On Sat, 2006-03-04 at 20:21 -0500, seth vidal wrote: > On Fri, 2006-03-03 at 13:06 -0800, David Lutterkort wrote: > > If I run 'yum install p1 p2 ...' I would define success as all the > > packages being installed on the system after yum finishes. > > > > If I run 'yum remove p1 p2 ...' success should be that none of p1, > > p2, ... are installed on the system after yum finishes. > > but it's not just discrete package names, either: > yum install p1* p2* Do you mean with that: what if there are no packages matching 'p1*' - should that be a failure or not ? I could go either way with that. > for a long while we questioned whether or not: > > yum install p1 p2 > > when p2 is already installed shouldn't be an error, if only b/c it is > useful to know that for dealing with that from a shell script. Good point. I think the basic question is: what is the user saying when they execute a yum command ? Are they specifying the actions yum should execute (in which case I would flag the above as an error) or are they specifying what they want their system to look like after yum is done (in which case the above shouldn't be an error) I can't really say which of those two views I would prefer. Does anybody see any clear downsides to either of them ? For the purposes of exit codes, we should stick with one or the other, so that people that rely on them have a clear understanding what yum will and won't signal them through exit codes. > We can definitely implement some sort of error code but we need to be > fairly strict about it lest we waltz into something horrible to figure > out. Agreed. What kind of horrors do you have in mind there ? Is it making it clear to users what the exit codes mean in a comprehensible way or is it more than that ? David
