[pkg-bacula-devel] Current status of merging development2 branch

Wed May 30 21:19:23 UTC 2012

On Wed, 30 May 2012 13:57:34 +0200, Luca Capello wrote:
> Hi Alexander!
>
> On Sun, 27 May 2012 02:40:22 +0200, Alexander Golovko wrote:
>> There is a list of changes in binary packages, builded from revision
>> 0f4409a468a6749837e6b852024b6a52f93c74cb (revision before merge) and
>> a0cfd0fd8d5d7cc420828f21d7ab4a30211f3b44 (current last commit).
>>
>> I try do not miss anything significant, but will see later...
>
> Well, that is fantastic, thanks.
>
>> 
>> ###############################################################################
>> Changes in installed files
>> 
>> ###############################################################################
> [...]
>> /var/lib/bacula/log:
>>  - removed (packages now builded with correct log dir)
>
> While at it, we should also considering migrating to /run, thus
> depending upon initscripts (>= 2.88dsf-13.3):
>
>   <https://wiki.debian.org/ReleaseGoals/RunDirectory>
>

yes, we should do this.

>> 
>> /usr/share/doc/bacula-common/examples/nagios/check_bacula/Makefile.gz:
>>  - added (but subject for remove)
>
> Would it be better to provide a separate nagios-bacula-plugin,
> coordinating that with the Debian Nagios Team?
>
>   <http://pkg-nagios.alioth.debian.org/>
>

Yes, but there are 3-rd party nagios plugins additional, that will be 
well to include in this package too.
It is subject to discuss too.

>> /etc/bacula/bconsole.conf:
>>  - added (subject for remove?)
>
> Why?  That file is mandatory, so we should provide it anyway.

Because this file generates by postinst script.

>
>> /usr/share/man/man8/bregex.8.gz,
>> /usr/share/man/man8/bwild.8.gz:
>>  - added manpages
>
> I would replace them with upstream's versions as quilt patches:
>
>   
> <http://www.bacula.org/git/cgit.cgi/bacula/tree/bacula/manpages/bregex.8>
>   
> <http://www.bacula.org/git/cgit.cgi/bacula/tree/bacula/manpages/bwild.8>
>

Yes, this my miss, i didn't check that upstream already add this 
manpages.

>> /usr/share/doc/bacula-common/README.Debian.gz:
>>  - reflect changes about removing bconsole wrapper
>
> See below for the bconsole path.
>
>> /usr/share/lintian/overrides/bacula-console,
>> /usr/share/lintian/overrides/bacula-console-qt:
>>  - added wrong strings (subject for remove)
>
> Do you mean that we can remove both the two overrides below?  I have 
> not
> recompiled it yet, so this is a pure speculation ;-)
>
> - debian/bacula-console-qt.lintian-overrides
>   pkg-not-in-package-test bat usr/share/menu/bacula-console-qt
>
>   This should have been fixed by:
>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=a0cfd0fd8d5d7cc420828f21d7ab4a30211f3b44>
>
> - debian/bacula-console.lintian-overrides
>   menu-command-not-in-package usr/share/menu/bacula-console:2
> usr/bin/bconsole
>
>   This should have been fixed by:
>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=1fab6a25a5af0806b09692e0a696f4a9d2d0f146>
>
>   ...but please see my comments below.

Yes

>
>> /usr/bin/bconsole, /usr/sbin/bacula-console:
>>  - bconsole was shell wrapper, but now vanilla bconsole search it
>> config in the same place,
>>    so this wrapper removed. for compatibility provided simlink, so 
>> no
>> changes for users
>
> I think that bconsole should stay in /usr/sbin given that upstream
> install it there:
>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=d92858c57ec47449e5ce49b81e27af026a77a098>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=1fab6a25a5af0806b09692e0a696f4a9d2d0f146>
>

But maybe we want to provide symlink in old place until post-wheezy?
Or maybe we change bconsole path when update to 5.2.x?

>> 
>> ###############################################################################
>> Changes in packages metadata:
>> 
>> ###############################################################################
> [...]
>> bacula-console/conffiles:
>>  - added /etc/bacula/bconsole.conf (i think, this is merge error)
>
> See above for the reason why it should stay there.

My opinion that we should generate configs on package install and 
manage them by ucf.
Will be better if we don't change configuration login until switching 
to ucf.

>
>> 
>> ###############################################################################
>> Changes in sources:
>> 
>> ###############################################################################
> [...]
>> A lot of. Most significant:
>>
>> debian/rules is now use debhelper
>
> Finally, thank you very much to Hauke and you.
>
>> 
>> ###############################################################################
>> Some info about merges:
>> 
>> ###############################################################################
>>
>> This is not very accurate info, for details see merge commits.
>>
>>
>> * 570e78a Deploy alternative backup scripts
>>     NOT YET PROCESSED
>> * 007c95d Sort debian/additions a bit
>>     NOT YET PROCESSED
>
> Should we process these two before any upload to unstable?
>

We can merge this commits later,

>> * 0e942e7 Adjust bat.conf and bconsole.conf as well
>>     REJECTED
>
> Is this because of f968df9 below?

yes. I reject all commits, related to switching to default configs, 
installed by dpkg.
Will be better to generate configs and manage them by ucf.

>
>> * d77e917 Fix errors in man pages
>>     APPLIED
>
> This should be pushed upstream, I will do later on.

Yes, i didn't do this, because manpages still contain errors. This 
commit only fix lintian errors.

>
>> * 112e36a Install bsmtp to /usr/sbin as intended by upstream
>>     APPLIED + fixes
>
> I still dislike bsmtp (see #519567 and #620825, both fixed on my 
> local
> repository) and I would have liked to switch to sendmail on Debian, 
> but
> I agree that if we want to provide an easy switch from users coming 
> from
> other OSes this is the Right Thing™ to do.
>
>> * 8b01282 Get rid of useless common-functions
>>     REJECTED
>
> I could not find any rationale for that, would you elaborate, please?

Because common-functions in not useless, i reject switching to configs 
installed by dpkg

>
>> * 6fdee4f Override missing templates files check
>> * 4b7fa18 Don't set passwords at build time but ask for them
>>     REJECTED
>
> Both because of the rejection of f968df9, see below.
>
>> * a4382f0 Properly set logdir
>>     APPLIED
>
> Only in development2 branch :-(

hmm. current master branch contain this changes.

>
>> * f968df9 Install conffiles, don't mess with them
>>     REJECTED
>
> Your Git commit shows:
>
>   Bug 585037 can be closed by properly escaping variables in sed 
> expressions.
>
>   Debian is not FreeBSD, package after installing must work :) so 
> best way
>   will be migrate to ucf usage for installing config files
>
> ...while Hauke's Git commit shows:
>
>   Instead of pushing sed statements over files in /etc that an admin 
> might
>   have changed, just install them and let dpkg handle everything 
> else,
>   including removal.
>
>   Currently this means that processes shouldn't be started after 
> install
>   since they don't have proper config. We might consider changing 
> that by
>   sourcing properly created random passwords but let's revisit that 
> later.
>
>   Closes: 585037
>
> At a first view Hauke's approach is more Debian-like: /etc files are
> conffiles (see `dpkg-query -s dpkg`), so they should be treated as 
> such.
>

You are right partially.
Debian-like is not "conffiles", but safe process of package updating.
Dpkg conffiles is only one of methods for solve config update problem.
In our case ucf is much more better for this purpose.

>> * 21014db Add missing build-dependency on procps
>>     REJECTED
>
> This is not needed because procps is pulled by other packages during 
> the
> build process, so we should track this down.  Hauke, do you remember 
> why
> procps was needed?
>
>> * 621a064 Update Vcs-* in debian/control
>>     APPLIED
>
> Updated to anonscm.debian.org:
>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=86d60666353c720b4f9aad6978d391cc5ecd907d>
>
> Testing packages available at (I have not tested them yet, sorry, I 
> am
> quite in a hurry in Real Life ATM):
>
>
> 
> <http://pkg-bacula.alioth.debian.org/tmp/merging-development2-branch/dev2.0109.1/>
>
> Now, except the questions I asked above, if I am right there is only 
> one
> commit left from the development branch, which however I have not
> checked yet if it is needed or not:
>
>
> 
> <http://anonscm.debian.org/gitweb/?p=pkg-bacula/bacula.git;a=commitdiff;h=bdd0d231ae1055f297d2ababb0e317f0227ddf80>
>
> Thx, bye,
> Gismo / Luca

-- 
with best regards,
Alexander Golovko
email: alexandro at ankalagon.ru
xmpp: alexandro at ankalagon.ru