On (02/25/09 10:22), Peter Memishian wrote: > I mean that "ipmpstat -P" is an error without also specifying "-o". This > is similar to "-P" with "-o all" being an erorr, but should be a distinct > error because we'd like a distinct error message.
Yes, but I am already catching that as a null str passed to ofmt_open(). since the convention is that not specifying "-o" implies "-o all". What else did you have in mind? > I'm not sure I understand how the finer-grained error codes would work. > Can you provide an example? not today. And as you point out: > In any case, this API is consolidation-private for now so we can always > change this (though it is tedious to update all the callers). so I've gone with the simpler boolean return model here. > > anyway, latest (final?) webrevs are available in the usual > > Given that I've only reviewed two files (ofmt.c and ipmpstat.c) and I've > seen very few comments on the other files, I'm surprised this is being > considered a possible final webrev. I don't think you are suggesting that this is inadequately reviewed, but, fyi, Mike Lim and Seb sent me comments separately. You're also welcome to review the other files, if you like. As I mentioned, both webrev and webrev.meem have been updated. --Sowmini
