[18:26] <dosaboy> tvansteenburgh1: can we please allow the charm-helpers Makefile to use an existing venv rather than build a new one everytime i clone the branch and run tests
[18:27] <dosaboy> it is quite prohibitve to enforce a new build everytime
[18:28] <dosaboy> e.g. a have a prebuilt venv that i activate whenever i want to run tests for any charm/charm-helpers
[18:28] <dosaboy> which lives outside of tree
[19:10] <rick_h_> dosaboy: I'd suggest a patch that allows setting a venv place on disk and then your scripts can set that path and the makefile should detect already exists and roll with it?
[19:11] <rick_h_> dosaboy: I've not looked at the charm-helpers makefile but normally we point to a bin/python path and if that doesn't exist run the venv command, so setting your own bin/python would be a good way to work with that
[19:13] <dosaboy> rick_h_: if you are running within an activated venv you should not have to specify paths at all
[19:13] <dosaboy> python works it out and uses the packages you have installed in your venv
[19:14] <rick_h_> dosaboy: ok, but the point I think is the same, to add the detection for an activated env
[19:14] <dosaboy> rick_h_: yeah that would be good
[19:14] <rick_h_> dosaboy: but it's pretty standard practice against all of our stuff to build a fresh makefile so that depsn, download caches, etc are setup to catch a bug/mistake early and often vs trusting an existing setup
[19:15] <dosaboy> i'll have to scratch my head and think of bect way to do that
[19:15] <rick_h_> we've all been bit too many times by reusing 'it works here' where it's never been cleaned and rebuilt from commit to commit.
[19:16] <dosaboy> rick_h_: both have their merits imo, but I agree that the final gate should definitely be using a fresh venv
[19:17] <dosaboy> i'm basically applying the same rules as with upstream openstack
[19:17] <dosaboy> if only tox support had landed in charm-helpers...
[19:17] <rick_h_> dosaboy: understand, and the way it is in kind of how we've done things around here for a long time.