Mailinglist Archive: zypp-devel (56 mails)
| < Previous | Next > |
Re: [zypp-devel] Question / remark about zypper solving dependencies
- From: Michael Matz <matz@xxxxxxx>
- Date: Wed, 12 Dec 2007 14:24:11 +0100 (CET)
- Message-id: <Pine.LNX.4.64.0712121420570.23011@xxxxxxxxxxxxx>
Hi,
On Wed, 12 Dec 2007, Dominique Leuenberger wrote:
Yes, I would expect that the newer version be installed. I don't know if
the current libzypp solver handles this case ideally (I mean, it obviously
doesn't, but I'm not sure if it was supposed to be implemented or not,
i.e. bug or missing feature), but I believe the new SAT solver will do
this as expected.
No, I don't think this should be worked around by the packager. It would
inevitably lead to problems down the road somewhen.
Ciao,
Michael.
--
To unsubscribe, e-mail: zypp-devel+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: zypp-devel+help@xxxxxxxxxxxx
On Wed, 12 Dec 2007, Dominique Leuenberger wrote:
zypper shell
zypper> in pkgB
in this case, zypper resolves the dependencies and offers to install
pkgA-0.1-1 Why doesn't it suggest now also to install pkgA-0.2-1
together with libpkgA0-0.2-1? Wouldn't that case be the logical one?
Yes, I would expect that the newer version be installed. I don't know if
the current libzypp solver handles this case ideally (I mean, it obviously
doesn't, but I'm not sure if it was supposed to be implemented or not,
i.e. bug or missing feature), but I believe the new SAT solver will do
this as expected.
Interesting case for any of the solver hackers around? Or do you think
such a case should be 'caught' by the RPM builder, with a Obsoletes:
pkgA < 0.2 in the libpkgA0 package? (would be possible I guess, but
having a newer pkgA makes it look a bit weird)
No, I don't think this should be worked around by the packager. It would
inevitably lead to problems down the road somewhen.
Ciao,
Michael.
--
To unsubscribe, e-mail: zypp-devel+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: zypp-devel+help@xxxxxxxxxxxx
| < Previous | Next > |