On February 22nd, 2013, 9:24 a.m. UTC, Martin Krizek wrote:

blockerbugs/util/bz_interface.py (Diff revision 1)
		48	url = 'https://bugzilla.redhat.com/xmlrpc.cgi'

The url could be loaded from the config, right?

Yeah and it probably should. When the code was first written, the bugzilla URL wasn't in the app config and while normally, I'd say "let's not load the whole app just to get a config value", we're already doing that implicitly when we load the DB models so it's not a big issue.

On February 22nd, 2013, 9:24 a.m. UTC, Martin Krizek wrote:

blockerbugs/util/bz_interface.py (Diff revision 1)
		185	self.get_bugdata()

Just curious, why is this not loaded in the constructor?

Part of it is a stylistic thing - I don't like constructors that load remote data. I've run into issues in the past with constructors that take too long or load a bunch of data wrt testability. It's not so much of an issue here since the loading can be overridden with a mock bz object but like I said, it's partially a stylistic thing on my part.

Do you think that the code would be easier to read or better if that was loaded in the constructor?

- Tim

On February 22nd, 2013, 2:41 p.m. UTC, Tim Flink wrote:

Initial code for blocker submission page