© Mark Ollig
In 1991, a group of academic researchers worked at the University of Cambridge, England, in a computer science study lab called the Trojan Room.
These researchers spent most nights busy on their computer keyboards writing and coding software programs.
Their computers operated over the university’s in-house local area network (LAN).
Like most coders and programmers, they consumed generous amounts of coffee.
Of course, having a freshly-brewed pot of coffee available at all times is of the utmost importance to most of us.
Walking up quite a few stairs to get to the coffee machine was required by the researchers working on the building’s lower floors, while those on the upper floors walked down several stairs.
They carried their empty coffee cup during the journey up and down the stairs, not knowing whether there would be any coffee available in the pot.
As you assumed, there was one coffee machine serving the whole building.
Understandably, frustration would set in whenever a researcher with an empty cup in hand discovered an empty coffee pot.
One practical solution would have been to install more coffee machines, but we must remember; the university and academic researchers were living within a nominal budget.
The researchers gathered, pooled their talents, and came up with a resourceful plan.
They decided each of their computer screens would show the amount of coffee remaining in the coffee pot in near real-time.
This innovative plan developed into what became known as XCoffee.
The Trojan Room contained several shelving racks containing new computer servers.
The researchers discovered a spare Acorn Archimedes computer server installed with the X Window System protocols, thus providing the name for XCoffee.
This computer contained a gray-scale video-frame grabber circuit card.
The frame grabber is a device that takes a picture, captures still-frame analog images, and saves them digitally.
The first frame grabbers could only grab and save one still-frame digital image at a time.
Using a retort stand for holding scientific equipment, the researchers mounted a video camera onto it and pointed the lens towards the coffee machine, zooming in on the coffee pot.
They then ran all the cabling under the floor from the coffee camera location to the Acorn Archimedes computer server in the Trojan Room.
Paul Jardetzky coded a server software computer program that would run on the Acorn Archimedes computer.
The computer’s video-frame grabber circuitry captured live still-frame images of the coffee pot about once every three seconds.
Quentin Stafford-Fraser, a researcher in the Trojan Room, worked on an ATM (Asynchronous Transfer Mode) communication switching network. He wrote the code for the “coffee client” software computing program.
Stafford-Fraser’s client software program operated on the researchers’ computers in the building connected to the university’s internal data network.
The client software then communicated with the Acorn Archimedes computer’s newly-written server software.
The client program displayed the latest image of the coffee pot onto a corner of the display screen on a researcher’s computer – so they saw a near real-time picture of the coffee pot and, thus, the amount of coffee in it.
It took about a day for the programmers to get the XCoffee camera software program fully up and running over the university’s LAN.
The program operated over an MSNL (Multi-Service Network Layer) protocol designed for ATM (Asynchronous Transfer Mode) networks.
At the push of a button, updated images of the coffee pot appeared inside a small insert display on the researcher’s computer screen.
This solution was accepted by everyone because, according to Stafford-Fraser, the coffee pot filled rather slowly, and since they were using a greyscale photo capture, the images looked fine.
The researchers working at the university could see the amount of coffee remaining in the coffee pot on their computer screen.
They could now see when the coffee was brewing and confidently leave their computer station with a coffee cup in hand, knowing there was coffee waiting for them in the pot.
Regularly updated 128-by-128 greyscale pixel images of the coffee pot broadcast onto the internet beginning in 1992 became popular within the online community.
A 1992 snapshot from a MOSAIC (one of the first web browsers) of the XCoffee webcam is at https://bit.ly/3x9gISs.
A night light was placed near the coffee pot so that while Cambridge slept, online users worldwide could still monitor the updated images of the coffee pot.
“The lights in the Trojan Room aren’t always switched on, but we try to leave a small lamp pointing at the coffee pot so you can see it at night,” a message on the XCoffee webcam displayed.
By 1994, remotely monitoring the Cambridge coffee pot on the internet went viral all over the world.
The same year, a reporter from a local BBC radio station interviewed the XCoffee university researchers to discuss “The Cambridge University Coffee Machine.”
Here is the audio from the 1994 BBC Radio Cambridge interview: http://bit.ly/2ooUg4G.
You can see a photo of the famous coffee pot taken from one of the researcher’s computer screens at https://bit.ly/3n0eRut.
Sadly, the XCoffee webcam broadcast over the internet ended on Aug. 22, 2001.
A University of Cambridge link with the last message about the XCoffee webcam exists at http://bit.ly/2o6mKP7.
I corresponded with one of the researchers who built the XCoffee webcam, Quentin Stafford-Fraser, who still lives in England. I expressed my admiration regarding his ingenuity and that of the other researchers.
“Thanks Mark, and best wishes from this side of the pond!” Stafford-Fraser replied to me.
 |
| The actual image of the coffee pot has it appeared on one of the researcher’s computer screens. |
 |
| Snapshot taken from a MOSAIC browser in about 1992. |