<p>The change was to fix a test isolation bug: tests that ran a command with -v left that flag set for the rest of the process life. Sorry I broke this. Yes, let's put it in to 2.3.1. </p>
<p>- Martin</p>
<div class="gmail_quote">On 05/02/2011 5:37 AM, "John Arbash Meinel" <<a href="mailto:john@arbash-meinel.com">john@arbash-meinel.com</a>> wrote:<br type="attribution">> -----BEGIN PGP SIGNED MESSAGE-----<br>
> Hash: SHA1<br>> <br>> <a href="https://bugs.launchpad.net/bzr/+bug/713258">https://bugs.launchpad.net/bzr/+bug/713258</a><br>> <br>> I just found out that bzr 2.3 no longer records how many bytes were<br>
> transferred for a given command. I'm a bit sad to say that this changed<br>> in bzr 2.3b2 and I didn't catch it until now.<br>> <br>> This happened in Martin's r5473: "reset global verbosity after each<br>
> command".<br>> <br>> I don't really remember what the point of that change was, so I don't<br>> have a great "we should do X to get this working again".<br>> <br>> One option, is a patch like this:<br>
> <br>> === modified file 'bzrlib/commands.py'<br>> - --- bzrlib/commands.py 2010-10-29 17:07:42 +0000<br>> +++ bzrlib/commands.py 2011-02-04 18:35:48 +0000<br>> @@ -691,7 +691,10 @@<br>> return self.run(**all_cmd_args)<br>
> finally:<br>> # reset it, so that other commands run in the same process<br>> won't<br>> - - # inherit state<br>> + # inherit state. Before we reset it, log any activity, so<br>
> that it<br>> + # gets properly tracked.<br>> + ui.ui_factory.log_transport_activity(<br>> + display=('bytes' in debug.debug_flags))<br>> trace.set_verbosity_level(0)<br>
> <br>> def _setup_run(self):<br>> @@ -1202,8 +1205,6 @@<br>> argv = _specified_or_unicode_argv(argv)<br>> _register_builtin_commands()<br>> ret = run_bzr_catch_errors(argv)<br>> - - bzrlib.ui.ui_factory.log_transport_activity(<br>
> - - display=('bytes' in debug.debug_flags))<br>> trace.mutter("return code %d", ret)<br>> return ret<br>> <br>> <br>> Thoughts? I'd really like to see us put something like this into a 2.3.1<br>
> release, though I don't think it needs to move up the actual release<br>> schedule.<br>> <br>> John<br>> =:-><br>> -----BEGIN PGP SIGNATURE-----<br>> Version: GnuPG v1.4.9 (Cygwin)<br>> Comment: Using GnuPG with Mozilla - <a href="http://enigmail.mozdev.org/">http://enigmail.mozdev.org/</a><br>
> <br>> iEYEARECAAYFAk1MR1IACgkQJdeBCYSNAAOU7QCdH9tdu3c82QJm4RfSBgVyGj87<br>> 7AgAnA2ATwC14zBksAYetb36UvvkytQ+<br>> =5gzq<br>> -----END PGP SIGNATURE-----<br>> <br></div>