rpmbuild: Bad Requireflags: qualifiers: Requires(posttrans)

Panu Matilainen pmatilai at laiskiainen.org
Tue Jan 25 13:50:21 UTC 2011


On 01/25/2011 02:54 PM, Panu Matilainen wrote:
> On 01/24/2011 11:16 PM, Matt McCutchen wrote:
>> On Mon, 2011-01-24 at 14:02 -0500, Bill Nottingham wrote:
>>> I believe folding any requirements for %posttrans scripts into
>>> 'Requires(post)' should be sufficient.
>>
>> I don't think so... IIUC, Requires(post) only applies until installation
>> is complete, but a %posttrans script also runs following uninstallation.
>> Granted, it may not be a problem for other reasons.
>
> %posttrans does NOT run on uninstallation. So in that sense
> Requires(post) is in fact more correct than plain Requires.
> It really does need its own dependency type to be handled correctly
> though: %posttrans dependencies should not be taken into account on
> ordering, and it should be possible to remove such dependencies after
> installation (provided that nothing else depends on them of course).
> %posttrans dependencies only need to guarantee the dependency doesn't
> get removed in the same transaction.

...and that said, Requires(posttrans) is now implemented in 
rpm-4.9.0-0.beta1.4.fc15 and can be used once the new version hits the 
builders (in an hour or so). Unlike some other qualifiers that have been 
used while rpm was looking the other way, posttrans reflects an existing 
mechanism and expecting Requires(posttrans) to work is perfectly 
reasonable (pretrans dependencies were already implemented in 
rpm-4.9.0-beta1)

	- Panu -


More information about the devel mailing list