Appendix 1 - The Machine List

...and the feed.

machine name OS CPU specs date obtained source comments
UPL-SS20-01 any SparcServer20 model 51        
UPL-SS20-02 any SparcServer20 model 51        
UPL-SS20-04 any SparcServer20 model 51        
UPL-SS20-05 any SparcServer20 model 51        
UPL-SS20-06 any SparcServer20 model 51        
UPL-SS20-08 any SparcServer20 model 51        
UPL-SS20-09 any SparcServer20 model 51        
UPL-SS20-10 any SparcServer20 model 51        
UPL-SS20-12 any SparcServer20 model 51        
arundel.upl.cs.wisc.edu NeXTSTEP 3.0 Color NeXTStation 25MHz 68048, 32MB RAM, 600MB disk 1994  

felix's personal machine.

Supposedly named after a castle that had never been breeched because felix felt NeXTstep was so secure. We gave him a lot of shit when SATAN (remember that?) found a couple issues with it." - mitch

"the machine was actually named Arundel because I liked the castle. NeXTSTEP's [in]security had nothing to do with it." - felix

brunt.cs.wisc.edu NT4.0 333Mhz Pentium II ATX Tower 64MB RAM, Permedia 2 Graphics, and a 4GB disk September 1998 Microsoft Used for NT development.
bynar.cs.wisc.edu BeOS BeBox with 2 Power-PC 66Mzh 603e chips 16MB RAM May 1996 Intel

Named because it was our first dual-processor box so we wanted something that implied "two."

I think we used part of our webspider nest egg.

"No, you're right, that was part of the WebCrawler gift (which they've still got some of) And someone should mention Brian Pinkerton as part of the history..

"Bart, in later years, would rub it in that we spent money on it." - epaulson

"It was supposed to come in the early spring but there was some snafu with getting the check cut and then there was a delay on the Be side." - mitch

"BeOS was fine, especially on Intel architecture, it's just that it came in at the wrong time (after Linux), after the geek niche was filled." - aselle

"The bebox was the coolest thing _every_ anyone who says different is clearly delusional.

"It had two, TWO! PPC processors, with sexy sexy bargraphs for
cpuload running up each side of the case. It had the best reset switch of any machine I have ever worked on, ever. It was a dual processor box for
like $2k, it was hyper killer phat. Remember the Geekport (tm)? We could have done anything... too bad Beos sucked so much, then died." - gulfie

candy.upl.cs.wisc.edu none MacintoshSE       upl candy dish.
cat5k-upl cisco Catalyst 5000 CHAS, SPRVSR, PS        
cisco-upl 11.2(2.5) CISCO 7010 Router        
crusher.upl.cs.wisc.edu MaxOS, A/UX Mac IIfx        
data.cs.wisc.edu       May 2000 HP Primary user machine.
dax.cs.wisc.edu HP-UX 10.20 HP 9000/735-99 48MB RAM, scavenged to 128MB RAM and "a neat graphics card." in 1996. January 1994 HP

"Dax was named because it was a hot and sexy (we didn't find out how hot until hard drives started to fry).

"The other name of 'odo' was up for grabs, but that went to the printer. William wanted the name 'Bashir' for the same reason that 'Dax' was chosen." - mturner

"As I recall, the fact that dax has a bunch of venting holes down the side that looked similar to the character's makeup had omething to do with it as well." - dsb

drizzt.upl.cs.wisc.edu slackware .96 486DX/2      

Craig Peeper's machine.

Mostly ran slackware although we also played with a preview of NT3.1 and some Nextstep-i386.

dukat.cs.wisc.edu Solaris/Red Hat x86/6.2 Pentium Pro 200Mhz 128MB RAM June1996 CSL

Primary user machine until we got data.

"Garak and Dukat were part of a department-wide upgrade that got rid of most of the DECStations and replaced them with PPro 200 desktops.

"We were supposed to turn yar and picard back into the department for tradein, but we wound up getting to keep them. They were kept around as actual machines for a while, then got turned into X-Terms near the end of their life. I think we bought a couple of upgrades for dukat.

"Somewhere in there dukat died and got replaced with garak, and then old-dukat
was fixed and returned as garak. That might have happened a couple of times." - epaulson

"BTW, this was originally why I ressurected harmon. The plan was that when we were ship them back picard and yar we'd actually ship them picard and harmon. Then yar would become the uber-decstation (40M of ram!) and remain as the web/mail server. However since they never bothered to shred all the old dec machines we got to keep them all anyways." - mitch

duras.upl.cs.wisc.edu BSD 4.3 HP Bobcat 300 series 4MB RAM (later upgraded to 8MB)    

"duras and gowron each had 40M external harddrives, connected over HP/IB.

"duras/gowron were 300 series machines -
they previously had been part of the old "catXX.cs.wisc.edu" instructional lab." - mitch

emh.cs.wisc.edu any HP Vectra PC 486/66N        
eta.upl.cs.wisc.edu BSDI BSD/386 UNIX, then Linux, then FreeBSD UNIX, and then finally back to Linux. 486DX/33*   September 28th, 1994 to August 29th, 1996.  

IP: 128.105.45.18

I remember it would reboot if you hit it with a nerf projectile (which we did a lot out of spite).

The sordid history of ETA.

garak.cs.wisc.edu Linux/NT4.0 Pentium Pro 200Mhz 64MB RAM June 1996 CSL

Linux development machine.

"Garak and Dukat were part of a department-wide upgrade that got rid of most of the DECStations and replaced them with PPro 200 desktops.

"We were supposed to turn yar and picard back into the department for tradein, but we wound up getting to keep them. They were kept around as actual machines for a while, then got turned into X-Terms near the end of their life. I think we bought a couple of upgrades for dukat.

"Somewhere in there dukat died and got replaced with garak, and then old-dukat
was fixed and returned as garak. That might have happened a couple of times." - epaulson

"BTW, this was originally why I ressurected harmon. The plan was that when we were ship them back picard and yar we'd actually ship them picard and harmon. Then yar would become the uber-decstation (40M of ram!) and remain as the web/mail server. However since they never bothered to shred all the old dec machines we got to keep them all anyways." - mitch

gowron.upl.cs.wisc.edu  BSD 4.3 HP Bobcat 300 series 8MB RAM    

 "Had problems with some X apps since not everyone was ready to cope with a 6-bit pseudocolor display. I adopted this box eventually - for a little while it was my primary home machine since it had a 21" (but only 800x600 or something gross) monitor. The monitor weighed over 100lbs and eventually got thrown away.

"duras and gowron each had 40M external harddrives, connected over HP/IB.

"duras/gowron were 300 series machines -
they previously had been part of the old "catXX.cs.wisc.edu" instructional lab." - mitch

hate.upl.cs.wisc.edu   Apple IIgs      

mitch's sister machine for muscle, ("muscle and hate", from a Nitzer Ebb lyric).

"I don't believe "hate" was ever a net-worthy box. I think it just refered to an apple 2gs I had scavanged from my sysadmin job in family resources [*] which later I gave to Gus.

"[*] My syadmin job was taking care of cyrus.famres.wisc.edu which was a SPARC LX(?) Solaris 2.1 (later 2.3). It became home to the original dmf web site." - mitch

DMF and the UPL

hugh.upl.cs.wisc.edu NeXTstep 3.2 NeXT Cube

16MB RAM

"standard 40M swap only disk and ran network-booted off of dax (although
later we found some ~100M drives to throw in them so some of them had some
local binaries)" - mitch

January 1995    "We actually had 4 NextCubes (sorta "borrowed" from the stat dept; they had been sitting unused under a table in the CSL lab for awhile)

"They included locutus, hugh and a couple others. They each started with 8MB of RAM but one of them had a motherboard failure (precipitating a polical shitstorm I don't want to relive at the moment) and locutus (I think) inherited the extra RAM." - mitch

"They came from Stat - CSL gave us the tip off, but we had to go get them from a supply closet in unit 2." - hartmann

"I think they were the only machines I ever saw that had floptical drives, too." - kilroy

"...although none of the four worked by the time we got them. Apparently it was pretty common for those drives to fail after a couple years." - mitch

hugh/locutus were named because we wanted to pick "borg names" (think 'big black cubes').

khan.upl.cs.wisc.edu Ultrix/Linux/NetBSD HP Bobcat 5000/200   June 1993 DEC "Kahn and Saavik where the user machines prior to picard and yar showing up in the lab. Saavik was the machine that 'work' was done on (I recall a few 'art' runs on there) because it was harder to type and so people logged into Kahn. I don't believe that dsb was there during the bobcat days (-92)." - mturner
kirk.upl.cs.wisc.edu            
locutus.cs.wisc.edu any HP Vectra PC 486/66N        
locutus.upl.cs.wisc.edu  NeXTstep 3.2 NeXT cube

8MB RAM

"standard 40M swap only disk and ran network-booted off of dax (although
later we found some ~100M drives to throw in them so some of them had some
local binaries)" - mitch

 January 1995  

"We actually had 4 NextCubes (sorta "borrowed" from the stat dept; they had been sitting unused under a table in the CSL lab for awhile)

"They included locutus, hugh and a couple others. They each started with 8MB of RAM but one of them had a motherboard failure (precipitating a polical shitstorm I don't want to relive at the moment) and hugh inherited the extra RAM." - mitch

"They came from Stat - CSL gave us the tip off, but we had to go get them from a supply closet in unit 2." - hartmann

"I think they were the only machines I ever saw that had floptical drives, too." - kilroy

"...although none of the four worked by the time we got them. Apparently it was pretty common for those drives to fail after a couple years." - mitch

lor.cs.wisc.edu Ultrix/Linux/NetBSD/IRIX 6.2 200 MHz DEC R5000/125, 64MB   June 1993 DEC  
mccoy.upl.cs.wisc.edu   PDP-8      

"meek tells me he got a PDP-8 from UW-Milw via the friedels and gave it to Mitch who put it in the lab" - smolstad

mccoy.upl.cs.wisc.edu   PDP-8      

"Kilroy's machine. I forgot the details of it, but once (twice?) dac had logged into to by cracking root or something and kilroy got super pissed." - psilord

"Nope, I don't recall getting pissed about anything but an OS upgrade making the hardware nonfunctional[1]. Other people got pretty pissed about it getting dual booted to GameOS and used for Mechwarrior, in one of
the cyclical "Games good!" "Games bad!" swings.

" [1] AFAIK. That was a hella long time ago though, in Chuck years. Also keep in mind that Dac was my roommate." - kilroy

kilroy's machine, (and the source of some bitterness in the upl)

minuet.upl.cs.wisc.edu MacOS9.2/OSX Mac 867MHz 1.2GB RAM, DVD-R December 2001 Apple/DoIT nonlinear video editing goodness.
mookie.cs.wisc.edu Windows 98 333Mhz Pentium II ATX Tower 64MB RAM, Permedia 2 Graphics, and a 4GB disk September 1998 Microsoft Used for NT development.
muscle.upl.cs.wisc.edu some early version of Slackware; previously SLS Linux 1.01 w/ a pre-1.0 kernel  486 DX/25, 8MB RAM, 80M MFM full-height disk      

"something mitch built out of spare parts and cardboard. It was amazing."

"Later we added "the maxtor" which was a 100M MFM full-height disk that Besh had scavenged from somwhere. The maxtor beeped loudly when it found a bad sector (which was all the time). Since the disks didn't fit in the case they both sat loose on top.

"Originally had hercules mono gfx but I think for awhile we made it dual-headed with a borrowed VGA monitor... I remember trying to play network doom on it but it only gave a fre frames every 20 seconds since it was wwapping hard to the MFM disk.

"After I left muscle.upl's IP got reused for someone else's (nwa?) personal box." - mitch

nothing.upl.cs.wisc.edu Slackware 2.x, Windows 3.1 multiboot 486 DX2/66 4MB memory, 250MB hard drive January 1994  

Dave Carley's machine. Started with Linux kernel 1.1.5x.

Named after Nine Inch Nails' record label.

odo.upl.cs.wisc.edu   HP Deskjet 1200c-ps printer        
pain.upl.cs.wisc.edu Linux 66MHz  

 

Gulfie's box. "I remember the word PAIN written on it in black coffee."

The 66Mhz machine that would boot into 2Mhz as a default speed if you didn't pay attention.

"Let me tell you, the linux kernel boots god damned slow at 2MHz." - psilord

picard.upl.cs.wisc.edu Ultrix xterm DEC 5000/25  24MB RAM, 300MB Hard Drive     Also the main user machines (with yar) for many years.
picard.upl.cs.wisc.edu   AT&T 3B1   1990  

 "Picard had a previous incarnation prior to the ultrix system. It was apparently a very nice system that was was earmarked for the upl, but something happened (political) and it didn't show up. Dunno, that was pre fall of '91 and I only heard rumors of it." - mturner

"quotas were later bumped to 1M after we
scored more space, and quotas had to be checked by hand, since Ultrix didn't
have quota support." - hartmann

The RAIV   15 SPARCStation 10s 32MB RAM, 1GB disk each     The RAIV is around for parallel development. They are interconnected with Lattice Net.
saavik.upl.cs.wisc.edu  BSD 4.3 HP9000/200 Bobcat        "The saavik name and several others may have been used for the pile-of-DECstation xterms that were in the lab as X chooser hosts to the real machines." - kilroy
scotty.upl.cs.wisc.edu   AT&T 3b1  2.5MB RAM, 40MB disk    AT&T

The UPL had the first 2 3b2s in the department, donated by AT&T, a long long time ago. -dparter

"Lots of hardware modifications by us. Had the ultra-rare 3b1 ethernet card and an incredibly ancient IP network stack (as in "no subnet mask support") Would usually crash the hardware if you tried to ftp a file bigger than about 40K. I think besh ended up adopting it." - mitch

"Besh says it came from the loading dock behind the math building. "It really will take three days, Captain" - hartmann

"Besh has Scotty. He recently disposed of the other machine." - hartmann

silly.upl.cs.wisc.edu Slackware Linux 486 DX33       Kicked out when Bart cleaned out all the machines that the UPL didn't have root on.
something.upl.cs.wisc.edu MacOS 7.6/Rhapsody PowerMacintosh 9500       Dave Turner's machine
sparc1.upl.cs.wisc.edu SPARCLinux SPARCStation 10       xterm
spock.upl.cs.wisc.edu  NeXTstep 3.2 NeXT Cube 68030

8MB RAM

"standard 40M swap only disk and ran network-booted off of dax (although
later we found some ~100M drives to throw in them so some of them had some
local binaries)" - mitch

 January 1995  

 One of four NeXT Cubes.

"It seems we named the first one "spock" because (as felix mentioned) there had been a personal nextcube in the lab a couple years earlier with that name. So it was basically the same machine but now was a lab machine." - mitch

"They came from Stat - CSL gave us the tip off, but we had to go get them from a supply closet in unit 2." - hartmann

"I think they were the only machines I ever saw that had floptical drives, too." - kilroy

"...although none of the four worked by the time we got them. Apparently it was pretty common for those drives to fail after a couple years." - mitch

"I believe one of the 3b1's was briefly renamed spock. See, the CSL had this script that changed the root passwords on all their machines, and somehow the UPL machines were on the script's list, and had the account that let
this happen. So one morning we couldn't get root. I forget exactly the chain of setuids and scripts and logins, but it ultimately hinged on a .rhost file that let root log in without a password from spock. So we pulled the 3b1 out of the closet, changed its name to spock, and hacked root. I tremble to think that this might have been the episode that Sam stumbled upon (it would have been in fall '93 or in '94 - possible?).

"Anyway, we got in, figured out what had happened, and a troup of us headed upstairs to confront a CSL employee who was on the phone at the time. He politely ended his phone conversation with "I have to go - there's a lynch mob outside my office."" - dsb

tribble.cs.wisc.edu   25MHz Mac IIci 80MB hard drive     "I did all of the ART (Abe's Ray Tracer) code on Saavik using vi and HPStarbase. I ported that code to Dax using p2c. Afterwards, I did most ofthe Microcosm coding on Tribble using Think Pascal. Tribble was a 25 MHz 65030 with a whopping 80 MB Hard Drive, but it was capable enough to write a few hundred thousand lines of code on." - abe
tribble.cs.wisc.edu HP-UX 10.20 HP9000/715 256MB RAM, 100MHz, 31MB SPU September 1995 SPEC

served as another User Machine ('96.)

"We scored tribble...and had no end of bad luck
with HP-UX versions. 10.30 is *not* like 10.20." - hartmann

troi.upl.cs.wisc.edu NetBSD Mac IIcx ran the webcam.    

Abe's machine

"There were two trois. The first one just stopped one day. It got a drive swap with another IIci that was running NetBSD and ran for another few years. It actually finally left the building about 3 months ago (9/01.)" - epaulson

tuvok.upl.cs.wisc.edu NeXTstep 3.2 NeXT Cube    January 1995  

"I'm not sure why we named the last one "tuvok". This is the one that had a motherboard failure after just a few weeks. Its name soon got reused for the Wang-from-hell that we built.

"When tuvok died its memory got put into hugh (not locutus like I guessed earlier)

The nexts all ran NeXTstep 3.2. The nexts probably arrived late-winter/early spring of 95 or thereabouts. The doomed next had died around mid march 1995. tuvok was up by around 5/95." - mitch

"That was the Wang we kept pulling random chips out of while it was running, and finally taunted it into flames by running a flathead screwdriver accross the ISA slots." - hartmann

"Correct. It's amazing how well MS-DOS would run while hot swapping the RAM
and with a Z-80 in the 8087 coprocessor slot :-)" - mitch

tuvok.upl.cs.wisc.edu Wangix (Linux port) 386DX  4.5MB RAM    

"It was a 386DX box built soley out of parts scavenged. had 4.5M of RAM. "Wangix" was the name we gave the linux/lilo tweaks I did in the middle of the night (with the help of still-in-high-school besh) to make linux 1.3.something boot on a machine with 512K base memory (instead of the normal 640K)

"The other 4M of memory was on two proprietary Wang ISA-like cards.

"The network card was a 3c505 that someone had liberated from CAE. I doubt
they ever missed it.

The monitor/video was from an old PC that wannis used to have. It was EGA. The monitor was "Casper" brand. We called it "Casper the unfriendly monitor" since if when turned on it would make nearby monitors go out of focus.

Hardrive was a half-height 5.25" 40M MFM. Not sure where we found that.

The 3c505 generated a lot of heat so we removed the front plastic facade and bolted a powerful fan to the front of the case (totally unprotected)

When dax's harddrive died in the summer of 95 tuvok acted as our web server (using a shell script running from inetd that just gave a "we're down" message to any query)

I still have tuvok.

Unfortunately casper seems to have gone missing in a recent move... I still hope to locate it. (Note we also had some other scavenged Wang machines including one that we started on fire)" - mitch

"Do recall the voice coil... that went 'beep' when it had to recalibrate after hitting a bad block on the drive. IIRC, there was a directory named "/beep" that contained files with bad blocks on them. Typing "cat /beep/*" made the drive make significant noise." - mturner

unlikely.upl.cs.wisc.edu   AT&T 3b1      AT&T

The UPL had the first 2 3b2s in the department, donated by AT&T, a long long time ago. (dparter)

"They had been dropped off by someone at the Hospital who thought we might have fun with them. They had the oriignal SysIV boot floppies and everything, and took a *really* long time to start up.

"IIRC, didn't someone from the hospital evetually come looking for their multi-thousand-dollar Unix workstations a few years later, not realizing that obsolensence and depreciation affects even fancy computers..

"I think Scotty briefly had residence in William's apartment. I don't beleive this lasted long." - wonko

upl-sparcs   Sparc10/30        
upl.hp2   C100/9000, 100Mhz        
upl.hp4   C100/9000, 100Mhz        
unnamed.upl.cs.wisc.edu   DEC 5000/125       xterms
unnamed.upl.cs.wisc.edu   HP C100       4 xterms
uplvax.cs.wisc.edu-   Vax 11/785    March-April 1995  

Holds lots of shit. Gus used to sleep in it.

"Mitch, Gulfie, and I rolled that thing from Engineering to CS, where we had to enlist all available hands to get it over the threshholds and up
the ramp. This was followed by the weeks spent gutting it - every screw unscrewed, every salvagable part used (the three phase concverter was given
to the CSL as a token of goodwill), and finally we played a great game in which people would climb into the nearly empty Vax, only to have the fans turned on. The pressure drop was unbelievable. Those of us with long hair had to be very careful." - hartmann

"Actually I remember the bulk of the work happening in one all-night orgy of
screwdrivers, leaving a deep pile of screws and wires all over the floor.

BTW, the main VAX cabinet had probably half of the 15 SBI slots filled.

It had 4 huge boards full of chips totalling 16M of RAM. We also scored that other 19" rack which originally held its tape drive.

Inside the rack was a never-opened board for hooking the drive up to a PDP-11 (it probably shipped standard with both a VAX and a PDP-11 interface I guess) which I still have.

I think it originally probably had some Fujitsu Eagles or similar 19" rack mount hard drives associated with it but we never had those.

Prior to us taking it the machine had sat for quite some time in a hallway in engineering while its old owners tried to scrape up the budget money to pay movers to scrap it. They were quite excited when we offered to take it
FOR FREE :-)" - mitch

"The external disk and tape cabinets did get taken - Besh still has them. Gulfie laid claim to them, left them when he left town, and then Besh and I had them at the Grotto." - hartmann

"It was a glorrious night of carnage. There was more time spent converting the vax to the bookshelf that is there now, but that is another (and much crappier story).

"Wonko, originaly against the entire thing, was amazing to watch, ripping memory boards out of there cage and taring the Vax to bits. Quite cathartic.

"Peeper, was also against the event, untell he was bistowed with a 12" extension cord. Which was perfect for what he needed. The Vax gave to everyone.

"There was a disk cabinet, and a tape cabenet (same size, different mounting stuff inside). The disk cabinet became my sterio / computer rack, and the tape drive became my dirty clothes hamper. Besh damn dirty sock hamper I've ever had.

"They became besh's when I pulled the ejection levers on my school career, and went to the bay of pigs. Besh moved them to Seatle, he can continute that story." - gulfie

vger.cs.wisc.edu Solaris 2.6 Ultra Sparc1 64MB RAM, 2GB disk June 1996 Sun Microsystems

"Vger is our primary "graphics machine" since it has such cool hardware like a Creator 3D graphics card and CrystalEyes viewing glasses."(9/95)

"I've got saved mail from 12/95 discussing the grant proposal. I'm pretty sure it arrived shortly after I left (around 6/96)" - mitch

"It was the first machine in the bulding to run Solaris 2.6, which gives a good idea of the time frame." - hartmann

wormhole.upl.cs.wisc.edu Switch Cayman GatorBox      

"Not sure of the exact model number but I guess it was a GatorBox CS or something. They were a ubiquitous little
appletalk<->ethernet standalone router from the early 90s. The box was actually lunatic's property. Cayman Systems later became Netopia." - mitch

"From what I remember, I really hated this thing because lunatic had
'fixed' the power cord with scotch tape." - kilroy

yar.upl.cs.wisc.edu   AT&T 3B2   1985-86  AT&T

First machine in the lab, Donated by AT&T.

"One of these was the original "yar" and was the prime nethacking machine when I started at the UPL (projects only, no games ;-) ). The other was probably scotty but was indisposed most of the time I was there - I think it was in someone else's office for a long time." - orn

yar.upl.cs.wisc.edu Ultrix DEC 5000/25  24MB RAM, 300MB Hard Drive    

Also the main mail server/web server.

"quotas were later bumped to 1M after we
scored more space, and quotas had to be checked by hand, since Ultrix didn't
have quota support." - hartmann

redshirt.upl.cs.wisc.edu