Hi,
Just changed the installer to supply a logger object for virtualenv when
it is missing.
Cheers,
Gary
On 03/13/2012 10:55 AM, [email protected] wrote:
Author: gjm
Date: Tue Mar 13 10:55:14 2012
New Revision: 1300071
URL: http://svn.apache.org/viewvc?rev=1300071&view=rev
Log:
maintain compatibility with older versions of virtualenv, fixing #4
Modified:
incubator/bloodhound/trunk/installer/installer.py
On 03/12/2012 10:25 AM, Gary wrote:
Hi Antony,
Thanks for that.
OK, I would prefer not to have to specify a minimum version of
virtualenv that is newer than that available for the default version
of a major linux distribution. So we can either resort to calling
virtualenv through subprocess or we can do to the following:
if not hasattr(virtualenv, 'logger'):
virtualenv.logger =
virtualenv.Logger([(virtualenv.Logger.LEVELS[-1],
sys.stdout)])
Cheers,
Gary
On 03/10/2012 12:36 PM, Antony Semonella wrote:
Hello,
I managed to set up a bloodhound development environment on Debian 6.0
following the instructions at:
https://issues.apache.org/bloodhound/wiki/BloodhoundContributing
You may want to update the wiki to highlight the following dependency I
encountered:
The version of virtualenv in the repo for Debian squeeze is: 1.4.9-3; this
version of virtualenv results in:
NameError: global name 'logger' is not defined
When running bloodhound/installer/installer.py.
logger is defined only in main() for this version of virtualenv.
This issue has been resolved by:
https://github.com/pypa/virtualenv/commit/ea1786
Using the latest version of virtualenv (1.7.1.2) resulted in no NameError
and so the bloodhound dev. environment was setup ok for me.
I therefore suggest a version of virtualenv>=1.7.1.2 be listed as a
dependency.
Cheers!
Antony
