Paul Ramsey

Now just a footnote as @dbaston continues his work

If you rebase to main do we get a green CI?

I'd have to go back and dig, but any tests we may have had (and maybe we had none) would have come out / been turned off at OverlayNG switchover....

Probably some benefit to this in PostGIS too, getting inside the MemoryContext stucture means things like configured memory upper limits can be enforced, and small allocations (should) be a little...

There shouldn't be anything in there that would crash the backend, so my suspicion is a bad mix'n'match install. A version of the extension that doesn't match the pgsql version...

Nope, it's not a config option or anything in the core library, it's hard coded into ogr2ogr. We'd have to copy that code into the fdw to have that feature.

My code review: - move the geometry handling from the Const handler to the OpExpr handler and only throw it into gear if you find a Const on one side...

No plans at this time. Still an interesting ticket though, so don't count it out :)

StringList seems to show up most commonly in GML and GeoJSON backed drivers (WFS, Couch, etc) IntegerList shows up in those as well as TIGER and FGDB/ArcObjects. Both show up...

Possible that the latest fixes also fix your problem, as they were a result of Pg14 changes to update/delete support.