Asterism Error

[swalker2m]

The Science Program model won't stop anybody from changing the instrument after the fact, and thereby potentially invalidating the Asterism / Instrument choice combination. Since we can't prevent it, we need to at least generate an error. This is the same issue we face when selected guiders no longer apply to the instrument in use. So, a P2 check error is needed at the very least.

[sraaphorst]

How does one change the instrument? I didn't even know this was possible.

When this introduces invalid combinations (e.g. as you mention, guiders, but what about certain details of the sequence), how is this currently handled?

[swalker2m]

How does one change the instrument? I didn't even know this was possible.

You can delete an instrument component and then add another, different one.

When this introduces invalid combinations (e.g. as you mention, guiders, but what about certain details of the sequence), how is this currently handled?

The OT stops (or at least tries to stop) you from mixing up instrument iterators and static components. If you end up with a guide star for a guide probe that is no longer available, we add a P2 check error and we show the guide star in the table in the target component with a red error icon.

[sraaphorst]

I see what you mean. I never even thought about the possibility of that in the OT. Again, that seems like a strange feature: is it actually used?

I have a gut feeling that this is not going to be so easy to handle.

[swalker2m]

Again, that seems like a strange feature: is it actually used?

I think it's more of a case of the model (and OT) allowing it to happen so we have to handle it. I don't think it comes up much in practice though.

[sraaphorst]

Honestly, I didn't even know you could cut the target, condition, and instrument nodes from an observation until this came up. It wouldn't have even occurred to me do to so.

[sraaphorst]

I hope this task won't take more than 1-2 days, tops: otherwise I suspect we should do something as simple as possible, like a P2 error or just chuck away a bunch of the target environment to make it compatible with other instruments.

[sraaphorst]

Right now, all we do is a P2 error check. I'll leave this task here since we have discussed something more robust.