Daniel Rall wrote:
>
> On Sep 13, 2007, at 10:29 AM, Daniel Rall wrote:
>
>> On Sep 12, 2007, at 10:48 PM, lgo@tigris.org wrote:
>> ...
>>> # make a property change and commit it, allowing lock to be released
>>> svntest.main.run_svn(None, 'propset', 'blue', 'azul', file_path)
>>> - svntest.main.run_svn(None, 'commit', '-m', '', file_path)
>>> + svntest.main.run_svn(None, 'commit',
>>> + '--username', svntest.main.wc_author,
>>> + '--password', svntest.main.wc_passwd,
>>> + '-m', '', file_path)
>> ...
>>
>> Is this necessary anywhere that we use run_svn()? I just added more
>> usage of it to actions.py...
>
> In r26582, I updated my recent changes with those arguments. However,
> should we instead modify run_svn() to add them when they aren't
> specified (or something of the kind)?
>
Dan,
there have been some discussions about this topic lately:
http://svn.haxx.se/dev/archive-2007-05/0388.shtml
http://svn.haxx.se/dev/archive-2007-06/0518.shtml
I think we should:
- disable creds caching for all tests (through the config file)
- make all functions in actions.py and main.py take a username and pwd
parameter
- let them default to 'jrandom' and '**********' (you know the passwd,
I'm not sharing it on a public mailing list ;), so we don't have to
specify those on all 1000+ usages of run_svn*.
- modify all tests using jconstant to explicitly set user & pwd.
I'm willing to work on this, but not before 1.5 is released. I only have
evenings available to work on svn and there is lots of (ra_)serf work
left to do.
Lieven
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Sep 13 21:02:02 2007