Not strictly VSX-related, but certainly relevant: Luis Balona in South Africa has posted a potentially very interesting paper on astro-ph this evening:
...which reports that he has classified >120,000 Kepler, K2, and TESS variable stars _by_eye_, not with some machine-learning algorithm or other automated method. The text is remarkably short, and describes succinctly what he proposes to be new/updated/revised GCVS categories based on physical data beyond simply the lightcurve morphologies.
Well, it is very relevant to VSX, although ironically it never mentions it or used any of its data, variability classifications, statistics, etc. It seems that we haven't been doing a good job advertising VSX...
Leaving that aside, it is a very interesting paper and the resulting table will be worth a VSX update (although the number of lists to add to VSX is enormous at this point).
One of the problems I see, related to the fact that VSX hasn't been checked, is that some variability types in VSX that are not in the GCVS, are named differently or defined differently. This will add confusion and create unnecessary duplication of classes in the literature (provided these types are going to be adopted elsewhere).
Also, several known classes are defined differently, and changing those definitions is not something that we will do on the fly. Consensus in the literature will have to be found. Different authours agreeing in new definitions.
The same for new variability classes, which are several in this paper.
The most surprising for me are the MAIA variables. I had been reading about them for years but never found either definite proof of their existence, or many authors supporting this classification.
This will all have to be checked against the literature over the upcoming years.
TESS and Kepler have surely opened a new world to variable star astronomy. In this regard, it is nice to see these classification attempts.
Thanks for calling our attention to it.