On 06/03/2014 09:14 AM, Pádraig Brady wrote: > Or just leave it as is? > I.E. true can return false in the edge cases where it needs to.
I'm liking this idea more - and we can extend it to make 'false --version' return a different status for write failure than its normal always-1 status for successfully reporting false. > > I can fix up the test issues I noticed separately. Yes, at this point, separating the test snafu of only covering the builtin from any change in exit status behavior is wise. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
signature.asc
Description: OpenPGP digital signature