Rsync shouldn't display a meaningless speedup on a dry run

[Matt McCutchen]

On Wed, 2007-11-07 at 06:54 -0800, Wayne Davison wrote:
> On Tue, Nov 06, 2007 at 10:22:56PM -0500, foner-rsync at media.mit.edu wrote:
> > I worry about those trying to write things that parse rsync's output;
> 
> I had a similar reaction, and will not be eliminating the speedup from
> the output.  If there is nothing to transfer, the speedup number in
> dry-run mode is even correct.  I'm inclined to just leave it alone
> rather than tweak it (e.g. force it to 0).

I am very unhappy with this decision, perhaps to the point of using my
own modified version of rsync that doesn't print a speedup in dry-run
mode.  I feel strongly that rsync should not print a bogus value,
especially when it constitutes false advertising.  As for output
parsers, I was under the impression that correctness trumps backward
compatibility when they are in conflict.  And yes, the value is correct
when there is nothing to transfer, but it could be omitted in other
cases.  I think forcing the speedup to zero (at least when there is
something to transfer) would be a reasonable compromise.

That's my two cents; do as you will.

Matt