Correct. We don't do these things and we never will.
When you 'nmap' an rsync.net storage array, you get:
22 TCP
... and that's it. There are no other services running, or offered. There are no interpreters in the environment. The filesystems are mounted noexec,nosuid.
We do one very simple thing and that's it.
But you don't.
To be competitive you either need to drop prices or do more stuff to be worth the money.
But having no other cost than the storage itself is relieving and having the service running for nearly 20 years gives you a peace of mind on their reliability. I also hear their support is good
I would want a bit more modern web interface though. BorgBase does look interesting to me, except only 2 years of operation can't tell about its reliability and longevity.
That's good but at the volume needed to reach it, I imagine that you're in the realm of "contact us and get a special deal" at other providers (e.g. AWS, GCP, Azure) too.
Plus, we haven't even got to the "you have to pay for capacity, not usage" aspect of rsync.net pricing.
Pricing is the big thing I wish you'd fix. Your service has a bunch of cool features but they're just not worth that price, not when I can get most of the same features from Hetzner for a fraction of the cost.
> As for bandwidth intensive protocols, we have HPN-SSH[2] patches built into our environment which allow for high bandwidth transfers, over SSH, over long WAN links.
That's cool (honestly, not being condescending) but my point was that for most of the reasons you'd be excited about free bandwidth, rsync.net isn't a viable solution. For example you can't host images, videos, software updates, machine learning datasets or other large binaries and serve them over HTTP on the public internet.