On Fri, Oct 22, 2010 at 09:44:22AM -0400, Jameson Rollins wrote: > Hi, Clint. Using "set -e" was a deliberate choice. The reasoning being > that we don't want to just pass over any errors that could potentially > be bad news. I personally think this is reasonable, and is somewhat in > line with how other programing languages work (e.g. if memory can't be > allocated, or a function call is incorrect, the program terminates). > This forces us to be vigil about catching the return of calls that we > expect to be non-zero. This seems like a more reasonable task than > catching the return of *all* calls, and checking that they are in fact > zero.
I think that that is a false analogy. I would liken it more to setting your bicycle to fall apart if ever it is ridden at a speed of exactly 5 miles per hour. > So is this just a theoretical concern, or did you actually run into a > bug caused by the use of "set -e"? Yes, monkeysphere update-authorized_keys failed because "set -e" considers ((nline=0)) to be an error. Thus I am running a monkeysphere with the "set -e" removed. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
