On Tue, Apr 17, 2018 at 04:32:50PM +0200, Markus Armbruster wrote: > Paolo Bonzini <pbonz...@redhat.com> writes: > > > On 17/04/2018 16:05, Daniel P. Berrangé wrote: > >>> For what it's worth, my "[RFC PATCH 00/32] Command line QAPIfication" > >>> replaces the original parsing art by getopt_long_only(). > >>> > >>> Completing that work will take some time, but once it's done, we can > >>> (and I think we should) prefer double-dash for consistency. > >> Since our existing parser accepts single & double-dash already, is it > >> worth explicitly deprecating single-dash usage right now. So that when > >> your code comes along ready to merge, we're already able to say > >> "i told you so" and drop single-dash support at that same time. > > > > Serious questions: is _anyone_ of you actually using double-dash support > > in daily invocations of QEMU? > > I use both. I readily concede that people familiar with QEMU probably > use single dash a lot more than double dash. I don't think we should > get rid of single dash, unless we decide to bite the bullet and clean up > our command line without regard for backward compatibility, as Thomas > proposed elsewhere in this thread, and others proposed before, multiple > times. > > I don't entertain the idea of compatibility breaks lightly. But our > command line is genuinely nasty. The closer you look, the nastier.
FWIW, I misread your earlier mail in this thread. I read it as 'getopt_long' rather than 'getopt_long_only', so was assuming your getopt conversion implied dropping the single-dash long form and allowing us to have distinct single-letter short options with single-dash like sane apps do ! Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
