Fredrik Tõnisson-Bystam

I have one test failure on Sqlite, and looking more closely at it, it feels light it might be a separate SQLite-oriented bug perhaps? The table/entity pair is declared like...

Bug has been filed and reproduced in tests here: - https://youtrack.jetbrains.com/issue/EXPOSED-405/SQLite-bugs-Table-with-custom-ID-behaves-weirdly-in-DAO-and-batchInsert - https://github.com/JetBrains/Exposed/pull/2116

> Thank you for PR, it looks interesting, I'll create an issue in YouTrack for the discussion of it, I have one here: https://youtrack.jetbrains.com/issue/EXPOSED-403/DAO-soft-delete-restriction-support Unless you meant something else :)

After getting deeper knowledge of the codebase I think this approach requires quite a lot more thought before it would be sensible to put it in the library. Therefore I...

Apologize for the failing PR title/commit message/detekt checks. I will make sure to get that sorted.

I obviously broke a bunch of other tests using this implementation 🙃 - looking into it

Okay - my attempt broke quite a lot of tests it looks like. Might have been a bit naive to simply blend them like I did. Maybe things need to...

Looking more closely at the failing tests - I got the sense that all the "extra stuff" that I cherry-picked from the `DataSourceTransactionManager` that manipulated the connection might have been...

> When I first proposed to re-implement the `SpringTransactionManager`, I thought about implementing an Exposed `TransactionManager` for Spring, but since it is hard work, I thought it would be more...

I created a copy of the existing `ExposedTransactionManagerTest` that runs basically the same tests but using springs `JdbcClient` to verify all the same cases. Some of them behaved differently, but...