Two small hg problems: test-bad-pull, --dry-run

[Thomas Arendsen Hein]

Hi Vadim!

1. In 2f5dbc24874a you changed the test output, but I still get the
old output, so the test fails for me.

2. You added -n/--dry-run to some commands, but external tools like
tailor don't set this in **opts, and so "not opts['dry_run']"
will fail. Using opts.get('dry_run') helps here.

But I'd rather have --dry-run as a global option, just as I'd like
with -I and -X, so they only have to be documented once. You could
argue that many commands don't use --dry-run or -X/-I, but the same
is true for -q, -v, -R and -y, too.

Unfortunately -n is already taken as a local option for some other
commands, so only --dry-run would survive this.

What do you (and the others) think?

Thomas

-- 
Email: thomas at intevation.de
http://intevation.de/~thomas/