I hate these posts that start with a huge, preachy paper about how “what you know/do is wrong and bad, there is a new and much better way” and then that new thing is actually the product the author is selling you.

Especially when they’re about databases because those are practically drawn from a template. “Databases are bad and confusing, every developer eventually learns this. What’s to be done? If only there was another database that promised you everything? Oh, here it is!”

I also don’t like how they use a Mastodon implementation as a comparison against Twitter. Twitter isn’t bad (on a technical level) because they “chose the wrong database.” Twitter invented a lot of the techniques this person talks about and builds on because when Twitter started they had an entirely different business model and requirements. And, again, none of the other technologies existed yet! A dedicated team could probably reimplement almost any mature company’s tech in a better way because you have the faulty original to use as a guide for what to make and what to avoid, and you have no legacy requirements to constrain your green field rewrite.