Closed dbenn closed 1 year ago
I noticed a strange behavior:
Now zoom the first part, say, from 2459103.35 to 2459103.39. You will see doubled V observations: both JD and HJD are visible.
This strange duplication existed before and is not directly connected with the current changes, but I only noticed it now. The duplication appears only if a "New series" is created.
I can reproduce this on the current branch but not with 2.22.0.
Step 1. b) is not needed BTW. You can just create a series via View -> Create Series using Johnson V (vs Filtered).
The duplicate observations must be entering through the new-series-in-presence-of-additive-load logic.
Okay, fixed! Geez! This one drove me crazy!
See last commit, but in the end, all the craziness I added/changed in Mediator
was not necessary. Which is good, because it felt wrong, had a bad smell. Seemed like an anti-pattern.
Instead, a 2 lines change in AbstractObservationRetriever.categoriseValidObservation()
was all that was required to properly handle user-defined series. Please check again @mpyat2.
Note that this last commit does not address the user-defined-series-does-not-get-HJD-converted bug you noted above. We can either do that here or in another issue.
Hi @mpyat2, I've fixed the HJD conversion bug too.
@mpyat2 There's a number of things going on in this pull request and the changes are deep in the guts of VStar, so feel free to ask questions. It took me awhile to work through and test this. A review from you is the key thing. Thanks.