adjust new_with_pool to return PGMQueue directly

[gruebel]

I'm not sure, if this was done on purpose, but new_with_pool(...) can't return an error, because the pool creation and initial connection is done beforehand. I changed it to just return itself 🙂

[ChuckHend]

I'm not sure, if this was done on purpose, but new_with_pool(...) can't return an error, because the pool creation and initial connection is done beforehand. I changed it to just return itself 🙂

I think you're right, good idea!

Quick note: the Rust lib will publish to crates.io on merge to main, so CI will fail until we bump the version in Cargo.toml

[gruebel]

@ChuckHend I don't think the failing test test_ext_create_list_drop has something to do with my changes. is it a known flaky test?

[ChuckHend]

@ChuckHend I don't think the failing test test_ext_create_list_drop has something to do with my changes. is it a known flaky test?

Yea it is flaky. I re-ran it and now it has passed.

Do you want to bump the version in Cargo.toml so that we can release these changes when it merges to main?

[gruebel]

Sure, can do. Should be version 0.27.0 or 0.26.2? The adjustment is not big, but it changes the signature of a public method.

[ChuckHend]

Sure, can do. Should be version 0.27.0 or 0.26.2? The adjustment is not big, but it changes the signature of a public method.

Good call. Let's make it a minor release -> 0.27.0

[gruebel]

@ChuckHend sorry for the delay, was not at my computer 😅 increased the version to 0.27.0

[ChuckHend]

Thanks. I'm taking a look at that migration check now.

[ChuckHend]

The order of columns returned by pgmq.list_queues() was changed in the v1.2.0 release. It's causing the sqlx cache check to fail. I'll push a fix for this quick.

old:

 queue_name | is_partitioned | is_unlogged |          created_at           
------------+----------------+-------------+-------------------------------
 x          | f              | f           | 2024-05-21 18:53:45.589962+00

new:

 queue_name |          created_at           | is_partitioned | is_unlogged 
------------+-------------------------------+----------------+-------------
 x          | 2024-05-21 18:56:11.791476+00 | f              | f

[ChuckHend]

I went ahead and merged the fix for the sqlx cache back into this branch. All looks good.

Thank you @gruebel !