Jens Schauder

After looking at this again: the better extension point would be `Dialect`. A request to add features to `Dialect` has probably way higher chance to actually happen.

The problem is, we don't really have a useful abstraction that we could use in such a callback. At that stage we are dealing with `ResultSet` which is way to...

Please provide a [Minimimal Reproducable Example](https://stackoverflow.com/help/minimal-reproducible-example), preferable as a Github repository. Make sure to include the database, either as an in memory database or if that is not possible using...

If I interpret that correctly you don't measure actual loads with JPA but only the lookup in the first level cache, which is probably not what you want. Invoke `EntityManager.clear()`...

Thanks for the reproducer. The main difference between the two benchmarks is, that you accidentally kicked out the Hikari connection pool by explicitly constructing the datasource. you can completely remove...

Results of modified benchmark. Benchmark Mode Cnt Score Error Units SpringDataJdbcBenchmarking.springDataJdbcQuery avgt 5 10103.975 ± 1140.990 ns/op SpringDataJpaBenchmarking.springDataJpaQuery avgt 5 8622.891 ± 1886.295 ns/op You can find the modified benchmark...

In general I would not expect better better performance from Spring Data JDBC compared to JPA implementations in the typical benchmark scenario. The benefit of Spring Data JDBC is that...

> Was it done intentionally? Yes and no. I intentionally added the annotation to demonstrate the effect of not caching the generated query. Changing the query semantics was a mistake...

Duplicate of #756

Is it possible, that you get mislead by the `toString` of the timestamp? If I save an entity with the instant as described, and load the timestamp using a `JdbcTemplate`...