(From a listserv post by Jeremy Elson in 1997)
When I was a freshman in the CS department of Johns Hopkins in '91, the undergraduate workstation cluster was a bunch of old DECstation 3100's that had strange names (poincare, syzygy, hull, whatever, emanon, and a few others). A couple of the undergrads at the CS department (Justin Chandler and Dave Reed) experimented with running a copy of DikuMUD around September of '91. Since they ran it on the DECstation named "whatever", they called it WhatMUD. WhatMUD was the first MUD I ever played, and pretty much the only MUD I've ever played as a mortal.
Afer a couple of months, WhatMUD died due to a series of hardware failures. Before it had died, though, I'd started tinkering with the code itself, mainly because I found the world files and a bunch of documentation in Justin's directory and realized that I could get the stats on any object in the game just by decoding the world format . (This was very exciting to me at the time!) Also I'd found the code itself fascinating, and was able to predict how the game would react to different situations because I could just look at the source! What power!! My friends were amazed.
Around December of '91 WhatMUD was completely dead and I needed a MUD fix badly. I got someone to send me a copy of DikuMUD (it was very hard to find at the time) and tried compiling it myself. I remember not being able to sit still while the thing compiled for the first time - I was bouncing off the walls, running around, pacing in the hallway outside my dorm room because seeing my VERY OWN MUD actually COMPILE FOR THE FIRST TIME was just so damned exciting!! (I guess I was easily excited back then.) Of course, since it was on a DECstation 3100, it was quite slow and the agonizing wait was like 30 minutes. And, since it was compiled with GCC v1.x which didn't work very well at the time, the thing was constantly spewing out hundreds of assembler errors. (I remember being so excited when gcc v2 came out because the MUD actually compiled without assembler errors for the first time!)
Now, I had not (of course) secured any sysadmin approval for running a MUD, so I had to run it covertly. The CS department had just added a new DECstation to the cluster -- called "circle". Since the name was new, none of the users knew it existed yet and the machine was usually idle. So I ran the MUD on circle.
Now, some of my freshman friends heard that I had a MUD running, and we spent all that night in the computer lab killing things to our heart's content, with me loading billions of weapons for everyone except this one guy who we didn't like. When we finally went home, one of my friends (Naved Surve), said to me, "So, do you think you'll open CircleMUD to the public?" He'd called it that because the old JHU MUD, which had run on whatever.cs.jhu.edu, was called WhatMUD -- so it seemed only logical that the MUD running on circle.cs.jhu.edu should be called CircleMUD. And the name stuck. Thanks, Naved.
circle.cs.jhu.edu (the machine) was finally decomissioned a few months ago. The whole DECstation cluster was replaced with a single Sparc 20 about a year ago (hops.cs.jhu.edu), and the DECstations that were still alive were turned into X-terminals.. until they died. Which, by now, most of them have.
circle.cs.jhu.edu still has an IP address.. but, if you ping it for old time's sake, no one is home any more. Sniff.