Kyle Conroy

Plan: - implement per-dataset finite integration for LCs for 2.0b (split into new issue: #68) - defer per-timepoint exposure times - defer finite integration for other dataset types

great, then I'll just switch from 2.0 to future since we decided to defer the "simple" reflection case from this release

@bpablo - can you check or remind us of the status of this? I know we changed the behavior slightly for 2.1 - was that enough to consider this closed...

This is coming from an internal call to [phoebe.list_online_passbands](http://phoebe-project.org/docs/development/api/phoebe.atmospheres.passbands.list_online_passbands). The "solution" here is probably to update the warning to be more intuitive when coming from an internal call in `run_compute`...

I'd prefer not to do that as it just complicates logic and we'd probably need to rename the kwargs to be more specific for this individual case ("refresh" and "overwrite"...

Great! It definitely would be nice to have additional built-in options to compare (and to easily choose whichever is most appropriate for a given situation). I looked a bit through...

reflection nosetests now pass (see https://github.com/phoebe-project/phoebe2/issues/39), so moving milestone to oc support

this should be implemented, but I don't know how well its been tested. We need to write a test (preferably nosetest) for this.

As of 2.1, there are only mesh datasets, with times explicitly defined. So the wrapper just needs to fill the mesh (without any dataset columns) at those times by faking...

and let's not forget to do the opposite direction as well so that from_legacy can properly set decoupled luminosities when phoebe1's model parameter is "unconstrained"