Marco Slot
marcoslot.com
Marco Slot
@marcoslot.com
Mostly posts about PostgreSQL, Snowflake Postgres, and PostgreSQL extensions.

Formerly Crunchy Data, Microsoft, Citus Data, AWS, TCD, VU
github.com
November 4, 2025 at 4:07 PM
Reposted by Marco Slot
No system hits the sweet spot of allowing for extensibility while maintaining systems safety. It would be nice if there was a standard plugin API (think POSIX) that allows compatibility across systems.

Thanks to @marcoslot.com + @daveandersen.bsky.social for their collaboration on this project
July 3, 2025 at 7:03 PM
Thanks Gunnar!

We generate regular position delete files for merge-on-read, so any Iceberg query engine can read them. Equality deletes would be more CDC friendly, but not supported in most engines.

We have some secret sauce around how we track/know positions, but being Postgres helps a lot there.
April 22, 2025 at 4:52 PM
Would be cool if Iceberg/Parquet had support for storing JSON as vectors.
March 26, 2025 at 9:26 PM
Generally by unwinding the JSON in the insert..select that processes the raw log files. JSON is broken down into columns by default, though nested JSON remains as jsonb. You can either store that directly (stored as string), unwind it manually, or convert to a composite type (stored as struct)
March 26, 2025 at 9:25 PM