On Tue, Jun 16, 2015 at 04:54:41PM -0400, Kenneth Gober wrote: > On Mon, Jun 15, 2015 at 7:20 PM, Michael McConville > <[email protected]> wrote: > > Of course, this breaks existing scripts that used the -a option - > > they'll just get the usage prompt. It will fail more dangerously > > (silently) with scripts intended for old-style tapes, so (if I > > understand correctly) it should probably go in "Following -current" and > > the release notes. > > I don't see any benefit to breaking scripts that use the -a option.
True. I wasn't sure what the project's opinion of placebo legacy flags was, so I thought I'd KIS unless someone suggested otherwise. > Optimally, behavior of dump(8) should only change if a user has > specified *none* of the following flags: > > -a > -B > -c > -d > -s > > A user who has specified any of the above flags presumably > knows what they want, and shouldn't have things changed out > from under them. Good point, I forgot to address that. We could drop the -A option entirely in that case, or rephrase it as "use default static volume sizes". The only downside I see is that it could cause subtle breakage (e.g. a script only occasionally uses default parameters and defaults to auto-sizing), whereas the simple flag addition would make everyone just acknowledge the change immediately. By the way, does anyone know how many people still use tapes without EOM indication? I've been told that they're very rare, but I have no idea personally.
