and {} issue came up since zsh posix compatible 
implementation is stricter than pdksh one. zsh has switch for that 
(no_ignore_brace) but... there is no single configuration file that is read 
by zsh in /bin/sh emulation mode).

Well, there is only bash left as a option for /bin/sh it seems ;-)

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


From arekm at pld-linux.org  Wed Nov  8 11:29:51 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Wed, 8 Nov 2006 11:29:51 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <200611081037.33841.arekm@pld-linux.org>
References: <200611081037.33841.arekm@pld-linux.org>
Message-ID: <200611081129.51844.arekm@pld-linux.org>

On Wednesday 08 November 2006 10:37, Arkadiusz Miskiewicz wrote:

> Well, there is only bash left as a option for /bin/sh it seems ;-)

Interesting test suite: 
http://www.feyrer.de/NetBSD/blog.html/nb_20050320_1408.html

[arekm at tarm ~]$ /bin/bash ifs.sh
# tests 6856 passed 6856 failed 0

/bin/sh == bash
[arekm at tarm ~]$ /bin/sh ifs.sh
# tests 6856 passed 6856 failed 0

[arekm at tarm ~]$ /bin/ksh ifs.sh | grep '^#'
# tests 6856 passed 2584 failed 4272

/bin/sh == pdksh
[arekm at tarm ~]$ /bin/sh ifs.sh | grep '^#'
# tests 6856 passed 2584 failed 4272

[arekm at tarm ~]$ /bin/zsh ifs.sh | grep '^#'
# tests 664 passed 72 failed 592 -- fundamental IFS error -- 6856 tests 
expected

/bin/sh == zsh
[arekm at tarm ~]$ /bin/sh ifs.sh | grep '^#'
# tests 6856 passed 3560 failed 3296

heirloom-sh
[arekm at tarm ~]$ /bin/jsh ifs.sh | grep '^#'
ifs.sh: bad substitution

dash
[arekm at tarm ~]$ /bin/dash ifs.sh 2>&1 | grep '^#'
# tests 6856 passed 3944 failed 2912 ordered 1336

posh
[arekm at tarm ~]$ /bin/posh ifs.sh 2>&1 | grep '^#'
# tests 6856 passed 2584 failed 4272


Welcome back bash as /bin/sh?

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


From qboosh at pld-linux.org  Wed Nov  8 11:49:38 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 11:49:38 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <200611081129.51844.arekm@pld-linux.org>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
Message-ID: <20061108104938.GA21795@gruby.cs.net.pl>

On Wed, Nov 08, 2006 at 11:29:51AM +0100, Arkadiusz Miskiewicz wrote:
> On Wednesday 08 November 2006 10:37, Arkadiusz Miskiewicz wrote:
> 
> > Well, there is only bash left as a option for /bin/sh it seems ;-)
> 
> Interesting test suite: 
> http://www.feyrer.de/NetBSD/blog.html/nb_20050320_1408.html

But it tests only IFS compliance.
What about all other functionality?
bash is not 100% POSIX compliant, although (at least) some of issues are
commonly accepted (e.g. no escape expansion in echo without -e switch).


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


From arekm at pld-linux.org  Wed Nov  8 11:54:33 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Wed, 8 Nov 2006 11:54:33 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <20061108104938.GA21795@gruby.cs.net.pl>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
	<20061108104938.GA21795@gruby.cs.net.pl>
Message-ID: <200611081154.33422.arekm@pld-linux.org>

On Wednesday 08 November 2006 11:49, Jakub Bogusz wrote:
> On Wed, Nov 08, 2006 at 11:29:51AM +0100, Arkadiusz Miskiewicz wrote:
> > On Wednesday 08 November 2006 10:37, Arkadiusz Miskiewicz wrote:
> > > Well, there is only bash left as a option for /bin/sh it seems ;-)
> >
> > Interesting test suite:
> > http://www.feyrer.de/NetBSD/blog.html/nb_20050320_1408.html
>
> But it tests only IFS compliance.
> What about all other functionality?

Haven't seen tests for other functionality.

> bash is not 100% POSIX compliant, although (at least) some of issues are
> commonly accepted (e.g. no escape expansion in echo without -e switch).

It seems that there is no single POSIX compilant shell. I wonder which is 
closest to posix compilance?

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


From glen at delfi.ee  Wed Nov  8 14:48:33 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 8 Nov 2006 15:48:33 +0200
Subject: php 5.2
In-Reply-To: <20061107234704.GC29880@fngna.oyu>
References: <200611030040.03319.glen@delfi.ee>
	<200611080106.38176.glen@delfi.ee>
	<20061107234704.GC29880@fngna.oyu>
Message-ID: <200611081548.33852.glen@delfi.ee>

On Wednesday 08 November 2006 01:47, Jakub Bogusz wrote:
> > but if to go with versions in virtuals, then it's going to be like :
> > php(pcre) = 5.2.0?
> >
> > ie no epoches around those virtuals?
>
> Sure, no epochs. That's the next advantage of using virtual name
> different than any real package name.
i decided not to use versions for php(extension) provides until it's 
neccessary.

but what about the already existing 'php' virtual which currently is used 
as 'any php engine in webserver':

$ ac-provides php
apache-mod_php-5.1.6-2
apache-mod_php4-4.4.4-5
apache1-mod_php-5.1.6-2
apache1-mod_php4-4.4.4-5
php-fcgi-5.1.6-2
php4-fcgi-4.4.4-5
$

why asking, is that it includes right now epoch

-- 
glen


From qboosh at pld-linux.org  Wed Nov  8 14:51:20 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 14:51:20 +0100
Subject: php 5.2
In-Reply-To: <200611081548.33852.glen@delfi.ee>
References: <200611030040.03319.glen@delfi.ee>
	<200611080106.38176.glen@delfi.ee>
	<20061107234704.GC29880@fngna.oyu>
	<200611081548.33852.glen@delfi.ee>
Message-ID: <20061108135120.GE21795@gruby.cs.net.pl>

On Wed, Nov 08, 2006 at 03:48:33PM +0200, Elan Ruusam?e wrote:
> On Wednesday 08 November 2006 01:47, Jakub Bogusz wrote:
> > > but if to go with versions in virtuals, then it's going to be like :
> > > php(pcre) = 5.2.0?
> > >
> > > ie no epoches around those virtuals?
> >
> > Sure, no epochs. That's the next advantage of using virtual name
> > different than any real package name.
> i decided not to use versions for php(extension) provides until it's 
> neccessary.
> 
> but what about the already existing 'php' virtual which currently is used 
> as 'any php engine in webserver':
> 
> $ ac-provides php
> apache-mod_php-5.1.6-2
> apache-mod_php4-4.4.4-5
> apache1-mod_php-5.1.6-2
> apache1-mod_php4-4.4.4-5
> php-fcgi-5.1.6-2
> php4-fcgi-4.4.4-5
> $
> 
> why asking, is that it includes right now epoch

Because it used to be a name of real package.
Maybe let's use (versioned without epoch) webserver-php?


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


From glen at delfi.ee  Wed Nov  8 14:56:53 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 8 Nov 2006 15:56:53 +0200
Subject: php 5.2
In-Reply-To: <20061108135120.GE21795@gruby.cs.net.pl>
References: <200611030040.03319.glen@delfi.ee>
	<200611081548.33852.glen@delfi.ee>
	<20061108135120.GE21795@gruby.cs.net.pl>
Message-ID: <200611081556.53345.glen@delfi.ee>

On Wednesday 08 November 2006 15:51, Jakub Bogusz wrote:
> > but what about the already existing 'php' virtual which currently is used
> > as 'any php engine in webserver':
> >
> > $ ac-provides php
> > apache-mod_php-5.1.6-2
> > apache-mod_php4-4.4.4-5
> > apache1-mod_php-5.1.6-2
> > apache1-mod_php4-4.4.4-5
> > php-fcgi-5.1.6-2
> > php4-fcgi-4.4.4-5
> > $
> >
> > why asking, is that it includes right now epoch
>
> Because it used to be a name of real package.
i know that. it happened to be me who split it into apache1-mod_php and 
apache-mod_php, php-fcgi came later for lighttpd ;)

> Maybe let's use (versioned without epoch) webserver-php?
maybe even webserver(php)  ?

as currently there exist:
$ grep ^Provides.*webserver apache1.spec
Provides:       webserver = apache
Provides:       webserver(access)
Provides:       webserver(alias)
Provides:       webserver(auth)
Provides:       webserver(cgi)
Provides:       webserver(indexfile)

$ ac-provides 'webserver(access)'
apache-mod_authz_host-2.2.3-5
apache1-mod_access-1.3.37-5
lighttpd-mod_access-1.4.13-2

and webserver(php) must be versioned.

-- 
glen


From qboosh at pld-linux.org  Wed Nov  8 15:00:32 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 15:00:32 +0100
Subject: php 5.2
In-Reply-To: <200611081556.53345.glen@delfi.ee>
References: <200611030040.03319.glen@delfi.ee>
	<200611081548.33852.glen@delfi.ee>
	<20061108135120.GE21795@gruby.cs.net.pl>
	<200611081556.53345.glen@delfi.ee>
Message-ID: <20061108140032.GF21795@gruby.cs.net.pl>

On Wed, Nov 08, 2006 at 03:56:53PM +0200, Elan Ruusam?e wrote:
> On Wednesday 08 November 2006 15:51, Jakub Bogusz wrote:
> > > but what about the already existing 'php' virtual which currently is used
> > > as 'any php engine in webserver':
> > >
> > > $ ac-provides php
> > > apache-mod_php-5.1.6-2
> > > apache-mod_php4-4.4.4-5
> > > apache1-mod_php-5.1.6-2
> > > apache1-mod_php4-4.4.4-5
> > > php-fcgi-5.1.6-2
> > > php4-fcgi-4.4.4-5
> > > $
> > >
> > > why asking, is that it includes right now epoch
> >
> > Because it used to be a name of real package.
> i know that. it happened to be me who split it into apache1-mod_php and 
> apache-mod_php, php-fcgi came later for lighttpd ;)
> 
> > Maybe let's use (versioned without epoch) webserver-php?
> maybe even webserver(php)  ?

OK.


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


From patrys at pld-linux.org  Wed Nov  8 20:21:01 2006
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Wed, 08 Nov 2006 20:21:01 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <200611081154.33422.arekm@pld-linux.org>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
	<20061108104938.GA21795@gruby.cs.net.pl>
	<200611081154.33422.arekm@pld-linux.org>
Message-ID: <1163013661.6292.1.camel@localhost>

Dnia 08-11-2006, ?ro o godzinie 11:54 +0100, Arkadiusz Miskiewicz
napisa?(a):
> On Wednesday 08 November 2006 11:49, Jakub Bogusz wrote:
> > But it tests only IFS compliance.
> > What about all other functionality?
> Haven't seen tests for other functionality.

Should we aim at perfect POSIX compatibility or perfect usability? Bash
does not seem to be a heavy dependency and is actively maintained (so
the compatibility issues will go away over time as opposed to pdksh).

> > bash is not 100% POSIX compliant, although (at least) some of issues are
> > commonly accepted (e.g. no escape expansion in echo without -e switch).
> It seems that there is no single POSIX compilant shell. I wonder which is 
> closest to posix compilance?

As above.

-- 
Patryk Zawadzki <patrys at pld-linux.org>
PLD Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: To jest cz??? listu	podpisana cyfrowo
URL: </mailman/pipermail/pld-devel-en/attachments/20061108/8a1897a0/attachment.sig>

From qboosh at pld-linux.org  Wed Nov  8 20:26:50 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 20:26:50 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <1163013661.6292.1.camel@localhost>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
	<20061108104938.GA21795@gruby.cs.net.pl>
	<200611081154.33422.arekm@pld-linux.org>
	<1163013661.6292.1.camel@localhost>
Message-ID: <20061108192650.GA7979@fngna.oyu>

On Wed, Nov 08, 2006 at 08:21:01PM +0100, Patryk Zawadzki wrote:
> Dnia 08-11-2006, ?ro o godzinie 11:54 +0100, Arkadiusz Miskiewicz
> napisa?(a):
> > On Wednesday 08 November 2006 11:49, Jakub Bogusz wrote:
> > > But it tests only IFS compliance.
> > > What about all other functionality?
> > Haven't seen tests for other functionality.
> 
> Should we aim at perfect POSIX compatibility or perfect usability? Bash
> does not seem to be a heavy dependency and is actively maintained (so
> the compatibility issues will go away over time as opposed to pdksh).

And POSIX-compatibility of scripts will go away as nobody will care
about.


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


From sparky at pld-linux.org  Wed Nov  8 20:54:23 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Wed, 8 Nov 2006 20:54:23 +0100
Subject: SPECS: dunelegacy.spec - updated to 0.94.1 - use scons in place
	of...
In-Reply-To: <SPECS_dunelegacy.spec.1.4@pld-linux.org>
References: <SPECS_dunelegacy.spec.1.3@pld-linux.org>
	<SPECS_dunelegacy.spec.1.4@pld-linux.org>
Message-ID: <20061108195423.GA27275@pld-linux.org>

On Wed, Nov 08, 2006 at 08:32:36PM +0100, lisu wrote:
> Author: lisu                         Date: Wed Nov  8 19:32:36 2006 GMT
> Module: SPECS                         Tag: HEAD

> +scons

I H8 that thing...

is there some easy, gelerally working way to pass CC, CXX, CFLAGS,
CXXFLAGS and LDFLAGS to scons ?

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From glen at delfi.ee  Wed Nov  8 21:02:08 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 8 Nov 2006 22:02:08 +0200
Subject: SPECS: dunelegacy.spec - updated to 0.94.1 - use scons in place
	of...
In-Reply-To: <20061108195423.GA27275@pld-linux.org>
References: <SPECS_dunelegacy.spec.1.3@pld-linux.org>
	<SPECS_dunelegacy.spec.1.4@pld-linux.org>
	<20061108195423.GA27275@pld-linux.org>
Message-ID: <200611082202.08339.glen@delfi.ee>

On Wednesday 08 November 2006 21:54, sparky at pld-linux.org wrote:
> I H8 that thing...
>
> is there some easy, gelerally working way to pass CC, CXX, CFLAGS,
> CXXFLAGS and LDFLAGS to scons ?
from scons.1:

       The user could specify the C compiler on the command line:

              scons "CC=my_cc"


that doesn't work?

-- 
glen


From sparky at pld-linux.org  Wed Nov  8 21:07:33 2006
From: sparky at pld-linux.org (sparky at pld-linux.org)
Date: Wed, 8 Nov 2006 21:07:33 +0100
Subject: SPECS: dunelegacy.spec - updated to 0.94.1 - use scons in place
	of...
In-Reply-To: <200611082202.08339.glen@delfi.ee>
References: <SPECS_dunelegacy.spec.1.3@pld-linux.org>
	<SPECS_dunelegacy.spec.1.4@pld-linux.org>
	<20061108195423.GA27275@pld-linux.org>
	<200611082202.08339.glen@delfi.ee>
Message-ID: <20061108200733.GA2043@pld-linux.org>

On Wed, Nov 08, 2006 at 10:02:08PM +0200, Elan Ruusam?e wrote:
> On Wednesday 08 November 2006 21:54, sparky at pld-linux.org wrote:
> > I H8 that thing...
> >
> > is there some easy, gelerally working way to pass CC, CXX, CFLAGS,
> > CXXFLAGS and LDFLAGS to scons ?
> from scons.1:
> 
>        The user could specify the C compiler on the command line:
> 
>               scons "CC=my_cc"
> 
> 
> that doesn't work?

normally not

and if CC is specified in SConstruct file it won't be overwritten

and inside there are those weird env[] things I don't know how to patch

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From patrys at pld-linux.org  Wed Nov  8 21:46:54 2006
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Wed, 08 Nov 2006 21:46:54 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <20061108192650.GA7979@fngna.oyu>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
	<20061108104938.GA21795@gruby.cs.net.pl>
	<200611081154.33422.arekm@pld-linux.org>
	<1163013661.6292.1.camel@localhost> <20061108192650.GA7979@fngna.oyu>
Message-ID: <1163018814.6292.5.camel@localhost>

Dnia 08-11-2006, ?ro o godzinie 20:26 +0100, Jakub Bogusz napisa?(a):
> On Wed, Nov 08, 2006 at 08:21:01PM +0100, Patryk Zawadzki wrote:
> > Dnia 08-11-2006, ?ro o godzinie 11:54 +0100, Arkadiusz Miskiewicz
> > napisa?(a):
> > > On Wednesday 08 November 2006 11:49, Jakub Bogusz wrote:
> > > > But it tests only IFS compliance.
> > > > What about all other functionality?
> > > Haven't seen tests for other functionality.
> > 
> > Should we aim at perfect POSIX compatibility or perfect usability? Bash
> > does not seem to be a heavy dependency and is actively maintained (so
> > the compatibility issues will go away over time as opposed to pdksh).
> 
> And POSIX-compatibility of scripts will go away as nobody will care
> about.

To me it sounds similar to the sparc/alpha problem. If someone uses a
POSIX-only shell (which does not exist), she will complain and things
will get fixed.

If noone complains then it is a sign that noone cares and POSIX
compatibility is done just for the sake of doing it.

-- 
Patryk Zawadzki <patrys at pld-linux.org>
PLD Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: To jest cz??? listu	podpisana cyfrowo
URL: </mailman/pipermail/pld-devel-en/attachments/20061108/80d6a1cc/attachment.sig>

From qboosh at pld-linux.org  Wed Nov  8 22:35:16 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 8 Nov 2006 22:35:16 +0100
Subject: /bin/sh shell - the posix shell and {} braces expansion case
In-Reply-To: <1163018814.6292.5.camel@localhost>
References: <200611081037.33841.arekm@pld-linux.org>
	<200611081129.51844.arekm@pld-linux.org>
	<20061108104938.GA21795@gruby.cs.net.pl>
	<200611081154.33422.arekm@pld-linux.org>
	<1163013661.6292.1.camel@localhost>
	<20061108192650.GA7979@fngna.oyu>
	<1163018814.6292.5.camel@localhost>
Message-ID: <20061108213516.GB7979@fngna.oyu>

On Wed, Nov 08, 2006 at 09:46:54PM +0100, Patryk Zawadzki wrote:
> Dnia 08-11-2006, ?ro o godzinie 20:26 +0100, Jakub Bogusz napisa?(a):
> > On Wed, Nov 08, 2006 at 08:21:01PM +0100, Patryk Zawadzki wrote:
> > > Dnia 08-11-2006, ?ro o godzinie 11:54 +0100, Arkadiusz Miskiewicz
> > > napisa?(a):
> > > > On Wednesday 08 November 2006 11:49, Jakub Bogusz wrote:
> > > > > But it tests only IFS compliance.
> > > > > What about all other functionality?
> > > > Haven't seen tests for other functionality.
> > > 
> > > Should we aim at perfect POSIX compatibility or perfect usability? Bash
> > > does not seem to be a heavy dependency and is actively maintained (so
> > > the compatibility issues will go away over time as opposed to pdksh).
> > 
> > And POSIX-compatibility of scripts will go away as nobody will care
> > about.
> 
> To me it sounds similar to the sparc/alpha problem. If someone uses a
> POSIX-only shell (which does not exist), she will complain and things
> will get fixed.
> 
> If noone complains then it is a sign that noone cares and POSIX
> compatibility is done just for the sake of doing it.

Yeah, sure, just like FHS, LSB and so on. Who cares beside few
developers? Why not be just like [put some big vendor name here] and
so on? But... then why be at all?

A script can expect that /bin/sh provides functionality defined by
POSIX, not bash extensions. bash script must use /bin/bash shebang,
period.

You said usability? So OK.
Do some benchmarks. If bash appears to be faster and have lower (or
just insignificantly higher) memory footprint, then it would be good
reason to provide bash as pdksh replacement for /bin/sh. Hint: libtool
scripts used to link some huge libraries (containing many, many
objects) are good stress tests.


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


From glen at delfi.ee  Wed Nov  8 22:36:21 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 9 Nov 2006 00:36:21 +0300
Subject: php 5.2
In-Reply-To: <20061107205330.GA29880@fngna.oyu>
References: <200611030040.03319.glen@delfi.ee>
	<20061107205330.GA29880@fngna.oyu>
Message-ID: <200611082336.21483.glen@delfi.ee>

On Tuesday 07 November 2006 22:53, Jakub Bogusz wrote:
> Maybe let's drop "php-module" Provides in php4 and use (versioned?)
> "php(module)" Provides in both versions (and update deps in other
> packages)?
> Using virtual Provides with other _real_ package name brings many
> troubles and should be avoided.

i believe those virtuals should be also added to php-pecl modules....


-- 
glen


From zswi at pers.pl  Mon Nov 13 15:18:37 2006
From: zswi at pers.pl (=?utf-8?q?Rafa=C5=82_Cygnarowski?=)
Date: Mon, 13 Nov 2006 15:18:37 +0100
Subject: lm_sensors - config files for motherboards
Message-ID: <200611131518.37226.zswi@pers.pl>

Hi!

I would like to create lm_sensors-config packages for apropriate motherboards 
and lm_sensors-config-default to contain our current configuration files. 
Also I want to move sensorsd init script to sensors package so it could load 
kernel modules without sensorsd installed (sensorsd could be started if two 
conditions are met: sensorsd binary exists and /etc/sysconfig/sensors want it 
(by default)). Has anyoune got sth against such change?

Regards,
-- 
Rafa? Cygnarowski
rafi at pers.pl


From qboosh at pld-linux.org  Tue Nov 14 01:40:27 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 14 Nov 2006 01:40:27 +0100
Subject: file 4.18 vs libtool dependencies autodetection
Message-ID: <20061114004027.GA30326@fngna.oyu>

file 4.18 began to recognize libtool files (*.la):

/usr/lib/librpm.la: libtool library file
(instead of "ASCII English text" like before)

and it appears to enable libtool dependencies autodetection in rpm.
Should be use it or disable it?

If we decide to use it, then:
- all packages with *.la will require rebuild to generate Provides
- there will be needed some _noauto* in some cases:
  - when some *-devel contains more than one library, some of them
    have additional dependencies, but are rarely used (e.g. libwmf, flac)
  - in packages with ltdl-loaded plugins, which must have their *.la,
    but don't require *.la shared libraries they use


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


From glen at delfi.ee  Tue Nov 14 12:57:58 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Tue, 14 Nov 2006 13:57:58 +0200
Subject: Fwd: [gentoo-announce] [ GLSA 200611-08 ] RPM: Buffer overflow
Message-ID: <200611141357.58661.glen@delfi.ee>



----------  Forwarded Message  ----------

Subject: [gentoo-announce] [ GLSA 200611-08 ] RPM: Buffer overflow
Date: Tuesday 14 November 2006 00:57
From: Raphael Marichez <falco at gentoo.org>
To: gentoo-announce at gentoo.org
Cc: bugtraq at securityfocus.com, full-disclosure at lists.grok.org.uk, 
security-alerts at linuxsecurity.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Gentoo Linux Security Advisory                           GLSA 200611-08
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                                            http://security.gentoo.org/
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

  Severity: Normal
     Title: RPM: Buffer overflow
      Date: November 13, 2006
      Bugs: #154218
        ID: 200611-08

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Synopsis
========

RPM is vulnerable to a buffer overflow and possibly the execution of
arbitrary code when opening specially crafted packages.

Background
==========

The Red Hat Package Manager (RPM) is a command line driven package
management system capable of installing, uninstalling, verifying,
querying, and updating computer software packages.

Affected packages
=================

    -------------------------------------------------------------------
     Package       /  Vulnerable  /                         Unaffected
    -------------------------------------------------------------------
  1  app-arch/rpm     < 4.4.6-r3                           >= 4.4.6-r3

Description
===========

Vladimir Mosgalin has reported that when processing certain packages,
RPM incorrectly allocates memory for the packages, possibly causing a
heap-based buffer overflow.

Impact
======

An attacker could entice a user to open a specially crafted RPM package
and execute code with the privileges of that user if certain locales
are set.

Workaround
==========

There is no known workaround at this time.

Resolution
==========

All RPM users should upgrade to the latest version:

    # emerge --sync
    # emerge --ask --oneshot --verbose ">=app-arch/rpm-4.4.6-r3"

References
==========

  [ 1 ] CVE-2006-5466
        http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2006-5466

Availability
============

This GLSA and any updates to it are available for viewing at
the Gentoo Security Website:

  http://security.gentoo.org/glsa/glsa-200611-08.xml

Concerns?
=========

Security is a primary focus of Gentoo Linux and ensuring the
confidentiality and security of our users machines is of utmost
importance to us. Any security concerns should be addressed to
security at gentoo.org or alternatively, you may file a bug at
http://bugs.gentoo.org.

License
=======

Copyright 2006 Gentoo Foundation, Inc; referenced text
belongs to its owner(s).

The contents of this document are licensed under the
Creative Commons - Attribution / Share Alike license.

http://creativecommons.org/licenses/by-sa/2.5

-------------------------------------------------------

-- 
glen


From aflinta at at.kernel.pl  Wed Nov 15 01:21:57 2006
From: aflinta at at.kernel.pl (Artur Flinta)
Date: Wed, 15 Nov 2006 01:21:57 +0100
Subject: SPECS: horde-webmail.spec (NEW) - new; needs lots of work
In-Reply-To: <SPECS_horde_webmail.spec.1.1@pld-linux.org>
References: <SPECS_horde_webmail.spec.NONE@pld-linux.org>
	<SPECS_horde_webmail.spec.1.1@pld-linux.org>
Message-ID: <455A5DA5.8060206@at.kernel.pl>

glen wrote:
> Author: glen                         Date: Tue Nov  7 11:13:16 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - new; needs lots of work

[aflinta at router SPECS]$ ./builder horde-webmail.spec
# $Revision: 1.2 $, $Date: 2006/11/14 21:10:39 $
Available branches:
horde-webmail-1.0-rc1.tar.gz having proper md5sum already exists
cvs server: nothing known about horde-webmail.conf
cvs server: nothing known about horde-webmail-prefs.patch
MD5 sum mismatch or 0 size.  Use -U to refetch sources,
or -5 to update md5 sums, if you're sure files are correct.
Error: some source, patch or icon files not stored in CVS repo.
(horde-webmail.conf)

Artur


From hawk at limanowa.net  Wed Nov 15 09:41:14 2006
From: hawk at limanowa.net (=?ISO-8859-2?Q?Marcin_Kr=F3l?=)
Date: Wed, 15 Nov 2006 09:41:14 +0100
Subject: distfiles notifications
Message-ID: <455AD2AA.1090308@limanowa.net>

Anyone is getting distfiles notification mails?

I've just commited openssh.spec with updated md5 and have no idea if
sources were downloaded or not. No mails from distfiles.

M.


From qboosh at pld-linux.org  Wed Nov 15 09:44:02 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 15 Nov 2006 09:44:02 +0100
Subject: distfiles notifications
In-Reply-To: <455AD2AA.1090308@limanowa.net>
References: <455AD2AA.1090308@limanowa.net>
Message-ID: <20061115084402.GA30852@gruby.cs.net.pl>

On Wed, Nov 15, 2006 at 09:41:14AM +0100, Marcin Kr?l wrote:
> Anyone is getting distfiles notification mails?
> 
> I've just commited openssh.spec with updated md5 and have no idea if
> sources were downloaded or not. No mails from distfiles.

Haven't they been already downloaded for HEAD?

BTW, storing plain text files in distfiles should be avoided.


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


From deejay1 at srem.org  Wed Nov 15 10:01:20 2006
From: deejay1 at srem.org (=?iso-8859-2?q?=A3ukasz_Jerna=B6?=)
Date: Wed, 15 Nov 2006 10:01:20 +0100
Subject: fusebase specs names
In-Reply-To: <454FBF97.8070603@gmail.com>
References: <454FBF97.8070603@gmail.com>
Message-ID: <200611151001.21270.deejay1@srem.org>

Dnia wtorek, 7 listopada 2006 00:04, Andrzej Zawadzki napisa?:
> Hello!
>
> I'd like to ask what is a proper name (maybe is some convection) for
> fusebase specs?

I would opt for fuse-<module>.spec, it would be easy to find them in poldek, 
etc..
-- 
?ukasz [DeeJay1] Jerna?


From hawk at limanowa.net  Wed Nov 15 10:13:47 2006
From: hawk at limanowa.net (=?ISO-8859-1?Q?Marcin_Kr=F3l?=)
Date: Wed, 15 Nov 2006 10:13:47 +0100
Subject: distfiles notifications
In-Reply-To: <20061115084402.GA30852@gruby.cs.net.pl>
References: <455AD2AA.1090308@limanowa.net>
	<20061115084402.GA30852@gruby.cs.net.pl>
Message-ID: <455ADA4B.2020001@limanowa.net>

> Haven't they been already downloaded for HEAD?

Hm. They were downladed for HEAD indeed. Sorry, my fault, I was sure
that I'll get message saying "already got blah blah".

> BTW, storing plain text files in distfiles should be avoided.

True. I'll move them to SOURCES.

M.


From qboosh at pld-linux.org  Wed Nov 15 10:19:47 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 15 Nov 2006 10:19:47 +0100
Subject: distfiles notifications
In-Reply-To: <455ADA4B.2020001@limanowa.net>
References: <455AD2AA.1090308@limanowa.net>
	<20061115084402.GA30852@gruby.cs.net.pl>
	<455ADA4B.2020001@limanowa.net>
Message-ID: <20061115091947.GB30852@gruby.cs.net.pl>

On Wed, Nov 15, 2006 at 10:13:47AM +0100, Marcin Kr?l wrote:
> > Haven't they been already downloaded for HEAD?
> 
> Hm. They were downladed for HEAD indeed. Sorry, my fault, I was sure
> that I'll get message saying "already got blah blah".
> 
> > BTW, storing plain text files in distfiles should be avoided.
> 
> True. I'll move them to SOURCES.

Already done.

I also moved connect to separate spec/package as I proposed some time
ago.


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


From glen at delfi.ee  Wed Nov 15 11:49:02 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 15 Nov 2006 12:49:02 +0200
Subject: SPECS: horde-webmail.spec (NEW) - new; needs lots of work
In-Reply-To: <455A5DA5.8060206@at.kernel.pl>
References: <SPECS_horde_webmail.spec.NONE@pld-linux.org>
	<SPECS_horde_webmail.spec.1.1@pld-linux.org>
	<455A5DA5.8060206@at.kernel.pl>
Message-ID: <200611151249.02976.glen@delfi.ee>

On Wednesday 15 November 2006 02:21, Artur Flinta wrote:
> glen wrote:
> > Author: glen                         Date: Tue Nov  7 11:13:16 2006 GMT
> > Module: SPECS                         Tag: HEAD
> > ---- Log message:
> > - new; needs lots of work
>
> [aflinta at router SPECS]$ ./builder horde-webmail.spec
> # $Revision: 1.2 $, $Date: 2006/11/14 21:10:39 $

well. i think the spec should be just:
R: horde-imp, horde-ingo, horde-kronolith, horde-turba, horde-nag, horde-mnemo

> Available branches:
> horde-webmail-1.0-rc1.tar.gz having proper md5sum already exists
> cvs server: nothing known about horde-webmail.conf
> cvs server: nothing known about horde-webmail-prefs.patch
> MD5 sum mismatch or 0 size.  Use -U to refetch sources,
> or -5 to update md5 sums, if you're sure files are correct.
> Error: some source, patch or icon files not stored in CVS repo.
> (horde-webmail.conf)
>
> Artur
> _______________________________________________
> pld-devel-en mailing list
> pld-devel-en at lists.pld-linux.org
> http://lists.pld-linux.org/mailman/listinfo/pld-devel-en

-- 
glen


From aflinta at at.kernel.pl  Wed Nov 15 12:58:01 2006
From: aflinta at at.kernel.pl (Artur Flinta)
Date: Wed, 15 Nov 2006 12:58:01 +0100
Subject: SPECS: horde-webmail.spec (NEW) - new; needs lots of work
In-Reply-To: <200611151249.02976.glen@delfi.ee>
References: <SPECS_horde_webmail.spec.NONE@pld-linux.org>	<SPECS_horde_webmail.spec.1.1@pld-linux.org>	<455A5DA5.8060206@at.kernel.pl>
	<200611151249.02976.glen@delfi.ee>
Message-ID: <455B00C9.5060700@at.kernel.pl>

Elan Ruusam?e wrote:
> well. i think the spec should be just:
> R: horde-imp, horde-ingo, horde-kronolith, horde-turba, horde-nag, horde-mnemo

OK, thx :)

Artur


From zswi at pers.pl  Thu Nov 16 08:44:51 2006
From: zswi at pers.pl (=?iso-8859-2?q?Rafa=B3_Cygnarowski?=)
Date: Thu, 16 Nov 2006 08:44:51 +0100
Subject: SPECS: lm_sensors-config.spec - more pl,
	removed invalid use of _s...
In-Reply-To: <SPECS_lm_sensors_config.spec.1.2@pld-linux.org>
References: <SPECS_lm_sensors_config.spec.1.1@pld-linux.org>
	<SPECS_lm_sensors_config.spec.1.2@pld-linux.org>
Message-ID: <200611160844.51263.zswi@pers.pl>

Dnia ?roda, 15 listopada 2006 23:53, qboosh napisa?:
> Author: qboosh                       Date: Wed Nov 15 22:53:18 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - more pl, removed invalid use of _sysconfdir

What is _sysconfdir for? I've seen that in rpm-4.4.6 _sysconfdir was defined 
as /etc and now (4.4.7) as %{_prefix}/etc. What about some std /etc 
directories like init.d, sysconfig? - maybe they should have their own 
defines?

-- 
Rafa? Cygnarowski
rafi at pers.pl


From glen at delfi.ee  Thu Nov 16 10:14:01 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 16 Nov 2006 11:14:01 +0200
Subject: ac-builder
Message-ID: <200611161114.01780.glen@delfi.ee>

please remove python-PyQt (solib mismatch), qscintilla from all builders so 
their update could be finished building.

also openssh has upgrade failed ok message on ppc (no further messages, but i 
debugged and it has openssh-clients installed twice)

and kdebase has upgrade failed ok on all arches (no further messages)

ps: maybe increase my rights so i could do that myself?

-- 
glen


From glen at delfi.ee  Thu Nov 16 11:41:28 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 16 Nov 2006 12:41:28 +0200
Subject: firefox 2 for ac
Message-ID: <200611161241.28877.glen@delfi.ee>

has anybody tried building it for ac? does it work properly too?

i have noticed two problems
- bookmarks menu is empty unless new bookmark is added or it's opened at 
sidebar
- the statusbar is 80px high.
- throbber keeps spinning if the tab has content downloaded
- searchbox doesn't activate on enter nor on mouse click on icon

-- 
glen


From glen at delfi.ee  Thu Nov 16 11:51:27 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Thu, 16 Nov 2006 12:51:27 +0200
Subject: glib2 G_GNUC_UNUSED
In-Reply-To: <200611030221.41731.glen@delfi.ee>
References: <200611030221.41731.glen@delfi.ee>
Message-ID: <200611161251.27888.glen@delfi.ee>

if somebody help solving this problem, could continue compiling ooo2 for ac

On Friday 03 November 2006 02:21, Elan Ruusam?e wrote:
> is glib2 defining the macro incompatible for gcc in Ac, or there's some
> other cause?
>
> $ g++ -c t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags gtk+-2.0`
> t.c:5: error: syntax error before `__attribute__'
>
> $ cat t.c
> #include <gtk/gtk.h>
> #include <glib.h>
>
> static void open_url_cb( GtkWidget *pButton G_GNUC_UNUSED, gpointer data )
> {
> }
>
> $ g++ -E t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags
> gtk+-2.0`>t.C $ grep open_url_cb t.C
> static void open_url_cb( GtkWidget *pButton __attribute__((__unused__)),
> gpointer data )
>
> $ q gcc glib2 gtk+2
> gcc-5:3.3.6-4
> glib2-1:2.10.3-1
> gtk+2-2:2.8.20-1
> $
>
> http://developer.gnome.org/doc/API/2.0/glib/glib-Miscellaneous-Macros.html#
>G-GNUC-UNUSED:CAPS

-- 
glen


From qboosh at pld-linux.org  Thu Nov 16 13:42:09 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 16 Nov 2006 13:42:09 +0100
Subject: SPECS: lm_sensors-config.spec - more pl,
	removed invalid use of _s...
In-Reply-To: <200611160844.51263.zswi@pers.pl>
References: <SPECS_lm_sensors_config.spec.1.1@pld-linux.org>
	<SPECS_lm_sensors_config.spec.1.2@pld-linux.org>
	<200611160844.51263.zswi@pers.pl>
Message-ID: <20061116124208.GB5647@gruby.cs.net.pl>

On Thu, Nov 16, 2006 at 08:44:51AM +0100, Rafa? Cygnarowski wrote:
> Dnia ?roda, 15 listopada 2006 23:53, qboosh napisa?:
> > Author: qboosh                       Date: Wed Nov 15 22:53:18 2006 GMT
> > Module: SPECS                         Tag: HEAD
> > ---- Log message:
> > - more pl, removed invalid use of _sysconfdir
> 
> What is _sysconfdir for? I've seen that in rpm-4.4.6 _sysconfdir was defined 
> as /etc and now (4.4.7) as %{_prefix}/etc.

%{_sysconfdir} is system-wide configuration directory. It can be
redefined per package, while /etc/sysconfig, /etc/rc.d/init.d etc. are
constant.

> What about some std /etc 
> directories like init.d, sysconfig? - maybe they should have their own 
> defines?

There is no need for it.


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


From n3npq at mac.com  Thu Nov 16 21:42:45 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Thu, 16 Nov 2006 15:42:45 -0500
Subject: epm-4.4.7 + poldek
Message-ID: <6FBBAC95-6992-4E8A-8ADD-EF19723C1E0E@mac.com>

Dunno if this still relevant. Any success with rpm-4.4.7 and poldek yet?

 > As you might have already guessed, Jeff dropped all the deprecated  
stuff
 > from this release of rpm. Now it's there in Th. Don't upgrade.

 > poldek makes use of most of the deprecated functions that are now  
gone
 > (Jeff says some of them were marked as deprecated for as long as 6  
whole
 > years).

Mea culpa. I apologize for breaking poldek.

Meanwhile, a patch to restore most of the functionality I ripped can  
be found at

     http://wraptastic.org/pub/jbj/rpm-revert.patch

I'll carry that code through rpm-4.4.8 (or as long as needed) if you  
want. Please note
that rpm-4.4.7 no longer reads /usr/lib/rpm/rpmrc by default, and so  
behavior with and without
reading rpmrc is well, one behavior too many.

My intent is to simplify choosing cpu-vendor-os, all that rpm needs  
is strings.
That should help make cross-packaging and multilib and xen and qemu  
and ...
easier with rpm.

But I certainly did not mean to break any code.

 > DEVEL branch contains my attempt at making poldek usable again.  
There is
 > a poldek-deprecated.patch that fixes the building process but
 > unfortunately, other things are not that easy to fix:

Your patch looks pretty sound from a rpm POV, I'm less familiar with  
poldek
sources,.

 > [EMAIL PROTECTED] SOURCES]$ LC_ALL=C poldek
 > Loading [pndir]th...
 > Loading [pndir]th...
 > Loading [pndir]th-test...
 > Loading [pndir]th-test...
 > 13041 packages read
 > Removed 11 duplicate packages from available set
 > error: db4 error(146746632) from db->set_cachesize: Unknown error  
146746632
 > error: db4 error(22) from db->close: Invalid argument
 > error: cannot open Packages index using db3 - Unknown error 146746632
(146746632)
 > error: //var/lib/rpm: open rpm database failed
 > error: Load installed packages failed


Hmmm, weird. If you send me the equivalent of rpm -vv output from  
poldek privately,
I might be able to guess what is happening. The errors are quite  
strange.

hth

73 de Jeff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </mailman/pipermail/pld-devel-en/attachments/20061116/3d883878/attachment.htm>

From n3npq at mac.com  Thu Nov 16 21:46:20 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Thu, 16 Nov 2006 15:46:20 -0500
Subject: epm-4.4.7 + poldek
In-Reply-To: <6FBBAC95-6992-4E8A-8ADD-EF19723C1E0E@mac.com>
References: <6FBBAC95-6992-4E8A-8ADD-EF19723C1E0E@mac.com>
Message-ID: <C6720AF5-6360-4EEC-9081-1DCB3D7782FE@mac.com>

Eeek, I meant rpm-4.4.7 of course.

Old tired eyes here ...

73 de Jeff



From qboosh at pld-linux.org  Fri Nov 17 09:14:59 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Fri, 17 Nov 2006 09:14:59 +0100
Subject: autoconf 2.60 in Ac
Message-ID: <20061117081459.GA3700@gruby.cs.net.pl>

More and more projects start using (and requiring) autoconf 2.60.
Isn't it time to upgrade autoconf in Ac?


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


From glen at delfi.ee  Fri Nov 17 13:03:57 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Fri, 17 Nov 2006 14:03:57 +0200
Subject: autoconf 2.60 in Ac
In-Reply-To: <20061117081459.GA3700@gruby.cs.net.pl>
References: <20061117081459.GA3700@gruby.cs.net.pl>
Message-ID: <200611171403.57128.glen@delfi.ee>

On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> More and more projects start using (and requiring) autoconf 2.60.
> Isn't it time to upgrade autoconf in Ac?
+1 from me :)

-- 
glen


From sparky at pld-linux.org  Fri Nov 17 16:47:17 2006
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Fri, 17 Nov 2006 16:47:17 +0100
Subject: SPECS: kernel-net-r1000.spec - use %build_kernel_modules,
	%install...
In-Reply-To: <SPECS_kernel_net_r1000.spec.1.3@pld-linux.org>
References: <SPECS_kernel_net_r1000.spec.1.2@pld-linux.org>
	<SPECS_kernel_net_r1000.spec.1.3@pld-linux.org>
Message-ID: <20061117154717.GA11768@pld-linux.org>

On Fri, Nov 17, 2006 at 04:22:32PM +0100, vip wrote:
> Author: vip                          Date: Fri Nov 17 15:22:32 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - use %build_kernel_modules, %install_kernel_modules
> - request for comments:

> what with "ifarch ppc"

It is old piece of code not used anymore, and should not be needed

> and "_target_base_arch != _arch" in %build section?

%build_kernel_modules includes cross-compilation support

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From glen at delfi.ee  Sun Nov 19 19:01:31 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Sun, 19 Nov 2006 20:01:31 +0200
Subject: Fwd: hal.spec
Message-ID: <200611192001.31862.glen@delfi.ee>

sorry, i don't know, forwarding to devel-en list

----------  Forwarded Message  ----------

Subject: hal.spec
Date: Sunday 19 November 2006 17:48
From: Karol Dowbecki <dowbecki at gmail.com>
To: glen at pld-linux.org

Hi,

I've got a problem with hal under PLD AC.

Since hal-0.5.8-1 requires Xorg >= 7.0 the most recent one for AC is
hal-0.5.7-3, and that's the one I'm now useing. Under KDE >= 3.5.4 it causes
problems with mounting removable devices 'cause of KDE new mounting
procedure. You can read about it here:

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=377689

Is the case mentioned in this bug report already included in one of varius
patches in CVS?


--
Pozdrawiam,
Karol Dowbecki <dowbecki at gmail.com>

-------------------------------------------------------

-- 
glen


From pluto at agmk.net  Sun Nov 19 19:41:33 2006
From: pluto at agmk.net (=?utf-8?q?Pawe=C5=82_Sikora?=)
Date: Sun, 19 Nov 2006 18:41:33 -0000
Subject: file 4.18 vs libtool dependencies autodetection
In-Reply-To: <20061114004027.GA30326@fngna.oyu>
References: <20061114004027.GA30326@fngna.oyu>
Message-ID: <200612191942.03131.pluto@agmk.net>

On Tuesday 14 November 2006 01:40, Jakub Bogusz wrote:
> file 4.18 began to recognize libtool files (*.la):
>
> /usr/lib/librpm.la: libtool library file
> (instead of "ASCII English text" like before)
>
> and it appears to enable libtool dependencies autodetection in rpm.
> Should be use it or disable it?

+1 for Th.


From twittner at o2.pl  Sun Nov 19 20:43:31 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Sun, 19 Nov 2006 20:43:31 +0100
Subject: file 4.18 vs libtool dependencies autodetection
In-Reply-To: <20061114004027.GA30326@fngna.oyu>
References: <20061114004027.GA30326@fngna.oyu>
Message-ID: <200611192043.32286.twittner@o2.pl>

On Tue 14. November 2006 01:40, Jakub Bogusz wrote:
> file 4.18 began to recognize libtool files (*.la):
> 
> /usr/lib/librpm.la: libtool library file
> (instead of "ASCII English text" like before)
> 
> and it appears to enable libtool dependencies autodetection in rpm.
> Should be use it or disable it?
Is any profit of not disabling them? IMO generating libtool(libfoo.la) P/R 
should be disabled. We could have lived without them till now, we can live 
further. ;)
> 
> If we decide to use it, then:
> - all packages with *.la will require rebuild to generate Provides
> - there will be needed some _noauto* in some cases:
>   - when some *-devel contains more than one library, some of them
>     have additional dependencies, but are rarely used (e.g. libwmf, flac)
>   - in packages with ltdl-loaded plugins, which must have their *.la,
>     but don't require *.la shared libraries they use
> 
> 

-- 
Tomasz Wittner


From n3npq at mac.com  Sun Nov 19 21:01:09 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Sun, 19 Nov 2006 15:01:09 -0500
Subject: file 4.18 vs libtool dependencies autodetection
In-Reply-To: <200611192043.32286.twittner@o2.pl>
References: <20061114004027.GA30326@fngna.oyu>
	<200611192043.32286.twittner@o2.pl>
Message-ID: <E0D10A76-C3E5-4F3C-B46A-2EA9085D4916@mac.com>


On Nov 19, 2006, at 2:43 PM, Tomasz Wittner wrote:

> On Tue 14. November 2006 01:40, Jakub Bogusz wrote:
>> file 4.18 began to recognize libtool files (*.la):
>>
>> /usr/lib/librpm.la: libtool library file
>> (instead of "ASCII English text" like before)
>>
>> and it appears to enable libtool dependencies autodetection in rpm.
>> Should be use it or disable it?
> Is any profit of not disabling them? IMO generating libtool 
> (libfoo.la) P/R
> should be disabled. We could have lived without them till now, we  
> can live
> further. ;)

Wise man or wuss, depending on your point-of-view.

The reason for using "libtool(...)" dependencies is to make builds more
reproducible. There are certain dependencies hidden within *.la files
that need to be tracked explicititly in rpm dependencies to make builds
more reproducible.

OTOH, you're absoutely correct that rpm has managed to live without
libtool dependencies for years.

>>
>> If we decide to use it, then:
>> - all packages with *.la will require rebuild to generate Provides
>> - there will be needed some _noauto* in some cases:
>>   - when some *-devel contains more than one library, some of them
>>     have additional dependencies, but are rarely used (e.g.  
>> libwmf, flac)
>>   - in packages with ltdl-loaded plugins, which must have their *.la,
>>     but don't require *.la shared libraries they use
>>

Note that there may be a problem in the /usr/lib/rpm/libtooldeps.sh
script distributed with rpm-4.4.7 as well. I'm seeing some problems
building rpm using libtooldeps.sh, I'm currently having to add
libtool(...) dependencies to /etc/rpm/sysinfo to mask some deeper
problem.

The problem is surely easy to fix, there's nothing complicated in
libtooldeps.sh.

However, my main development platform atm is FC6, which is (ahem) not  
the
best platform to try to debug extracted *.la dependency chains since  
*.la
files are being haphazardly removed from FC6 packaging.

Patches gratefully accepted ;-)

hth

73 de Jeff


From qboosh at pld-linux.org  Sun Nov 19 23:44:53 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sun, 19 Nov 2006 23:44:53 +0100
Subject: file 4.18 vs libtool dependencies autodetection
In-Reply-To: <200611192043.32286.twittner@o2.pl>
References: <20061114004027.GA30326@fngna.oyu>
	<200611192043.32286.twittner@o2.pl>
Message-ID: <20061119224453.GB11675@fngna.oyu>

On Sun, Nov 19, 2006 at 08:43:31PM +0100, Tomasz Wittner wrote:
> On Tue 14. November 2006 01:40, Jakub Bogusz wrote:
> > file 4.18 began to recognize libtool files (*.la):
> > 
> > /usr/lib/librpm.la: libtool library file
> > (instead of "ASCII English text" like before)
> > 
> > and it appears to enable libtool dependencies autodetection in rpm.
> > Should be use it or disable it?
> Is any profit of not disabling them? IMO generating libtool(libfoo.la) P/R 
> should be disabled. We could have lived without them till now, we can live 
> further. ;)

Yes, we could and we can.
But we also spent some time on tracking where some build dependencies
come from when they weren't noticed at their true source before (see
all those threads about adding to wrong BRs some end package because
of *.la dependencies).
With autogenerated *.la dependencies we can avoid this in future.
 

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


From havner at smtp.kamp.pl  Mon Nov 20 16:00:15 2006
From: havner at smtp.kamp.pl (havner)
Date: Mon, 20 Nov 2006 16:00:15 +0100
Subject: Fwd: hal.spec
In-Reply-To: <200611192001.31862.glen@delfi.ee>
References: <200611192001.31862.glen@delfi.ee>
Message-ID: <20061120150014.GA12205@sirius.livecd.pl>

> I've got a problem with hal under PLD AC.
> 
> Since hal-0.5.8-1 requires Xorg >= 7.0 the most recent one for AC is
> hal-0.5.7-3, and that's the one I'm now useing. Under KDE >= 3.5.4 it causes
> problems with mounting removable devices 'cause of KDE new mounting
> procedure. You can read about it here:
> 
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=377689

Of-topic: this "new procedure" i kinda step back for KDE.
pmount/gnome-mount was invented not to have fstab entries for removeable
devices. (after fstab-sync was boycotted). For cdroms it's not that
crucial, but f.e. pendrives or any other usb storage may have different
/dev/ nodes assigned everytime you plug them. And now in KDE you have to
fill fstab again for them to be automounted. This doesn't look good.

-- 
Regards    Havner                      {jid,mail}:havner(at)pld-linux.org
PLD developer                                    http://www.pld-linux.org
PLD LiveCD author                             http://livecd.pld-linux.org
                   "Quis custodiet ipsos custodes?"


From qboosh at pld-linux.org  Tue Nov 21 10:00:23 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 21 Nov 2006 10:00:23 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <200611171403.57128.glen@delfi.ee>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<200611171403.57128.glen@delfi.ee>
Message-ID: <20061121090023.GA5556@gruby.cs.net.pl>

On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > More and more projects start using (and requiring) autoconf 2.60.
> > Isn't it time to upgrade autoconf in Ac?
> +1 from me :)

So... any objections?
I'm going to sent a request today.


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


From baggins at sith.mimuw.edu.pl  Tue Nov 21 10:18:11 2006
From: baggins at sith.mimuw.edu.pl (Jan Rekorajski)
Date: Tue, 21 Nov 2006 10:18:11 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <20061121090023.GA5556@gruby.cs.net.pl>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<200611171403.57128.glen@delfi.ee>
	<20061121090023.GA5556@gruby.cs.net.pl>
Message-ID: <20061121091811.GH22176@sith.mimuw.edu.pl>

On Tue, 21 Nov 2006, Jakub Bogusz wrote:

> On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> > On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > > More and more projects start using (and requiring) autoconf 2.60.
> > > Isn't it time to upgrade autoconf in Ac?
> > +1 from me :)
> 
> So... any objections?

If it doesn't screw up ac/am regeneration as usual then none ;)

Janek
-- 
Jan Rekorajski            |  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 qboosh at pld-linux.org  Tue Nov 21 10:26:16 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 21 Nov 2006 10:26:16 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <20061121091811.GH22176@sith.mimuw.edu.pl>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<200611171403.57128.glen@delfi.ee>
	<20061121090023.GA5556@gruby.cs.net.pl>
	<20061121091811.GH22176@sith.mimuw.edu.pl>
Message-ID: <20061121092616.GA6970@gruby.cs.net.pl>

On Tue, Nov 21, 2006 at 10:18:11AM +0100, Jan Rekorajski wrote:
> On Tue, 21 Nov 2006, Jakub Bogusz wrote:
> 
> > On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> > > On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > > > More and more projects start using (and requiring) autoconf 2.60.
> > > > Isn't it time to upgrade autoconf in Ac?
> > > +1 from me :)
> > 
> > So... any objections?
> 
> If it doesn't screw up ac/am regeneration as usual then none ;)

It does, as usual ;P

However most of failing packages have been fixed already, and remaining
ones (not known yet) need to be fixed for Th anyway.


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


From arekm at pld-linux.org  Tue Nov 21 10:28:10 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Tue, 21 Nov 2006 10:28:10 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <20061121091811.GH22176@sith.mimuw.edu.pl>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<20061121090023.GA5556@gruby.cs.net.pl>
	<20061121091811.GH22176@sith.mimuw.edu.pl>
Message-ID: <200611211028.10290.arekm@pld-linux.org>

On Tuesday 21 November 2006 10:18, Jan Rekorajski wrote:
> On Tue, 21 Nov 2006, Jakub Bogusz wrote:
> > On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> > > On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > > > More and more projects start using (and requiring) autoconf 2.60.
> > > > Isn't it time to upgrade autoconf in Ac?
> > >
> > > +1 from me :)
> >
> > So... any objections?
>
> If it doesn't screw up ac/am regeneration as usual then none ;)

Won't apache/php stop building like it's in Th now? (due to -o posix afaik)

> Janek

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


From qboosh at pld-linux.org  Tue Nov 21 10:44:58 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 21 Nov 2006 10:44:58 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <200611211028.10290.arekm@pld-linux.org>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<20061121090023.GA5556@gruby.cs.net.pl>
	<20061121091811.GH22176@sith.mimuw.edu.pl>
	<200611211028.10290.arekm@pld-linux.org>
Message-ID: <20061121094458.GB6970@gruby.cs.net.pl>

On Tue, Nov 21, 2006 at 10:28:10AM +0100, Arkadiusz Miskiewicz wrote:
> On Tuesday 21 November 2006 10:18, Jan Rekorajski wrote:
> > On Tue, 21 Nov 2006, Jakub Bogusz wrote:
> > > On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> > > > On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > > > > More and more projects start using (and requiring) autoconf 2.60.
> > > > > Isn't it time to upgrade autoconf in Ac?
> > > >
> > > > +1 from me :)
> > >
> > > So... any objections?
> >
> > If it doesn't screw up ac/am regeneration as usual then none ;)
> 
> Won't apache/php stop building like it's in Th now? (due to -o posix afaik)

apr and apache have been fixed on HEAD recently (I hope on x86_64 too).
I'll check php.


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


From arekm at pld-linux.org  Tue Nov 21 11:34:48 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Tue, 21 Nov 2006 11:34:48 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <4561B2A9.1020300@agmk.net>
References: <4561B2A9.1020300@agmk.net>
Message-ID: <200611211134.48181.arekm@pld-linux.org>

On Monday 20 November 2006 14:50, Pawe? Sikora wrote:
> $ rpm -Uhv kdelibs-libs-3.5.5-3.x86_64.rpm --test
> error: Failed dependencies:
> (...)
>      libtool(/usr/lib64/libkdefx.la) is needed
>            by kdelibs-libs-3.5.5-3.x86_64
>
> $ rpm -qlp --provides kdelibs-libs-3.5.5-3.x86_64.rpm|grep kdefx
> libkdefx.so.4()(64bit)
> /usr/lib64/libkdefx.la
> /usr/lib64/libkdefx.so.4.2.0

rpm is doing stupid thing:

#define 
_LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' | /usr/lib/rpm/libtooldeps.sh -P /"

and Provides are broken.

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


From arekm at pld-linux.org  Tue Nov 21 11:36:54 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Tue, 21 Nov 2006 11:36:54 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611211134.48181.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611211134.48181.arekm@pld-linux.org>
Message-ID: <200611211136.54966.arekm@pld-linux.org>

On Tuesday 21 November 2006 11:34, Arkadiusz Miskiewicz wrote:

> #define
> _LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' |
> /usr/lib/rpm/libtooldeps.sh -P /"
>
> and Provides are broken.

Or not. Sounds like some weird default setting only.

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


From arekm at pld-linux.org  Tue Nov 21 12:22:50 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Tue, 21 Nov 2006 12:22:50 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611211136.54966.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611211134.48181.arekm@pld-linux.org>
	<200611211136.54966.arekm@pld-linux.org>
Message-ID: <200611211222.50282.arekm@pld-linux.org>

On Tuesday 21 November 2006 11:36, Arkadiusz Miskiewicz wrote:
> On Tuesday 21 November 2006 11:34, Arkadiusz Miskiewicz wrote:
> > #define
> > _LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' |
> > /usr/lib/rpm/libtooldeps.sh -P /"
> >
> > and Provides are broken.
>
> Or not. Sounds like some weird default setting only.

rpmbuild:
161 /home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/libDCOP.la ASCII 
English text, with very long lines

file:
/home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/libDCOP.la: libtool 
library file

no idea why

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


From n3npq at mac.com  Tue Nov 21 12:28:26 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Tue, 21 Nov 2006 06:28:26 -0500
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611211134.48181.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611211134.48181.arekm@pld-linux.org>
Message-ID: <4180ACAA-562C-4659-9AF8-DBD5EABCFC21@mac.com>


On Nov 21, 2006, at 5:34 AM, Arkadiusz Miskiewicz wrote:

> On Monday 20 November 2006 14:50, Pawe? Sikora wrote:
>> $ rpm -Uhv kdelibs-libs-3.5.5-3.x86_64.rpm --test
>> error: Failed dependencies:
>> (...)
>>      libtool(/usr/lib64/libkdefx.la) is needed
>>            by kdelibs-libs-3.5.5-3.x86_64
>>
>> $ rpm -qlp --provides kdelibs-libs-3.5.5-3.x86_64.rpm|grep kdefx
>> libkdefx.so.4()(64bit)
>> /usr/lib64/libkdefx.la
>> /usr/lib64/libkdefx.so.4.2.0
>
> rpm is doing stupid thing:
>
> #define
> _LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' | /usr/lib/ 
> rpm/libtooldeps.sh -P /"
>
> and Provides are broken.
>

The __LIBTOOL_PROVIDES is used in /usr/lib/rpm/rpmdeps to
verify closure of "libtool(...)" requires in packages against
*.la files actually installed on the file system.

That code path is not traversed by rpmbuild or rpm itself, only
by invoking
     /usr/lib/rpm/rpmdeps --libtool

73 de Jeff



From n3npq at mac.com  Tue Nov 21 12:30:35 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Tue, 21 Nov 2006 06:30:35 -0500
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611211222.50282.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611211134.48181.arekm@pld-linux.org>
	<200611211136.54966.arekm@pld-linux.org>
	<200611211222.50282.arekm@pld-linux.org>
Message-ID: <C7ECA900-BBD2-48BC-B951-CB6EC5E6C78F@mac.com>


On Nov 21, 2006, at 6:22 AM, Arkadiusz Miskiewicz wrote:

> On Tuesday 21 November 2006 11:36, Arkadiusz Miskiewicz wrote:
>> On Tuesday 21 November 2006 11:34, Arkadiusz Miskiewicz wrote:
>>> #define
>>> _LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' |
>>> /usr/lib/rpm/libtooldeps.sh -P /"
>>>
>>> and Provides are broken.
>>
>> Or not. Sounds like some weird default setting only.
>
> rpmbuild:
> 161 /home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/ 
> libDCOP.la ASCII
> English text, with very long lines
>
> file:
> /home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/ 
> libDCOP.la: libtool
> library file
>
> no idea why
>

Sounds like the magic pattern is failing for certain libtool files.

Hacking in lib/rpmfc.c to classify libtool files by suffix "*.la" is the
expedient fix.

73 de Jeff



From sparky at pld-linux.org  Tue Nov 21 17:11:16 2006
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Tue, 21 Nov 2006 17:11:16 +0100
Subject: SPECS (LINUX_2_6_16): kernel.spec - keep headers/sources in
	separa...
In-Reply-To: <SPECS_kernel.spec.1.441.2.1558.2.44@pld-linux.org>
References: <SPECS_kernel.spec.1.441.2.1558.2.43@pld-linux.org>
	<SPECS_kernel.spec.1.441.2.1558.2.44@pld-linux.org>
Message-ID: <20061121161116.GA23446@pld-linux.org>

On Tue, Nov 21, 2006 at 02:52:34PM +0100, hawk wrote:
> Author: hawk                         Date: Tue Nov 21 13:52:34 2006 GMT
> Module: SPECS                         Tag: LINUX_2_6_16
> ---- Log message:
> - keep headers/sources in separate folders for default/grsecurity/xen
>   kernels so they don't overlap and to make alt_kernel usage possible
> - release 1.1, testing in progress

>  %post headers
> -rm -f /usr/src/linux
> -ln -snf linux-%{version} /usr/src/linux
> +rm -f /usr/src/linux%{_subname}
> +ln -snf linux%{_subname}-%{version} /usr/src/linux%{_subname}

so now all the modules will have to be build with e.g.
'--define alt_kernel grsecurity', well, I really lika that, but what the
people will say ? And how about builders ?

oh, and most modules which doesn't have %{_alt_kernel} support will
build for correct kernel if alt_kernel was defined. It just won't have
any information about it in the package name. So, don't worry.

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From qboosh at pld-linux.org  Tue Nov 21 19:52:24 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Tue, 21 Nov 2006 19:52:24 +0100
Subject: autoconf 2.60 in Ac
In-Reply-To: <20061121094458.GB6970@gruby.cs.net.pl>
References: <20061117081459.GA3700@gruby.cs.net.pl>
	<20061121090023.GA5556@gruby.cs.net.pl>
	<20061121091811.GH22176@sith.mimuw.edu.pl>
	<200611211028.10290.arekm@pld-linux.org>
	<20061121094458.GB6970@gruby.cs.net.pl>
Message-ID: <20061121185224.GA23828@fngna.oyu>

On Tue, Nov 21, 2006 at 10:44:58AM +0100, Jakub Bogusz wrote:
> On Tue, Nov 21, 2006 at 10:28:10AM +0100, Arkadiusz Miskiewicz wrote:
> > On Tuesday 21 November 2006 10:18, Jan Rekorajski wrote:
> > > On Tue, 21 Nov 2006, Jakub Bogusz wrote:
> > > > On Fri, Nov 17, 2006 at 02:03:57PM +0200, Elan Ruusam?e wrote:
> > > > > On Friday 17 November 2006 10:14, Jakub Bogusz wrote:
> > > > > > More and more projects start using (and requiring) autoconf 2.60.
> > > > > > Isn't it time to upgrade autoconf in Ac?
> > > > >
> > > > > +1 from me :)
> > > >
> > > > So... any objections?
> > >
> > > If it doesn't screw up ac/am regeneration as usual then none ;)
> > 
> > Won't apache/php stop building like it's in Th now? (due to -o posix afaik)
> 
> apr and apache have been fixed on HEAD recently (I hope on x86_64 too).
> I'll check php.

php fixed on AC-branch, will be fixed on HEAD soon.

So autoconf 2.60 is on AC-branch now, sent to ac-ready.
And autoconf 2.61 goes to HEAD :P


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


From deejay1 at srem.org  Tue Nov 21 20:35:50 2006
From: deejay1 at srem.org (=?iso-8859-2?q?=A3ukasz_Jerna=B6?=)
Date: Tue, 21 Nov 2006 20:35:50 +0100
Subject: Weird behaviour of %find_lang?
Message-ID: <200611212035.52454.deejay1@srem.org>

Hello.

While working on kerry.spec and beagle.spec I've noticed a strange behaviour 
of the %find_lang macro. It seems to leave behind some locale files in those 
spec ending with:
warning: Installed (but unpackaged) file(s) found:
...
/usr/share/locale/ar/LC_MESSAGES/beagle.mo
in case of beagle.spec and:
/usr/share/locale/sk/LC_MESSAGES/kcmbeagle.mo
/usr/share/locale/sk/LC_MESSAGES/kerry.mo

for kerry.spec.
Am I doing something wrong or is it just the script?
[deejay1 at betty SPECS]$ rpm -q rpm-build
rpm-build-4.4.6-1.17.i686

-- 
?ukasz [DeeJay1] Jerna?


From patrys at pld-linux.org  Tue Nov 21 21:04:23 2006
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Tue, 21 Nov 2006 21:04:23 +0100
Subject: file 4.18 vs libtool dependencies autodetection
In-Reply-To: <200612191942.03131.pluto@agmk.net>
References: <20061114004027.GA30326@fngna.oyu>
	<200612191942.03131.pluto@agmk.net>
Message-ID: <1164139463.6698.11.camel@localhost>

Dnia 19-12-2006, wto o godzinie 19:42 +0100, Pawe? Sikora napisa?(a):
> On Tuesday 14 November 2006 01:40, Jakub Bogusz wrote:
> > file 4.18 began to recognize libtool files (*.la):
> >
> > /usr/lib/librpm.la: libtool library file
> > (instead of "ASCII English text" like before)
> >
> > and it appears to enable libtool dependencies autodetection in rpm.
> > Should be use it or disable it?
> 
> +1 for Th.

I say we should use it in Th.

-- 
Patryk Zawadzki <patrys at pld-linux.org>
PLD Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: To jest cz??? listu	podpisana cyfrowo
URL: </mailman/pipermail/pld-devel-en/attachments/20061121/82112234/attachment.sig>

From patrys at pld-linux.org  Tue Nov 21 21:44:49 2006
From: patrys at pld-linux.org (Patryk Zawadzki)
Date: Tue, 21 Nov 2006 21:44:49 +0100
Subject: Weird behaviour of %find_lang?
In-Reply-To: <200611212035.52454.deejay1@srem.org>
References: <200611212035.52454.deejay1@srem.org>
Message-ID: <1164141889.6698.27.camel@localhost>

Dnia 21-11-2006, wto o godzinie 20:35 +0100, ?ukasz Jerna? napisa?(a):
> Hello.
> 
> While working on kerry.spec and beagle.spec I've noticed a strange behaviour 
> of the %find_lang macro. It seems to leave behind some locale files in those 
> spec ending with:
> warning: Installed (but unpackaged) file(s) found:
> ...
> /usr/share/locale/ar/LC_MESSAGES/beagle.mo
> in case of beagle.spec and:
> /usr/share/locale/sk/LC_MESSAGES/kcmbeagle.mo
> /usr/share/locale/sk/LC_MESSAGES/kerry.mo
> 
> for kerry.spec.
> Am I doing something wrong or is it just the script?
> [deejay1 at betty SPECS]$ rpm -q rpm-build
> rpm-build-4.4.6-1.17.i686

It leaves empty .mo files unpackaged. Isn't that the case?

-- 
Patryk Zawadzki <patrys at generatedcontent.com>
GeneratedContent.com
-- 
Patryk Zawadzki <patrys at pld-linux.org>
PLD Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: To jest cz??? listu	podpisana cyfrowo
URL: </mailman/pipermail/pld-devel-en/attachments/20061121/36869119/attachment.sig>

From deejay1 at srem.org  Tue Nov 21 22:09:40 2006
From: deejay1 at srem.org (=?iso-8859-2?q?=A3ukasz_Jerna=B6?=)
Date: Tue, 21 Nov 2006 22:09:40 +0100
Subject: Weird behaviour of %find_lang?
In-Reply-To: <1164141889.6698.27.camel@localhost>
References: <200611212035.52454.deejay1@srem.org>
	<1164141889.6698.27.camel@localhost>
Message-ID: <200611212209.41540.deejay1@srem.org>

Dnia wtorek, 21 listopada 2006 21:44, Patryk Zawadzki napisa?:
> Dnia 21-11-2006, wto o godzinie 20:35 +0100, ?ukasz Jerna? napisa?(a):
> > Hello.
> > for kerry.spec.
> > Am I doing something wrong or is it just the script?
> > [deejay1 at betty SPECS]$ rpm -q rpm-build
> > rpm-build-4.4.6-1.17.i686
>
> It leaves empty .mo files unpackaged. Isn't that the case?

Yes it is... mea culpa

-- 
?ukasz [DeeJay1] Jerna?


From hawk at limanowa.net  Tue Nov 21 23:36:53 2006
From: hawk at limanowa.net (=?ISO-8859-1?Q?Marcin_Kr=F3l?=)
Date: Tue, 21 Nov 2006 23:36:53 +0100
Subject: SPECS (LINUX_2_6_16): kernel.spec - keep headers/sources
	in	separa...
In-Reply-To: <20061121161116.GA23446@pld-linux.org>
References: <SPECS_kernel.spec.1.441.2.1558.2.43@pld-linux.org>	<SPECS_kernel.spec.1.441.2.1558.2.44@pld-linux.org>
	<20061121161116.GA23446@pld-linux.org>
Message-ID: <45637F85.8090200@limanowa.net>

> so now all the modules will have to be build with e.g.
> '--define alt_kernel grsecurity', well, I really lika that, but what the
> people will say ?

Modules for normal kernel doesn't require --define alt_kernel, so people
will not see the difference. Grsecurity enabled kernel _will not be_
default one in PLD.

> And how about builders ?

They don't handle --define alt_kernel. I'll have to find some workaround.

M.


From qboosh at pld-linux.org  Wed Nov 22 09:42:48 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Wed, 22 Nov 2006 09:42:48 +0100
Subject: too verbose update-mime-database
Message-ID: <20061122084248.GA10829@gruby.cs.net.pl>

Using %update_mime_database causes unnecessary message to be displayed
during installation:

***
* Updating MIME database in /usr/share/mime...
***

I don't see an option for quiet operation, so ">/dev/null" should be
added, right?


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


From blues at pld-linux.org  Wed Nov 22 15:41:13 2006
From: blues at pld-linux.org (=?ISO-8859-2?Q?Pawe=B3_Go=B3aszewski?=)
Date: Wed, 22 Nov 2006 15:41:13 +0100 (CET)
Subject: SPECS: kernel-desktop.spec - cxacru.c patch metged to small_fixes
	...
In-Reply-To: <SPECS_kernel_desktop.spec.1.86@pld-linux.org>
References: <SPECS_kernel_desktop.spec.1.85@pld-linux.org>
	<SPECS_kernel_desktop.spec.1.86@pld-linux.org>
Message-ID: <Pine.LNX.4.64N.0611221539490.6875@fgnss.gnfx.tqn.cy>

On Wed, 22 Nov 2006, sparky wrote:
> Author: sparky                       Date: Wed Nov 22 13:47:54 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - cxacru.c patch metged to small_fixes - please, don't use separated
>   patches for so small changes
[...]
> -Patch101:	kernel-desktop-cxacru.c.patch
[...]
> -%patch101 -p1
[...]

Why?
It's small, nice-stand-alone change. It's easy to consider drop/fix after 
some upgrade...

-- 
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 sparky at pld-linux.org  Wed Nov 22 17:31:41 2006
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Wed, 22 Nov 2006 17:31:41 +0100
Subject: SPECS: kernel-desktop.spec - cxacru.c patch metged to small_fixes
	...
In-Reply-To: <Pine.LNX.4.64N.0611221539490.6875@fgnss.gnfx.tqn.cy>
References: <SPECS_kernel_desktop.spec.1.85@pld-linux.org>
	<SPECS_kernel_desktop.spec.1.86@pld-linux.org>
	<Pine.LNX.4.64N.0611221539490.6875@fgnss.gnfx.tqn.cy>
Message-ID: <20061122163141.GA4166@pld-linux.org>

On Wed, Nov 22, 2006 at 03:41:13PM +0100, Pawe? Go?aszewski wrote:
> On Wed, 22 Nov 2006, sparky wrote:
> > Author: sparky                       Date: Wed Nov 22 13:47:54 2006 GMT
> > Module: SPECS                         Tag: HEAD
> > ---- Log message:
> > - cxacru.c patch metged to small_fixes - please, don't use separated
> >   patches for so small changes

> Why?
> It's small, nice-stand-alone change. It's easy to consider drop/fix after 
> some upgrade...

After adding some more I won't be able to find %build section.

I think small_fixes is just perfect for such changes because after any
upgrate it should be revised.


-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From glen at delfi.ee  Wed Nov 22 19:11:20 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 22 Nov 2006 20:11:20 +0200
Subject: firefox 2 for ac
In-Reply-To: <200611161241.28877.glen@delfi.ee>
References: <200611161241.28877.glen@delfi.ee>
Message-ID: <200611222011.20342.glen@delfi.ee>

there problems descibred below ceased to exist after using release >= 1 and 
removing /usr/lib/mozilla-firefox/chrome/chrome/app-chrome.manifest, which i 
don't know exactly how it appeared to my system.

if somebody wishes to dig into it then the file is available at 
http://glen.alkohol.ee/pld/app-chrome.manifest

On Thursday 16 November 2006 12:41, Elan Ruusam?e wrote:
> has anybody tried building it for ac? does it work properly too?
>
> i have noticed two problems
> - bookmarks menu is empty unless new bookmark is added or it's opened at
> sidebar
> - the statusbar is 80px high.
> - throbber keeps spinning if the tab has content downloaded
> - searchbox doesn't activate on enter nor on mouse click on icon

-- 
glen


From tomek at uninet.com.pl  Thu Nov 23 09:46:37 2006
From: tomek at uninet.com.pl (Tomasz Trojanowski)
Date: Thu, 23 Nov 2006 09:46:37 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611211222.50282.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611211134.48181.arekm@pld-linux.org>
	<200611211136.54966.arekm@pld-linux.org>
	<200611211222.50282.arekm@pld-linux.org>
Message-ID: <1164271597.2771.5.camel@localhost.localdomain>

On Tue, 2006-11-21 at 12:22 +0100, Arkadiusz Miskiewicz wrote:
> On Tuesday 21 November 2006 11:36, Arkadiusz Miskiewicz wrote:
> > On Tuesday 21 November 2006 11:34, Arkadiusz Miskiewicz wrote:
> > > #define
> > > _LIBTOOL_PROVIDES  "/usr/bin/find /usr/lib -name '*.la' |
> > > /usr/lib/rpm/libtooldeps.sh -P /"
> > >
> > > and Provides are broken.
> >
> > Or not. Sounds like some weird default setting only.
> 
> rpmbuild:
> 161 /home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/libDCOP.la ASCII 
> English text, with very long lines
> 
> file:
> /home/users/arekm/tmp/kdelibs-3.5.5-root-arekm/usr/lib64/libDCOP.la: libtool 
> library file

I report that bug some time ago on pld-devel-pl and
rpm-list(at)redhat.com. File seems be broken because:


$ file /usr/lib/*.la
/usr/lib/libFLAC.la:                      libtool library file
/usr/lib/libICE.la:                       ASCII English text
/usr/lib/libIDL-2.la:                     libtool library file
/usr/lib/libORBit-2.la:                   libtool library file
/usr/lib/libORBit-imodule-2.la:           libtool library file
/usr/lib/libORBitCosNaming-2.la:          libtool library file
/usr/lib/libOggFLAC.la:                   ASCII English text
/usr/lib/libSDL.la:                       ASCII English text
/usr/lib/libSM.la:                        ASCII English text
[...]

but

$ file /usr/lib/libICE.la
/usr/lib/libICE.la: libtool library file
$ file /usr/lib/libSDL.la
/usr/lib/libSDL.la: libtool library file
$ file /usr/lib/libSM.la
/usr/lib/libSM.la: libtool library file

Tomek




From qboosh at pld-linux.org  Thu Nov 23 21:12:49 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Thu, 23 Nov 2006 21:12:49 +0100
Subject: SPECS: kernel-misc-nforce.spec - release 5
In-Reply-To: <SPECS_kernel_misc_nforce.spec.1.29@pld-linux.org>
References: <SPECS_kernel_misc_nforce.spec.1.28@pld-linux.org>
	<SPECS_kernel_misc_nforce.spec.1.29@pld-linux.org>
Message-ID: <20061123201249.GA5286@gruby.cs.net.pl>

On Thu, Nov 23, 2006 at 07:53:06PM +0100, hawk wrote:
> Author: hawk                         Date: Thu Nov 23 18:53:05 2006 GMT
> Module: SPECS                         Tag: HEAD
> ---- Log message:
> - release 5
> 
> ---- Files affected:
> SPECS:
>    kernel-misc-nforce.spec (1.28 -> 1.29) 

This spec is obsolete. NVIDIA noticed and acknowledged existence of
open-source drivers for nForce chipsets:

http://www.nvidia.com/object/linux_nforce_1.11.html


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


From arekm at pld-linux.org  Thu Nov 23 22:26:50 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Thu, 23 Nov 2006 22:26:50 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <1164271597.2771.5.camel@localhost.localdomain>
References: <4561B2A9.1020300@agmk.net>
	<200611211222.50282.arekm@pld-linux.org>
	<1164271597.2771.5.camel@localhost.localdomain>
Message-ID: <200611232226.50836.arekm@pld-linux.org>

On Thursday 23 November 2006 09:46, Tomasz Trojanowski wrote:

>
> $ file /usr/lib/*.la
> /usr/lib/libFLAC.la:                      libtool library file
> /usr/lib/libICE.la:                       ASCII English text
> /usr/lib/libIDL-2.la:                     libtool library file
> /usr/lib/libORBit-2.la:                   libtool library file
> /usr/lib/libORBit-imodule-2.la:           libtool library file
> /usr/lib/libORBitCosNaming-2.la:          libtool library file
> /usr/lib/libOggFLAC.la:                   ASCII English text
> /usr/lib/libSDL.la:                       ASCII English text
> /usr/lib/libSM.la:                        ASCII English text
> [...]
>

libmagic misshandles /80 in the pattern. Drop it and it will detect all files 
as libtool one (as temporary workaround).

> Tomek

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


From grzegorz at art4u.pl  Fri Nov 24 11:02:44 2006
From: grzegorz at art4u.pl (Grzegorz 'Dzikus' Sterniczuk)
Date: Fri, 24 Nov 2006 11:02:44 +0100 (CET)
Subject: eventum-db.patch missing
Message-ID: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>

[grzegorz at pld_ac ~]$ builder -bb eventum
# $Revision: 1.217 $, $Date: 2006/11/24 09:58:11 $

No conditional flags passed

from available:
--with   :       order_patch postfix qmail
--without:

Available branches: iss_sort localization AC-branch
eventum-20061119.3143.tar.bz2 having proper md5sum already exists
cvs server: nothing known about eventum-db.patch
Budowanie dla platform: noarch
Budowanie dla noarch
b??d: Plik /home/users/grzegorz/rpm/SOURCES/eventum-db.patch: Nie ma 
takiego pliku ani katalogu
Error: package build failed. (no more info)

[grzegorz at pld_ac SOURCES]$ cvs stat eventum-db.patch
cvs server: nothing known about eventum-db.patch
===================================================================
File: no file eventum-db.patch          Status: Unknown

    Working revision:    No entry for eventum-db.patch
    Repository revision: No revision control file

-- 
Grzegorz 'dzikus' Sterniczuk  grzegorz at tyski.pl

From glen at delfi.ee  Fri Nov 24 12:50:14 2006
From: glen at delfi.ee (Elan =?utf-8?q?Ruusam=C3=A4e?=)
Date: Fri, 24 Nov 2006 13:50:14 +0200
Subject: eventum-db.patch missing
In-Reply-To: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
Message-ID: <200611241350.14511.glen@delfi.ee>

well. to use that patch you must first use version before that one, so the 
previous trigger can be ran, otherwise the other db changes won't be done 
(that patch just removes duplicate update queries)

and that version is not good. it gives lots of vsprintf errors due unescaped % 
in gettext strings. i suggest to use instead 
eventum-1.7.1-4.128.20061107.3132 or rpm from AC-main.

but anyway i added the patch to cvs.

On Friday 24 November 2006 12:02, Grzegorz 'Dzikus' Sterniczuk wrote:
> [grzegorz at pld_ac ~]$ builder -bb eventum
> # $Revision: 1.217 $, $Date: 2006/11/24 09:58:11 $
>
> No conditional flags passed
>
> from available:
> --with   :       order_patch postfix qmail
> --without:
>
> Available branches: iss_sort localization AC-branch
> eventum-20061119.3143.tar.bz2 having proper md5sum already exists
> cvs server: nothing known about eventum-db.patch
> Budowanie dla platform: noarch
> Budowanie dla noarch
> b??d: Plik /home/users/grzegorz/rpm/SOURCES/eventum-db.patch: Nie ma
> takiego pliku ani katalogu
> Error: package build failed. (no more info)
>
> [grzegorz at pld_ac SOURCES]$ cvs stat eventum-db.patch
> cvs server: nothing known about eventum-db.patch
> ===================================================================
> File: no file eventum-db.patch          Status: Unknown
>
>     Working revision:    No entry for eventum-db.patch
>     Repository revision: No revision control file

-- 
glen


From sparky at pld-linux.org  Fri Nov 24 17:30:46 2006
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Fri, 24 Nov 2006 17:30:46 +0100
Subject: SOURCES: zaptel-make.patch - re-enabled ztdummy module
In-Reply-To: <SOURCES_zaptel_make.patch.1.9@pld-linux.org>
References: <SOURCES_zaptel_make.patch.1.8@pld-linux.org>
	<SOURCES_zaptel_make.patch.1.9@pld-linux.org>
Message-ID: <20061124163046.GA30097@pld-linux.org>

On Fri, Nov 24, 2006 at 05:19:57PM +0100, hawk wrote:
> Author: hawk                         Date: Fri Nov 24 16:19:57 2006 GMT
> Module: SOURCES                       Tag: HEAD
> ---- Log message:
> - re-enabled ztdummy module

AFAIR ztdummy requires HZ=1000, so it's useless with our kernel.spec,
where HZ=250

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From wrobell at pld-linux.org  Fri Nov 24 18:36:39 2006
From: wrobell at pld-linux.org (wrobell)
Date: Fri, 24 Nov 2006 17:36:39 +0000
Subject: eventum-db.patch missing
In-Reply-To: <200611241350.14511.glen@delfi.ee>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee>
Message-ID: <20061124173639.GA23564@borg>

On Fri, Nov 24, 2006 at 01:50:14PM +0200, Elan Ruusam?e wrote:
> well. to use that patch you must first use version before that one, so the 
> previous trigger can be ran, otherwise the other db changes won't be done 
> (that patch just removes duplicate update queries)

please, stop top-posting.

    wrobell <wrobell at pld-linux.org>


From arekm at pld-linux.org  Sat Nov 25 00:10:01 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Sat, 25 Nov 2006 00:10:01 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611232226.50836.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<1164271597.2771.5.camel@localhost.localdomain>
	<200611232226.50836.arekm@pld-linux.org>
Message-ID: <200611250010.01660.arekm@pld-linux.org>

On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
> On Thursday 23 November 2006 09:46, Tomasz Trojanowski wrote:
> > $ file /usr/lib/*.la
> > /usr/lib/libFLAC.la:                      libtool library file
> > /usr/lib/libICE.la:                       ASCII English text
> > /usr/lib/libIDL-2.la:                     libtool library file
> > /usr/lib/libORBit-2.la:                   libtool library file
> > /usr/lib/libORBit-imodule-2.la:           libtool library file
> > /usr/lib/libORBitCosNaming-2.la:          libtool library file
> > /usr/lib/libOggFLAC.la:                   ASCII English text
> > /usr/lib/libSDL.la:                       ASCII English text
> > /usr/lib/libSM.la:                        ASCII English text
> > [...]
>
> libmagic misshandles /80 in the pattern. Drop it and it will detect all
> files as libtool one (as temporary workaround).

Fix (not sure if best one; at least working):

http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.1

>
> > Tomek

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


From arekm at pld-linux.org  Sat Nov 25 00:35:02 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Sat, 25 Nov 2006 00:35:02 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611250010.01660.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611232226.50836.arekm@pld-linux.org>
	<200611250010.01660.arekm@pld-linux.org>
Message-ID: <200611250035.02995.arekm@pld-linux.org>

On Saturday 25 November 2006 00:10, Arkadiusz Miskiewicz wrote:
> On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:

> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.1

More hacky but better :)
http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.2

> > > Tomek

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


From tomek at uninet.com.pl  Sat Nov 25 01:10:18 2006
From: tomek at uninet.com.pl (Tomasz Trojanowski)
Date: Sat, 25 Nov 2006 01:10:18 +0100
Subject: [th/rpm] R: libtool(*.la) nie widzi P: z tego samego pakietu?
In-Reply-To: <200611250010.01660.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<1164271597.2771.5.camel@localhost.localdomain>
	<200611232226.50836.arekm@pld-linux.org>
	<200611250010.01660.arekm@pld-linux.org>
Message-ID: <1164413418.2791.8.camel@localhost.localdomain>

On Sat, 2006-11-25 at 00:10 +0100, Arkadiusz Miskiewicz wrote:
> On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
> > On Thursday 23 November 2006 09:46, Tomasz Trojanowski wrote:
> > > $ file /usr/lib/*.la
> > > /usr/lib/libFLAC.la:                      libtool library file
> > > /usr/lib/libICE.la:                       ASCII English text
> > > /usr/lib/libIDL-2.la:                     libtool library file
> > > /usr/lib/libORBit-2.la:                   libtool library file
> > > /usr/lib/libORBit-imodule-2.la:           libtool library file
> > > /usr/lib/libORBitCosNaming-2.la:          libtool library file
> > > /usr/lib/libOggFLAC.la:                   ASCII English text
> > > /usr/lib/libSDL.la:                       ASCII English text
> > > /usr/lib/libSM.la:                        ASCII English text
> > > [...]
> >
> > libmagic misshandles /80 in the pattern. Drop it and it will detect all
> > files as libtool one (as temporary workaround).
> 
> Fix (not sure if best one; at least working):
> 
> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.1

Great. I'm looking for fix since few months.

Regards
    Tomasz Tojanowski




From glen at delfi.ee  Sat Nov 25 13:46:09 2006
From: glen at delfi.ee (Elan =?utf-8?q?Ruusam=C3=A4e?=)
Date: Sat, 25 Nov 2006 14:46:09 +0200
Subject: eventum-db.patch missing
In-Reply-To: <20061124173639.GA23564@borg>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee> <20061124173639.GA23564@borg>
Message-ID: <200611251446.09593.glen@delfi.ee>

what for? :)

On Friday 24 November 2006 19:36, wrobell wrote:
> On Fri, Nov 24, 2006 at 01:50:14PM +0200, Elan Ruusam?e wrote:
> > well. to use that patch you must first use version before that one, so
> > the previous trigger can be ran, otherwise the other db changes won't be
> > done (that patch just removes duplicate update queries)
>
> please, stop top-posting.
>
>     wrobell <wrobell at pld-linux.org>

-- 
glen


From qboosh at pld-linux.org  Sat Nov 25 14:38:48 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sat, 25 Nov 2006 14:38:48 +0100
Subject: unwanted LDFLAGS propagation
Message-ID: <20061125133848.GA19800@fngna.oyu>

We stopped experiencing such problems since dropping "-s" from
rpmldflags, but now it hits again:

$ mysql_config --libs
-Wl,--as-needed -L/usr/lib -lmysqlclient -lz -lm -L/usr/lib -lssl -lcrypto

$ pg_config --ldflags
-Wl,--as-needed

...and so on.
It shouldn't be propagated by all *-config scripts --{libs,ldflags}
options.


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


From wrobell at pld-linux.org  Sat Nov 25 16:07:51 2006
From: wrobell at pld-linux.org (wrobell)
Date: Sat, 25 Nov 2006 15:07:51 +0000
Subject: eventum-db.patch missing
In-Reply-To: <200611251446.09593.glen@delfi.ee>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee> <20061124173639.GA23564@borg>
	<200611251446.09593.glen@delfi.ee>
Message-ID: <20061125150751.GB23529@borg>

On Sat, Nov 25, 2006 at 02:46:09PM +0200, Elan Ruusam?e wrote:
> what for? :)

guess :P

    wrobell <wrobell at pld-linux.org>

> On Friday 24 November 2006 19:36, wrobell wrote:
> > On Fri, Nov 24, 2006 at 01:50:14PM +0200, Elan Ruusam?e wrote:
> > > well. to use that patch you must first use version before that one, so
> > > the previous trigger can be ran, otherwise the other db changes won't be
> > > done (that patch just removes duplicate update queries)
> >
> > please, stop top-posting.
> >
> >     wrobell <wrobell at pld-linux.org>
> 
> -- 
> glen
> _______________________________________________
> 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  Sat Nov 25 16:31:42 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sat, 25 Nov 2006 16:31:42 +0100
Subject: eventum-db.patch missing
In-Reply-To: <200611251446.09593.glen@delfi.ee>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee> <20061124173639.GA23564@borg>
	<200611251446.09593.glen@delfi.ee>
Message-ID: <20061125153142.GB19800@fngna.oyu>

- Like that.
- How?
- Because it reverses the usual way we read information.
- Why topposting is bad?

:P

On Sat, Nov 25, 2006 at 02:46:09PM +0200, Elan Ruusam?e wrote:
> what for? :)
> 
> On Friday 24 November 2006 19:36, wrobell wrote:
> > On Fri, Nov 24, 2006 at 01:50:14PM +0200, Elan Ruusam?e wrote:
> > > well. to use that patch you must first use version before that one, so
> > > the previous trigger can be ran, otherwise the other db changes won't be
> > > done (that patch just removes duplicate update queries)
> >
> > please, stop top-posting.


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


From sparky at pld-linux.org  Sat Nov 25 16:54:59 2006
From: sparky at pld-linux.org (Przemyslaw Iskra)
Date: Sat, 25 Nov 2006 16:54:59 +0100
Subject: unwanted LDFLAGS propagation
In-Reply-To: <20061125133848.GA19800@fngna.oyu>
References: <20061125133848.GA19800@fngna.oyu>
Message-ID: <20061125155459.GA9156@pld-linux.org>

On Sat, Nov 25, 2006 at 02:38:48PM +0100, Jakub Bogusz wrote:
> We stopped experiencing such problems since dropping "-s" from
> rpmldflags, but now it hits again:
> 
> $ mysql_config --libs
> -Wl,--as-needed -L/usr/lib -lmysqlclient -lz -lm -L/usr/lib -lssl -lcrypto
> 
> $ pg_config --ldflags
> -Wl,--as-needed
> 
> ...and so on.

Well, I've got some *-config and I've found "-Wl,--as-needed" string in
only one of them: nspr, which does not display it anyway.

> It shouldn't be propagated by all *-config scripts --{libs,ldflags}
> options.

Of source it shouldn't, but I see no general solution for it, those
*-config files may be shell code, python, C binary and probably more
kind of things. So, each of them has to be treated separately.

-- 
 ____  Sparky{PI] -- Przemyslaw _  ___  _  _  ........... LANG...Pl..Ca..Es..En
/____) ___  ___  _ _ || Iskra  |  | _ \| |  | : WWW........ppcrcd.pld-linux.org
\____\| -_)'___| ||^'||//\\// <   |  _/| |  | : JID......sparky<at>jabberes.org
(____/||   (_-_|_||  ||\\ ||   |_ |_|  |_| _| : Mail....sparky<at>pld-linux.org


From grzegorz at art4u.pl  Sat Nov 25 17:11:07 2006
From: grzegorz at art4u.pl (Grzegorz 'Dzikus' Sterniczuk)
Date: Sat, 25 Nov 2006 17:11:07 +0100 (CET)
Subject: eventum-db.patch missing
In-Reply-To: <200611241350.14511.glen@delfi.ee>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee>
Message-ID: <Pine.LNX.4.64N.0611251702300.30404@ybpnyubfg>

On Fri, 24 Nov 2006, Elan Ruusam?e wrote:

> well. to use that patch you must first use version before that one, so the 
> previous trigger can be ran, otherwise the other db changes won't be 
> done (that patch just removes duplicate update queries)

This isn't explanation why this patch wasn't in repo :).

> but anyway i added the patch to cvs.

thx

By the way does anybody working on polish translation of eventum?

-- 
Grzegorz 'dzikus' Sterniczuk  grzegorz at tyski.pl

From glen at delfi.ee  Sat Nov 25 17:37:10 2006
From: glen at delfi.ee (Elan =?iso-8859-15?q?Ruusam=E4e?=)
Date: Sat, 25 Nov 2006 18:37:10 +0200
Subject: eventum-db.patch missing
In-Reply-To: <Pine.LNX.4.64N.0611251702300.30404@ybpnyubfg>
References: <Pine.LNX.4.64N.0611241100400.5481@ybpnyubfg>
	<200611241350.14511.glen@delfi.ee>
	<Pine.LNX.4.64N.0611251702300.30404@ybpnyubfg>
Message-ID: <200611251837.10527.glen@delfi.ee>

On Saturday 25 November 2006 18:11, Grzegorz 'Dzikus' Sterniczuk wrote:
> By the way does anybody working on polish translation of eventum?
afaik no, at least if somebody is then he forgot to "register":
http://eventum.mysql.org/wiki/index.php/Localization#Planned_Languages

and i'm not aware somebody actually using Eventum in pld but me :)

-- 
glen


From twittner at o2.pl  Sat Nov 25 18:30:53 2006
From: twittner at o2.pl (Tomasz Wittner)
Date: Sat, 25 Nov 2006 18:30:53 +0100
Subject: SPECS (AC-branch): courier.spec - ac 2.60 complains: autoheader:
	e...
In-Reply-To: <SPECS_courier.spec.1.141.2.2@pld-linux.org>
References: <SPECS_courier.spec.1.141.2.1@pld-linux.org>
	<SPECS_courier.spec.1.141.2.2@pld-linux.org>
Message-ID: <200611251830.53790.twittner@o2.pl>

On Sat 25. November 2006 17:57, glen wrote:
> Author: glen                         Date: Sat Nov 25 16:57:49 2006 GMT
> Module: SPECS                         Tag: AC-branch
> ---- Log message:
> - ac 2.60 complains: autoheader: error: AC_CONFIG_HEADERS not found in 
configure.in
> - rel 3
> 
> ---- Files affected:
> SPECS:
>    courier.spec (1.141.2.1 -> 1.141.2.2) 
> 
> ---- Diffs:
> 
> ================================================================
> Index: SPECS/courier.spec
> diff -u SPECS/courier.spec:1.141.2.1 SPECS/courier.spec:1.141.2.2
> --- SPECS/courier.spec:1.141.2.1	Sat Nov 25 16:57:10 2006
> +++ SPECS/courier.spec	Sat Nov 25 17:57:43 2006
> @@ -14,7 +14,7 @@
>  Summary(pl):	Serwer poczty Courier
>  Name:		courier
>  Version:	0.53.3
> -Release:	2
> +Release:	3
>  License:	GPL
>  Group:		Networking/Daemons
>  Source0:	http://dl.sourceforge.net/courier/%{name}-%{version}.tar.bz2
> @@ -285,9 +285,10 @@
>  </Directory>
>  " > apache.conf
>  
> -%build
>  # we don't want fax module
>  rm -rf courier/module.fax
> +
> +%build
>  cp -f /usr/share/automake/config.sub webmail
>  
>  # Change Makefile.am files and force recreate Makefile.in's.
> @@ -302,7 +303,9 @@
>  	%{__libtoolize}
>  	%{__aclocal}
>  	%{__autoconf}
> -	%{__autoheader}
> +	if grep -q AM_CONFIG_HEADER configure.in; then
> +		%{__autoheader}
> +	fi
??? 
1. not AM_CONFIG_HEADER but AC_CONFIG_HEADER
2. maybe adding AC_CONFIG_HEADER(config.h) right after AC_INIT is better 
solution. See `info autoconf --index-search=AC_CONFIG_HEADERS'
[...]

-- 
Tomasz Wittner


From arekm at pld-linux.org  Sun Nov 26 21:50:27 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Sun, 26 Nov 2006 21:50:27 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611250035.02995.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611250010.01660.arekm@pld-linux.org>
	<200611250035.02995.arekm@pld-linux.org>
Message-ID: <200611262150.27146.arekm@pld-linux.org>

On Saturday 25 November 2006 00:35, Arkadiusz Miskiewicz wrote:
> On Saturday 25 November 2006 00:10, Arkadiusz Miskiewicz wrote:
> > On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
> >
> > http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.1
>
> More hacky but better :)
> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.2

Of course there are still problems.

 libdb.la symlink to real file libdb-4.5.la

Now other packages get R: libtool(libdb.la) but nothing provides it since it's 
not detected.

Wouldn't be it better to generate Requires: /usr/lib/file.la instead of 
libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild all packages 
containing .la files.

> > > > Tomek

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


From n3npq at mac.com  Sun Nov 26 22:02:52 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Sun, 26 Nov 2006 16:02:52 -0500
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611262150.27146.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611250010.01660.arekm@pld-linux.org>
	<200611250035.02995.arekm@pld-linux.org>
	<200611262150.27146.arekm@pld-linux.org>
Message-ID: <6074A94F-E671-4B5E-A4C8-FECE1C237F55@mac.com>


On Nov 26, 2006, at 3:50 PM, Arkadiusz Miskiewicz wrote:

> On Saturday 25 November 2006 00:35, Arkadiusz Miskiewicz wrote:
>> On Saturday 25 November 2006 00:10, Arkadiusz Miskiewicz wrote:
>>> On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
>>>
>>> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch? 
>>> rev=1.1
>>
>> More hacky but better :)
>> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch? 
>> rev=1.2
>
> Of course there are still problems.
>
>  libdb.la symlink to real file libdb-4.5.la
>
> Now other packages get R: libtool(libdb.la) but nothing provides it  
> since it's
> not detected.
>
> Wouldn't be it better to generate Requires: /usr/lib/file.la  
> instead of
> libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild  
> all packages
> containing .la files.
>

The intent is to isolate libtool dependencies within a "libtool(...)"  
name space
so that libtool dependencies can be checked independently of other  
dependencies.

But sure, dependencies are just strings, just like file paths ;-)

73 de Jeff



From tomek at uninet.com.pl  Sun Nov 26 22:12:31 2006
From: tomek at uninet.com.pl (Tomasz Trojanowski)
Date: Sun, 26 Nov 2006 22:12:31 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611262150.27146.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611250010.01660.arekm@pld-linux.org>
	<200611250035.02995.arekm@pld-linux.org>
	<200611262150.27146.arekm@pld-linux.org>
Message-ID: <1164575551.2842.0.camel@localhost.localdomain>

On Sun, 2006-11-26 at 21:50 +0100, Arkadiusz Miskiewicz wrote:
> On Saturday 25 November 2006 00:35, Arkadiusz Miskiewicz wrote:
> > On Saturday 25 November 2006 00:10, Arkadiusz Miskiewicz wrote:
> > > On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
> > >
> > > http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.1
> >
> > More hacky but better :)
> > http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch?rev=1.2
> 
> Of course there are still problems.
> 
>  libdb.la symlink to real file libdb-4.5.la
> 
> Now other packages get R: libtool(libdb.la) but nothing provides it since it's 
> not detected.
> 
> Wouldn't be it better to generate Requires: /usr/lib/file.la instead of 
> libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild all packages 
> containing .la files.

Or simply:

--- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
+++ libtooldeps.sh      2006-11-26 22:06:56 +0100
@@ -30,6 +30,7 @@
            do
                case "$dep" in
                /*.la)
+                   $dep = `readlink -f $dep`
                    echo "libtool($dep)"
                    ;;
                esac

Tomek




From n3npq at mac.com  Sun Nov 26 22:21:41 2006
From: n3npq at mac.com (Jeff Johnson)
Date: Sun, 26 Nov 2006 16:21:41 -0500
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <1164575551.2842.0.camel@localhost.localdomain>
References: <4561B2A9.1020300@agmk.net>
	<200611250010.01660.arekm@pld-linux.org>
	<200611250035.02995.arekm@pld-linux.org>
	<200611262150.27146.arekm@pld-linux.org>
	<1164575551.2842.0.camel@localhost.localdomain>
Message-ID: <C28C8596-91A6-4837-9C6F-602B324A8D46@mac.com>


On Nov 26, 2006, at 4:12 PM, Tomasz Trojanowski wrote:

> On Sun, 2006-11-26 at 21:50 +0100, Arkadiusz Miskiewicz wrote:
>> On Saturday 25 November 2006 00:35, Arkadiusz Miskiewicz wrote:
>>> On Saturday 25 November 2006 00:10, Arkadiusz Miskiewicz wrote:
>>>> On Thursday 23 November 2006 22:26, Arkadiusz Miskiewicz wrote:
>>>>
>>>> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file- 
>>>> offset.patch?rev=1.1
>>>
>>> More hacky but better :)
>>> http://cvs.pld-linux.org/cgi-bin/cvsweb/SOURCES/file-offset.patch? 
>>> rev=1.2
>>
>> Of course there are still problems.
>>
>>  libdb.la symlink to real file libdb-4.5.la
>>
>> Now other packages get R: libtool(libdb.la) but nothing provides  
>> it since it's
>> not detected.
>>
>> Wouldn't be it better to generate Requires: /usr/lib/file.la  
>> instead of
>> libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild  
>> all packages
>> containing .la files.
>
> Or simply:
>
> --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> @@ -30,6 +30,7 @@
>             do
>                 case "$dep" in
>                 /*.la)
> +                   $dep = `readlink -f $dep`
>                     echo "libtool($dep)"
>                     ;;
>                 esac
>

Nice.

FWIW, libdb.la within rpm is a sick Makefile hack. Patches cheerfully  
accepted.

73 de Jeff



From arekm at pld-linux.org  Sun Nov 26 22:22:59 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Sun, 26 Nov 2006 22:22:59 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <1164575551.2842.0.camel@localhost.localdomain>
References: <4561B2A9.1020300@agmk.net>
	<200611262150.27146.arekm@pld-linux.org>
	<1164575551.2842.0.camel@localhost.localdomain>
Message-ID: <200611262222.59287.arekm@pld-linux.org>

On Sunday 26 November 2006 22:12, Tomasz Trojanowski wrote:

> > Of course there are still problems.
> >
> >  libdb.la symlink to real file libdb-4.5.la
> >
> > Now other packages get R: libtool(libdb.la) but nothing provides it since
> > it's not detected.
> >
> > Wouldn't be it better to generate Requires: /usr/lib/file.la instead of
> > libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild all
> > packages containing .la files.
>
> Or simply:
>
> --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> @@ -30,6 +30,7 @@
>             do
>                 case "$dep" in
>                 /*.la)
> +                   $dep = `readlink -f $dep`
>                     echo "libtool($dep)"
>                     ;;
>                 esac
>

That won't work. The dep against libdb.la is correct one. The problem is with 
provides not requires.

> Tomek

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


From tomek at uninet.com.pl  Sun Nov 26 22:57:33 2006
From: tomek at uninet.com.pl (Tomasz Trojanowski)
Date: Sun, 26 Nov 2006 22:57:33 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611262222.59287.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611262150.27146.arekm@pld-linux.org>
	<1164575551.2842.0.camel@localhost.localdomain>
	<200611262222.59287.arekm@pld-linux.org>
Message-ID: <1164578253.2764.4.camel@localhost.localdomain>

On Sun, 2006-11-26 at 22:22 +0100, Arkadiusz Miskiewicz wrote:
> On Sunday 26 November 2006 22:12, Tomasz Trojanowski wrote:
> 
> > > Of course there are still problems.
> > >
> > >  libdb.la symlink to real file libdb-4.5.la
> > >
> > > Now other packages get R: libtool(libdb.la) but nothing provides it since
> > > it's not detected.
> > >
> > > Wouldn't be it better to generate Requires: /usr/lib/file.la instead of
> > > libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild all
> > > packages containing .la files.
> >
> > Or simply:
> >
> > --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> > +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> > @@ -30,6 +30,7 @@
> >             do
> >                 case "$dep" in
> >                 /*.la)
> > +                   $dep = `readlink -f $dep`
> >                     echo "libtool($dep)"
> >                     ;;
> >                 esac
> >
> 
> That won't work.

Will. Ie. some package R: libtool(/usr/lib/libdb.la), with this patch
will be converted to R: libtool(/usr/lib/libdb-4.5.la), which is
provided by db4.5-devel

>  The dep against libdb.la is correct one. The problem is with 
> provides not requires.



> 
> > Tomek
> 



From qboosh at pld-linux.org  Sun Nov 26 23:25:27 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sun, 26 Nov 2006 23:25:27 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <1164578253.2764.4.camel@localhost.localdomain>
References: <4561B2A9.1020300@agmk.net>
	<200611262150.27146.arekm@pld-linux.org>
	<1164575551.2842.0.camel@localhost.localdomain>
	<200611262222.59287.arekm@pld-linux.org>
	<1164578253.2764.4.camel@localhost.localdomain>
Message-ID: <20061126222527.GA24779@fngna.oyu>

On Sun, Nov 26, 2006 at 10:57:33PM +0100, Tomasz Trojanowski wrote:
> On Sun, 2006-11-26 at 22:22 +0100, Arkadiusz Miskiewicz wrote:
> > On Sunday 26 November 2006 22:12, Tomasz Trojanowski wrote:
> > > > Of course there are still problems.
> > > >
> > > >  libdb.la symlink to real file libdb-4.5.la
> > > >
> > > > Now other packages get R: libtool(libdb.la) but nothing provides it since
> > > > it's not detected.
> > > >
> > > > Wouldn't be it better to generate Requires: /usr/lib/file.la instead of
> > > > libtool(/usr/lib/file.la) ? That way we wouldn't need to rebuild all
> > > > packages containing .la files.
> > >
> > > Or simply:
> > >
> > > --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> > > +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> > > @@ -30,6 +30,7 @@
> > >             do
> > >                 case "$dep" in
> > >                 /*.la)
> > > +                   $dep = `readlink -f $dep`
> > >                     echo "libtool($dep)"
> > >                     ;;
> > >                 esac
> > >
> > 
> > That won't work.
> 
> Will. Ie. some package R: libtool(/usr/lib/libdb.la), with this patch
> will be converted to R: libtool(/usr/lib/libdb-4.5.la), which is
> provided by db4.5-devel

But in fact it requires /usr/lib/libdb.la, not /usr/lib/libdb-4.5.la.


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


From arekm at pld-linux.org  Sun Nov 26 23:26:15 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Sun, 26 Nov 2006 23:26:15 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <1164578253.2764.4.camel@localhost.localdomain>
References: <4561B2A9.1020300@agmk.net>
	<200611262222.59287.arekm@pld-linux.org>
	<1164578253.2764.4.camel@localhost.localdomain>
Message-ID: <200611262326.15695.arekm@pld-linux.org>

On Sunday 26 November 2006 22:57, Tomasz Trojanowski wrote:

> > > --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> > > +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> > > @@ -30,6 +30,7 @@
> > >             do
> > >                 case "$dep" in
> > >                 /*.la)
> > > +                   $dep = `readlink -f $dep`
> > >                     echo "libtool($dep)"
> > >                     ;;
> > >                 esac
> >
> > That won't work.
>
> Will. Ie. some package R: libtool(/usr/lib/libdb.la), with this patch
> will be converted to R: libtool(/usr/lib/libdb-4.5.la), which is
> provided by db4.5-devel

What happen if there will be new db4.6? You will need to rebuild things again 
to get libtool(libdb4.6.la) which wouldn't be required if the dep be 
libtoo(libdb.la).

But for now you solution is easiest.

> > > Tomek

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


From qboosh at pld-linux.org  Sun Nov 26 23:35:45 2006
From: qboosh at pld-linux.org (Jakub Bogusz)
Date: Sun, 26 Nov 2006 23:35:45 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611262326.15695.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611262222.59287.arekm@pld-linux.org>
	<1164578253.2764.4.camel@localhost.localdomain>
	<200611262326.15695.arekm@pld-linux.org>
Message-ID: <20061126223545.GD24779@fngna.oyu>

On Sun, Nov 26, 2006 at 11:26:15PM +0100, Arkadiusz Miskiewicz wrote:
> On Sunday 26 November 2006 22:57, Tomasz Trojanowski wrote:
> > > > --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> > > > +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> > > > @@ -30,6 +30,7 @@
> > > >             do
> > > >                 case "$dep" in
> > > >                 /*.la)
> > > > +                   $dep = `readlink -f $dep`
> > > >                     echo "libtool($dep)"
> > > >                     ;;
> > > >                 esac
> > >
> > > That won't work.
> >
> > Will. Ie. some package R: libtool(/usr/lib/libdb.la), with this patch
> > will be converted to R: libtool(/usr/lib/libdb-4.5.la), which is
> > provided by db4.5-devel
> 
> What happen if there will be new db4.6? You will need to rebuild things again 
> to get libtool(libdb4.6.la) which wouldn't be required if the dep be 
> libtoo(libdb.la).

It would be anyway, because of libdb* soname change.


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


From tomek at uninet.com.pl  Sun Nov 26 23:48:48 2006
From: tomek at uninet.com.pl (Tomasz Trojanowski)
Date: Sun, 26 Nov 2006 23:48:48 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611262326.15695.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net>
	<200611262222.59287.arekm@pld-linux.org>
	<1164578253.2764.4.camel@localhost.localdomain>
	<200611262326.15695.arekm@pld-linux.org>
Message-ID: <1164581329.2617.7.camel@localhost.localdomain>

On Sun, 2006-11-26 at 23:26 +0100, Arkadiusz Miskiewicz wrote:
> On Sunday 26 November 2006 22:57, Tomasz Trojanowski wrote:
> 
> > > > --- libtooldeps.sh.orig 2006-11-26 00:16:19 +0100
> > > > +++ libtooldeps.sh      2006-11-26 22:06:56 +0100
> > > > @@ -30,6 +30,7 @@
> > > >             do
> > > >                 case "$dep" in
> > > >                 /*.la)
> > > > +                   $dep = `readlink -f $dep`
> > > >                     echo "libtool($dep)"
> > > >                     ;;
> > > >                 esac
> > >
> > > That won't work.
> >
> > Will. Ie. some package R: libtool(/usr/lib/libdb.la), with this patch
> > will be converted to R: libtool(/usr/lib/libdb-4.5.la), which is
> > provided by db4.5-devel
> 
> What happen if there will be new db4.6? You will need to rebuild things again 
> to get libtool(libdb4.6.la) which wouldn't be required if the dep be 
> libtoo(libdb.la).

foo package requires libdb-4.5.so (not libdb.so), so IMO foo-devel
should require libtool(/usr/lib/libdb-4.5.la) not
libtool(/usr/lib/libdb.la).

For libdb-4.6 you must rebuild foo because of soname change.

Tomek




From arekm at pld-linux.org  Mon Nov 27 11:05:13 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Mon, 27 Nov 2006 11:05:13 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <20061126223545.GD24779@fngna.oyu>
References: <4561B2A9.1020300@agmk.net>
	<200611262326.15695.arekm@pld-linux.org>
	<20061126223545.GD24779@fngna.oyu>
Message-ID: <200611271105.13876.arekm@pld-linux.org>

On Sunday 26 November 2006 23:35, Jakub Bogusz wrote:

> > What happen if there will be new db4.6? You will need to rebuild things
> > again to get libtool(libdb4.6.la) which wouldn't be required if the dep
> > be libtoo(libdb.la).
>
> It would be anyway, because of libdb* soname change.

So the problem will be only if symlink changes while soname major number is 
the same. Anyway I've commited changes that use readlink (buggy at this 
moment; will fix ;)

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


From arekm at pld-linux.org  Mon Nov 27 18:20:53 2006
From: arekm at pld-linux.org (Arkadiusz Miskiewicz)
Date: Mon, 27 Nov 2006 18:20:53 +0100
Subject: [th/rpm] R: libtool(*.la) - good or bad
In-Reply-To: <200611271105.13876.arekm@pld-linux.org>
References: <4561B2A9.1020300@agmk.net> <20061126223545.GD24779@fngna.oyu>
	<200611271105.13876.arekm@pld-linux.org>
Message-ID: <200611271820.53976.arekm@pld-linux.org>

On Monday 27 November 2006 11:05, Arkadiusz Miskiewicz wrote:
> On Sunday 26 November 2006 23:35, Jakub Bogusz wrote:
> > > What happen if there will be new db4.6? You will need to rebuild things
> > > again to get libtool(libdb4.6.la) which wouldn't be required if the dep
> > > be libtoo(libdb.la).
> >
> > It would be anyway, because of libdb* soname change.
>
> So the problem will be only if symlink changes while soname major number is
> the same. Anyway I've commited changes that use readlink (buggy at this
> moment; will fix ;)

Commited few hours ago.

Anyway there is a new problem. Some la-files are included in main packages 
(kde as example) because these are needed for libltdl to work correctly. Due 
to that main packages start requiring tons of -devel subpackages :|

Solutions?
- ignore these deps on per spec basis
- only generate such Requires in -devel subpackages (checking by subpackage 
name)
- other?

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


From glen at delfi.ee  Wed Nov 29 00:10:14 2006
From: glen at delfi.ee (Elan =?iso-8859-1?q?Ruusam=E4e?=)
Date: Wed, 29 Nov 2006 01:10:14 +0200
Subject: glib2 G_GNUC_UNUSED
In-Reply-To: <200611030221.41731.glen@delfi.ee>
References: <200611030221.41731.glen@delfi.ee>
Message-ID: <200611290110.14223.glen@delfi.ee>

anyone? qboosh?


anyway i've looked how linux uses __used__:

/usr/src/linux/include/linux/compiler.h:

#if __GNUC__ > 4
#error no compiler-gcc.h file for this gcc version
#elif __GNUC__ == 4
# include <linux/compiler-gcc4.h>
#elif __GNUC__ == 3
# include <linux/compiler-gcc3.h>
#else
# error Sorry, your compiler is too old/not recognized.
#endif

/usr/src/linux/include/linux/compiler-gcc3.h:
#if __GNUC_MINOR__ >= 3
# define __attribute_used__ __attribute__((__used__))
#else
# define __attribute_used__ __attribute__((__unused__))
#endif

and /usr/src/linux/include/linux/compiler.h continues:
#ifndef __attribute_used__
# define __attribute_used__ /* unimplemented */
#endif

maybe this will help.

On Friday 03 November 2006 02:21, Elan Ruusam?e wrote:
> is glib2 defining the macro incompatible for gcc in Ac, or there's some
> other cause?
>
> $ g++ -c t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags gtk+-2.0`
> t.c:5: error: syntax error before `__attribute__'
>
> $ cat t.c
> #include <gtk/gtk.h>
> #include <glib.h>
>
> static void open_url_cb( GtkWidget *pButton G_GNUC_UNUSED, gpointer data )
> {
> }
>
> $ g++ -E t.c `pkg-config --cflags glib-2.0` `pkg-config --cflags
> gtk+-2.0`>t.C $ grep open_url_cb t.C
> static void open_url_cb( GtkWidget *pButton __attribute__((__unused__)),
> gpointer data )
>
> $ q gcc glib2 gtk+2
> gcc-5:3.3.6-4
> glib2-1:2.10.3-1
> gtk+2-2:2.8.20-1
> $
>
> http://developer.gnome.org/doc/API/2.0/glib/glib-Miscellaneous-Macros.html#
>G-GNUC-UNUSED:CAPS

-- 
glen


From arkadiusz.machol at gmail.com  Thu Nov 30 10:16:46 2006
From: arkadiusz.machol at gmail.com (=?ISO-8859-2?Q?Arkadiusz_Macho=B3?=)
Date: Thu, 30 Nov 2006 10:16:46 +0100
Subject: [SPECS] bzr.spec to latest rc1
Message-ID: <5ac9d7250611300116t1b7e3027ra22559e9ebc77f19@mail.gmail.com>

I've attached a patch upgrading bzr.spec to the latest version (rc1).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </mailman/pipermail/pld-devel-en/attachments/20061130/e814105a/attachment.htm>
-------------- next part --------------
--- bzr.spec.orig	2006-11-29 09:57:41.000000000 +0100
+++ bzr.spec	2006-11-29 09:55:29.000000000 +0100
@@ -2,12 +2,13 @@
 Summary:	Bazaar-NG - a changeset oriented revision control system
 Summary(pl):	Bazaar-NG - system kontroli wersji zorientowany na zestawy zmian
 Name:		bzr
-Version:	0.12
-Release:	1
+Version:	0.13
+%define		_rc	rc1
+Release:	0.%{_rc}.1
 License:	GPL v2
 Group:		Development/Version Control
-Source0:	http://bazaar-vcs.org/releases/src/%{name}-%{version}.tar.gz
-# Source0-md5:	68bf4e425a93bdf5961af9c3c35229bf
+Source0:	http://bazaar-vcs.org/releases/src/%{name}-%{version}%{_rc}.tar.gz
+# Source0-md5:	018787a350e423ca8d8da3bd11f77f72
 Patch0:		%{name}-FHS.patch
 URL:		http://bazaar-vcs.org/
 BuildRequires:	python >= 1:2.4
@@ -72,15 +73,16 @@
 Bazaar-NG obs?uguje tak?e wsp??dzielenie ga??zi mi?dzy programistami.
 
 %prep
-%setup -q
+%setup -q -n %{name}-%{version}%{_rc}
 %patch0 -p1
 
 %build
-python setup.py build
+%{__python} setup.py build
 
 %install
 rm -rf $RPM_BUILD_ROOT
-python setup.py install --optimize=2 \
+%{__python} setup.py install \
+	--optimize=2 \
 	--root=$RPM_BUILD_ROOT
 
 %py_postclean