[MERGE][Bug #115990] status should handle the -q (quiet) option like svn

[Aaron Bentley]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Daniel Watkins wrote:
> Attached is a bundle that modifies 'bzr status' such that the quiet
> option is shorthand for 'bzr status --versioned --short', as suggested
> in the bug report.

!abstain

I'm not entirely comfortable with this.  First, there's the meaning of
- -q.  Right now, it suppresses informational messages.  I'm not sure it
makes sense to extend it to changing the behavior in other ways.

Second, there's precedence.  Standard command arguments have their
arguments processed by optparse, and this means that the last-given
option takes effect.

So, "bzr status --short --long" selects the long log formatter.  But
- --quiet is processed before other command processing is done, so we lose
positional information-- we just have a boolean.  "bzr status -q --long"
and "bzr status --long -q" ought to have different effects, but they can't.

Third, there's the documentation issue.  Since "--quiet" is global, it's
not listed in the command help (it's in "global-options").  That means
there's no place to document this status-specific behavior, and it's not
very discoverable.

In terms of code, Alexander is right about make_branch_and_tree('.'),
but also, you use use wt.add('hello.txt') and wt.commit('added').  You
should only use run_bzr for the command you're testing.  (There is some
older code that doesn't follow this rule, but new code should.)

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGrzdJ0F+nu1YWqI0RAoEXAJoDAuq3oq7S/2qrYNU1Y+kdEjk39wCcCE9+
lN+p6Wo8nGeV7JL5W3/7C3w=
=gFbo
-----END PGP SIGNATURE-----