Starting in 1997 I have much more precise records of events, so from here I will be denoting things by specific months.
November 1996 was so much fun that I flew out to Buena Park California to attend ConFurence 8 with the UniGraphix team on January 16-19, 1997. It too was a lot of fun and we decided that we should do this every year.
Rumors begin to circulate at Intelligent Electronics about the company being sold.
On March 28th (Saturday) the rumor becomes true; Ingram micro is going to make an offer to buy I.E., and if Ingram Micro does buy the company, I’ll have a choice of moving to Memphis or finding another job.
That Monday I take a week’s vacation and fly out to Maryland where Scott agrees to put me up while I check the job market out there. It looks good, and Scott tells me I’m free to use their spare room as long as needed. So, I fly back to Denver and start the process of moving.
Over the next month I sell my VW Beetle and my Porsche 914, box up and ship anything I deem essential to Scott, and sign over the trailer to Ken.
While closing down operations at I.E., my supervisor hands me two trays of 200Mhz 512K Pentium Pros when I’m at my car one evening, stating these are my severance being as the “assholes in corporate got their payday and no one outside of C-level was going to get anything”. I start to respond and he just holds up a hand and walks away.
200Mhz Pentium Pros sold for about a thousand dollars each in 1997, which means I had about $16,000 in CPUs. So, I send a bunch of them to Scott who sells them in D.C. and uses that money to acquire a lease on a 40 acre farm in Rhoadesville Virginia.
I land at Dulles International where Scott picks me up, and over the next two weeks I help them move from Maryland to our new home in the back woods of Virginia. Everything I’d shipped from Colorado to Maryland is now in Virginia, and I quickly get all of my systems back up and running… The problem is the new farm is so far out in the boonies the best we can get is a couple of phone-lines and 14.4kbps on those… So, I start figuring out how to fix this.
I quickly land a job at Amerind in Alexandria where Scott already works.
Scott and I are working at Amerind in Alexandria and Johnna, Scott’s wife, is working for “American Photobooth” (APBI) also in Alexandria.
APBI made these Photo Booth things that you used to see everywhere, and they had locations in a lot of neat places; the world trade center observation decks, the Air and Space Museum, the Grand Canyon, The Smithsonian Museum, Disney Parks, and several high-end malls all over the US. At the time I’m told they have 750 Gen1 sticker making machines pulling in over 10 million a year.
Not bad for stickers…
I wind up doing things like converting the coast guard’s internal systems at the Pentagon from Banyan Vines to Novell. Or flying out to Ohio to work on a database project for the US Army.
This is the month where Scott lets me borrow their Ford Taurus, said Taurus doesn’t have the tax sticker in the window, and I get pulled over for it in Alexandria (we don’t have tax stickers in Colorado, so I’m clueless) – this starts the epic many-year tale to get Virginia to let me have a driver’s license again.
All told it was Scott and Johnna, their two children Kevin and Kim, myself, three horses, two dogs, two cats, and assorted chickens on a 40-acre bit of paradise in the deep woods of Virginia.
Scott and I are approached to see if we can fix the issues APBI is having with their Gen1 “Sticker Station”. I get turned loose on the spaghetti code in the Gen1 and have it literally re-written in a weekend and create “Gen2”.
The owner of APBI, Sam Attenberg, is impressed and sets up a meeting with Scott and I and asks “What could you do with an unlimited budget?”
I start going on about live chromakey of a full-motion subject based on the stuff I used to do on the Amiga with a video toaster, and using a PC with network connectivity to allow for the rotation of new backgrounds without having to actually go to the machine… Scott chimes in with miniaturized dye-sublimation printing technology being demoed by Fujitsu.
We leave that meeting with employment contracts and a virtually unlimited budget to create the Gen3 “Sticker Station” … A system that, by the time we’re done, will briefly become the most advanced vending machine in the world.
The hitch is we will have one year to build it starting January 1, 1998.
Given the numbers involved (and the prospect of not having to drive 150 miles a day round trip to Alexandria), Scott and I tender our resignations from Amerind.
With things looking like the farm would become a hardware and software development lab and we’d all be working from there, I work on getting a T1 installed at the farm… This entails about ten miles of fiberoptic work though the forest from Orange and an install cost of $8000. When done the connection will cost a mere $1800 a month with a 24 month contract. I figure that into the budget for the project and it comes up as a no-brainer, so I flip the switch.
I buy myself a car: a lemon yellow 1967 Mustang with the high performance 289 in it. It’s fairly ‘hotrodded’ when I buy it, but in a lot of wrong ways — so I start fixing it.
Scott buys Johnna a car as well; a much more practical 1996 Geo Prizm.
For November 18 – 22, Scott and I are sent to the “International Association of Amusements Parks and Attractions” (IAAPA) show in Florida to support the Gen2 machines and do research on what other vending machine companies are up to.
Scott makes connections with a cabinet manufacturer and other parts vendors while I do a couple of live bug fixes on the Gen2 codebase before the show.
I did get to see the show though. IAAPA is basically the tradeshow for Willey Wonka – it’s amazing.
December is spent creating the broad strokes for the Gen3 system in the sun room of the main house in Rhoadesville.
Scott and I put on our game faces by taking two weeks in January to drive from Virginia to California and back to run UniGraphix one last time at “ConFurence 9”.
We take Johnna’s Prizm for this trip and it turns out to be a truly miserable little car for any trip longer than about twenty miles…
CF9 takes place from January 15th to the 18th in Buena Park again. We arrive a day early and walk Disney Land as well.
We return to the farm by way of Colorado – stopping by my old trailer in late January. I pick up a lot of the stuff I’d left behind in my rushed move six months ago.
The T1 install is finished for my birthday in February and I get the touchscreen UI and some of the core control portions of Gen3 running by mid-February.
In March Scott starts spending time in St. Louis working with the cabinet maker to get the cabinet design and mechanical parts of it into a prototype state.
Conveniently, a fellow Scott knows from FurryMuck – Ken McCleaft (Pegasus) – is working on selling his ISP in Tennessee and might be looking for something new to do. He’s an amazing low-level programmer and circuit designer, and we offer him something new to do.
Ken drives to St. Louis while Scott is there and helps out with driver code for the mechanicals.
On March 25th, “Widget” is born.
The above photos were taken with an Apple QuickTake 150…
Frencheska was a retired sulky racer and, according to her owners, was barren – so having no more use for her it was either find a new home or off to the knackers. Scott and Johnna picked her up to give her a nice retirement home, but once moved she and Thing decided to give it one more go – and 11 months later Widget was born.
Widget was awesome; all of the easy going of her Clydesdale dad with the spunky nature of her Standardbred mom.
April – August 1998
Work-wise we collectively bust our asses for the next five months; Scott working on the mechanical aspects of the system like case, touchscreen, camera tilt mechanism, and internal wiring and layout. Ken and I work on the code – me from there on the farm and Ken from Tennessee when he’s not visiting.
We run into a few issues here and there and have to work some deep magic to create the technology we need for Gen3. One issue was with the cellular modems we were using for the systems to ‘call home’. It was a revolutionary setup from little company called “Research In Motion” (RIM) who made the 801D ARDIS radio modems we were using. This is what allows the Gen3 to be the world’s first fully networked vending machine.
You might know RIM by another product they created, “Blackberry”.
Another issue was a mysterious bill validator reboot issue every time the StickerStation project manger would use the demo system there in Alexandria… It took a couple of weeks to figure out that the combination of his thousand dollar leather loafers and his thousand dollar leather office chair, mixed with his thousand dollar silk suits would imbue him with about a billion volts of static charge when he stood up… And as soon as he touched the machine the bill validator would reboot.
This was fixed by isolating it with a restive drain to ground that we named after him – the anti-Hugh circuit…
Fun fact: The Gen3 machines used Dallas Semiconductor “iButtons” for access – these are 1-wire devices the size of a watch battery that hold a per-device serial number that can be read just by touching a sensor. In addition to the ten user-defined access buttons per machine, the base code had three hard-coded iButtons that would instantly grant complete access to any Gen3 machine on the planet…
I still have that last iButton on the list:
Roni is involved in the graphic design of the cabinet as it nears completion in July and comes up with the logotype and mascot. Hugh is responsible for the garish purple and yellow color-scheme.
Ken comes out and visits a few times, and plans to move to the farm next year (1999).
I’m also still doing art on occasion as well. This one is from August of 1998:
We’re 99% done with Gen3. The mechanicals are in production and the code is waiting on a driver from Sony.
Jeff, Roni, Scott and I take a breather and head to ConiFur NW on September 4-6, 1998. The convention is run by Dan (Flinters), an old friend, so it’s nice to go bum around with him for a weekend and decompress.
According to my code archive I sent the final 1.0 code to production on September 23rd.
Gen3 systems are in production! In the first two weeks about a hundred of the $8000 machines are running. About twenty of those are outside of the US.
Jeff, Roni, Scott and I head to SciCon 20 in Virginia Beach November 6-8, 1998 to unwind.
There’s a small tweak to the code taking it to 1.01 and sent off to production on November 9th.
By Thanksgiving there’s a rumor that APBI is being prepped for sale to Foto Fantasy up in New Hampshire.
I run the email server for APBI at the farm, so it’s elementary to BCC certain emails in the server to figure out if the rumors are true — they are. We discover that Hugh has been prepping the company for acquisition during the Gen3 project, and was actively trying to thwart us to prevent an increase in company worth.
But, because we were successful anyway Scott and I are listed as required company assets as part of the sale. Foto Fantasy must keep us on for at least a year to manage the Gen3 systems.
While this is interesting, I’m frankly tired of having my work sold out from under me.
The 1998 holiday season passes uneventfully and we look warily towards 1999…