Ralf S. Engelschall wrote:

>> Maybe the iosm needs a new explicit "use-runtime-owner-group" option,
>> than have rpmio call out to the rpmdb function to check for a SRPM ?
>>
>> Like the --no-same-owner option in tar (or some other bad analogue).
>> But what I meant was that maybe it should be made into more generic ?
>
> Ok, but the problem is that the "use-runtime-owner-group" should be
> really just be done for SRPMs. For binary RPMs this would be silly.  
> So,
> I think I still have to be able to _detect_ in the code whether it  
> is a
> source or binary RPM. The remaining question just is: how can this be
> done (now) if headerIsEntry() should be no longer used?

My thinking here was that if there was such a toggle, then it could  
be set
to "on" for the SRPM and "off" for the RPM - before the I/O was started.
I do see the reason behind no-owner-group-on-srpm-install, just assumed
that the check/setting would have to move around a bit in the code now ?

AFAIK, headerIsEntry is perfectly OK in higher level code (like lib/)
just that there is some new rpmio/rpmdb/rpmlib dependency refactoring

--anders