Bryce Plunkett
Bryce Plunkett
Yes, that is an unfortunate limitation of postgres. It doesn't actually build any hard dependencies with pgplsql functions, so pg-schema-diff is kind of operating in the dark. We could manually...
I added a flag to disable plan validation in the cli via `--disable-plan-validation`. The way plan validation works: 1. Spins up a temporary database on your postgres instance 2. Places...
> > We could manually try to parse dependencies from the function but that is a lot of complexity: not only would the function itself need to be parsed, but...
Awesome! pgmigrate looks pretty neat. We use cobra already, so viper will probably slot in pretty nicely. Hopefully can get some bigger features out soon. This project is a bit...
This is effectively blocked by #131, which will enable us to more easily build dependencies on statements like column additions/deletions.
Create wrapper around schema name maps such that they can only be queried with a SchemaQualifiedName
Hmmm this issue is bit of a tough first issue. Maybe this one would [be better](https://github.com/stripe/pg-schema-diff/issues/121). Does that interest you?
Hi provokateurin! In what scenarios would you like to not build an index concurrently? I realize it does consume more CPU than a normal index build. Is that the primary...
I see! I'd argue transactional DDL isn't super critical because you can always roll it back (with exceptions of deletes) via plan(oldSchema). Concurrent index builds can't silently fail, so you'll...
> We are only using pg-schema-diff for generating the migration statements, so we are not able to use the automatic reindexing. As in, you're persisting your migration to your codebase?...
>Yes, the are saved as a file. The process of generating the migration and applying the migration are not the same. The migration is generated while the changes are developed...