Today’s story is complicated, as it involves multiple threads converging at a central moment during the history of computing.
Also, this image via NASA of the “Pillars of Creation” (aka M16, Eagle Nebula, NGC 6611) is relevant; we’ll get back to it later.
THREAD 1: Edmund Berkeley
You might think the first personal computer in history would be a moment of deep significance. Not exactly.
There is, first off, what even “counts” as a personal computer; if we bypass all that argument to the minimum, the first personal computer was Simon, which could not be purchased in a store but where plans for making one were printed in the magazine Radio-Electronics starting in October 1950. It was the brainchild of Edmund Berkeley (co-founder of the ACM, Association for Computing Machinery) and cost north of several hundred dollars to make (in 2025 currency, at least $3500). It only had 32 bits of memory.
Based on the price, having the plans be meted out in a magazine, and having the entire thing require self-manufacture: it did not have much historical impact. It was, however, only Berkeley’s first attempt bringing something resembling computing to the home, as he had a much better price point with the Geniac (co-designed with Oliver Garfield) of $20 in 1950s money.
Eventually, this sort of “physical computer” was made even cheaper at $5 with the Digi-Comp I, a finite state machine powerful enough it also has been dubbed “the first personal computer”.
Certainly both devices could be programmed as if they were real computers; both devices included guides to programming Nim, an obligatory rite of passage for any 50s/60s computer.
From Supplementary Wiring designs for the Geniac, 1955.
These later devices had relatively widespread use — the Digi-Comp I had an estimated 100,000 in sales — unlike the Simon which remained a novelty. All these products were the logical outgrowth of Berkeley’s attempts to reach the masses with computing, which started even before Simon, in 1949, with his book Giant Brains, or Machines that Think.
These new machines are important. They do the work of hundreds of human beings for the wages of a dozen. They are powerful instruments for obtaining new knowledge. They apply in science, business, government, and other activities. They apply in reasoning and computing, and, the harder the problem, the more useful they are. Along with the release of atomic energy, they are one of the great achievements of the present century. No one can afford to be unaware of their significance.
This is all relevant for today’s story…
THREAD 2: Joseph Weisbecker
…as one of the people who read the book was Joseph Weisbecker, where (according to his daughter Joyce), “he saw for the first time what an electronic computer could do, but, more importantly, how it worked. Binary logic, flip-flops, switching circuits – very simple elements combined in subtle, clever ways resulted in surprisingly sophisticated behavior from a machine.”
Joseph Weisbecker was only a teenager when he read the book; by age 19 (in 1951) he had built his own Tic-Tac-Toe machine. During the 50s he joined RCA, not only working on chip and memory design projects but making lower-end educational toys (akin to the Digi-Comp) intended to bring computers to the masses. He had a special contract with RCA that let him sell his inventions to outside companies, like Think-a-Dot (sold by E.S.R, same company who made the Digi-Comp).
He was in the odd position of being involved with a vast number of the RCA computing initiatives all the way through the 1970s but also being ideologically opposite in a way; RCA cared mostly about large business where Weisbecker kept the flame alive for smaller computing. He put forward a proposal for mini-computers in 1960 (a level between giant mainframes and personal-computers) that was ignored (when this market emerged with the DEC PDP-8 in 1964, it became huge). Where this really became clear is when he went on to make his own personal computer system called FRED, developing it from his home in New Jersey.
Picture of FRED, aka Model 00. Source.
Knowing RCA’s apathy to the idea, he didn’t even bother pitching FRED (which eventually became the basis of the 1802 chip) until after RCA had a collapse of their mainframe computer business in 1971; according to Joyce Weisbecker he’d already been working on it for two years on the side. Later in the 70s he bypassed RCA entirely and wrote a series for Popular Electronics in 1976 and 77 that laid out the design for a personal computer, the Cosmac Elf, with the full 1802 chip design. This computer was essentially the fully-developed version of the FRED.
Via Jim Kearney‘s recent build of a Cosmac Elf.
It isn’t like the 1802 would have gone to waste without the personal computer connection; the chip was the first CMOS (Complementary metal–oxide–semiconductor) chip and consequently had low noise and low power consumption. Both are essential aspects in spacecraft and so an 1802 has found its way both in the Galileo probe launched in 1989 (giving a close-up view of Jupiter and its moons) as well as the famous Hubble space telescope launched a year later (see image at the top of this post).
Mosaic of Europa, from the Galileo probe.
The 1802 also found its way into the short-lived RCA Studio II console, off and on the market in a year. It is notable for having none other than Joyce Weisbecker (as quoted earlier) implement some games, making her one of the first female programmers in videogames.
For our purposes, the important thing to take away is that despite RCA being heavily corporatized, the Cosmac Elf was in a way “liberated” from it, as part of the movement to bring computing to the masses. Speaking of bringing computing to the masses…
THREAD 3: Tom Pittman
…we now need to move from New Jersey to California and the Homebrew Computing Club of Menlo Park.
Are you building your own computer? Terminal? T V Typewriter? I/O device? or some other digital black-magic box?
Or are you buying time on a time-sharing service?
If so, you might like to come to a gathering of people with likeminded interests. Exchange information, swap ideas, talk shop, help work on a project, whatever…
They were founded quite shortly after the launch of the Altair computer, another candidate for “first personal computer” (more properly here, first commercially successful personal computer). While plenty of hobbyists had already made their own systems through arcane means, here was a computer kit that seemed to break things open, 256 bytes of default memory and all–
If it was even possible to get a set. The makers, Micro Instrumentation and Telemetry Systems (MITS) out of New Mexico, had a story featured in the January 1975 issue of Popular Electronics; it was a Hail Mary by the founder, Ed Roberts, after they got their calculator business destroyed by Texas Instruments and friends (it cost more to make a ship a calculator than it did to buy it). They were not prepared for the influx of orders, some made to products mentioned that didn’t even exist yet.
Steve Levy’s book Hackers mentions how one person (Steve Dompier) drove all the way to Alberqueue and “had left the office only after Roberts had given him a plastic bag of parts he could begin working with, and over the next couple of months more parts would arrive by UPS, and finally Dompier had enough parts to put together an Altair with a serial number of four.”
This atmosphere of just touching the edge of the technological revolution was when the Homebrew Computing Club kicked off in the garage of Gordon French, co-founder of the group with Fred Moore. They both had links through the People’s Computing Company which for a brief time had one of the only working Altairs at the time (sent directly to the director, Bob Albrecht, who sang its praises in their newsletter) and that Altair landed in the garage for the first meeting.
We arrived from all over the Bay Area — Berkeley to Los Gatos. After a quick round of introductions, the questions, comments, reports, info on supply sources, etc., poured forth in a spontaneous spirit of sharing. Six in the group already had homebrew systems up and running. Some were designing theirs around the 8008 microprocessor chip; several had sent for the Altair 8800 kit.
Even with hardware there was the problem of software; coding in assembly was quite error-prone and slow, and the PCC already had good experience with using BASIC. Salvation seemed to come in the form of a version of BASIC made by Paul Allen and Bill Gates (with Monte Davidoff) which was sold by MITS…
Via the Computer History Museum, paper tape provided by Bill Gates.
…but it started selling at $397, and eventually landed on (after price reduction!) a price of $200, far more than what many of the computer enthusiasts were used to paying for software ($0).
A “Caravan” also known as the “MITS-Mobile” was travelling from town to town demonstrating the wonders of the Altair, and in June 1975 the Caravan arrived in Palo Alto, California.
The Homebrew Computer Club visited and there was an (early, buggy) version of Allen/Gates/Davidoff BASIC running on one of the computers, expanded to 4K. “Someone” swiped a paper tape which turned out to be a copy, which eventually landed in the hands of Steve Dompier, and from there this copy spread to the community.
All this led to this to the “Open Letter to Hobbyists” printed in the newsletter of the Homebrew Computing Club, February 1976, written by Bill Gates, outlining how it seemed nearly everyone had the pirated BASIC, and given the numbers of how with royalties paid, their time spent developing the product was “worth less than $2 an hour”.
Many hobbyists groused about this; a follow-up letter in a later issue opined that perhaps Gates was directing his ire at the wrong people, and that
I’m sure that if I were MITS, I’d be chuckling all the way to the bank over the deal I got from you.
Some of the Homebrew Computer Club decided the best response was to make their own BASIC. Tom Pittman, in particular, had been a member since the first meeting, and he was one of those who had made his own computer prior to the Altair — using the Intel 4004 chip, with only 4 bits as opposed to the 8-bit chips that came after (the 8008 which was used in the Canadian MCM/70 and the French Micral N, and the 8080 used in the Altair). He took up the challenge. While not the first to do so, Tom Pittman wanted to try charging for it, but a nominal fee only:
Gates was moaning about the ripoffs, and people were saying, ‘If you didn’t charge $150, we’d buy it.’ I decided to prove it.
Tiny BASIC was a variation of BASIC developed to be as simple as possible to fit in small-capacity computers; Pittman made his BASIC conform to the Tiny BASIC standard (and then added in some extra just because he could), and importantly, only charged $5. Rather than for the Altair this was for a different chip (the Motorola 6800) and he eventually sold the interpreter to a company for $3,500 (while retaining the rights to sell to $5 to hobbyists).
The 6800 wasn’t his passion, though, nor the MOS Technology 6502 he also wrote Tiny BASIC for (even though the 6502 showed up in everything from the Apple II to the BBC Micro to the Nintendo Entertainment System). The chip he truly loved was the 1802.
…the microprocessor is even more elegant than Joe Weisbecker intended. This microprocessor is so good that even RCA is not really aware how good it is. The 1802 is a complete and symmetrical microprocessor.
He liked the computer so much he eventually wrote a short course in programming for the commercial-kit follow-up, the Elf II. When he wrote his First Book of Tiny BASIC Programs in 1981 he clearly had his beloved Cosmac Elf in mind.
The Elf II was the commercial-kit version of the original Cosmac Elf design, sold a year later for $100. Source.
And now we finally get to why the threads all tie together, and why they are here on All the Adventures. Pittman’s book of Tiny BASIC programs includes Tiny Adventure, source for a full adventure game.
The instructions specifically mention Crowther/Woods Adventure “provided the inspiration”, but this game has significant differences. It feels very alien to play.
INTERLUDE
Some quick notes if anyone else wants to try this out in an “authentic” way. You need a Cosmac Elf emulator; I used Emma 02. Under File -> Configuration -> Load I picked Netronics Tiny Basic -> Serial I/O and then bumped the clock speed up slightly before starting the emulator (I used 6.5, any farther and BASIC has trouble loading).
I then took the source code for Tiny Adventure, copied the whole thing to clipboard, and pasted it to the emulator screen. This is very slow. I let it run in the background for 15 minutes before it was finished, and then played. There’s some “save” buttons on the emulator which I would assume makes the process faster thereafter but I was getting corruptions trying to get them to work, so I had to cut and paste every single time I was starting the emulator anew (which given this game took me multiple days to beat … well, let’s just say I feel like I was getting the authentic 70s/80s experience).
I tried finding another BASIC interpreter that would work, but even the one marked as TinyBASIC compatible gave me issues. I think there are some unique aspects to the Elf implementation of BASIC that haven’t been ported over. (Despite there being a “standard”, there are quite a few variants as discussed here.) I have no doubt there’s ways to clear up the issue but playing on a historical emulator gave the 1981 flavor, and so worked for my purposes.
THE GAME
Tiny Adventure is set in a fantasy world. There is no quest given. (“…unlike the original game, TA keeps no score; you play for the pleasure of exploring, or set your own goals.”) There isn’t even a specific treasure goal mentioned. We are just told to wander.
So far that’s unusual but not shockingly so, although if you study the instructions above carefully, they also specify you are only allowed to carry one item at a time in your hands. You can store items in your knapsack, but you have to juggle items and put them in and out again if you are trying to use something that’s stored.
Commands are not given in a regular parser fashion. Initial letters are used instead of words. (Usually. Often the game gets fussy if you go past one letter, sometimes it doesn’t.) There’s Take, Putdown, Keep (put in knapsack), Go, Look, Inventory, Help, Open, Close, Attack, Drink.
Look and go do not work like you normally expect. This game involves relative direction. Not only that, it involves relative direction where the paths you travel along don’t necessarily go straight back and forth. This is absolutely unprecedented so let me clarify.
In a game we played recently, The Maze, while it had relative direction, it also gave a first-person view of a maze so it wasn’t confusing. Still, it meant that rather than going north, south, east, or west, the directions were generally left, right, and forward (with “A” for “turn around”). The “tank controls” that happened in the late 90s for some games like Resident Evil were a similar concept.
There’s also been relative direction with text-only games but it has been much rarer. Mystery Mansion had the inside of the mansion start out with relative directions until you found a compass; you’d see in the room description what was to the left, forward, and right, and if you turned to the right and went forward, you would expect to return the way you came by turning around 180 degrees and moving forward again.
Map from Mystery Mansion, showing turning right and going forward, followed by turning around and going forward. The design on this part of the map is in a grid to make this a little easier to manage. It still was a hassle to play.
Tiny Adventure has relative direction, and one-way exits, and directions that turn. It took me a very long time to work out what was going on. An example from the very start of the game:
Essentially, what happened above is
a.) I went forward from the starting room. (G F = “go forward”, and the game requires you to use letters like that)
b.) I used LOOK to turn to the left twice. (L L = “look left”, which both turns the player and describes what is ahead of them, it took me a long while to even realize LOOK doubled as a turn command)
c.) I went forward again, landing me in an entirely different room (G F), except it doesn’t appear to be that way and the only way to realize this issue is to rotate around all four directions and spot something is different.
You might also expect the turning-passage to rotate the direction the player is facing, but no, if you’re facing “north” you’ll still stay facing north no matter what when you arrive at the next location. In the end this makes things easier to map but it was difficult for me to realize this was how the game was working. (You can imagine a player sashaying sideways as their head stays fixed in the same direction.)
To make a map, upon arriving at a new room I would “L R” (look right) four times to get a description of what was in each direction, notating all four on the map. To move around, in order to be careful, I always looked in the direction I wanted and did G F (“go forward”); while you can go back, left, etc. and essentially skip a step, I found it extremely easy to get disoriented if I did any shortcuts.
Perhaps the issue could be mitigated with dropping items? Alas no: there is, for example, a rock to the “west” of the start, where the same rock is in two rooms at once. I think the idea is the rock is equidistant “between” them so the rock could be taken at either place, but goodness the game is already confusing enough as it is. For extra inconsistency, there’s also items you can also see while looking in a particular direction.
I tried my best to map the outside but I honestly gave up trying to make it accurate and just made it accurate enough for me to get through. The really important object on the outside is the sword, which you can use to whack at the two enemies (dragon and troll).
The “fall” drops you in a dark place and I never got around to experimenting with the lantern there.
There’s a cottage with a locked door; the way to get inside is to open the window.
I gave up here on any kind of tracking of left/right. Only the connectivity is accurate. I was making full spins every time I stepped in a new room.
The starting room (bedroom) has a chest with keys, as shown in an earlier screenshot. You can also go in farther to find a flask of “dragon’s tears” and a “lantern”. (The dragon’s tears turn the player invisible. I never found a good place to use them, but since this game is a language tutorial with no set goal the author likely was just tossing in what he thought was neat.)
Down some stairs is a wine cellar with a locked door; using the keys from the bedroom on the door leads to a tunnel. (I’m making this all sound straightforward, but I didn’t find the keys right away because of the look-relative-position issue — I didn’t realize until very late it applied to a chest that could hold an object.)
At the far end of the tunnel is a dragon. If the dragon is sleeping it is easy to dispatch with a sword. According to the source code the dragon can be awake (and wander between rooms) but I never experienced that.
Part-way up the tunnel is another locked door leading to a “troll’s den”. There is a “maiden” in the den that you can rescue.
Another exit in the tunnel leads to a cave with an axe (presumably an alternate weapon — again weird for a regular game but not for a tutorial one), and then out to an island with a boat. When I reached the island the game crashed.
Again, the game gives no specific goal; I figured killing the troll and rescuing the maiden was good enough for me, but the book gives some interesting suggestions:
Can you rescue the maiden and her jewels without killing the troll (leave him locked in his den)? What is the least number of turns to do this?
There are two ways into the dragon’s lair, but you cannot get back out by one of them. Can you find it?
Can you discover what the “magic dragon tears” do for you? Can you undo it? Can you get more, after you use them up?
This is a hard one: If you get lost in the forest, can you get out? Hint: You need to head off in the direction of the ravine, but you must get your bearings before you get lost. Crashing through the underbrush of the forest tends to get you turned around, and you usually end up going around in circles.
Once you solve the forest problem, you might want to take the maiden on a moonlight boat ride around the island. Watch out for the riptide!
How many turns does it take you to visit every place? There are 17 places in all, counting both ends of the tunnel as one place. Usually you can tell you’re in a different place if the scenery is different, or if something you Putdown is no longer visible.
The troll will under certain circumstances, wander around on his own. Can you coax him into the bedroom? Harder yet, can you lock him in the bedroom without the maiden being there to look on?
The relative-movement system is so much like wading through sludge I’m not going to make an attempt at these, but others are welcome to try. That does leave one open question I am intrigued by…
How did this happen?
…by which I mean, why did what is essentially tutorial code in a book end up being designed like it was from an alternate universe? (Not just the movement style, but the lack of goals, and the inventory where you can only hold one item at a time and need to specifically say you want to stuff items in your backpack.) I think there’s some flavor of The Hobbit here (made by a quartet of computer scientists) where seeing how systems play out was considered more interesting than any kind of destination. Regarding relative movement, though, there’s a strong hint in the book:
One common complaint I’ve heard from several people who played this game is that it does not follow standard Euclidean geometry. That is not true. A map (on a flat piece of paper) was drawn of the area before a single line of code was written, and it is faithful to the map. What happens is that in crawling, climbing, or otherwise moving from one place to another, you got turned around, and the way out may not be behind you. Or, the divisions between places (such as rooms) may not fall on cartesian boundaries. This is true to life, and the game is consistent.
That is, the author was trying to create a modeled universe, again with an engineer/computer science bent, and if the player doesn’t have a compass, of course they would be confused and turned around sometimes! And of course you realistically wouldn’t be holding that much, just like a real person! This game was written with the realism-model approach without the consideration that because we are being conveyed the model via text, no matter what happens there is an element of unrealism anyway. Certainly my stumbling around a tunnel felt very different than any kind of being lost in real life I’ve ever experienced; this game is what would be like if you completely bypassed all thought of player convenience. As he states on his own webpage, he “doesn’t play games”; this was a true outsider work.
Which is interesting! I feel like I stumbled across a microcosm of innovation that started and ended where it landed. Pittman is still around and working in computers, but after one more game (Grand Slam Tennis for the Emerson Arcadia 2001) he went on to teach and write about compiler design, work on an automated Bible translation project, and finally (in the present day) teach programming to middle and high school students.
Coming up: I’m taking a break for the remainder of June! Sort of. I have a number of behind-the-scenes things to finish, including some posts that won’t show up until a future date (mystery!). I can say for now the game I have next on my list when I return is a graphical game for Apple II.
CREDITS NOTE: Very special thanks to Kevin Bunch, who is working on a book on RCA and graciously shared some of his research. If you’d especially like to hear him talk about the RCA Studio II at length, he did an interview with the Hagley Museum you can find here.
Renga in Blue 
