[CommanderData]

I hate to be the person but I've seen complicated dynamic applications push much higher bandwidth and serve millions of concurrent users with similar if not smaller h/w requirements.

Would be interesting seeing Twitters complete backend and while mastodon might not be apples to apples also interested cost per user to infrastructure analysis too.

[lzooz]

Big difference being that Twitter uses Java, Scala, etc. Twitter used to use RoR also and it went down literally every day. I'm talking 2012 or so I think, bad memory haver here.

[vidarh]

Twitters primary problem was that they had not build a system that was designed to shard, not Rails. They'd have needed a rewrite no matter which framework they'd started with.

I have no love for Rails, but blaming it for Twitters old problems is not fair.

That said, Mastodon has much of the same problem, and is only "saved" by the combination of federation and ten years of hardware advances. Thankfully, the federation means there's plenty of opportunity for people to experiment with other implementations of ActivityPub (or even implementations of the full Mastodon API), or fixes to it.

[lzooz]

Rails doesn't scale.

[vidarh]

And yet it runs sites magnitudes larger than Hachyderm just fine.