The Dover

Xerox PARC didn't really believe in writing journal papers. They built stuff. If they wanted you to learn about something, they'd build it for you and dump it on your doorstep. Around 1980 or 81, they built three huge laser printers out of Xerox's largest copying machine. Each one came with a dedicated minicomputer to organize the jobs. One of them showed up one day at the AI Lab and was duly installed on the 9th floor. It was magic. To this day (1996), I have never seen a faster laser printer. You stood next to it and 80 pages/minute came out in a steady ka-chuh, ka-chuh, ka-chuh. A little more than one "ka-chuh" a second.

I was taking a French Lit. class at MIT and had to write my first paper. Gregor showed me how to use the LispM to generate Dover output and I spent the whole evening tarting up my 10-page paper with fonts, italics, and bold type. I was convinced that my humanities professor would never have seen such professionally-formatted work from anyone. He would be awed.

The paper came back a week later. I got a D.

-- Philip Greenspun, January 4, 1998

Answers

I was a wee lad in one of the Learning Research Groups at Xerox Parc, back in 1976. I distinctly remember seeing an *even older* version of the FAX machine. I only saw the receiving end, but suffice it to say, it was truly hilarious: It consisted of a spinning drum with a piece of paper strapped to it, with some kind of (I guess pretty fast) printing "needle" tracking slowly down the length of the drum, printing the bits on the paper as it went. Apparently, the needle wasn't quite fast enough or the PLL was a bit funky, because I examined the output from this thing one time, and some raster lines were crazily skewed relative to one another. Thank heaven for semiconductor lasers and those good 'ole spinnin' mirrors, eh!

-- Geo Homsy, January 4, 1998

It was nice, very fast and a pain in de ass! That's why I removed my phone from Inquire. Like most newborns,the Dover needed constant care & feeding... Don't look such, from "sum ah dem" 'round here, when that Thesis or Paper is due..

-- TYrone Sealy, January 4, 1998

In the olden "Dover" days, you used to truck up to the 9th floor to pick up your output and find yourself standing on-line w/ dozens of folks from across the street - I guess they didn't HAVE printers over in Bldgs 36 and 38!

-- Rebecca Bisbee, January 4, 1998

Before the Dover there was the XGP (Xerox Graphics Printer), a 180 dpi laserprinter that printed to continuous rolls of 8 1/2" paper and then cut pages with the resounding "chunk". It was a lot slower and the print quality wasn't quite as beautiful, but the contrast to contemporary line-printer output was even more astonishing. My wife, an English major at BU at the time, printed a paper for a Dickens class and was asked by the professor how much of the paper she had actually written, and how much was written by the computer!

The XGP technology was based, incidentally, on Xerox's LDX (Long Distance Xerography) system, an early attempt to build very expensive fax machines out of a scanner at one end and what is essentially a laser printer without the resterizer at the other. Someone realized that you could replace the scanner by a computer-based synthetic source of bits and have a high-quality computer output device, and the laser printer was born.

-- Peter Szolovits, January 4, 1998

Between 1974 and 1980 or so I was part of the Real-Time Systems group on the 4th floor. We were the "have not'" of LCS since we didn't have a PDP-10 and weren't part of Multics (which was winding down). Instead we had PDP-11's running UNIX. Although there was an XGP up on the 9th floor this whole time, it was a pain to use since our machines weren't on the Arpanet. So we made do with daisy-wheel printout from troff, and later a wet-process electrostatic printer (Varian, I believe). While the resolution was OK (200 dpi), the paper just felt carcinogenic and the liquid toner will spill all over the place. Preparing papers for class was a two-step process: print something on the Varian, then photocopy it so you could pretend it came off the XGP.

The toner for this was really nasty, and the machine used some sort of filter with "Fram" on the side that Topher Eliot (stuck with keeping the beast running) would occasionally get replacements from in autoparts store. Topher once complained to a Varian technician how hard it was to get toner out of his shirt.

"It's easy to get the toner out," said the tech.

"How?" asked Topher. "I've got lots of otherwise new shirts ruined by that stuff."

"Use a pair of scissors."

By the time the Dover arrived we had a share of a VAX and put it on the Chaosnet and were able to tap into the Dover. This was a huge improvement. At that time, the Dover was enough of a secret that you could start your print out, take the elevator to the ninth floor, and by the time you arrived, your print-out was ready. It did encourage bad habits of the form "Where did I put that UNIX manual? Oh, well, I'll just print another..."

-- Tom Teixeira, January 8, 1998

The worst thing about the XGP was that the paper curled as it was coming out! (especially at the end of a roll, I think). When it curled enough, it started rolling onto the floor, so you might end up with 50 unsorted pages spewn in front of it if you didn't watch them print out....

I guess the XGP was also responsible for the flowering of typographic exuberance in Carl Hewitt's Actor papers....

-- Stavros Macrakis, May 24, 1998

Around 1980, I worked for MIT's left-wing paper, "The Link". The text for it was all printed out on the Dover, in narrow columns of 12-point Times Roman. The output was then taken back to our squalid darkroom on the other side of campus, and I reduced it in size by about 25% to disguise the low resolution of the Dover. The columns were then cut and pasted (literally) to form the final paper.

I learned an important lesson working on the Link: since I was usually the only person there the night before deadline who admitted knowing how to use the enlarger and copy camera, I was the guy who ended up stumbling around in the darkroom while everyone else sat around the office, smoked dope, and hacked layout. Moral: do not become the local expert in a boring but essential activity. To this day I disclaim all knowledge of SQL.

-- Carl Feynman, October 8, 1998