spp > blog > party_like_it's_1981

1,001 Things to Do With Your PicoCalc! (Ideas not included)

Like a lot of people I have a particular weakness for well-marketed electronic toys.

The particular route to my wallet these days involves any combination of two or more of the following: interesting form factors, retro computing, and Linux. Because of this, I've thrown more money than is really necessary at products made by ClockworkPi which tend to put Raspberry Pi PCs into small, cyberpunk-styled cases. Their most recent offering is the PicoCalc, which looks like a scientific calculator with a Raspberry Pi Pico microcontroller inside. The PicoCalc differs from ClockworkPi's other offerings in that the stock model does not offer any wireless connectivity. I'm not sure how other people reacted to it, but glancing through the specs I thought it sounded like it might be able to emulate an early, 8-bit home computer, but faster and with a lot more memory. So, y'know, I had to buy it.

One of the non-problems with the PicoCalc is that unlike other ClockworkPi machines, it was delivered in a reasonable timeframe. They use preorders to fund the development and production of their devices which has meant in the past that delivery has taken around ten months (as opposed to their promised delivery time of "90 business days"). It's annoyingly long to wait, but it always given me plenty of time to think about what I'm going to do with the device whenever it shows up.

This time I was promised 30 days, and it showed up in twelve, so I had barely enough time to consider possible applications for another portable computer that I don't really need.

ClockworkPi's forums are always buzzing with developers and hardware hackers who are doing interesting things, and the PicoCalc already has a NES emulator, dialogs of LISP and BASIC, and a stripped-down version of Unix. I have played with all of these things and think I'll probably spend the most time messing around with MMBasic (it being similar to the TI-BASIC I cut my first programming teeth on), but...what am I actually going to do with it?

Based on forum posts, it's clear that lots of people had the same initial thought I did: the PicoCalc looks like Douglas Adams' description of The Hitchhiker's Guide to the Galaxy. How might such a thing be implemented? Where would you get the content? Well, as it turns out, the entirety of Wikipedia can be downloaded and fits onto a reasonably-sized SD card... As useful as that might be, I thought of a different source of information: Project Galactic Guide.

Nobody remembers Project Galactic Guide, apparently, because it's a part of the early, lawless days of the consumer Internet, back when we still capitalized Internet. PGG was a very loosely organized attempt to create a real Hitchhiker's Guide to the Galaxy: anybody could submit their own entries as a "researcher," and the entries were stored in a regularly-updated database which you could download and browse with special software that was available for a variety of platforms. I was aware of PGG when it was active, and I kick myself every time I consider that I could have been involved. The only barrier to entry was my own insecurity; I was afraid to reach out to the people who maintained the project.

PGG was most active in the early to late '90s, and I was totally enthralled by the idea when I came across it in 1995 (or so). Remember, this was significantly before the launch of Wikipedia, and information at the public library was only as current as this month's periodicals and limited to mainstream areas of interest. The idea of a monolithic repository of vetted, user-submitted information was very attractive. I was less excited when I discovered that PGG's entries were classified as real, semi-real, and unreal. All three classes contained a lot of entries trying unsuccessfully to sound like Douglas Adams. The semi-real and unreal classifications were added by popular demand but were full of trashy entries that were simply not as funny as their authors hoped (and therefore not worth my time). The scope of the real entries wasn't as broad as I'd have liked (this was a small, amateur attempt, after all) and contained a lot of unhelpful, semi-real information.

Still, I kept up on on PGG's updates for several years. I'm not sure when I lost track, but coming back to it more recently I found quotes more than a decade old from PGG founder Alex McLintock promising that PGG is still alive and well. I suspect the truth is that it was limping along as a nostalgia piece by then, and is now, as I said, almost entirely forgotten.

Google will still serve you a smattering of PGG pages, but very little of this preserved information is helpful. A lot of those websites are full of broken links, and what still exists paints a sorely inadequate picture of what the project used to be.

Anyway, the PicoCalc is a great candidate for implementing a real-life Hitchhiker's Guide, but in considering how I'd accomplish it I've argued myself to a standstill. The problem is compromise, or rather, the fact that I don't think a compromise can be reached. In a compromise you're supposed to get something better than what you started with, and the work this would require combined with the limitations of the PicoCalc make the whole project feel like a step backward when compared with using the supercomputer in my pocket for on-the-go reference.

How would I go about programming the thing? My plan is to provide a searchable, alphabetical list of entries. When an entry is chosen, then its text is read into memory and displayed to the user who uses the keyboard to navigate up and down, follow "see also:" links, or return to the menu. The most obvious way to accomplish this would be to store the each menu item and each line of entry text in arrays, however, an array eats up a lot of RAM and there's not enough RAM to store the whole list of entries. No sweat; I can see how a rolling array that jettisons old lines of text in order to read new ones might work. The real problem is displaying the entries themselves, and this, as Hamlet might say, is the rub. The screen can only display 40 columns x 25 lines of plain text. Images are not out of the question, but I think it probably is out of the question to display them inline with the text. That problem, combined with the fact that I'd have to sacrifice a lot of Wikipedia's formatting means I'd be making a crappier version of Wikipedia than the one I can access at any time on my phone. In the case of PGG, the entries consist of 80-column lines of text and include a fair amount of ASCII art. Centered text and bullet lists would have to be re-formatted and ASCII drawings would have to be redone! Do I want to do all this work just to have a tongue-in-cheek reference guide that won't impress my less technically-inclined friends? No, I do not.

And that, really, is the hurdle I can't clear: the limitations of the PicoCalc's display mean that the most obvious application I have for it won't do justice to the data it would need to display. So, I guess PGG and Wikipedia are both out of the question.

A third reference work occurred to me, and that's the Jargon File, colloquially- (though not better-) known as The Hacker's Dictionary. Doing this would at least get the basic engine up and running, and I could adapt it to Wikipedia or PGG in the future. But what are the chances that I'll want to do that while I have a smartphone? And what are the chances that I'll get rid of my smartphone? I have no plans to do either of those things.

Programming projects, for me, are a way to unwind and escape from other things I should be doing. I think it's significant that I've done more programming since the beginning of the pandemic than in the decade before it started; I work from home now, and a Linux terminal is just a flip of my KVM switch away. But that's fun in a way that wouldn't apply to designing a computerized reference work that I'll never use. So, right now my PicoCalc plays Hangman, Hammurabi, and Super Star Trek, sometimes Wumpus.

Maybe I could write a port of Wordle.