I like the idea of being able to add `IDENTITY` columns. I like Variant B
because I don't see the point of an identity column without a `UNIQUE`
constraint, or even PK. It seems if you're okay with the overhead of adding a
surrogate key, you'd take it the next step too.
I have a hard time imagining where variant C has much value. That seems like a
pretty niche edge case and I suspect any instances with that setup already have
code to do exactly what they need. If Postgres had `IDENTITY` support long ago
I don't think sequences would ever have been talked about.
--
Reply to this email directly or view it on GitHub:
https://github.com/osm2pgsql-dev/osm2pgsql/issues/2274#issuecomment-2591575149
You are receiving this because you are subscribed to this thread.
Message ID: <osm2pgsql-dev/osm2pgsql/issues/2274/[email protected]>
_______________________________________________
Tile-serving mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/tile-serving
