On Tue, Dec 23, 2008 at 08:50:36PM -0800, [email protected] wrote: > >For example, my prototype for adapting SVR4 package scripting to IPS > >completely ignores CAS and post* script failures. What else could it > >do? Go into maintenance? But this service runs scripting for any SVR4 > >pkg, so to stop running all of them on account of one failure seems too > >blunt a technique. Better to update a pkg self-assembly status DB that > >some UI(s) can interface with. > > You could implement this with a separate instance per-package that is > under this serivce - that way, each package could individually go into > maintenance. But again, this service seems more suited for third-party > packages rather than the ones associated with the legacy WOS concept.
While the paragraph you quoted was about the SVR4 thing, the issue is generic: how to communicate self-assembly failures. Are you saying that we're not going to have any pkgs in opensolaris.org/release that rely on SMF services to do the self-assembly, that every such action will actually be a native IPS action, such that there will be no self-assembly asynchronous to pkg installation/removal? Nico -- _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
