On Tue, 2013-03-12 at 23:52 +0100, Nicolas Mailhot wrote: > Le Mar 12 mars 2013 19:04, Peter Jones a écrit : > > > Obviously we need to do a good job of making sure we tolerate failures, > > and there are multiple ways to do this - if you reboot N times within M > > seconds or somesuch might be a worthwhile heuristic. > > By definition an heuristic is unreliable. The current mechanism, while > not-pretty, is reliable. Reliability is the major property you want in any > rescue system. (that's why safety jackets use flashy unfashionable colors) Just for the record, (non/less) deterministic != (un)reliable. Take the following numbers with a grain of salt because I obviously pulled them out of a hat: If we have error detection that "only" catches say 95% of error conditions, why shouldn't we use it to make these cases bearable (for power users anyway), additionally to say 80% of normal boots "awesome" for everyone at the cost of making the remaining 1% a little less easy -- e.g. having to push a keycombo which you have to know at that point. Making this knowledge discoverable is something else and IMO has no place in the boot process (because it just confuses, annoys in 99% of cases). People who want to be power users will find this out, even if we fail to make that information easily discoverable. People who don't have these ambitions won't magically want to turn into power users because we advertise "here's where power users turn right" during every boot. Nils -- Nils Philippsen "Those who would give up Essential Liberty to purchase Red Hat a little Temporary Safety, deserve neither Liberty nils@xxxxxxxxxx nor Safety." -- Benjamin Franklin, 1759 PGP fingerprint: C4A8 9474 5C4C ADE3 2B8F 656D 47D8 9B65 6951 3011 -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel
