Re: Delayed ldconfig execution in postinst step

2006-03-22 Thread Goswin von Brederlow
Bill Allombert [EMAIL PROTECTED] writes: On Sat, Mar 18, 2006 at 10:56:14PM +0100, Eduard Bloch wrote: #include hallo.h * Eduard Bloch [Sat, Mar 18 2006, 10:52:53PM]: #include hallo.h * Bill Allombert [Sat, Mar 18 2006, 02:56:27PM]: On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch

Re: Delayed ldconfig execution in postinst step

2006-03-18 Thread Bill Allombert
On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not cost much time because VFS cache has the relevant file parts. However,

Re: Delayed ldconfig execution in postinst step

2006-03-18 Thread Eduard Bloch
#include hallo.h * Bill Allombert [Sat, Mar 18 2006, 02:56:27PM]: On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not

Re: Delayed ldconfig execution in postinst step

2006-03-18 Thread Eduard Bloch
#include hallo.h * Eduard Bloch [Sat, Mar 18 2006, 10:52:53PM]: #include hallo.h * Bill Allombert [Sat, Mar 18 2006, 02:56:27PM]: On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, ... ld.so manpage imply the opposite. I would not be that sure. Following that

Re: Delayed ldconfig execution in postinst step

2006-03-18 Thread Bill Allombert
On Sat, Mar 18, 2006 at 10:56:14PM +0100, Eduard Bloch wrote: #include hallo.h * Eduard Bloch [Sat, Mar 18 2006, 10:52:53PM]: #include hallo.h * Bill Allombert [Sat, Mar 18 2006, 02:56:27PM]: On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, ... ld.so

Re: Delayed ldconfig execution in postinst step

2006-03-17 Thread Eduard Bloch
#include hallo.h * Goswin von Brederlow [Wed, Mar 15 2006, 03:17:29PM]: Eduard Bloch [EMAIL PROTECTED] writes: #include hallo.h * Goswin von Brederlow [Tue, Mar 14 2006, 10:11:43PM]: What is a depends? Do you mean dependency or dependents? I think he means dependents: If package

Re: Delayed ldconfig execution in postinst step

2006-03-17 Thread Goswin von Brederlow
Eduard Bloch [EMAIL PROTECTED] writes: #include hallo.h * Goswin von Brederlow [Wed, Mar 15 2006, 03:17:29PM]: Eduard Bloch [EMAIL PROTECTED] writes: #include hallo.h * Goswin von Brederlow [Tue, Mar 14 2006, 10:11:43PM]: What is a depends? Do you mean dependency or dependents?

Re: Delayed ldconfig execution in postinst step

2006-03-15 Thread Goswin von Brederlow
Eduard Bloch [EMAIL PROTECTED] writes: #include hallo.h * Goswin von Brederlow [Tue, Mar 14 2006, 10:11:43PM]: What is a depends? Do you mean dependency or dependents? I think he means dependents: If package foo depends on library foobar, dpkg/apt can unpack and

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Pierre Habouzit
Le Lun 13 Mars 2006 19:46, Eduard Bloch a écrit : #include hallo.h * Pierre Habouzit [Mon, Mar 13 2006, 07:16:22PM]: What is a depends? Do you mean dependency or dependents? Further, I would not depend on package installation operations but instead invent something like dpkg-hook

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Brian May
Goswin == Goswin von Brederlow [EMAIL PROTECTED] writes: - there are no dangerous transitions (libc5-libc6) that would require having updated ld.so.cache immediately - all applications should follow the ld.so.conf paths if something is not in the cache. No problem here. I am

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Frank Küster
Pierre Habouzit [EMAIL PROTECTED] wrote: * delayed execution * instantaneous execution (purges cache) * instantaneous execution if an operation is pending, nothing else (purge the cache). e.g. If a package uses tex to build some sort of documentation at install time (I know it's

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Ralf Wildenhues
[ Please Cc: me on replies ] Therefore it is crucial that only commands - that cannot sensibly be assumed to fail, und any circumstances including f'cked up filesystems, can be allowed. That would exclude ldconfig then. If someone happens to install sonamed libfoo.so only but rely on

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Eduard Bloch
#include hallo.h * Frank Küster [Tue, Mar 14 2006, 09:52:06AM]: Registering the TeX fonts, on the other hand, is a good candidate for delayed execution: It takes long each time it is done, many packages do it in their postinst, and it will fail with the first package that is configured, no

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Bill Allombert
On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not cost much time because VFS cache has the relevant file parts. However,

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Henning Makholm
Scripsit Ralf Wildenhues [EMAIL PROTECTED] If someone happens to install sonamed libfoo.so only but rely on ldconfig to create the libfoo.so.N symlink Ralf (no I don't know whether Debian policy already forbids this setup) It takes about 20 seconds to find out that Debian policy does

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Brendan O'Dea
On Tue, Mar 14, 2006 at 04:20:28AM -0600, Bill Allombert wrote: I offer to implement a update-ldconfig program that would work the same way update-menus work, by checking a lock and forking in the background and waiting for the dpkg lock. It's more than just update-menus and ldconfig.

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Ralf Wildenhues
* Henning Makholm wrote: Scripsit Ralf Wildenhues [EMAIL PROTECTED] If someone happens to install sonamed libfoo.so only but rely on ldconfig to create the libfoo.so.N symlink It takes about 20 seconds to find out that Debian policy does forbid such a packaging - see the last paragraph

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Frank Küster
Brendan O'Dea [EMAIL PROTECTED] wrote: dpkg would then run any pending hooks (with diagnostics as proposed by Eduard in [EMAIL PROTECTED]) by calling: I would like to add a wish WRT diagnostics: There should be a commandline option (handled with --add-option) that has the effect that the

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Eduard Bloch
#include hallo.h * Bill Allombert [Tue, Mar 14 2006, 04:20:28AM]: On Sun, Mar 12, 2006 at 01:56:13PM +0100, Eduard Bloch wrote: Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Goswin von Brederlow
Frank Küster [EMAIL PROTECTED] writes: Eduard Bloch [EMAIL PROTECTED] wrote: dpkg-hook /usr/lib/man/update-manpages - run only once in total dpkg-hook --on-depends foobar ldconfig - run once before depends of foobar What is a depends? Do you mean dependency or dependents? I think he

Re: Delayed ldconfig execution in postinst step

2006-03-14 Thread Eduard Bloch
#include hallo.h * Goswin von Brederlow [Tue, Mar 14 2006, 10:11:43PM]: What is a depends? Do you mean dependency or dependents? I think he means dependents: If package foo depends on library foobar, dpkg/apt can unpack and configure-without-ldconfig all packages that don't depend on

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Eduard Bloch
#include hallo.h * Goswin von Brederlow [Sun, Mar 12 2006, 03:35:42PM]: I think it would be a good idea to have a general dpkg hook to register a command to be run at the end of dpkg. The syntax would be something like this: dpkg-hook /usr/lib/man/update-manpages - run only once in total

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Frank Küster
Eduard Bloch [EMAIL PROTECTED] wrote: dpkg-hook /usr/lib/man/update-manpages - run only once in total dpkg-hook --on-depends foobar ldconfig - run once before depends of foobar What is a depends? Do you mean dependency or dependents? I think he means dependents: If package foo depends on

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Eduard Bloch
#include hallo.h * Pierre Habouzit [Mon, Mar 13 2006, 07:16:22PM]: What is a depends? Do you mean dependency or dependents? Further, I would not depend on package installation operations but instead invent something like dpkg-hook --execute ldconfig to run outstanding tasks noted under

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Eduard Bloch
#include hallo.h * Frank Küster [Mon, Mar 13 2006, 07:28:42PM]: Eduard Bloch [EMAIL PROTECTED] wrote: dpkg-hook /usr/lib/man/update-manpages - run only once in total dpkg-hook --on-depends foobar ldconfig - run once before depends of foobar What is a depends? Do you mean dependency or

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Frank Küster
Eduard Bloch [EMAIL PROTECTED] wrote: Ehm... all the time I talk about non-critical things! Dealing with success checking in merged commands would be insane, we cannot assign the failure to a certain package (except when there is just one, of course). [...] First, if dpkg-hook is

Re: Delayed ldconfig execution in postinst step

2006-03-13 Thread Pierre Habouzit
Le Lun 13 Mars 2006 18:38, Eduard Bloch a écrit : #include hallo.h * Goswin von Brederlow [Sun, Mar 12 2006, 03:35:42PM]: I think it would be a good idea to have a general dpkg hook to register a command to be run at the end of dpkg. The syntax would be something like this: dpkg-hook

Delayed ldconfig execution in postinst step

2006-03-12 Thread Eduard Bloch
Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not cost much time because VFS cache has the relevant file parts. However, if memory is limited and there are other applications running, the

Re: Delayed ldconfig execution in postinst step

2006-03-12 Thread Goswin von Brederlow
Eduard Bloch [EMAIL PROTECTED] writes: Hi people, I just wondered why exactly my laptop uses that much time for updates and I think that calling ldconfig is a main problem. In theory, it should not cost much time because VFS cache has the relevant file parts. However, if memory is limited