On Monday 25 May 2009 11:31:43 Marco van Wieringen wrote:
> On Mon, 2009-05-25 at 10:38 +0200, Kern Sibbald wrote:
> > On Monday 25 May 2009 00:27:23 Arno Lehmann wrote:
> > > Hi packagers (Felix in particular :-),
> > >
> > > I wonder if it's intentional that ACL support is not included in (at
> > > least some of) the packages.
> > >
> > > I have installed bacula-postgresql-3.0.1-1.el5.x86_64.rpm from the
> > > sourceforge download page, and found this to not include ACL support.
> > > (This is a CentOS 5 system.) Now I'm wondering if this is intentional,
> > > and I have to build my own packages, or if this is an oversight, and I
> > > will find fresh packages soon?
> >
> > I suspect that this was not intentional, and was a rather easy oversight
> > to make -- probably needing some programming to help packagers.
>
> On Linux we have a dependancy on additional acl libs so those need to be
> on your system when building the bacula sources or it will not be
> included. So adding this dependancy also means adding it as a dependancy
> to to the rpm build.
>
> > If I am not mistaken, the ACL support (at least the most recent extended
> > stuff) is automatically configured during ./configure based on whether or
> > not the appropriate ACL packages have been installed.  If they are
> > installed on the original build (packager's) machine, then the code will
> > be generated.  If they are not installed, then no code will be configured
> > in Bacula.
>
> Correct
>
> > Perhaps what we should do is default to ACL support (this may not be so
> > easy), and if the packages are not available the ./configure should fail,
> > and the builder would either be required to install the needed libraries
> > or explicitly disable ACL support on the ./configure line.
> >
> > Marco: what do you think about this idea?  (I'm not asking for any
> > programming right at this point, just trying to decide the best action to
> > take).
>
> Uh I would say we could it a bit different, we check if --enable-acls is
> set (so thats different then now) and then complain if things are not
> installed. Then we can easily change the configure used in the rpm building
> (SPEC file) to include --enable-acl. We could do the same for xattr
> support. So we only act when someone explicitly enabled acl and xattr
> support.

Yes, I think this would be a very reasonable solution.

>
> As far as I know its now either enable it when found or disable it when
> --disable-acl is given.

That is a very reasonable behavior, but in addition, if we *require* it if the 
user explicitly puts --enable-acls the ./configure should fail if the 
libraries are not available -- I think that is what you suggested above.

Is that something you would like to do, or should I do it?

Kern

>
> Marco
>
>
>
> ---------------------------------------------------------------------------
>--- Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT is
> a gathering of tech-side developers & brand creativity professionals. Meet
> the minds behind Google Creative Lab, Visual Complexity, Processing, &
> iPhoneDevCamp asthey present alongside digital heavyweights like Barbarian
> Group, R/GA, & Big Spaceship. http://www.creativitycat.com
> _______________________________________________
> Bacula-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bacula-devel



------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, & 
iPhoneDevCamp asthey present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://www.creativitycat.com 
_______________________________________________
Bacula-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-devel

Reply via email to