Robins

This is interesting - good to hear another view to the approaches mentioned. I see that you mention that ideally the goal was / is to always stay on latest...

> On second thoughts, looking at the in-builts tests - probably it'd be a good idea to run at least these 4 tests and come back with results. To wrap...

`SET LOCAL` is an aspect that is covered in #1046 (although keeping it open doesn't hurt) - but @PradeepFSTdhane123 - [this comment](https://github.com/sqlancer/sqlancer/pull/1091#issuecomment-2708717489) may help explaining where this issue stands today.

Thanks for working on this! @ahmed-kamal2004 Before diving deeper, I see that the image in section "JSON in DDL" actually shows a DML. Is that intentional?

@guptapratykshh - A few things. 1. We don't want to go the route of `server_version_num` here. The goal is that ideally SQLancer doesn't need to check the version and should...

@codebreaker32 Could you point me to how you ensured that the float numbers are exact?

@PradeepFSTdhane123 yes - Please see this comment for more context https://github.com/sqlancer/sqlancer/issues/1045#issuecomment-2716334071

- TLP - Repro stack trace below. - Full file attached - [db-1734222303.txt](https://github.com/user-attachments/files/18209083/db-1734222303.txt) ``` --java.lang.AssertionError: The size of the result sets mismatch (10 and 11)! --First query: "SELECT t0.c0 FROM...

Here is a simplified repro. It seems that this test is failing because of the documented Postgres caveat that `INHERITANCE` doesn't work well with PKs. See this answer - https://stackoverflow.com/a/56637523/1378768...

To answer, yes this (and other vN issues) are still open as of today. To add, (and this isn't fixed in stone but to explain the progress of these issues),...