@austindrenski do you think it's necessary to release a 4.0.4.2 for this? The 4.0.4.1 was a pretty exceptional case, since we had just released 4.0.4 but found an incompatibility with EF Core 2.0.2. I think it's better to hold off with this until 4.0.5 (which can be relatively soon). In any case I'd rather have at least some users complain before doing such a sub-patch release.

@roji That'll work too I suppose. Though that'll block the 2.1.4 patch, but that seems fine.

Though that'll block the 2.1.4 patch, but that seems fine.

We can release 2.1.4 against 4.0.4.1, just like 2.2.0 is. It's never nice to release something with a known bug but this seems relatively minor and it seems better to have versions aligned etc...

Would we just disable the failing tests for the release then?

Would we just disable the failing tests for the release then?

Not sure I understand... 2.1.4 would depend on Npgsql.NetTopologySuite 4.0.4.1 and there wouldn't be any failing tests there. On the Npgsql side we can merge your fix to both dev (4.1) and hotfix/4.0.5, no?

Not quite... npgsql/efcore.pg#732 has failing tests when set to target Npgsql.NetTopologySuite 4.0.4.1. This is the patch to fix those tests.

Am I misunderstanding your question?