Kyle Conroy

Thanks! I opened that branch as a draft PR (#702) so that it's easier for anyone to update, and also add a list of things that I think we'll need...

Ok, will do. It's probably not terribly high on our priority list if the version you have locally works for what you need, but we'll try to wrap it up...

Thanks for the detailed report! Can you compare against phoebe's dynamical RVs (`rv_method='dynamical'`) to see if that improves the remaining small discrepancy? Otherwise, I think it still might be worth...

Your changes that you described to the code seem like a reasonable fix to me. If I understand your last comments correctly (and sorry - I haven't had a chance...

We might need to have the output from jktebop stored somewhere in the tests (since installing jktebop on the CI machinery might not be simple). But if you want to...

That's a good point - to test the wrapper itself we will either need to get jktebop installed on the CI system or test the temporary jktebop input files that...

Awesome! Let's move the fti discussion over to #700 just so the conversations aren't all mangled together (or feel free to also message us on the old workshop slack if...

I think that's probably fine myself. `FlatTrace` actually has a `set_position` method that achieves the same thing. In that case, just making the internal `trace` and `trace_pos` attributes as readonly...

After some discussion in #147, it has become clear that this is a more general "problem" than just `FlatTrace`, so I'm copying some thoughts from that thread here so that...

This is a clever workaround, and in general I like it. But before moving forward in this direction, I think we need to open a wider discussion over whether we...