@LinusTech I can totally feel you. I would love to resolve all the server issues and give people a nicer experience. And for some point in time we had that. I don't know what changed.
I work as a software engineer on a software as a service. There are 8 full time software engineers, 1 full time server admin and 2 business experts working on it. The whole thing is running on a managed kubernetes cluster and costs shitload.
Compared this to FAF: We run a software-ecosystem that I would estimate as at least 2 or 3 times as complex, with not a single full time dev, on a single dedicated server with a budget of ~100€. The problems we are facing are in the software or in the server config or a comination of both. And unfortunately we lack the manpower to investigate it.
Yes we could throw more money at problems (maybe we should), but that only helps to some extend. What we need is people's time. And not random people but people who exactly know these systems. Long term contributors with a high amount of free time per single developer. Battle proven senior engineers who know how to search for errors. These people are rare. We have a tiny set of them, but they are all very time limited. We'll never be able to put them on a payroll. And FAF is not attractive enough to get more of them.
So I totally understand if you or others leave FAF because it's unreliable. In that case I would just ask you to be indulgent with us. Try again a few weeks or months later. We will not give up trying to fix it.