http://bugzilla.wpkg.org/show_bug.cgi?id=262

Steve <steve.kers...@keble.ox.ac.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |steve.kers...@keble.ox.ac.u
                   |                            |k

--- Comment #9 from Steve <steve.kers...@keble.ox.ac.uk>  ---
(In reply to comment #3)
> However, I disagree with "I still believe this makes most sense since packages
> are the most detailed "leaf"-entity in WPKG.". This way, variables defined in
> host or profile level, with the same names as in the package level, do not 
> have
> sence.
> It's naturally, when I define variable in profile or host level - I define it
> for override.

I'd agree with this too - I ran into this problem this morning.  I've written a
number of my packages such that I can perform test installs of new versions of
software by overriding the package-defined version number with a profile or
host-defined version before I update the version number in the package to push
the software out to all computers.  Since this is the main example use case
given on the 'Variables' wiki page and had been working with 1.2, I had assumed
that this was the way it was intended to work.

At least I can stop pulling my hair out now trying to work out why it was
installing the wrong version and look at rewriting my packages to use
conditional variable definitions and see whether that works.

-- 
Configure bugmail: http://bugzilla.wpkg.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
-------------------------------------------------------------------------
wpkg-users mailing list archives >> http://lists.wpkg.org/pipermail/wpkg-users/
_______________________________________________
wpkg-users mailing list
wpkg-users@lists.wpkg.org
http://lists.wpkg.org/mailman/listinfo/wpkg-users

Reply via email to