Re: [Help] How to fix a buggy prerm script in sarge? (was: Bug#338638: tetex-base: dies when /usr/local/ read-only)

[Daniel Burrows]

On Sun, Nov 13, 2005 at 12:25:01PM +0100, Frank Küster <[EMAIL PROTECTED]> was 
heard to say:
> we just received a bug report that is caused by a buggy prerm script
> in the package in sarge (it fails because it doesn't handle read-only
> /usr/local properly).  Is there any way to fix this, except documenting
> it in the release notes?

  If your old prerm fails, the new prerm should be called with
"failed-upgrade" as its first argument (see Policy 6.4 and 6.5),
so you can do any necessary workarounds there.  Errors in postrm
are handled similarly.

  Daniel


signature.asc
Description: Digital signature

Re: [Help] How to fix a buggy prerm script in sarge? (was: Bug#338638: tetex-base: dies when /usr/local/ read-only)

[sean finney]

hi frank,

On Sun, Nov 13, 2005 at 12:25:01PM +0100, Frank Küster wrote:
> we just received a bug report that is caused by a buggy prerm script
> in the package in sarge (it fails because it doesn't handle read-only
> /usr/local properly).  Is there any way to fix this, except documenting
> it in the release notes?  

assuming that this is only a problem with removals (and not upgrades too,
in which case the problem is quite a bit more complicated), i think the
best way to fix such a package would be to work with the release managers
to make sure that a fixed version makes it into a sarge point-release
via proposed-updates.

> And on a related thought, if the bug was in the postrm script, would it
> be acceptable that the preinst script of the new version edits
> /var/lib/dpkg/info/tetex-base.postrm? 

i think you will probably get a resounding "no" from the release teams
in question, from my experience with a similar problem wrt the
woody->sarge path for mysql.


sean

-- 


signature.asc
Description: Digital signature