[14:05:11] <J1m> benji, implied that he now knew what it did (and was wrong before :)
[14:06:44] <benji> heh, yeah. Before I thought it was recording HTTP traffic with the intent on replaing the client side (as a test), when it is in fact recording the server side so it can be a stand-in for the server for future test runs (so they will be fast)
[14:08:47] <aclark> J1m: did you decide anything on the buildout.* namespace? I hope that works out…
[14:09:01] <benji> it looks like a nice optimization, but not earth-shattering; it might even be an attractive nuisance if the server changes very often (e.g., "Darn, my tests are failing. I guess I should debug my client. [three days pass] Oh no! My client was fine, it was just that VCR captured a bit of the response that it really shouldn't have and I was getting wierd results.")
[14:11:32] <J1m> benji, so it sounds like it *was* doing what you thought. I'm confused, but not quite interested enough to dig. :)