[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

other weirdness with rpm and the python bindings



Hi,
 I noticed that anaconda works around this "Feature" but I'd like to
know why.

I'm building a transaction set, I've got 40 pkgs that COULD be available
to the transaction set.

I am explicitly requesting that one pkg be installed, but I don't know
what things it will need, but I know that everything it will need is
among the 40 pkgs.
so I:
ts.add(hdr,data,'i') the pkg I want to install
then I 
ts.add(hdr,data,'a') the other pkgs so they are available.

I then do a depcheck and iterate over the dependency responses.

If I find I need a pkg I do not have installed I do a ts.add(hdr, data,
'i') for that pkg.

now, you would think that I could then do a new depcheck and not have
that dependency problem listed, but I do not, instead the ts is never
updated.

I have to del the ts and restart it and instead of ts.add'ing the pkg as
'a' I have to add it as 'i'.

then the depcheck will no longer have it available.

It seems like the transaction set is not allowing for the overwriting or
modification of the add state of a pkg in the ts.

anaconda works around this by tearing down and recreating the
transaction set each time.

This seems incredibly inefficient to me.

Can anyone explain this to me?

Thanks
-sv

-- 
GPG Public Key: http://www.phy.duke.edu/~skvidal/skvidal.gpg

Attachment: signature.asc
Description: This is a digitally signed message part


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index] []