I think this is just a consequence of #96. The unit test for the merger tree code checks that if a subhalo continues to be resolved and contains the same particles then it will have DescendantTrackId == TrackId. But we've decided we don't want that behaviour any more. I'll try to think of a more appropriate test.
(this is another situation where #101 would help: I forgot to run the tests before approving the PR)
I think this is just a consequence of #96. The unit test for the merger tree code checks that if a subhalo continues to be resolved and contains the same particles then it will have DescendantTrackId == TrackId. But we've decided we don't want that behaviour any more. I'll try to think of a more appropriate test.
(this is another situation where #101 would help: I forgot to run the tests before approving the PR)