News for hackers. Stuff that matters.
Re:Medium-hard SQL interview questions by jandrewrogers on Saturday May 02, 2020 @03:40PM (#23054323)
Behaviors around locking and concurrency are not standard. Not only do they vary widely across RDBMS implementations, they also vary depending on how you configure a specific RDBMS implementation in many cases.

Some databases use lock structures that automatically detect and resolve deadlocks, so from a user standpoint there are no deadlocks but deadlock resolution has visible side effects that are implementation defined.

Re:Medium-hard SQL interview questions by hobs on Saturday May 02, 2020 @04:51PM (#23054810)
Yeah but the functions and style used in these answers vary across implementations (date_trunc, etc.)

I highly recommend the SQL Cookbook for newer SQL dev as a quick reference comparison to see how often this is the case for even trivial problems in any RDBMS.

Also, for your language I wouldn't expect all mid tier SQL devs to be able to write a recursive CTE from memory (though its useful, you can just look it up again), but something like breaking apart a query plan on your platform of choice is way more important as that muscle allows you to tell if your CTE was crap or not.