"r" stands for "remote". They predated ssh and were how things were done remotely in the (distant) past. You can pretty much just replace the "r" with an "s" and that's what they did, only without the "secure" provided by ssh/scp/etc.
25 years ago I was using an Amiga 3000 with a 25 MHz 68030 CPU and 8 MB of RAM. It feels pretty distant to me. A lot has changed in that time.
And, OpenSSH isn't even 20 years old, so it's actually not been that long. SSH was 1995, that's probably the beginning of the ssh epoch and the beginning of the end of the rsh age.
Despite using FreeBSD I had the same question as I had never really touched them. From reading up on it, it looks like they're old remote versions of various commands for interacting with a server/mainframe from a thin client. For example, rcp copies to a remote host, rlogin starts a terminal session on a remote host, etc.
I could well be missing nuance though. My experience with FreeBSD is using it for my server for ZFS' gloriousness.
They're not exactly insecure, they're just used for different things, like networks that are already secure. You can secure existing networks pretty well against things like MitM, and then it doesn't matter if you use rsh or ssh. (part of why so many routers still use telnet is it's supposed to only be enabled on management interfaces)
Quite a few years ago I helped build the public terminal cluster for HOPE. That year we used a token ring network for all of the hosts. Nobody anywhere near the conference had any token ring gear whatsoever. It was impossible to break into the network, so you had to use existing nodes. And the nodes' adapters had their promiscuous modes disabled in hardware. That plus a hardened terminal server and an extremely limited thin client meant nobody could successfully hack the network. You could have used telnet all day and been totally secure.
Though in conference environments, the network is not the biggest risk for targeted attacks. Carefully watching your target enter their password is simpler and more effective.
I think the engineers setting up red/black networks for the military might disagree with you, but even in the enterprise 802.1X can result in a secure local network.
I would say trust is a result of the operation of the network, as you may need to do something to establish it. And in this vein you're right, "networks you already trust" definitely would have been better wording.
No. You can't use rlogin on a 802.1X network or a red/black network. Passwords are transmitted in plain text, which means any other terminal on the network can sniff the password.
Basically unencrypted straight remote connections, the predecessor of telnet, scp, etc. I always wondered back when FreeBSD was main daily why it was still there.
Well they may not have technically predated then but what I meant is that as far as I know for a while r(etc) was used and then was generally replaced in actual usage by telnet, etc.
It was the other way around. Rsh and rlogin appeared well after telnet, and displaced some telnet use because they were able to do password-free connections via .rhosts
