[Foresight-devel] Re: Pre-update script maintenance (was: pinned kerneloops)

[Jack Doerner]

That's why I suggested using group/package versions as a trigger
instead of just a number in the script - I agree exactly your last
statements. As for making every problem look like a nail, you may be
right there... if this does happen, it certainly should only be used
in cases where no other option is available.

Another thing that hasn't been brought up - whatever scripts, if any
are run, would be running with root priveledges, I assume. It would
have to be made sure that whatever system is implemented in absolutely
secure, and that whatever scripts are run are perfect - giving this
power to an automated system could be an issue

On Wed, Jun 18, 2008 at 1:46 PM, Michael K. Johnson <johnsonm at rpath.com> wrote:
> I frankly think that this is not a good idea.  It would become a hammer
> that would make every problem look like a nail.
>
> I know that the pre script bail-and-background is ugly.  It's really
> ugly.  And that makes it unlikely that we'll resort to it unless it
> is really necessary.
>
> We now have a pattern for cases where we absolutely, positively have
> to deal with the problem.  Starting to run arbitrary net-fetched
> scripts feels to me like a descent back into the nether regions of
> RPM trigger scripts...  It's not testable, it's not associated with
> particular group versions; I have trouble finding anything good to
> say about it...
> _______________________________________________
> Foresight-devel mailing list
> Foresight-devel at lists.rpath.org
> http://lists.rpath.org/mailman/listinfo/foresight-devel
>



-- 
 Jack Doerner

 There are worse crimes than burning books
 One of them is not reading them.
 -Joseph Brodsky