Re: Quality control - was Re: lmatter[1,6] = '011191' does not work, but ...

Peter Lancashire wrote: > June Tong wrote: > > > > Bugs are prioritized for fixing based on "Priority" & "Severity" (and some other > > things). Priority means how important it was to the customer. Severity means > > how nasty the consequences are. Core dumps and wrong results are severity 1. > > But a "cosmetic" bug, which would be severity 3, could be very serious to a > > customer if it caused the user to hit the wrong key, for instance, hence the > > addition of customer-assigned importance. I've seen misplaced messages treated > > as "down system"-level bugs. I think you'll appreciate that if an Informix > > employee finds a misplaced error message, it isn't as important as if a > > customer's system fails because a user misinterpreted a message and killed a job. > > Yet, without the "priority" weighting, they would be treated the same. If > > Jonathan reported the bug, it would have no priority at all (but it would still > > have a severity). Incidentally, it has nothing to do with the size of the > > customer. > > Yes, that makes sense to me. I hope fix_urgency != severity * priority. It's usually described as Order By Priority, Severity, but this is overly simplistic, because using this formula, severity 1 bugs detected by employees would never get fixed (or only after all customer-reported bugs were fixed). I think it's something like: All Priority 1 and Severity 1 bugs are fixed in the next release (I think this is a firm commitment) All other bugs are sorted by Priority and then Severity. Hmmm. Actually, I wonder whether it isn't more like: Priority 1 & Severity 1 Priority 2 Severity 2 Priority 3 Severity 3 etc. I don't speak for Informix, so everything I say could be false. This is just my interpretation, based on what I've been told in the past, plus what I've observed. June -- june_t@hotmail.com Grounded in Palo Alto, living on Pepperidge Farm Double Chocolate Milano Cookies