On Mon, Oct 5, 2009 at 9:52 PM, Bart Smaalders <[email protected]> wrote:
>  Deleting or disabling a publisher
> causes it to be removed from the search order. Re-enabling a publisher
> causes it to added at the end of the search order.

That doesn't seem right. I would expect disable/enable to retain the
position in the search order - I see the disable as a transient operation
and a disable/enable cycle would be a no-op.

> Each selection of the publisher for a package is made independently
> according to the algorithms above; there is no implicit inheritance
> of publisher across dependencies of any type.

I would expect that dependencies for a package from a given
publisher would first be searched for from that publisher.

Consider the following scenario: Package p1 from publisher X
depends on package p2, currently only available from publisher X.
A week later, the preferred publisher also releases package p2.
(Publisher X may be completely unaware of this.) If I then install p1,
it results in p2 being obtained from the preferred publisher, resulting
in a different system. This sort of unpredictable behaviour over
time drives sysadmins up the wall.

I can image whether a publisher being sitcky for dependency
resolution might be a configurable property; alternatively, packages
themselves might declare their dependency resolution policy.

-- 
-Peter Tribble
http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to