It's really simple: MySQL / MariaDB do not function correctly for the feature set they supposedly support, while PostgreSQL works correctly for that which it supports and people recognised this eventually, which is thankfully why PostgreSQL is the leading open source relational database management system.
There is not much philosophy here: MySQL should have never been used since it's always been busted, and thankfully now it's lost so much market share that it's become all but irrelevant. If you want to build bullet proof, robust systems and don't want or can't afford Oracle, one of PostgreSQL variants is a critical piece for building such a system.
I like sleeping through my nights instead of wasting my life away troubleshooting MySQL. How about you?
> I like sleeping through my nights instead of wasting my life away troubleshooting MySQL. How about you?
I waste them working on postgres. Doesn't mean I've to denigrate other systems.
At least you’re investing your time in a product which works correctly and isn’t throughly broken by design. Sometimes it’s necessary to point out MySQL’s flaws lest someone get the “bright” idea to put it in production, then I get the job and then get stuck with that disaster of a database (that actually happened to me several times). Raising awareness is the best prevention.
I'm fine with arguing that $pro{je,du}ct is better for most cases. But that doesn't mean the other alternatives have to be denigrated as being pure shit. That's not going to convince anybody that's not already convinced, and stirs um enmity.
FWIW, I think both postgres and mysql/mariadb benefit from the competition.
Do I think that postgres is overall the better project? Yes. Do I think mysql is shit? No.