Derrick Stolee <stolee@xxxxxxxxx> writes: > Thanks for double-checking. I also think that dropping the > "hide the error" patch is prudent. Thanks again for a quick and straight-forward fix. As I mentioned elsewhere in the thread, it appears that we invented duplicate API with parallel implementation in get_reachable_subset(), which seems to be a strict superset of in_merge_bases_many(), and that may be what led to an initial and incorrect "get_reachable_subset() is not broken the same way as in_merge_bases_many() so use it instead" response. I haven't paid attention to the quality of implementation of get_reachable_subset() as much as in_merge_bases_many() (e.g. I know of an obvious way to optimize the latter) so far, but it would be wonderful if we can eventually rewrite in_merge_bases_many() as a very thin wrapper() around get_reachable_subset() without any downside.
