On Fri, Jan 25, 2013 at 10:35:26AM +0100, Tomas Forsman wrote:
> >i guess the easiest workaround is to remove either package from system-model,
> >but can someone explain what is actually going wrong here?
> So both issues is that those packages aren't in our group anymore.

thanks, so it is pure coincidence that the two machines fail with the
same type of error.

i am wondering though, what would be the best way to avoid these kinds
of errors? redirects of course for packages where it makes sense, but
what about the lives case? asuming that the previous version was working,
the user is now forced to decide to remove the packacge manually or not
upgrade. 

could conary detect this case, and would it be helpful if it was more
verbose about it?
or is that the job of packagekit?

could the groups be adjusted to keep the old binary around?
(might work in some cases, but could get messy in others i guess)

should the package redirect to an empty package?
that would fix updateall, but it might surprise the user.

i am interested to figure out what is the most user-friendly way to deal
with such cases...

greetings, martin.
-- 
eKita           -           the online platform for your entire academic life
--
chief engineer                                                       eKita.co
pike programmer      caudium.net                            societyserver.org
foresight developer  community.gotpike.org                 foresightlinux.org
unix sysadmin        pike.lysator.liu.se                           realss.com
Martin Bähr          working in china        http://societyserver.org/mbaehr/

_______________________________________________
Foresight-devel mailing list
[email protected]
https://lists.foresightlinux.org/mailman/listinfo/foresight-devel

Reply via email to