Danek Duvall wrote: > On Fri, Sep 05, 2008 at 03:50:32PM +0100, Michal Pryc wrote: > >> > > packagemanager.py: > > - line 1686: isn't there some sort of synchronization you can do here, > rather than sleeping for an arbitrary amount of time that may or may > not be enough, depending on the characteristics of the machine you're > running on? > > Danek
We were updating package status after we receive the act_output_done from the progress, that is why this sleep was needed. I was thinking to do this step after index_done, but sometimes index is made before removing/installing so this approach is not good. Currently I've moved for remove.py the part for updating list of packages at the end of __remove_stage(). For installupdate.py this is not necessary, as we have a little bit different approach (in the duplicated code actions_done is called at the end of all actions stuff and after postexecute stage), so there is no more need for time.sleep(0.5) and I've removed that part: http://cr.opensolaris.org/~migi/04_09_2008_sync_with_ips_bug_3193_v5 I think that the best approach is either the one which is the current one, or the progress should have something like operations_done() which simply would be invoked after everything is done. best Michal _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
