The SQL engine part of osquery is backed by SQLite's parser, lexer, query planner, etc. MySQL would be awesome as well though, mostly because of how easy it might be to integrate the MySQL remote access protocol to do remotely authenticated queries!
Postgres would probably be even easier, since it already has the framework to do this (foreign data wrappers) and a nice API to write said wrappers in python[1] - 200 lines to use your SQL client as an email reader[2], I imagine that wrapping /proc would be considerably less :)
Turns out postgres is even easier. Took 15 minutes to go from "never used FDWs before, but how hard can it be?" to "got access to the process table as SQL" :)
Get a load of Mr. "Im-too-l33t-to-use-MySQL" over here.
Otherwise it sounds like you're insinuating that MySQL is somehow languishing or limping along, which is ridiculous.
MySQL is going to SELECT almost any query faster than Postgres in a single-user/desktop installation, it's going to be easier (although granted less strict) to import data, plus it has Sequel Pro.
Without endorsing the GP's trolling, Postgres generally substantially outperforms MySQL on queries with any complexity to them. Its query optimiser is enormously superior.
The idea that MySQL is much faster than PG is largely a relic of the era of MyISAM, which was very fast for trivial queries, but also terrible in basically every other way.
Interesting. Apparently my received wisdom is out-of-date. I usually use MySQL on the laptop when I have to clean data for analysis because it's already installed, but I will have to look into replacing it.
