ObHack: (Well, kinda) After years of curiosity, I finally tried to decode
the holes in the acoustic (asbestos?) ceiling tiles. The holes seemed to make
up a binary code. No luck though. Just gibberish, but I tried dammit! Maybe
it's in EBCDIC. Anyone else get results?
Well, one exists at dr...@csh.rit.edu, but I think I've heard of others.
What I want to know is how it works?
>ObHack: (Well, kinda) After years of curiosity, I finally tried to decode
>the holes in the acoustic (asbestos?) ceiling tiles. The holes seemed to make
>up a binary code. No luck though. Just gibberish, but I tried dammit! Maybe
>it's in EBCDIC. Anyone else get results?
Try rot13 :)
Shawn Ledbetter || Even a mistake may turn out to be the one thing
sled...@nyx.cs.du.edu || necessary to a worthwhile achievement.
Apple II Infinitum! || -- Henry Ford
finger co...@cs.wisc.edu (Coke machine at U of Wisconsin)
finger @coke.elab.cs.cmu.edu (Coke machine at Carnegie-Mellon U)
finger m...@coke.elab.cs.cmu.edu (M&M machine at Carnegie-Mellon U)
finger barg...@coke.elab.cs.cmu.edu (Visual map of CMU coke machine)
finger dr...@csh.rit.edu (Soda machine at Rochester Institute
ObHack: Getting 1.5 Amperes out of a 1-amp 7805 VREG by keeping its body
submerged in water. Yeah, I know. EE hacks don't count, but this VREG
is running my homebrew TTL tester complete with 8 inputs/8 outputs,
reset output, and variable speed clock output (manual to 1 MHz).
Interesting note: fingering @pepsi.elab.cs.cmu.edu gives the same stat
results of the Coke machine.
Lon Lowen Jr. | Wayne State University
Netcom Online Communications | Detroit, Michigan, USA
lll...@netcom.com | (lll...@mts.cc.wayne.edu)
>Some time ago, I remember reading about a coke machine on the internet. You
>could 'finger' it to check it's inventory. Does anyone have it's address?
The best known is at CMU (try finger co...@cmu.edu). This is apparently
their second machine, after the first died a some years ago. I've also
heard that MIT had one, but couldn't find any confirmation beyond two
machines called coke and pepsi. They were not dispensing machines of any
The Computer Club at the University of Western Australia also has a coke
machine. As one of the chief perpetrators, I can tell you something about
It all started way back in '92 (remember that year?). We had heard the
rumours about the CMU and MIT machines, and decided that this was something
we just had to do. A few phone calls to coke, we found someone sympathetic
to our cause, and explained what we wanted to do. It took a bit of
explaining (these people had never heard of the internet), but when they
understood what we were babbling about, they were (somewhat) enthusiastic
about the project. A few rules were bent, and we got an old machine on
Hack hack drill solder. Our coke machine got a 68000 based board and power
supply and relay drivers bolted to the insides. Some quick test software
was writted (a very simple monitor with commands to drop a can from slot
n), and a terminal was attached. In case you're interested, the controller
board is an 8MHz 68000 with 16K EPROM, 64K SRAM, and a 68681 DUART. A few
I/O ports were attached later (74LS373 and 244).
Months passed. People got told 'The hardware works. Come on - we need
software'. The software didn't happen. Phototransistors were added to sense
the 'sold out' lights on the front. The software didn't happen.
Blinkenlights were added (a bit-scrambled grey code. the scrambling changes
every 10 seconds or so). The software didn't happen.
Then at last, the software was written. It still uses the old test monitor,
but the UNIX end (running on our Sun) has accounting and various other nice
features. You can still get a free can by unplugging the Sun, sticking a
terminal into the plug, and typing 'D6' (slot 6 is where the coke usually
lives. We number them from right to left).
Anyway. finger co...@gu.uwa.edu.au will give you very little information on
what is in the machine at the moment, and is probably wrong (the
optosensors are playing up). We have a program called 'dispense' which will
drop cans and remove credits from your account. 'dispense coke' searches
for a non-empty slot containing coke, and sends the appropriate command to
the machine. 'dispense' without any parameters displays an evil curses menu
and lets you select the drink from that.
ObHack: Apart from the coke machine you mean? Umm... How about the Z80
board with programmable clock speed? A few unused address lines were fed
into the clock generator. Jump to a different shadow of the ROM, and the
machine changed speed. Andrew wanted a speaker, microphone, and PLL on it
instead. Whistle to get it started.
For the humour impaired: Insert a :-) after every third word
I did a similar thing once with a 7805 mounted on an indented heatsink
-----------\ ---|*****|___ /--------
The heatsink was getting up to the boiling point of water, [ which was OK
as a 7805 operates best at such temperatures :-) ], but I decided to
try to cool it down a little...by pouring good old (fluoridated) water
into the indentation. Worked fine for a month or so, though each time the
water quality got a little less due to miscellaneous junk (dust,
dissolved salts, etc), until one day when I overfilled it, and the water level
reached the leads. By this time the water was dirty enough to be conductive...
and the leads are pretty close, so resistance was low, and the water
was hottest at that point...whoosh. Bubbles of hydrogen, oxygen and steam.
ObScureBug: On the same project, I needed to take power in and power out.
Well, +8V in just goes straight to the 7805 pin. similarly +5V out.
However, the (free floating) pins are getting a little crowded now...I
don't want to connect both to the middle 7805 pin. So, one connection
to the middle pin, one screwed onto the heatsink (the little heatsink
inherent in the 7805 is internally connected to ground). However, the
graphics card I had it connectged to was behaving intermittently. So, get out
the (borrowed) CRO, connect it to the circuit ---- and, by Murphy's law, it
works perfectly. OK, give up for the day in disgust. Turn the modified TV-cum
monitor off --- the traces on the CRO suddenly go mad, showing the fault.
Turn the TV on...the CRO traces are normal again. Remove the CRO, the
TV picture goes.
WHY DOES MY CIRCUIT ONLY WORK WHEN THE CRO IS CONNECTED AND THE TV IS ON?
Investigate the output drivers. No luck. Get an inspiration. Check the
screw on my power supply --- it is loose, and with the CRO disconnected there
is a 1V drop accross it. When the CRO is connected and the TV is on,
there is a connection Power supply -> CRO -> mains -> TV -> Graphics card
that bypasses the loose screw.
Gee life was fun back then when I would spend two days to save 50c.
This reminds me of a trick we played on the users of a PDP system I used to
work on. Some friends and I had successfully hacked ourselves a privileged
account on the system and decided to create a new node on the network. This
node did not physically connect to anything but it started appearing on the
System Status reports and some of the more observant users tried to find out
what it was.
We wrote a small daemon that checked for people trying to assign the port or
write to it, and when it found someone doing this is would send back a random
message from a list of replies. This went on for a couple of weeks with really
obsure replies like "banana skins" and "mouldy sandwiches", and then we finally
revealed what the port was supposed to be by giving the port an alias.
And what was this mystery port ??? None other than the computer room bin :)
RealLifeObHack: Finding just the right place to stick the screwdriver in my
mates' electricity meter to stop the wheel turning ;)
:: INTERNET: d...@sound.demon.co.uk ::: Disclaimer: I have no connection with ::
:: FIDONET: VF10 at node 2:254/14 ::: any other users at sound.demon.co.uk. ::
Even more interesting note: When the server was first rereleased a
couple of months ago (it had been out of commission for the previous
5 years), about 3/4 of the requests to the server in its first week
of operation came from OUTSIDE cmu. Apparently, someone posted
about it to alt.folklore.computers.... It'll be interesting to
see what the traffic for this weekend looks like. Reading the log
files might be an interesting way to see who reads this group :-)
Obhack: collecting the money from the aforementioned soda machine and
actually getting the books to balance.
-- Jay Sipelstein
Well, you better refill the thing! It's out of Dew!
So, how the heck do these things work, anyway?
ObHack: making a script to make color PostScript images from an ascii
database file. I've got my tapes (with info, song titles, etc) listed
in an ascii file, so it goes through and makes up a color PS tape
insert... Fun fun fun...
- lan...@wpi.wpi.edu - - Redundant in Nu-Wave Black -
- Cockroaches determined to survive -
- Microwave me - watch me come alive -
Son, all the pretty, intelligent, healthy # Ken Johnson, AIAI, Edinburgh
young women are taken. It's a basic rule of # tel 031 650 2756, fax 650 6513
the universe, and if you don't like it, go # Seeking job: C, Prolog, Unix,
somewhere else. -- my dad 1906-1992 # Dos, training, documentation.
: I did a similar thing once with a 7805 mounted on an indented heatsink
Wouldn't it have been easier to do what everyone else does - put a resistor
in parallel with the 7805 which handles the minimum load current. This
doesn't work too well if you have a large range of current to supply, but
normally this isn't the case.
ObHack: Building a GAL programmer (someone elses design), getting some GALs,
only to find the software for the GAL programmer required GALs with a
slightly different fuse map. Oh well, with no way of getting the data over
Easter, and nothing better to do, I programmed it bits at a time and
observed how the GAL behaved. After a few hours I had worked out where all
the fuses had moved to (only the order of the fuses had changed in the
"architecture control" line). I then changed the software (source was
provided!) so now I can program these effortlessly. Does this count as an EE
or a CS hack?
I was driving a video board which could have current changes during
vertical retrace, (50Hz in Australia), so I wasn't sure how much the
current changed. Besides, why bother? I would then just have to dissipate
the heat from the resistor instead...and I didn't have any 5W, heatsinkable
>ObHack: Building a GAL programmer [...]
Can I borrow this thing? My wife has been nagging me a lot lately, and
I'd like to reprogram her not to do it anymore...
(sprinkle with smilies to taste)
ObHack: How about a medical hack?
We've had a rash of patients on ventilators lately, to the point that
we've had to dig out some old models we thought were (finally)
retired. One of them developed an air leak, which, you might guess,
would have a detrimental effect on the patients respiratory support. I
traced the air leak to the humidifier (dry air from a respirator will
screw up the lungs big time) and of course, no parts were available to
fix it. I kludged it with a couple tourniquets wrapped around the worn
seal and held together with an IV dressing. It worked long enough for
the courier to get us the parts we needed (as well as a couple more
respirators) and the patient is fine.
ObHack: Got a pile of 360K PC disks you want to read? Only working computer
around that old Commodore 64? No problem! Well. Almost. Someone else
already came up with a hack to double the bit rate on the 1541 drive,
so it is capable of reading 360K disks. But how do you read a double
sided disk on a single sided drive? Simple. Turn the disk upside-down. Now
all the bits come off backwards. My 1541 now has a little switch in the top
to reverse the direction of the motor. Getting a GCR drive to read MFM
disks is, of course, a software problem.