As someone who has used MySQL with absolutely no problems on over 100 projects over the past 20 years, I really don't get the hate I see for it on HN and other places. I've never used Postgres so cannot comment on that, but unless someone can show me that MySQL kills kittens with every query and Postgres moves us one step closer to a cure for cancer with every SELECT, then I'm just going to carry on over here in my corner...
Why no. If you do not get why PostgreSQL is so much better than MySQL after reading so many people telling you otherwise, I guess I will not spend time on this task. There are some people prefering PHP over Python or Golang, too.
Just a hint: if you did not try PostgreSQL, just maybe give it a try?
And yes, MySQL kills kittens (i.e. potentially silently truncates data) at every UPDATE it receives.
I have several websites with a reasonable number of users running on both MySQL and Postgres. On the balance I think I prefer MySQL - I find the replication in MySQL quite a lot simpler and the issues are not that big of a deal, and well known. Although I will say that both databases are fixing their faults as time goes on. For example Postgres is getting easier to configure replication and as of MySQL 5.7 the silent truncate default STRICT_TRANS_TABLES/STRICT_ALL_TABLES is fixed (although generally implementations I've used set SQL_MODE='STRICT_ALL_TABLES' by default, so it hasn't been an issue).
Well, I'll give you credit for going that little bit further than the post I was responding to. At least you gave a hint as to ONE issue that MySQL can have (which, by the way, I have never come across in the millions or so UPDATES that my projects have done over the years). A link to a discussion about the problem would have been handy, but, oh well.
And I actually did try PostgreSQL once or twice many years ago. If I remember correctly, it was for a project that required full text search capability, and I had no end of trouble trying to configure and optimise PostgreSQL to do it back then, and simply reverted back to MySQL which worked out of the box. Or possibly I even went to SQLite with FTS3 back then to get it to work. I, like most developers would rather be creative on my code rather than fight my tools.
I'd argue that the majority of people are supporting legacy rather than building something new, or something that has been built in the past couple years.
Actually longer than that. It effectively is what Ingres was a long time ago before Oracle basically killed them off. I believe Ingres was a product of some IBM people? Way way ahead of its time.
That's not that interesting to most of us. That's a huge edge case that I've seen used as fodder in every MySQL vs Postgres argument on HN since it's come up since that blog article was published.
That shows MySQL can be used as a base for building your own key-value store on. It doesn't say anything about it being reasonable (or even good) relational database.
Google is a MySQL shop. It's what their RDS-like cloud offering runs, and it is (or at least, was) the supported RDBMS for building apps internally. (Sadly)
I notice that Google has consistently built support for MySQL in all its products...but never postgres.
I don't understand why - is there a massive amount of engineering involved ?