Hi Lewis,

Personally I'm a big fan of Kyle's works with Jepsen, but we have to keep in mind that this is analytical work and that in real life we are able to work around those inefficiencies or constraints.
MySQL and derivatives are DB systems which are used mainly by websites, small or big. Although I did have some customers back in the day at MariaDB in payment systems but they are generally aware of the complexity and are able to retrace any errors, financial institutions will often use commercial database systems and/or have some transaction processing and checking in place.
I have been involved personally in many cases of using and supporting MySQL and MariaDB in e-commerce situations and even in cases where incorrectness of data could be observed, those have remained extremely rare and never have led to business issues.
You could compare this situation to pharmaceutical studies. Medicines are rarely 100% efficient or have side effects but still, people do use them :)

Le lun. 1 janv. 2024 à 05:55, Lewis Chan via discuss <discuss@lists.mariadb.org> a écrit :
> In the end, it is an evolutionary compromise between getting flexibility and
performance on the one hand, and sufficiently correct behaviour on the
other. My personal opinion is that the success of the database is in large
part due to the constant focus on ensuring the flexibility and performance
needed by practical use

This makes sense to me. However, something puzzles me. If mysql places more importance to flexibility and performance, aren't there any customers(especially financial fields) greatly complaining the subtle incorrectness of mysql,  leading to money loss and business failure ? If there are, is it enough to push mysql team to correct some incorrectnesses that Kyle found ?
_______________________________________________
discuss mailing list -- discuss@lists.mariadb.org
To unsubscribe send an email to discuss-leave@lists.mariadb.org