I have been interested in a query that returns a batch of results filtered by a subset of the first column of an index and ordered by the second.
I created a simple (hopefully) reproducible example of the issue, the two queries describe the same data but have very different costs (explain output included in the attached file).
server_version 12.8
On slack #pgsql-hackers channel, @sfrost, it was suggested that what I described is achieved by index skip scan. How can I get a development build to test this feature?