👉📝 Feedback Wanted: Share Your Use Case & Experience

[tianzhou]

I'd like to collect some customer testimonials to feature on our website and better understand your use cases. Please share your comments with the following information (all fields are optional):

• Are you using it in prod (Y/N):
• Use Case:
• What I Love:
• What Could Be Better:
• Other Tools I've Tried:
• Company (skip this if you're not comfortable to disclose):

[Dulanic]

Use Case: Playing! I like testing new things.
What I Love: Don't know yet
What Could Be Better: It feels like a default dump isn't a good way to export the data as the plan always fails. I was just looking for test with a existing data set for another project I've worked on Teslamate. However, I couldn't find a easy way to even try to test this. Maybe we did a bad job designing the schema in the first place! I posted the dump results, teslamate-schema.sql

Other Tools I've Tried: N/A

[abarreir]

Use Case: Lots of views & materialized views, and some of them are cascaded, so when we want to change a simple condition we have to drop / recreate all views, the migration to write is error prone and code review is counterproductive

What I Love: The promise, allowing devs to only review the change on the schema and not the generated migration

What Could Be Better: At the moment it does not work with my schema, crashes when generating a diff because of an incorrect syntax in the generated schema dump. Can't really share my schema. Also, I'm still on a Mac Intel, would be great to have an amd build on homebrew :)

Other Tools I've Tried: Atlas, but too expensive and comes with too many other features i don't need

[marcintustin]

Use Case: Capturing actual changes made in development and minimizing drift between development and production environments. Specifically we're using this in the context of modernising an ad hoc database schema incrementally into a more designed schema. I generally only use the dump functionality to see the current state of my databases, and manually prepare migrations. It also gives me and others visibility into the final state after a migration runs.

What I love: Filtering, ease of use out of the box, responsiveness of @tianzhou to bug reports, quality of documentation. Dumping to separate files makes viewing diffs much nicer.

What could be better: It still has bugs :) The migration generation functionality doesn't generate the most parsimonious changes (but for me this is not that important; it's also a problem with terraform, so I assume it's a largely intractable problem). Would be nice to have more debugging output available to see what's happening, which would probably help me submit more precise bug reports.

What would be really lovely would be if pgschema could use postgres' built in ability to get the text of objects ddl as much as possible to ensure consistent formatting and use some kind of parsing (e.g. a treesitter grammar) to identify discrepancies between what pgdump is seeing and what postgres is outputting.

Other tools: Pgdump for dumping postgres schemas, and tooling built on top of that. OK, but frankly much more fiddly than this.