Recently I posted about a new HTML-based CoCo 1/2 emulator named Varispeed. The author has since set it up online so you can try it out from a web browser. This may be easier for you than downloading the files and opening them in your browser locally. Check it out:
Well, not at all a vocoder, but it did sorta remind me of the way the Daleks sounded on Doctor Who.
During my years of isolation in deep East Texas in a tiny town with a population of 225, I had the only CoCo around. In fact, I may have had the only home computer that wasn’t at the school. I remember discovering (or maybe reading about this in Rainbow) how you could use the AUDIO command to create some weird effects as you played a tape through the cassette drive (I did not have a disk drive at this point).
Knowing what I now know about BASIC optimization, I present this bit of code:
0 AUDIOON:AUDIOOFF:GOTO
Stick a tape in the tape drive, press PLAY then type “MOTOR ON” to get the tape rolling, then RUN.
I learned I could open the tape door on my Sears (I think) cassette recorder and then hold down something inside and click REC-PLAY to put the tape player in RECORD mode. I could then speak into the microphone and it would send the output to the earphone jack into the CoCo, allowing me to use it like a microphone. I could turn up the volume and talk into the microphone and hear the distorted audio come out of my TV speaker.
Ever tried this?
Behold, in all its VHS glory:
This screen grab came from the intro to a video project I worked on with a high school friend that year. I had only captured the intro titles, and not the rest of the content, but a few frames of me doing an intro were at the end of it.
I was living in Pineland, Texas at the time (population 1111 if I recall). While I had been active on the BBSes in Houston before moving to East Texas at the end of 1984, there were no BBSes in the small towns I lived at during my East Texas high school years. Thus, no modem. Just a CoCo, a TV set, and floppy disk drive. I am not sure what else I had for my CoCo back then other than joysticks and some software.
The wooden desk was a kit I put together in late 1984. That desk made the move to Iowa with me in 1995 and went into storage temporarily while I was looking for a place to live. I remember hauling it from storage to my new apartment, and trying to drag it down the sidewalk towards the building. It feel apart. So, while it made it with me to Iowa, it never made it to the new computer room.
I also see Mickey Mouse ears on my TV, and now I want to find those. I recall buying some as a 20-something “adult” visiting the parks during work trips to California, but did not recall having some when I was in high school. Neat.
More fascinating posts like this to come… You have been warned.
Or perhaps 1985. I moved from Houston, Texas to a tiny town in East Texas in 1984. This VHS framegrab is from a “music video” I made back then. Ah, fun times.
MTV never played my music video. :-(
Even by the late 1980s, most households did not have camcorders. Most households did not have a home computer. Even fewer had both. Thanks to my dad giving me some old gear as he upgraded, my household did.
Behold, a VHS video intro I did in 1987 on my Radio Shack Color Computer 2. The CoCo only had RF (TV) output and my VHS recorder was an editing deck and did NOT have a TV tuner. Since the CoCo did not have audio/video out, and the recorder did not have RF inputs, I just pointed the camera at my color television and recorded the titles that way.
Oh yeah. State of the art.
I did my original post as a joke, showing a funny (to me) mockup of what a retro BACKROOMS video game on the CoCo might look like ;-)
But if we really are going to try to make some silly game, we need to understand what the game is based on. First, BACKROOMS comes from a photo that was shared online and folks started telling stories about it. You can read the history here:
https://en.wikipedia.org/wiki/The_Backrooms
And here is that original photo and the description that started it all:
Folks did not know the origin of this photo, at that time, but have since figured it out. It was a photo from a former furniture store (which explains the odd walls) with poor color correction (making it yellow) before that location was turned into a HobbyTown. Apparently.
And the description of this image read:
“If you’re not careful and you noclip out of reality in the wrong areas, you’ll end up in the Backrooms, where it’s nothing but the stink of old moist carpet, the madness of mono-yellow, the endless background noise of fluorescent lights at maximum hum-buzz, and approximately six hundred million square miles of randomly segmented empty rooms to be trapped in
God save you if you hear something wandering around nearby, because it sure as hell has heard you”— Anonymous, 4chan (May 13, 2019)
And thus, the fandom began. Folks wrote stories, created fan wikis, and even made video games based on the lore that was developing around this photo and paragraph.
One such creator was the Kane Pixels channel on YouTube where a teenager began creating his own take on the BACKROOMS lore. His first video was uploaded on January 22, 2022 and has had over 84 million views. The story is that it is a “found footage” video taken from a camcorder that recorded in 1990. You an watch it here to get the idea:
This led to a series of other first-person camcorder type videos recorded at different points in the timeline of this version of the BACKROOMS. The lore Kane Parsons (the real name behind the Kane Pixels channel) created the mysterious corporation A-SYNC Research.
The popularity of his particular take on this meme ended up having the A24 movie company approach him about a theatrical film based on his version. That was the BACKROOMS movie that opened in 2026 with the highest opening release of any A24 movie to date, and making Kane Parsons the youngest director (at age 20) to have a film make that kind of money ;-)
So obviously, any game based on this lore would feature a yellow maze with a first person view of someone wandering around in it.
But not on a 1980s CoCo ;-) On this era hardware, we could convert it into a top-down view of the yellow maze, similar to the 1980 Atari VCS game Adventure. Longtime readers of this blog may recall I have played with doing this type of game on the CoCo in the past by literally translating the game level maps into the CoCo’s 32×16 text screen using the semi graphics block characters:
And, frankly, if someone has not already done an Atari VCS/2600 BACKROOMS game (possibly as a hack to Adventure itself), I would be surprised.
Maybe we can dig into this together and figure out what the game will be. Basically, a player wandering through an endless yellow maze, and entities chasing you… Not much to it. But maybe we can spend some time and flesh out the concept a bit into something fun to play, at least for a few minutes.
To be continued? Maybe…
Sometime in June 2026, I saw a random post on LinkedIn where someone said they created an emulator for the Texas Instruments TI-99 computer using AI. Could AI really create a full emulator that actually worked? And if so, could it do that for the CoCo?
Turns out, yes, and yes. A CoCo emulator, mostly crafted by AI, has been released. It runs in a web browser! You can find the code here:
https://github.com/star-fs/varispeed-coco2
I cloned (downloaded) that code to my Mac and then opened the index.html file:
Wow! It actually works. Scrolling down shows some neat diagnostics stuff:
That “Active Disassembly” shows what 6809 instructions are running at the moment. It’s fun to watch.
I have not done anything useful with it yet, but thank you star-fs on GitHub for your efforts in creating this!
A few years ago, both the 1986 and 1987 versions of the Tandy Color Computer 3 GIME chip were de-capped and photographed under a microscope. This work was done thanks to the efforts of Sean Riddle, Erik Gavriluk, and Roger Taylor.
When the first CoCo 3 emulator was released by Jeff Vavasour, he would have had to recreate the CoCo 3 font manually by looking at the screen and counting the pixels. Over the years, we have seen updates to various emulators in regards to the fonts as folks figured out some pixels were wrong here and there.
I believe we know the full font of the CoCo 1/2 MC6847 VDG so those should be accurate these days, but we really never had confirmation of the font in the CoCo 3 until these de-caps were done.
Here is a low-resolution image of the 1986 and 1987 GIMEs:
Quite different! I do not know what the contents of the two ROM areas are, but I suspect the font data is part of it. In the GIME archive (available on my Dropbox or as a torrent), there is an image of the bits from those ROM areas:
And also from the archive is the font data represented in a text file that Roger Taylor (or one of the others) extracted:
----------
* * *
* *
*
*
*
* *
* * *
*
----------
* *
* *
* *
* *
* * *
* * *
----------
*
*
* * *
* *
* * * * *
*
* * *
----------
*
* *
* * *
*
* * * *
* *
* * * *
----------
* *
* * *
*
* * * *
* *
* * * *
----------
*
*
* * *
*
* * * *
* *
* * * *
----------
*
* * *
*
* * * *
* *
* * * *
----------
* * *
* *
*
* *
* * *
*
----------
*
* *
* * *
* *
* * * * *
*
* * *
----------
* *
* * *
* *
* * * * *
*
* * *
----------
*
*
* * *
* *
* * * * *
*
* * *
----------
* *
* *
*
*
*
* * *
----------
*
* *
* *
*
*
* * *
----------
* *
* *
* * *
* *
* *
* * *
*
----------
* *
*
* *
* *
* * * * *
* *
* *
----------
*
*
* *
* *
* * * * *
* *
* *
----------
*
*
* * *
* *
* *
* *
* * *
----------
* * *
* *
* * * *
* *
* * * *
----------
* * * *
* *
* *
* * * *
* *
* *
* * * *
----------
*
* *
* * *
* *
* *
* *
* * *
----------
* *
* * *
* *
* *
* *
* * *
----------
* * *
* * *
* * * *
* * *
* * *
----------
*
* *
* *
* *
* * *
* * *
----------
*
*
* *
* *
* *
* * *
* * *
----------
* * *
* * *
* * *
* * *
* * *
* * *
* * *
----------
* *
* * *
* *
* *
* *
* *
* * *
----------
* *
* *
* *
* *
* *
* *
* * *
----------
* * *
*
* * *
* *
* * *
*
* * *
----------
*
* *
*
* * *
*
* *
* * * *
----------
*
*
* * * * *
*
*
* * * * *
----------
*
* *
*
----------
*
* *
*
* * *
*
*
* *
*
----------
----------
*
*
*
*
*
*
----------
* *
* *
* *
----------
* *
* *
* * * * *
* *
* * * * *
* *
* *
----------
*
* * * *
* *
* * *
* *
* * * *
*
----------
* *
* * *
*
*
*
* * *
* *
----------
*
* *
* *
*
* * *
* *
* * *
----------
*
*
*
----------
*
*
*
*
*
*
*
----------
*
*
*
*
*
*
*
----------
*
* * *
* * *
* * *
* * *
*
----------
*
*
* * * * *
*
*
----------
*
*
*
----------
* * * * *
----------
*
----------
*
*
*
*
*
----------
* * *
* *
* * *
* * *
* * *
* *
* * *
----------
*
* *
*
*
*
*
* * *
----------
* * *
* *
*
* * *
*
*
* * * * *
----------
* * *
* *
*
*
*
* *
* * *
----------
*
* *
* *
* *
* * * * *
*
*
----------
* * * * *
*
* * * *
*
*
* *
* * *
----------
* * *
*
*
* * * *
* *
* *
* * *
----------
* * * * *
*
*
*
*
*
*
----------
* * *
* *
* *
* * *
* *
* *
* * *
----------
* * *
* *
* *
* * *
*
*
* * *
----------
*
*
----------
*
*
*
*
----------
*
*
*
*
*
*
*
----------
* * * * *
* * * * *
----------
*
*
*
*
*
*
*
----------
* * *
* *
*
*
*
*
----------
* * *
* *
*
* * *
* * *
* * *
* * *
----------
*
* *
* *
* *
* * * * *
* *
* *
----------
* * * *
* *
* *
* * *
* *
* *
* * * *
----------
* * *
* *
*
*
*
* *
* * *
----------
* * * *
* *
* *
* *
* *
* *
* * * *
----------
* * * * *
*
*
* * *
*
*
* * * * *
----------
* * * * *
*
*
* * *
*
*
*
----------
* * *
* *
*
*
* * *
* *
* * *
----------
* *
* *
* *
* * * * *
* *
* *
* *
----------
* * *
*
*
*
*
*
* * *
----------
*
*
*
*
*
* *
* * *
----------
* *
* *
* *
* *
* *
* *
* *
----------
*
*
*
*
*
*
* * * * *
----------
* *
* * * *
* * *
* * *
* *
* *
* *
----------
* *
* *
* * *
* * *
* * *
* *
* *
----------
* * *
* *
* *
* *
* *
* *
* * *
----------
* * * *
* *
* *
* * * *
*
*
*
----------
* * *
* *
* *
* *
* * *
* *
* * *
----------
* * * *
* *
* *
* * * *
* *
* *
* *
----------
* * *
* *
*
* * *
*
* *
* * *
----------
* * * * *
*
*
*
*
*
*
----------
* *
* *
* *
* *
* *
* *
* * *
----------
* *
* *
* *
* *
* *
*
*
----------
* *
* *
* *
* *
* * *
* * * *
* *
----------
* *
* *
* *
*
* *
* *
* *
----------
* *
* *
* *
*
*
*
*
----------
* * * * *
*
*
*
*
*
* * * * *
----------
* * *
*
*
*
*
*
* * *
----------
*
*
*
*
*
----------
* * *
*
*
*
*
*
* * *
----------
*
* * *
* * *
*
*
*
*
----------
*
*
* * * * *
*
*
----------
*
* *
* *
----------
* * *
*
* * * *
* *
* * * *
----------
*
*
* * *
* * *
* *
* * *
* * *
----------
* * *
* *
*
* *
* * *
----------
*
*
* * *
* * *
* *
* * *
* * *
----------
* * *
* *
* * * * *
*
* * *
----------
*
* *
*
* * *
*
*
*
----------
* * *
* * *
* * *
* * *
*
* * *
----------
*
*
* * *
* * *
* *
* *
* *
----------
*
* *
*
*
* * *
----------
*
*
*
*
* *
* * *
----------
*
*
* *
* *
* *
* *
* *
----------
* *
*
*
*
*
*
* * *
----------
* * *
* * *
* * *
* * *
* * *
----------
* * *
* * *
* *
* *
* *
----------
* * *
* *
* *
* *
* * *
----------
* * * *
* *
* *
* * * *
*
*
----------
* * * *
* *
* *
* * * *
*
*
----------
* * *
* * *
*
*
*
----------
* * * *
*
* * *
*
* * * *
----------
*
*
* * *
*
*
* *
* *
----------
* *
* *
* *
* * *
* * *
----------
* *
* *
* *
* *
*
----------
* *
* * *
* * *
* *
* *
----------
* *
* *
*
* *
* *
----------
* *
* *
* *
* * * *
*
* * *
----------
* * * * *
*
*
*
* * * * *
----------
*
*
*
*
*
*
*
----------
*
*
*
*
*
*
----------
*
*
*
*
*
*
*
----------
*
* * *
*
----------
* * * * *
----------
I believe the archive also has a binary file that is the raw data for the font.
Neat.
UPDATES:
NOTE: The author of this work will be referred to as Abraham Lincoln ;-) per this comment: “Thanks! But I’m terribly modest and shy away from attention. So, I humbly request that you refer to me as “Abraham Lincoln”.” ;-)
TRSDOS was the name of the disk operating system that Radio Shack used on their TRS-80 computers.
https://en.wikipedia.org/wiki/TRSDOS
There were alternative DOSes available, but TRSDOS is where it started, and what shipped with the original machines. As I write this post, I see that the last version of TRSDOS was released 42 years ago.
When the TRS-80 Color Computer came out in 1980, it was a T.I.N.O. – “TRS-80 In Name Only”. The 80 in TRS-80 came from the Zilog Z-80 processor, and the Color Computer used a Motorola 6809 processor. I assume Tandy just wanted to leverage the popularity of the TRS-80 brand and continue using that name. The use of TRS-80 continued even into 1983 when Tandy brought out its first PC-compatible-ish machine, the Tandy 2000. Officially called the “Tandy” 2000, the badge on the front of the machine still read “Tandy TRS-80.” This was, according to the always-accurate Wikipedia, the final use of TRS-80 branding on a Radio Shack computer. When the more-PC-compatible-ish Tandy 1000 came out in 1984, it was solely a Tandy-branded machine.
But I digress.
TRSDOS was never a thing for the CoCo. There was no disk operating system. Instead, the CoCo booted into a ROM BASIC. When a Disk Controller was added, it contained another ROM that extended the basic with new disk-related commands. DISK EXTENDED BASIC was not an operating system, but most of us referred to it as “RS-DOS” (Radio Shack DOS, one would presume). I am unsure when we started calling it RS-DOS. This would be a good side quest to research. I assume by the time OS-9 (an actual disk operating system) was released, we would have needed some way to tell if a program ran from DISK BASIC or required OS-9.
But I digress, again.
Color TRSDOS did exist, but perhaps it was another T.I.N.O. – this time meaning “TRSDOS In Name Only”. Color TRSDOS was included with the DISK EDTASM 6809 assembler. In the manual was a source code listing and on the disk was a DOS.BIN binary and DOS.BAS loader program that was Color TRSDOS.
But what was Color TRSDOS?
If you do a RUN”DOS” form the Disk EDTASM disk, you are greeted with a menu:
The name “R.G. Kilgus” is instantly recognizable to early adopters of the CoCo. R.G. created some of the earliest Radio Shack programs for to CoCo, including the infamous Dino Wars!
https://www.lcurtisboyle.com/nitros9/dinowars.html
As you may be able to tell from the screenshot, Color TRSDOS was not a DOS like we might expect. Normally, a DOS boots up into a command line that lets you run programs. MS-DOS is probably the most iconic DOS that comes to mind for me:
Instead, Color TRSDOS seemed to be more of a library of disk routines that one could use with their own 6809 assembly programs. BASIC documented a few “official” ROM calls that could be called from assembly routines to do things like print a character to the screen or input a character from the keyboard. DISK BASIC added one that could read or write raw data to a disk. This was very low level, and there was no ROM call to do things like read a directory, create a new file, delete an existing file, or anything else. This means assembly language programmers had to figure that out and write all that code themselves.
Color TRSDOS appears to have been Radio Shack’s official solution for this. You could use the Color TRSDOS assembly language routines in your own programs and get higher level disk functionality, such as the things shown in the Color TRSDOS menu.
Side Note: What was the license agreement for Color TRSDOS? Could that code be included in commercial sold software, or was it only allowed to be used for personal programs? Anyone know?
Maybe I ran into Color TRSDOS in some CoCo program I used over the years. There were certainly many programs that could load/save files to disk and show directories. Maybe some of these made use of the Color TRSDOS routines. Does anyone know of any program that ever used the Color TRSDOS routines?
It appears Color TRSDOS has a way to load and execute a binary program, as demonstrated from menu option #2:
There was also support for a software clock, as demonstrated by menu option #3. It would put an incrementing Hour/Minute/Second value in the top corner of the screen:
The Color Computer used the video circuitry to generate a 60Hz signal (also used by the TIMER command in BASIC). It was common for software clocks to make use of this and track time, though this time would drift any time interrupts were masked — such as when accessing a disk.
I have not explored the code myself, but I assume Color TRSDOS just provided a simple software clock for programs to use.
Option #4 would display the Disk Allocation Map. This display showed which granules on the disk were in use. From looking at the display, you can see it represents the 68 granules (9 256-byte sectors, 2304 bytes) that make up an RS-DOS disk. It looks like “.” indicates the granule is free, “X” may indicate all of the 9 of the 256-byte sectors in that granule are used, and a number may indicate how many sectors are used (if less than 9).
I have discussed DISK EXTENDED BASIC’s format and this Disk Allocation Map in previous blog posts. Check that link to learn more about granules and such.
Option #5 showed a simple interface to copy a file, either between the same disk or to separate disks. To copy a file to the same disk in Disk BASIC, you’d do something like this:
COPY "FILE.BAS" TO "NEWFILE.BAS"
This basically duplicated the file, and required a different name for the copy.
To copy a file from Drive 0 to Drive 1:
COPY "FILE.BAS:0" TO "FILE.BAS:1"
But BASIC had no way to copy between separate disks if you only had one disk drive. It looks like this Color TRSDOS program may have a way to do that, where it would read data from the disk, then tell you to swap in the second disk, then write it out there. (I did not test this.)
Last, option #6 performed a disk directory. I am unsure how this works, since it initially shows nothing and has some fields where you can type. Perhaps this supported some level of wild cards?
UPDATE: After I wrote this, I went back to try it. You can arrow over to the extension field and enter “BAS” then it only showed the “DOS.BAS” file. Neat! Perhaps wildcard support is built in to the Color TRSDOS routines?
And, of course, error handling…
It looks kinda neat. Had I realized this back then, I could have been adding disk support to my own assembly programs. I really didn’t know it was a thing.
Well, maybe I get a second chance now. Thanks to the tireless efforts of Abraham Lincoln, Color TRSDOS has been archived to GitHub. In a Facebook CoCo group post dated June 20 2026, he wrote:
“. . . I found an OCR scan of the Disk EDTASM manual and was able to copy-paste the source code from that. There were LOTS and LOTS and LOTS of errors introduced via the OCR scanning process, but I was able to resolve everything, and even correct a bug that was in the original manual (as determined from the non-OCR version) but subsequently fixed in the DOS.BIN executable that shipped with Disk EDTASM.
Now we can generate the exact DOS.BIN executable from source. I went through this exercise so that anyone who is interested can study, learn from, and incorporate the code into their projects, if they’d like. For me, in particular, I wanted to be able to explore it to understand the interface that Tandy expected assembly language developers to use for accessing the disk files so that I could create a command shell that provides the same interface to the programs that it launches.”
– “Abraham Lincoln” on Facebook
Thank you for your service, Abraham! You can find the result of Abraham’s work, as well as progress on a new command line interface, on his GitHub page:
https://github.com/DarkChocoholicDev/ColorTRSDOS
Abraham has also shown the beginning of a command line TRSDOS interface. I look forward to seeing how this turns out. Maybe we will end up with a “real” TRSDOS for the CoCo thanks to his efforts.
To be continued?
Hat tip to L. Curtis Boyle for pointing this one out to me. Published in The World of ’68 Micros back in 1995 was an article by John “SockMaster’ Kowalski demonstrating how to do smooth horizontal scrolling on the CoCo 3 in the 320×200 16-color graphics mode.
In BASIC.
With no assembly.
You can find this demo in Volume 2, Number 5 in an article called “The Seven-line demo: an amazing achievement with DECB!” I must have been fully into OS-9 by this point since this article does not seem familiar to me. I was not really keeping up with any BASIC stuff by that point, having moved to OS-9 assembly and C programming.
Here is that program as I typed it in and adjusted some spacing, with a few comments added by me. The only crucial part is line 70 which must have 11 colons and five spaces after that POKE Q,G line.
0 'The World of '68 Micros
1 'Volume 2, Issue 5
2 REM ** 2WAYSCRL.BAS
4 REM **bySockMaster
10 POKE 65497,0:HSCREEN 2:FOR G=0 TO 15:READ A:PALETTE G,A:NEXT
20 DATA 0,19,22,50,54,52,38,37,44,45,41,13,11,25,27,26
30 FOR G=0 TO 319 STEP .5:HSET(G,RND(191),RND(15)):NEXT
40 C=1:S=40:FOR G=15 TO 1 STEP-1:HCOLOR G
50 HCIRCLE(310-G*5,48),S:HPAINT(310-G*5,48):S=S-2.6:HLINE(RND(110)+210,RND(80)+104)-(RND(110)+210,RND(80)+104),PSET,BF:NEXT
60 Q=65439:W=0:E=127
65 '11 COLONS, 5 SPACES
70 FOR G=0 TO 127:PALETTE W,W:POKE Q,G::::::::::: POKE Q,E-G:NEXT:GOTO 70
I am unsure if the other spaces in line 70 are critical. The original listing looked like it had a lot of spaces between keywords and such, but that should not matter. LINE 70, being a loop, would be made a bit slower with extra spaces. Here is how it was presented in the magazine:
Here is what it does when ran in the XRoar emulator:
Spiffy!
Now I am off to read the full article to better understand how this works.