Comment 10 for bug 1227901

> What is the consequence of committing the heat change without the keystoneclient change?

Any attempt to enable trusts would make the engine crash, as we'd be attempting to use a non-existing argument to the v2 keystonclient authenticate().

>If trusts currently has issues anyway, then we can still release rc1 and recommend users not enable trusts until the keystoneclient change is out.

It looks like we'll have to release rc1 anyway, and recommend users not enable trusts until this bug is fixed (which will include a bump to a keystoneclient release containing the v2 authenticate() change mentioned above, probably)

I'll post an updated patch tomorrow, and we can make a final call then, but IMO we shouldn't block RC1 on this, we'll just have to handle it via a stable-branch backport after the fix is proven and the keystoneclient change is available.