On 9/11/25 5:36 AM, Sebastiaan Couwenberg via postgis-users wrote:
On 9/10/25 11:36 PM, Regina Obe wrote:
Never mind the test above I proposed, I tried it and it still decided to use
the spatial index instead of the primary key on n.id
Which I found really surprising as I thought primary keys trump everything.
Anyway Paul traced the commit that caused the issue and it was introduced in
3.4.
Which commit is that?
Nevermind, I'm blind. It's in the ticket:
https://trac.osgeo.org/postgis/changeset/31bcb7d414/git
Kind Regards,
Bas
--
GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146 50D1 6750 F10A E88D 4AF1
