4 ms·
I'm quite sure you could do a breakdown of any technology and find enough problems that it looks like a bad choice. What tech you use is always a compromise be
by onion2k 11y ago
I'm quite sure you could do a breakdown of any technology and find enough problems that it looks like a bad choice.
What tech you use is always a compromise between the benefits of what it's good for and the problems you need to work around. In the case of MySQL it's ubiquitous across hosting services, easy to use, very fast once you've tuned it a bit, and there's clients for pretty much every language. You need to weigh those things against all its problems, and often you'll find it's actually a good option if you're doing web work that you need to deploy to environments you have little control over - less so now than 5 years ago admittedly, but that is the reason MySQL is so popular - because before the devops trend that's got web teams running their own servers most sysadmins made MySQL the only database available.
- smt88 11y ago> it's ubiquitous across hosting services So? No one needs 10 different hosting services. You only need one, and you've been able to choose a (cheap!) one that offers Postgres for many years now. > easy to use I strongly disagree, and the entire article is basically rebutting this assertion. It seems easy to use, but silently failing, breaking promises, and being noncompliant with standards are all issues that make MySQL harder to use. > very fast once you've tuned it a bit You can tune Postgres as well. It actually optimizes its queries much better than MySQL does, so you can go longer without starting to optimize. > and there's clients for pretty much every language Unless you're using some fringe-y language, this isn't an issue either. Some languages have gotten a Postgres client before MySQL, simply because their communities have a culture that's more similar to Postgres.