I can remember when I first met Gary in 1988 in the same building as the Cray Computer Apple had bought. Gary was so pumped showing a beta version of Photoshop. At this time Adobe had not bought Photoshop yet and it was written by the original developers, Thomas and John Knoll. Talking to Gary was always a pleasure.
The laser printer story is the story everyone wants to hear and you can find plenty of that on the web.
With Gary’s passing I have been trying to recreate in my mind what Gary was going through being the solo inventor of the laser printer. Reading the Computer History interview located here it reminds of what few know about Gary. https://archive.computerhistory.org/resources/access/text/20...
Gary is a physics major who knew how to write software. Gary’s speciality was optics which led to him getting a job in Bausch and Lomb’s division working on lenses for high end cameras for Hollywood. This story is explained in more detail on page 7 of the 53 page document above.
Gary being in the Rochester was also where Kodak was located would often say “The Sky is blue, and the grass is green. No matter what you do with colors don’t break those rules.” So even when he was at Xerox he thought about color. It is part of being an expert on optics.
What I think Gary realized in 1967 when he had the “eureka” moment of the laser printer is if he used a laser to paint a drum the optics totally changes to a more precise addressing. He knew he could use the laser for CMYK for color printing, but first step is to do monochrome printing.
On page 6 of the above document, Gary discussed the use of computers to improve the optics in lens. Why can’t computers use the optics in a laser printer to print anything you want?
In 1992, I took the bold step of leaving Apple Computer to go work at Microsoft on Windows 3.1, being program manager for all the Far East TrueType fonts. But even though I was now in Redmond I would still see Gary in Cupertino when I would visit family and friends in the bay area. I would visit Gary to discuss imaging, fonts, DTP, and color. Five years later in 1997, Gary says he thinks it is time for him to leave Apple. I said why not come to the Windows Imaging team. He says isn’t really wet there, and he would rather stay in Saratoga. Luckily Microsoft made an offer Gary could not turn down and now we were both in Redmond in Windows group. 2-3 years later Gary moved to Microsoft Research and I think that is the happiest I have seen him when at work.
Gary explaining his move from Apple to Microsoft is on page 38 of the above link.
One nice picture that I think will get out there soon is a picture of Bill Gates at Gary Starkweather’s retirement party. Bill is smiling and Gary is too. That’s a nice way to end your working career.
Thanks for these wonderful details. If you do get that picture online, please edit your comment to add a link to it! I've made it so that your post will stay editable. Or just email us at hn@ycombinator.com and we'll put it in there for you. I'm sure people would love to see it.
Gary was a wonderful person, an engineer's engineer, with an inquisitive scientific bent and warm sense of humor about and love of invention.
His move from Rochester to Xerox Parc in its earliest days was his "last chance" according to Xerox management. There he found kindred spirits who welcomed him and would up quickly loving him for his fearless approach to invention, no matter how difficult.
He was a great guy to work with and be with (one of the raft of things I did in the early days of Parc was to experiment with the design and making of high quality display fonts using an allout video system* that could reach the limits of video). We realized that it could barely do the much larger characters needed for the first laser printing system and rigged a coax from "the old character generator room" down the hall to Gary's lab to provide test pages for Gary's early experiments.
The story below about the use of Edmund Scientific "hobby" reflecting telescopes is more or less the way it happened -- except that the front part that says he was "annoyed" is not. That was not Gary's style; he just moved forward, and what happened is very similar in spirit to the computer researchers at Parc building their own simulated mainframe (MAXC) also in the first year because Xerox wouldn't allow us to buy a PDP-10 which was made by a competing company.
I also object to him being called "a badass" (I realize it is suppose to be a compliment, but it quite misses what really top talents are like in its attempt to suggest some kind of pop-culture teenage aggressiveness. Gary was an artist who simply transcended difficulties put in his way.)
He now joins another great engineer's engineer at Parc -- Chuck Thacker -- in our memories of truly great people who could do truly great things.
---------
* designed by Butler Lampson, Bill English and Roger Bates, and mostly built by Roger, with an excellent interactive font design program made by Ben Laws.
This guy was a badass. He Mcgyvered a 1 kilometer data link using LIGHT because he was annoyed his employer, PARC, opened up a new office up the road and he couldn't test his printer.
Excerpt from Dealers of Lightning:
------
Starkweather and Rider worked together on coordinating the SLOT and character generator until early 1972, when they were stymied not by a technical obstacle but one entirely man-made. This was the relocation of more than twenty of PARC’s seventy scientists up the hill to a building newly rented from the Singer Company and known as Building 34 (because its address was 3406 Hillview). The Computer Science Lab, including Rider, got bundled off to the new quarters while everyone else, including Starkweather, temporarily stayed behind on Porter. The move separated the two by a kilometer of real estate—too far to string an overhead line and, with the four-lane Foothill Highway in the way, impossible to link via a ground cable.
“The administrators said, ‘Don’t worry. You’ll be back together in another year,’” Starkweather recalled. “I said, ‘Great, what are we supposed to do in the meantime?’”
But one Sunday afternoon shortly after the move Starkweather got a brainstorm while sitting at home. He immediately jumped in his car, drove to Porter Drive, and mounted a stairwell to the roof.
Just as he had thought, he could take line-of-sight aim from where he stood to the rooftop of Building 34. He might not be able to span the distance by cable or wire—but he could do it by laser beam.
The next day he ordered four telescopes from Edmund’s for about $300 apiece. He and Rider replaced the eyepieces of two with low-power lasers and the others with sensitive photodetectors. They bolted one laser scope and one detector on each roof, aiming each at its complement across the way, to create a visible light data link.
The circuit worked flawlessly in almost any weather, even fog, although minor adjustments were often necessary after a rainstorm, when the weight of accumulated water made the roofs sag slightly.
“When SLOT was running I’d send a pulse of light up the hill to signal the character generator to send a line of data down to the detector on my roof, which would send it down to this laser and then to the printer,”
Starkweather recalled. “After all, we were only encoding ones and zeros. It was like sending binary data on a long wire made out of light, instead of copper.”
One book frequently overlooked when it comes to computer and internet history is The Dream Machine. It tells the story of J.C.R. Licklider, who was a psychology and computer science professor, and a director at ARPA. He had the vision for the "Intergalactic Computer Network" which became the Internet, and either directed or came into contact with nearly every project that created fundamental computing technologies.
The recent book by Brian Kernighan, "UNIX: A History and a Memoir" [1], was a fascinating personal memoir about the birth of Unix as well as how things used to work at AT&T Bell Labs.
Where Wizards Stay Up Late: The Origins Of The Internet
The Idea Factory: Bell Labs and the Great Age of American Innovation
The Chip: How Two Americans Invented the Microchip and Launched a Revolution
That last one is a little dry, I still found it interesting. It's about Jack Kilby and Robert Noyce. It's currently available for sale for $1.99 on the Kindle.
I'd add Eccentric Orbits and Skunk Works, both of which are about other fields (satcom and aero, respectively), but have similar themes of innovation and its mother, necessity. (And mother-in-law, politics.)
Along those lines, a personal favorite is Empires of Light, covering the race for electrification and lighting. It critically includes the role of George Westinghouse, which a lot of books leave out (typically focusing more on Tesla & Edison instead).
Dealers of Lightning is an order of magnitude more interesting than any other book in the genre I've ever read (mainly because of Xerox PARC of course). But for 90's stuff, Chapman: "In Search of Stupidity: Over Twenty Years of High Tech Marketing Disasters" is fun and really good as well :)
Seconded! In search of stupidity is a huge amount of fun. It can be a little mean-spirited/bitter at times, but it felt like a "Computer Chronicles: After Dark" kinda thing.
I owe that man a lot. In the 80's I co-founded a company that printed photographs on ID cards (if you have a European driving license, that's a long ago descendant of that project). First we did the local tennis clubs, then student travel cards and finally many more important ID cards. The whole thing was acquired by Johan Enschede, a specialty printing company (money, IDs, valuables).
Without the early commercial laser printers that whole project would not have been possible, for us it was an enabling technology.
I never knew about Gary Starkweather but I'm super grateful for what he did. There are countless other mostly invisible people from that era who created such enabling technologies.
Are there any stories you'd care to add for the audience here? HN readers love personal/inside accounts, and of course are comfortable with technical detail.
Easily the most trivial $1T inventions of all time! Spinning hexagonal mirror, laser, modulated by a serial port running at 14mhz, roughly. A 2 week hardware project and 3 month software project! I am ex Xerox Office Systems Division, 1984, I work at 3406 Hillview Ave, and heard it from work! It created a whole industry of repurposing hardware for new purposes like rockem sockem robots, I did it for our Internet radio service, 1994, at UBC ECE Dept. Just repurposed a portable AM/FM CD player and digitizer board.
Is there a more meaningful article that we can change to? One that includes the information that he died? (Edit: a more meaningful article appeared and we changed to it—see https://news.ycombinator.com/item?id=22002822)
With little "voop voop" noises in between when it moved without printing or fed more paper. And then you get to pull the tractor holes off the side of the paper. :)
I remember the smell of carbonless copy paper from an Okidata dot matrix printer hooked up to an AT&T 6300 (same as Olivetti M24) at my dad’s garage (shop). It’s like an earthy “new car smell,” probably cancer-causing VOCs, but it is what it is.
Did they rat-tat? I must have just missed those. I remember a lot of high-pitched _reeeee_ sounds, though certainly accompanied by a fascinating crunchy texture. :-)
rat-tat was impact printers (Daisy Wheel, or Character printer), rather than dot-matrix. :)
Good grief, that brings back memories. I had one of these as my first printer as a kid, hooked up to an Apple ][e. It shook the table so hard things would fall off. :D
> It shook the table so hard things would fall off.
One of my earliest "computer memories" as a kid was being in a Radio Shack in the early 1980s and seeing a large dot-matrix printer outputting a maze on greenbar paper; the printer sat on a cheap printer cart thing that Radio Shack sold, and as the head moved back and forth, it caused the cart to also "flex and sway" with the movement (again - cheap particle board and screws, bad construction, age - all factored in).
I think the printer was connected to a TRS-80 Model 3 or 4, IIRC.
At the time, I "owned" (parents purchased - and I still have it) a TRS-80 Color Computer 2; which is probably why I was in the store with my Dad, probably buying something for it. It was my first "real computer" (my first "computer" - and introduction to programming - was the controller of the Milton Bradley Big Trak; I still have it, too).
OK, nostalgia time. Growing up behind the iron curtain, my access to electronics was limited. I did own a ZX Spectrum in the early eighties but a printer was out of reach. So I hooked up a russian-built teletype machine to it through a homebrew serial interface. My "printer driver" controlled one bit of a Z80 PIO, and baud rate timing was done by instruction cycle counting. The machine had cyrillic and latin characters but some ASCII special symbols where missing. I fondly remember filing down the peripheral parts of the cyrillic "sh" character (Ж) type to obtain something quite similar to an asterisk. The noise this thing made can only be compared to a jackhammer, but it worked.
Well I also attempted building a punched card reader (interfaced to the same PIO) as the university I studied at had a mainframe with punchcards and I had written some Pascal code there which I would have loved to transfer to the Speccy.
Unfortunately I never got it to work reliably - probably the DC brush motor I used for transporting the cards produced just too incoherent movement, or perhaps the photo diodes caught too much ambient 50Hz noise from flourescent lights. Anyway I gave up, re-typed my most important Pascal programs and compensated my frustration by extending Hisoft Pascal, which just displayed numeric error codes on the ZX Spectrum, with actual on-screen text messages. Ha!
I used to work for a company that had a band printer. The characters were on a metal band like that in a band saw with duplicates of the letters around the band. The printer would spin that band around and make an imprint. It was incredibly fast and incredibly noisy. The printer had to be in a noise dampening enclosure.
IBM mainframes in the 70's invariably came with the IBM 1403 high-speed band printer, which pretty much represented the apex of band-printing technology. They printed 132 column, 66 line pages of EBCDIC text at a few seconds per page (depending on how many lines were blank, which characters were being printed, and what band you had installed). They were about the size of a washer and dryer pair, and made a pretty big racket. The computer operators (!) loaded them with special "green band" paper stock that came in boxes about the size of a 12-ream box of printer paper you'd buy at OfficeDepot today, but it was all one big continuous z-fold piece, with sprocket holes down each side for the printer to grab and power it through the paper path. Jobs were printed with special "separator pages" to help to operators find where to split the perforated printouts and deliver them to the right users' collection bins.
Then in the 1980s, IBM came out with the 3800 laser printer, which was just insane for the time -- continuous form printing at speed, like a newspaper printing press. I got to operate them (my department had two).
> The computer operators (!) loaded them with special "green band" paper stock that came in boxes about the size of a 12-ream box of printer paper you'd buy at OfficeDepot today, but it was all one big continuous z-fold piece, with sprocket holes down each side for the printer to grab and power it through the paper path.
Loading paper onto those printers wasn't an easy or fun task; there were typically two tractor mechanisms, one below the printing area (usually inside the sound deadening enclosure), and one above it (still inside, but under the "lid" of the enclosure). You had to get things lined up just right, and then snap the feeds closed on the bottom (while kneeling and looking - sometimes with a flashlight - at an awkward angle under the printer, inside the enclosure, with a large box of paper in your way), then advance the paper past the print area, and align it properly with the second set of tractors, then close those - then run a test job. If all went well (which it usually did), you had things loaded and could continue with your other tasks.
Oh - and the cleaning of those printers - so much paper dust, plus the "chads" from the holes (not all were cleared on a run of paper, so the tractors would pop them out and they would get lost all over and inside the machine).
> Jobs were printed with special "separator pages" to help to operators find where to split the perforated printouts and deliver them to the right users' collection bins.
Fun times to find your "job" and split it off to take back to your cubicle; also fun to find someone else's job and deliver it to them on the way back. Sometimes they'd be thankful or surprised, but it was also fun to do it when they had gotten up to go to the printer - they'd get there, couldn't find their job (but could find the ones "around" it), go back to their desk in a huff, then sit down and find the printout on their desk...hehe.
I kinda miss those days (but I don't miss having to build and format reports - talk about a nightmare and thankless task).
The first company I worked for that started my career as an SWE was a small mom-n-pop shop; the wrote software for insurance companies.
I applied for a position (any position; at the time, I was still in school) and they hired and trained me as an "operator" (because I didn't have a degree, they told me, they couldn't hire me as a programmer). They put me in a small cold room with the main computer they worked on (via serial terminals) - an IBM RS/6000 AIX box. There was also a vacuum column 9-track tape drive (which they taught me how to load and use), and a large Genicom 4440 printer.
IIRC, the printer was a dot-matrix line printer, so it could fire one row of "dots" out, eventually forming the characters as the paper moved. Even in it's enclosure, it was loud, but it could rip thru a box of greenbar in no time flat.
The printer was something they sold as VAR to their clients, because it could handle multi-part carbon paper forms, like checks. We'd write programs to print to these custom forms (holding the paper up to the lights and lining up the characters with the form positions - ugh!). So, when the client used a program that printed, they were also expected to check if the printer was in the middle of a printing job, and had the write paper loaded.
Sometimes, though, we'd "dial in" to their machines to test certain code and get feedback via phone; sometimes that code was for a printer (imagine trying to debug proper printing formatting and such over a phone - hair pulling!). Well, sometimes they wouldn't check the printer before we'd run the software (even after we told them to do so). That printer would take off...
...and then we'd hear a scream over the phone for us to "stop the printer" because it would be ripping thru the box of multi-part check paper (actual physical checks with carbon copy) with whatever we were printing (generally, not checks).
Clients sometimes did this themselves as well - and eat thru an entire box of checks before they knew it; multipart check forms were (from what I understand) stupid-expensive (because they were custom printed for an individual account, not too mention the special paper they used, plus the carbon copy capability).
Ah - those were the days.
And yes, I eventually got hired as a programmer (aka SWE), before I turned 19; in between me loading tapes or running print jobs for the programming teams and such, I would also play around with learning the software they used (a variant of PICK called UniVERSE), writing small games and other such things. One of the sysadmins gave me a copy of a book on PICK to help me along, and they were "monitoring" my account. Eventually, they started handing me small tasks - "fix this bug" or "code this custom feature" - to see if I could do it.
I could, I did, and I eventually was hired, sans degree, as a programmer - and started my career that I still am doing today, as an SWE, nowadays doing SPAs using Node.js, React, and other technologies (I've lost track of the technologies and software over the years I have known, used, programmed in, etc - suffice to say, I haven't done it all, but I've carved a big chunk out, certainly).
Oh - and that place was also my introduction to Unix (via AIX on that IBB RISC platform); by the time I left that job, I knew terminfo like nobody - I had to, because we were always needing to set up clients systems with it properly to handle their serial terminals, and do it in a way that was also usable with terminal emulators on PCs, and such - it felt pretty crazy at the time).
Yes, many mainframe shops replaced their line printers with early lasers (which were super expensive, the price only came down after a couple of years).
Yup. At Georgia Tech, they attached a massive Xerox 9700 series laser printer to one of the CDC Cyber 180s (it may have been more complicated connection). It was a few hundred square feet of floor space for the whole install, it could spit out vast amounts of paper, and was pretty finicky to keep running. But (most) everyone liked it better than the equally massive, vastly louder, band printers.
I usually buy brother because they didn't do that to me so far. But the last ones I bought are still working fine so haven't seen any new developement they did. Are they going the HP way now?
When the toner light goes on I just open up, shake it carefully from side to side and stuff it back in. It happily prints another 200 pages before it complains again. HL-1430 is my latest one, it just refuses to die. But all brothers we bought for work is also running way past their expected lifetime. The HP:s on the other hand breaks down all the time.
> Going back in time 50 years ago the resistance to commit to the laser printer for high performance office printing was so high Gary’s only option to finish his invention was to move from corporate to a supportive environment at Xerox PARC.
I think this is the kind of thing Sam Altman was talking about related to important not always being popular, and the need to persevere.
There's a rarer and more special consideration in a case like this as well. Taking out "inventor of the laser printer" adds more information than it removes. It adds the information that Starkweather is significant enough that HN readers ought to know who he is. If that information were hard to obtain, there might be a tradeoff here, but since it's literally one click away, I think the benefit of this edit outweighs the cost.
The laser printer story is the story everyone wants to hear and you can find plenty of that on the web.
With Gary’s passing I have been trying to recreate in my mind what Gary was going through being the solo inventor of the laser printer. Reading the Computer History interview located here it reminds of what few know about Gary. https://archive.computerhistory.org/resources/access/text/20...
Gary is a physics major who knew how to write software. Gary’s speciality was optics which led to him getting a job in Bausch and Lomb’s division working on lenses for high end cameras for Hollywood. This story is explained in more detail on page 7 of the 53 page document above.
Gary being in the Rochester was also where Kodak was located would often say “The Sky is blue, and the grass is green. No matter what you do with colors don’t break those rules.” So even when he was at Xerox he thought about color. It is part of being an expert on optics.
What I think Gary realized in 1967 when he had the “eureka” moment of the laser printer is if he used a laser to paint a drum the optics totally changes to a more precise addressing. He knew he could use the laser for CMYK for color printing, but first step is to do monochrome printing.
On page 6 of the above document, Gary discussed the use of computers to improve the optics in lens. Why can’t computers use the optics in a laser printer to print anything you want?
In 1992, I took the bold step of leaving Apple Computer to go work at Microsoft on Windows 3.1, being program manager for all the Far East TrueType fonts. But even though I was now in Redmond I would still see Gary in Cupertino when I would visit family and friends in the bay area. I would visit Gary to discuss imaging, fonts, DTP, and color. Five years later in 1997, Gary says he thinks it is time for him to leave Apple. I said why not come to the Windows Imaging team. He says isn’t really wet there, and he would rather stay in Saratoga. Luckily Microsoft made an offer Gary could not turn down and now we were both in Redmond in Windows group. 2-3 years later Gary moved to Microsoft Research and I think that is the happiest I have seen him when at work.
Gary explaining his move from Apple to Microsoft is on page 38 of the above link.
One nice picture that I think will get out there soon is a picture of Bill Gates at Gary Starkweather’s retirement party. Bill is smiling and Gary is too. That’s a nice way to end your working career.