I had this exact APC 30 years ago. The green phosphor, screen resolution, and sound of the drive in the video takes me back in time. I loved that machine. I'm glad this video exists, because my machine does not anymore.
Little trick for vintage computers that don't support 4 year dates (and that also don't have some mechanism for dealing with year 2000 and beyond) that works well enough for most things: dial back the calendar 112 years (i.e. 2022 becomes 1910). The days of the week repeat every 28 years (112 is a multiple of that), so anything that does those calculations will still line up with the current year, you just have to add 12 (or 112 if that software assumes a 1900 century) to the year.
Until I saw Part 1 and 2 of this, I always thought the slowly fading CRT text was just some fancy CGI of some Hollywood movies. This is the first time I see long persistence phosphor on a computer.
1993?? That might explain why this thing is so well taken care of if the owner used it for that long. They either loved this machine or were too stubborn to upgrade or both lol. Thanks for the awesome video!
pretty sure most of us would have enjoyed a glimpse at the productivity software. even if it's not that interesting, it'd still be cool to see it running in that high resolution.
15:55 What you demonstrate with the index hole position here is excellent and very valuable knowledge...crucial to any 8" disk flux recovery system that I may develop here, so again, I and the vintage community thank you for this!!!
Yeah it's an interesting thing I didn't realize until recently!! I didn't say it in my video but I modded my Qume disk drive to actually allow me to format single sided disks as double sided. The circuit normally looks at which index hole is being read and then if it's the Single Sided hole, it will disallow use of the second head. I installed a jumper on the PCB to allow me to force it to ignore that part of the circuit. (This wasn't documented in the technical docs but I saw in the schematics the designers made a provision for this.) Anyway, yeah if you are going to be archiving 8" disks, it is essential to see the difference between single and double sided media since it's physically different.
@@adriansdigitalbasement Thanks for this! I'm doing some experiments as a result with my 8" drives, and I'll be certain to share my results with you as soon as I have any...VERY cool and thanks again for this AWESOME preservation work and great video documentation!
I remember using an industrial machine in the early 80's that used 8" hard sector disks. You could see multiple holes in the magnetic disk punched to indicate the location of each sector through the index hole in the sleeve. I'm guessing these never made it onto the modern PC's of that time era, but the head loading clunk takes me back. We used to joke that it was punching holes out of the disk like a paper tape!
Neon lamps get kind of weird as they age. Adrian, I’ll bet that if you shine a light on the neon lamp it’ll light and stay lit as long as there is sufficient ambient light, but if you turn down the lights, it will start flickering something awful.
Bizarre but true; the neons were made with a tiny amount of a radioactive gas in the neon to lower the strike voltage by providing an initial ionisation. I used to enjoy watching apparatus rooms full of equipment with neon power indicators which flickered like mad in the dark but as soon as the room lights were turned on, mysteriously got better and lit up. It took me a long time to find out the reason; that the light is sufficient to help the ionisation along.
Yeah I didn't know this! Someone mentioned Mr Carlson talked about this in a video. Fascinating thing! That switch I used was old and came out of something I ewasted long ago ... so it was probably quite spent.
23:18 “I really didn’t find anything particularly interesting.” 23:33 “There was a dev suite for the C programming language that doesn’t seem to be archived anywhere.”
As you noted, and many folks commented, not much software was created for this machine. Yet I know of at least one interesting use case for the colour version of the first APC, which I can explain with a bit of reminiscing. :) One of my consulting jobs in the early 80s was for a small electronics engineering firm that wanted to port a printed-circuit board CAD program to their brand-new colour APC computer. I ported it from, if I remember right, a CP/M-80 assembly-language program they got from somewhere beyond my memory now. Oh, and your videos reminded me how on each day I sat down to that machine I'd slide the keyboard out of the way and wonder whyyyy NEC decided to put the power switch there! The similarities between 8080 assembly-language CP/M programs and 8086 programs on CP/M-86 were vital to making my task possible since I didn't know much about 8086s at the time and I really needed that help. I wish I could remember if the converted program worked well in the end or not, though I *do* remember the APC being a vastly improved computer over whatever 8080/Z80 CP/M machine they were using before!
Back in the day, you could install a "flippy floppy" drive modification, which added two sensors (one for the timing, and one for the write protect), allowing you to use both sides of a 5 1/4" diskette on your single sided drive.. There were also factory 5 1/4" diskettes with two write slots and two timing holes, and a company that offered a punch that punched the write slot and timing hole. In general, it's not a great idea to use flippy floppies, since the inside of your 5 1/4" diskette has a pad inside to collect dirt. When the diskette is flipped, the dirt can be pulled back off the pad.
Abacus North, who wrote the Color Program, was a computer store in Anchorage. We price shopped there when buying our kaypro II a few years before they wrote that. Flashback to the smell of new computers and soldering of circuits....
Ah! The NEC APC series. First PC that I used that could make good-looking plots on the screen that could be easily dumped to the dot matrix printer. Wonderful blast from the past.
This is a genuinely gorgeous and novel machine. I definitely have a soft spot for ambitious hardware releases that never got to see their full potential.
Great video! Man this brings back memories, I started my first computer company in Ohio and we were selling these to all the former engineers from the Tire companies in Akron OH. We had a light fixture company come to us and asked for an inventory control system and these machines came bundled with dBase II Ashton Tate software. I ended up writing the program for the customer in dBase II on the NEC APC. That customer's company stayed and business well through the 90s and they were still using my program and the NEC APC through the mid 90s. One final note, if you logged into the program with the wrong username or password more than three times my program called the POW command, which you talk about in this video, and shut the computer off. 🙂
another great video Adrian. in part 1, you had mentioned using xmodem or kermit as a possibility for transferring files between kinds of machines. In grad school, I used Kermit-80, Kermit-CMS, and whatever the DOS version was called to move files between a DEC VT-180, the university's IBM mainframe, and the then-rare PCs that two staff engineers were fortunate to have. maybe part 3 can be a dive into serially transferring files between different machine?
Kermit was/is indeed the protocol of choice for transferring files between semi-obscure systems. There appears to be a version of kermit available for the NEC APC on kermit project website. From the txt file: MSKermit on the APC supports both the standard serial port (as port 1) and the optional (H14) add-on serial port (as port 2). Port selection is performed using the SET PORT command. Any baud rate up to 38400 is legal although 38400 has never been tested and may not work well. The port is always configured as 8 data bits, no parity, and 1 stop bit. Any necessary parity is supplied by Kermit.
Yeah I wish I had found a copy of KERMIT on the MSDOS disks. It seemed to exist (based on what I can see on the internet) but no copies existed. The non standard Serial cable would be a bit of a problem too. (Centronics type connector.)
@@adriansdigitalbasement Your copy of MS-DOS 2.11 should have ctty, so if you get RS232 working there's the option of pasting a program from your regular PC into a debug session on the APC.
@@adriansdigitalbasement The serial port is connected to the cpu board with a ribbon cable, which seems to be right pinout for D25 ribbon cable connector.
@@adriansdigitalbasement I have the parts (including the clunky CN-36 soldertail connectors), the APC and PC tech manuals, and docs on interconnecting two DTEs together, so I'm gonna work on a null-modem cable so that the APC and an IBM-PC will be able to talk to one another. Also investigating a native Kermit solution. At first blush, there doesn't seem to be an APC-specific version of MS-Kermit, even though the Kermit docs say one was created. There *IS* source for building Kermit-86 (looking at it now) that has explicit support for the APC, but the chicken-egg challenge persists: how to get the source files ONTO the APC to build them? with that in mind, I had a thought: while the IBM-PC + Qume 8in drive didn't seen to work with NFormat or the disk archiver, there's a discussion on Vintage Computer Federation forums titled "NEC APC: RDCPM and a CP/M-86 formatted diskette?" that might have some useful information for accessing 8in disks using 22DISK, listed in your resource list. Also keep in mind that there _is_ a universal CP/M 8in disk standard (the only disk standard, actually) that might be "more compatible" than the 1.2ish MB format native to the APC's CP/M-86 system. the SSSD format is the same as the IBM 3740 (IIRC, known as "IBM Disk 1") and it stores a whopping 250k or so per disk. It might be possible to use this least-common-denominator to move files between CP/M on the APC and the PC+8inFDD running 22DISK?. Additionally/alternatively, a possibility to to see if DOS and/or CP/M on the APK offer options when disk-fomatting; perhaps a format less-sophisticated than the ones native to APC would let you create a disk that's more directly compatible with ImageDisk/Nformat? Maybe creating an 8in clone of a 360k/720k diskette, for example? as an adjunct to experimenting with disk formats on the APC, there is MS-DOS utlity on the APC called "RDCPM.COM" which (should) allow its MS-DOS environment to access content on a CPM (I"m giving up on the '/') diskette. maybe in conjunction with a different DOS disk clone (see prior paragraph) that would be easier to deal with on the PC side. in the meantime, I'll start work on the null-modem cable (who knows, maybe there's a terminal program in those 8in disks somewhere that support zmodem or something) and continue to burn out my grey cells poring over 30-year-old documents for Kermit. PS: reading now about 8FORMAT ...
That reminded me that I have here a couple of Apricot machines (ACT), these ran MSDOS but the disk formatting was very slightly different to make them incompatible with IBM PC's.
I started in the mid 80´s with an IBM 386 in my dad´s bakery office, and I get such a nerd satisfaction in watching your videos, and I almost learn or remember something, thanks Adrian, keep them coming...
With 1.2MB per-disk, that's pretty good for 1983. I believe the 1.2MB high-density 5.25" floppy format only came out with the IBM AT in 1984, so the 8" disk would have been around with that size for at least a year before IBM went with the 5.25" version. I vaguely remember reading somewhere that MS-DOS did support 8" floppy disks at some stage, but support was removed pretty early on, so you might be able to read them with an early version (like what was running on the APC, although the IBM compatible version). But I could be remembering incorrectly, so take that with an arbitrary number of NaCl molecules. When you formatted the disk in NFORMAT!, the format ID was $F9, which is for 1.2MB HD 5.25". $FE is for 1.2MB DSDD 8". It was definitely formatting with 15 sectors of 512B instead of the 8 sectors of 1024B that the APC would be expecting. And that would be the difference in size you saw.
I found this machine a very interesting one for that time. Powerful. The battery backup RAM, the char generator on the RAM, etc. I worked with TRS-80 I and II, Super Brain, Osborne, Commodore PET, TI99 and Apple II at that time and I didn't know about the 8'' disks write protection (maybe because I never really used a TRS-80 model II, just once). Thanks for explaining that.
The DeSmet is a C compiler suite for 8088/86 CPUs. Creates really small code and is just perfect for low ram machines like this one. if you google for desmet C you will find the manual and all.
The difficulties of incompatibility is just something of that era, given everyone was competing for people to use their standards, usually claiming better this, faster that, etc., it just seemed to boil down to who could get the most computers out there, and that ended up as being the IBM standard for MS-DOS type computers given companies started cloning their stuff on mass...
It is indeed how much crazy corporate espionage type of moves to wrestle a platform away from company like IBM where Intel and Microsoft became much bigger controllers of the platform is something that companies octotuple check against, No third party Playstations not in any form not even legal emulators and so many other examples of extremely closed off platforms where even Microsoft is trying to make "their machines" more and more of a platform to protect and sell software on, though windows app store and their proprietary UWP format for programs are slightly failures that Microsoft eased on, most new gamepass games being more conventional computer programs, though the strong control from "the store" is there, which is a scary prospect to some like valve making Steam Deck to combat this seemingly one-way future.
If I remember correctly Linux floppy support allows you to customize the disk format. You might be able to read the APC floppies on a PC running Linux. Linux can also write to MSDOS hard disk volumes so it could be used to transfer your files.
This machine and software bundle is very typical of the early 1980s. If the user wanted to do something, outside of the few programs bundled they had to write the programs themselves. In the mid-1980s, I worked for Visual Technology who made their V-1050 CP/M Plus system. As an employee, I got a system at a steep discount and paid well under the list price of nearly $2500. That system too came bundled with a similar package including WordStar, DR Graph, MultiPlan, C-BASIC, and the Z-80 assembler. The V-1050 came with the built-in capabilities of reading other CP/M system and program disks by running a utility. Among the many the system supported was the DEC Rainbow and the Kaypro systems. I learned Z-80 assembly on my machine and also wrote numerous programs in C-BASIC. Instead of using the line-editor, I used WordStar in non-document mode. The line editor is similar if not the same as Edlin, making viewing and editing code very frustrating. Once my program was written and compiled in Z-80 Assembler, I'd execute it and view the memory using Z-Sid. Z-Sid is similar to debug on a DOS system. In many ways, I miss these days. Today, computers are blah, same, similar and not very exciting. This is the same with the operating systems. Linux comes close with its roll your own environment but that too is becoming more of the same from version to version.
John, guess what? Sitting here in my office is a Visual 1050! I haven't fired it up yer, waiting to make a video about it. I think I saw the drives on the 1050 are 96TPI versus the normal 48TPI on other systems, so you get 2x the storage. I figured it was going to have a similar issue to the APC though! Lack of the customized software it needs to really take advantage of the 1050's capabilities. I got this machine without disks or manuals. I haven't yet looked to see if scans and images exist, but I hope they do or it won't be a very interesting video.
@@adriansdigitalbasement Hi Adrian, I can't wait to see the system power up and bring back some memories. 🙂 There are disks up on Bit Savers for the system. I recommend downloading both the V1.3 and V.14 contents. Google for the V-1050. There's a couple of sites dedicated to the system and I think one of them has the schematics for the circuit boards. Visual was great that way. The systems came with tech docs as well as a user manual. Yes, the disks are 96 TPI and the TEAC floppy drives are really noisy and so is the fan inside the case. The 9-inch green or amber CRT is a separate unit. The keyboard to be warned is a Key-Tronic sponge and foil affair. Visual used the same display and keyboard for their V50, V55, V60 and V65 terminals from those components by placing a small circuit board inside the monitor. I agree, there's not much software to take advantage of the graphics and other capabilities of that system. The system was very powerful for its time. The V-1050 has a 6502 and 32K of RAM dedicated for the graphics portion of the system. The Z80 gets its own bank-switched set. Unfortunately, I can't remember the total amount of RAM on board any longer, but they had plans to come release some add-on cards including some memory boards mounted on some very long-sharp headers which used to gouge my hands whenever I was fixing the board. I tested some early memory boards but they never made it past that point. The system does support a hard drive through the Centronics port in the back.
Yes, the clones really didn't survive but to me these machines are some of the most interesting. For example someone ported DOS 1 to the Commodore B128 (CBM-II) with the 8088 add-on.
Hats off once again. A great visit to an early computer. My computer experience pretty much started with my boss dumping two brand new IBM-XTs on my desk (FDD-only models) and telling me to do something with them. Back then I bought Peter Norton's Guide to IBM PCs and was encouraged to 'play' so that later I could help other users. I spent a few days with MS-DOS 'debug' just peeking and poking in memory and on the floppies. At the time I was transferring data from a CP/M system via the machines' serial ports and with the help of the Norton book I navigated my way through the directories of several MS-DOS floppies that I'd used. Of course 'delete' doesn't delete necessarily, first it marks files as deleted, then they are overwritten if clusters and/or sectors need to be re-used. Do you think that your blank disks may have erased data and software on them or have they been formatted but not used? There might be something there. I think CP/M has a means by which floppy sectors can be accessed direct, but I can't remember how
Hi Adrian. You can actually use the “one-line maze” program with the regular slash and backslash, you just have to multiply the Rnd(1) result by the difference in ASCII values between the “/“ and “\” characters. So when it gets a “tails” (which I’ll set = 0) it multiplies that times the offset, but 0 times anything is 0, so you just get the ASCII code for the slash. When it gets a “heads” (= 1) it multiplies 1 by the offset, then adds that to the ASCII code for the slash, to get the backslash. It’s not as pretty as the ones that have the “bigger” slash and backslash, but it’ll work. Variations on this should work in other languages, like C or Python, that can be run at a command-line prompt on modern systems.
I came here to make this exact comment. Note also that the rnd function generates a floating point number between 0 and the provided parameter, so it has to be rounded if the difference between characters is greater than one. That single line maze program looks like: 10 ?chr$(47+(cint(rnd(1))*45));:goto 10 (Tested in GWBasic 2.0)
@@michaelcalvin42 Indeed. I came here to see if I needed to comment similarly. I knew someone must have beat me to it, since it's such an obvious point for anyone that knows anything about programming. I assume Adrian just had a brain fart--or perhaps he doesn't know much about programming, even in BASIC?
@@michaelcalvin42 If your BASIC doesn't have a function for rounding (like C64 BASIC doesn't), you can use something like INT(RND(1) + 0.5); essentially add 0.5 and then convert it to an integer; then multiply by the offset and add the base like you already have.
I like the styling of it because it looks so similar to a VT100 terminal. If you could find a terminal emulator to run on it, you could use the computer as a dedicated terminal for other projects (I know you already have dedicated terminals, but why not one more).
Thank you for Archiving the C Compiling Software , Love this NEC , all the Great early 80's Computers . My brain can understand a Lot of Late 70's and Early 80's Computers and it has taken me 39 years to Learn What CPM and DOS that I Know , I sure Hope the collections are found by Young People and Put up on the Net for future Study , I Miss Gary Kildall , I Miss old BBS's , I Collect Modems , Books , Magazines and Computers , Cheres :) QC
Actually the NEC '765 floppy disc controller (and the Intel 8272A that's pin-compatible) doesn't need the index pulse to read. It only needs it to format a track; the formatting process writes a sector number into each sector on every track, which means it doesn't need the index pulse when reading or even writing- only when formatting.
Interesting! I guess the OS implementation still looks for the pulse anyway. Kind of rude! I know IMD even doesn't work if the index pulse can't be seen (while reading) and it's controlling the disk controller directly.
Sir: I love your videos ! I know essentially nothing about computers, but I learned more about floppy dicks in a few minutes of watching your running comments, than I ever previously knew. You have a great knowledge of the machines from years gone by, a knack for teaching, & excellent articulation. I am in awe & admiration of your work. Cheers!
I've never used 8 inch drives but I love 5 inch drives. Can't explain why. I think because when I was a child with the family Sinclair Spectrum I thought floppy drives were the coolest things ever. Haha. We are all of our time.
No, Adrian. OLD is the original command to load a program, in Dartmouth BASIC (DTSS),the command is specified in the ANSI BASIC standard, which both TI BASIC, and DRI Personal BASIC implement.
I can fully understand the owner using it well into the 90ties. With wordstar / DB3 and a C compiler, I would have been hard to move to a different setup until the Internet broke out :-)
He said the computer was made in '83. To state the obvious, '93 is "only" ten years later. That'd be like using Windows 8 now, which doesn't seem that absurd. (Okay, that's not a great analogy.)
The pace of advancement back then was huge compared to now. You can easily use a 10 year old computer today and run Windows 10 on it with all modern software. (Except maybe the most demanding stuff.) Compare that to a 1981 IBM PC 5150 versus a 1991 clone with a 486 chip. The performance improvement in 10 years back then was like 100x faster in some cases. Now they just add more cores but the single thread speed hasn't increased that much.
@@adriansdigitalbasement Yeah, I was sort-of joking. That said, non-GUI DOS applications were still pretty common up to the early '90s, especially in a business context. If whatever company owned this computer had their records on it starting from '83, there had better be a good reason to migrate that data to a new system (and you discovered that it's very hard just to copy bytes, even ignoring potentially translating that data to a new application) or abandoning that data. If I was running a business using this machine for my records in the early '90s, I'd definitely think hard about what advantages existed on a new system that would outweigh the cost of new hardware and data migration. I wouldn't be surprised to find that the stuff you found marked "1993" were the artifacts of a migration effort. Why would a company using a word processor and a database (IIRC) suddenly have a C compiler?
Yes i clearly remember many people in 93 using pure text programs like wordstar, wordperfect and such on pc clones for business use, something graphical like Windows was seriously alienating and made things much slower, so it was a net loss on productivity. I clearly remember one person late 90ies still using her Wordperfect with pure MSDOS, she didn't want Windows or mouse, was in the way. All she needed was to type legal documents and print them, had everything configured margins, fonts, etc. that was it. She was using like a Pentium class computer with that exact same configuration, simply because the older ones died and since MSDOS did still work with later CPUs, it didn't matter. In parallel i have a friend using a Z80 QX10 for almost the exact thing until it broke, in addition to a newer PC for games 🙂
@@wbfaulk early 90s? Longer..I was doing data entry on vt100 terminals hooked to a mini computer early 2000s and my step-dad was doing dbiv programming up until about 2002 at least..lots of DOS stuff hung around longer than it probably should have..heck only a few years ago I saw a business still using what looked like maybe a 386 at best on dos for a bunch of business shit..I offered to trade them for a modern pc no cost because I wanted the retro machine and they said not a chance (was some weird fabric shop)
Even if I’d bought that disk way back then, brand new, I absolutely WOULD have duplicated it! I only use duplicates of purchased stuff and safely store the originals.
@@ax14pz107 Actually when they see you sticking an original disk in the drive they are saying: “Don’t! Copy that floppy!” So the message is definitely saying make a backup.
When I was a kid, I took an after school computer class that had Apple II+ computers. The teacher said he had a black Bell & Howell version of the Apple II with 8" disk drives that had a 1.2MB capacity on their floppies. When I got older and into MS-DOS PCs, I remember hearing the 5-1/4" High Density disk capacity of 1.2MB was to match the capacity of 8" disks. I also remember seeing a 5-1/4 Quad Density disk drive in some PC's in the computer lab, which used disks with a capacity of 720KB.
AFAIK the IBM PC only requires the index hole during formatting. For reading and writing data it is not necessary. So it would be entirely possible to create a flippy disk for the IBM PC, as long as the formatting was specially created on a double-sided drive.
Don't sound so shocked Adrian... I used my CP/m system well into the early 90s! Wordstar was perfect for me and worked wonderfully... Only purchased an IBM clone in the mid-90s...
That's some pretty interesting computer history there! I never knew any of this... I guess this machine has the Y2K bug... Hahaha... Thanks for the edumication Adrian...
@Adrian's Digital Basement 17:26 Does the disk labeled MSDOS 211 Practice Disk actually have ASM88 on it? or is that just ASM86. If it's ASM88, can you please upload that seperately with MASM, CSTUDIO.S and DEBUG? I couldn't find ASM88.
DeSmet compilers have been open-sourced as GPL. Only a few have source+manual though. Super interesting to have a read through, so simple. Hardly any semantic analysis and no optimization! It's nothing like the more advanced C compilers from the 80's and 90's such as DigitalMars that is for sure.
One of my favorite videos you’ve done in awhile. I have a pile of disks from my A500 sitting in the basement I haven’t touched in 30 some odd years. Someday I’ll go through them as you did with this computer. I hope.
DeSmet made compilers for several popular languages. They ran on CP/M and later MS-DOS. I think they later changed the name to Utah. I used the MS-DOS version of Utah COBOL and Utah Fortran 66 in college. Very inexpensive.
When you were randomly pressing keys on the keyboard, it seemed to register the presses of the CAP, ALT and GR keys, though I guess that doesn't matter if you can't press anything else. Just an oddity I noticed.
The APC/PC-98 was certainly a much cleaner, better thought-through design than the IBM PC in many ways. It's a shame that it didn't catch on outside Japan. Speaking of PC-98, I wonder if it's disks are logically compatible with this machine's. IIRC, it used 5.25 inch 1.2 MB floppies but with a non-standard format that sounds a lot like the APC 8 inch layout. Maybe tools for reading/writing PC-98 disks on PC might help for this machine.
The disk format on the 8" disks seems to be the same format the NEC PC-98 used for its 3.5" disks (which, incidentally, is the third supported format on USB floppy drives in addition to 1.44M and 720K). Maybe you can find software meant for working with PC-98 images or PC-98 disks on DOS/V?
Oh really that's interesting. I didn't know about the 3rd disk format support on USB 3.5" drives. I did some googling for PC-98 DOS support but don't see anything. Just lots of talk about the USB drive thing
I remember a "3 mode floppy support" BIOS option on some socket 7 boards, which is apparently for a 1.2 MB (not 1.44) 3.5" disk format used in Japan. Is that the same thing as the PC-98 format? The Wikipedia page for the PC-9800 says "Early PC-9801 models supported 1232 KB 8-inch floppy drives and/or 640 KB 5-1⁄4-inch floppy drives. [...] High density 5-1⁄4-inch and 3-1⁄2-inch floppy disks use the same logical format and data rate as 1232 KB 8-inch floppy disks."
@@awd42 Aye that is the same format. PC-98 was the biggest computer platform in Japan before Windows 95 made its differences from PC-compatibles more or less moot.
I found CPM-86 and Concurrent CPM-86 on WinWorld. Concurrent CPM-86 is a multitasking version of CPM . In case you need it for archival or restoration purposes.
Adrian , thank you for the video. Wow.. what a beautiful monochrome computer. I liked the monitor. On this monitor, I was taught in school, basic) Even a little sad that almost can't do anything on this computer...But for some reason it seems to me that all the same all is not hopeless )
about the power switch and its Neon lamp; the bulb is failing- and the reason it "sometimes lights up" is based on a few things: 1 the little electrodes are dipped in a slightly radioactive coating which helps the Neon "fire" and this combined with it being exposed to photons (light) also will cause it ti fire at random. a way to "test" this is very simple, simply turn it on in a darkened room and then shine a light on the switch and you'll then see it light up!
@@adriansdigitalbasement Hmm maybe we can reach out to the MAME people? I always have trouble using MAME for microcomputers. It looks like most of the coding for the APC in MAME was done 8-10 years back.
There were about twelve games published by Infocom alone for NEC APC CP/M-86. Including "Zork", of course. So it's strange the one you tried did not work... There was no Zork for IBM PC CP/M-86 as far as I know - maybe that copy of the game is for some other CP/M machine, like DEC Rainbow?
Your year 2000 clock bug had me wondering if anyone had approached a fix for the eventual year 2038 problem. I know DOS will probably keep working, but it would be nice to have the correct dates.
MS-DOS doesn't have a year 2038 problem. That issue is when a signed 32-bit count of seconds past 12:00am, Jan 1, 1970 rolls over. MS-DOS records dates as 5 bits to record the day of the month, 4 bits to record the month, and 7 bits to record the number of years past 1980. That takes us through 2107, though apparently the APIs that access those dates have a problem with dates past 2099.
Hey, Dear Adrian, could you please do a video on how you go about creating an archive old floppy disk, particularly ones which belong to systems no longer in use. Among a bunch of 5.25" floppies I bought, here in the UK, I came across an "HH Tigerbyte" "Full CP/M" version 1.1 (or 1.2 if you go by the hand-writing rather than the printed label) for the HH Tiger computer - this was, apparently, a computer developed by Tangerine of Cambridge in the UK. However, they sold the design to HH (Audio) Electronics of Cambridge, and went on the develop the Oric 1, instead. By all accounts, that HH Tiger is an extra-ordinarily rare machine. Sadly, of course, I have absolutely no way of discovering if the original operating system is preserved on this disk, or simply over-written (as quite a lot of the other disks I picked up were). If it does survive, and as far as I can tell from a brief Google-search, this will not be archived and some collectors have these computers and cannot boot them. Maybe, just maybe…? So, how is it done???
Unless you have a machine old enough to run DOS and have a floppy drive controller, you might want to look at getting a flux-level disk interface, such as a Greaseweazle, Kryoflux, Applesauce or FluxEngine (disclaimer: I made that last). This will allow you to archive pretty much anything which is physically compatible using a modern computer.
Once every couple of years, I need to boot up an old PC 8088 to run a program called BoxCar that for some reason only works on an 8088 processor. The software designs Box Culverts. It is the only program available to do some particular calculations for the hunches. Fun DOS stuff.
Unfortunately disk formats for MSDOS 2.11 on the various machines were not standardised. So, this NEC has it's own format as do other machines such as the ACT Apricot (which has a non-standard header before the FAT), which has the same software problems.
Indeed. I think it was the same as Digital Research did with CP/M. The FAT filesystem format itself seemed to be the same but the underlying encapsulation could change.
It's more that it's system specific and Microsoft didn't specify the disk format. In the case of the Apricot the first sector on the drive contains the fully drive parameters including the number heads, sectors and tracks on the medium, the sector offset to the FAT and what would be on a PC the BIOS settings. Goodness knows what the NEC has there.
Too bad Zork didn't work. I wish I was better in coding so I could try to rewrite that Infocom interpreter to make those games playable on different systems. By the way, I really like your new way of doing the split screen display while working on the computer connected to your capturing device.
The Infocom interpreter has already been rewritten many times and Infocom games can be played on virtually any computer. "Frotz" is the most notable Z-machine interpreter.
Wasted potential having ram limit of 128kb and mentioned in last video highest was 256kb. Game may have entered in to a loop waiting for keyboard bios call to returning scan code. Not trying to be full pc compatible unless only dos functions called, common for clones.
It may only have a 2 digit year but it had the day of the week correct. October 30, 1922 was a Monday, so I'd wager it is Y2K compatible, at least with CP/M.
Neon bulbs - even newer ones - require a bit of light to "strike"; if this one is old and tired, it might need more light. Try flashing a flashlight into the power switch when it doesn't start right away to verify... There is a "strike in darkness" type, but I doubt that these were ever put into power indicators.
When i was at school in the 80s we used to punch a second hole in the 5 1/4" disks to make them double sided using a small single hole punch. I think there were commercial ones on sale that prompted this called a "banana" disk.
Nearly all early PC software ran on non compatible MS-DOS machines. The two notable exceptions were Microsoft Flight Simulator and Lotus 1-2-3, but that was enough to tip the balance towards full clones. In the early days we didn't know if IBM was going to be as aggressive against clones as Apple had been, but once manufacturers saw that it was ok to make clones it didn't make sense to insist on non compatibles (like the TI Professional, the DEC Rainbow or the ones mentioned in the video)
I think early machines recreated the IBM PC architecture in hardware pretty much identically, but didn't have the IBM BIOS. This is what led to some incompatibility, but if the software tried to talk to hardware directly it would work. But this APC is completely different than the IBM PC with regards to the hardware architecture, so the software would have to rely 100% on DOS calls only to work.....
0:48: _Adrian:_ "So I wanna see if I can get any MS-DOS software that was designed for the IBM PC actually working on this thing." _Me, an idiot, but with related experience:_ Not without porting, you can't!
@32:15 There were disk drives for the ADAM? Or is 22 disk just imaging the funky not-quite a normal philips standard compact cassette cassete tapes? I'm genuinely curious now, considering there's a complete (well everything but any old composite monitor.) ADAM up in the attic, next to the //C the c128 and a ][e... with the ADAM having the least sopporting software, other than 2600 cartridge compatability....
When you booted ms dos there was a program called rdcpm , maybe you could read cpm disks from ms dos on the nec apc, and get ms dos programs into it that way?
There was an oscilloscope card for the PET during the late 70s, a cheap way to replace an expensive Tektronix. If I remember the PET was also used to drive a size A or B x-y plotter about the same time. Back in the day a full blown large size (size E?) flatbed plotter CAD system would cost in the 10s of thousands of dollars.
Love this video, super calc2 for cp/m was one of my first spreadsheets. I recall my cp/m machine had a 3” (not 3 1/2”) drive and a 5 1.4” disk drive and I had to use tdos for cp/m to read 360k pc dos disks.
Really awsome machine. Maybe if you find a CP/M program that can write MS/DOS disks similarily like the tool for DOS that can write CP/M disks you can make files move with two hoops? Also maybe MAME can emulate this? I do not clearly see it if they have proper emulation, but in case they do, there must be some way to mount images of the disk, add files and write back whole images (or someone could make software for that who is capable enough). This machine must have been very good for software development because of the added space on the disk drives. The disk that was labelled like "excercise" or what also seem to have some C compilers on it and such - maybe different ones?
I grew up in a college town. My dad was a NEC authorized dealer. When the university opened their own computer store with discounts all the other computer stores went under. That's how I ended up with a NEC APC in my bedroom as a six year old. He eventually found a buyer and I was given a NEC 8201A.
I had this exact APC 30 years ago. The green phosphor, screen resolution, and sound of the drive in the video takes me back in time. I loved that machine. I'm glad this video exists, because my machine does not anymore.
Little trick for vintage computers that don't support 4 year dates (and that also don't have some mechanism for dealing with year 2000 and beyond) that works well enough for most things: dial back the calendar 112 years (i.e. 2022 becomes 1910). The days of the week repeat every 28 years (112 is a multiple of that), so anything that does those calculations will still line up with the current year, you just have to add 12 (or 112 if that software assumes a 1900 century) to the year.
Very interesting workaround. Never thought about doing that even though I knew about the date cycles. Have you had use cases for that in the past?
You are better with numbers than me.
The screen showed oct 30th demonstrating the system was able to correctly calculate the 2022 date!
Although apparently MS DOS didn't like it lmao
@@forbiddenera MS-DOS 2 is not Y2K compatible. Think you need 4 or 5 to use any dates past 1999
Until I saw Part 1 and 2 of this, I always thought the slowly fading CRT text was just some fancy CGI of some Hollywood movies. This is the first time I see long persistence phosphor on a computer.
1993?? That might explain why this thing is so well taken care of if the owner used it for that long. They either loved this machine or were too stubborn to upgrade or both lol. Thanks for the awesome video!
The high res screen allows for a beautiful terminal font! Very interesting machine! Well made!
As neon bulbs age, the trigger voltage can increase. That might explain the seeming random activation of the power light.
pretty sure most of us would have enjoyed a glimpse at the productivity software. even if it's not that interesting, it'd still be cool to see it running in that high resolution.
15:55 What you demonstrate with the index hole position here is excellent and very valuable knowledge...crucial to any 8" disk flux recovery system that I may develop here, so again, I and the vintage community thank you for this!!!
Yeah it's an interesting thing I didn't realize until recently!! I didn't say it in my video but I modded my Qume disk drive to actually allow me to format single sided disks as double sided. The circuit normally looks at which index hole is being read and then if it's the Single Sided hole, it will disallow use of the second head. I installed a jumper on the PCB to allow me to force it to ignore that part of the circuit. (This wasn't documented in the technical docs but I saw in the schematics the designers made a provision for this.) Anyway, yeah if you are going to be archiving 8" disks, it is essential to see the difference between single and double sided media since it's physically different.
@@adriansdigitalbasement Thanks for this! I'm doing some experiments as a result with my 8" drives, and I'll be certain to share my results with you as soon as I have any...VERY cool and thanks again for this AWESOME preservation work and great video documentation!
I remember using an industrial machine in the early 80's that used 8" hard sector disks. You could see multiple holes in the magnetic disk punched to indicate the location of each sector through the index hole in the sleeve. I'm guessing these never made it onto the modern PC's of that time era, but the head loading clunk takes me back. We used to joke that it was punching holes out of the disk like a paper tape!
Sold these machines in the 80’s for business. They just worked and worked. Colour graphics was amazing, but very little software.
Neon lamps get kind of weird as they age. Adrian, I’ll bet that if you shine a light on the neon lamp it’ll light and stay lit as long as there is sufficient ambient light, but if you turn down the lights, it will start flickering something awful.
That is bizarre.
Bizarre but true; the neons were made with a tiny amount of a radioactive gas in the neon to lower the strike voltage by providing an initial ionisation. I used to enjoy watching apparatus rooms full of equipment with neon power indicators which flickered like mad in the dark but as soon as the room lights were turned on, mysteriously got better and lit up. It took me a long time to find out the reason; that the light is sufficient to help the ionisation along.
I have a similar vintage switch like that. It has the same problems where it doesn't light up immediately.
@@neilbarnes3557 Huh. Is that still true? You can still get neon lamps. I recently installed one as the indicator in a guitar amplifier.
Yeah I didn't know this! Someone mentioned Mr Carlson talked about this in a video. Fascinating thing! That switch I used was old and came out of something I ewasted long ago ... so it was probably quite spent.
I made a NEC APC disk today and was able to copy files to the DOS disks using Uniform PC version 3.0. It worked really well.
That is such a crisp and aesthetically pleasing font on that screen.
Such a cool looking machine. It would be great for a retro-sci fi movie set.
I totally agree Ken.
Came here to say, we will probably see this on the set of a upcoming movie or show. Beautiful looking machine.
Like "WarGames"?
23:18 “I really didn’t find anything particularly interesting.” 23:33 “There was a dev suite for the C programming language that doesn’t seem to be archived anywhere.”
Haha -- well at least it's archived now :-) And some other manuals that had never been scanned.
As you noted, and many folks commented, not much software was created for this machine. Yet I know of at least one interesting use case for the colour version of the first APC, which I can explain with a bit of reminiscing. :) One of my consulting jobs in the early 80s was for a small electronics engineering firm that wanted to port a printed-circuit board CAD program to their brand-new colour APC computer. I ported it from, if I remember right, a CP/M-80 assembly-language program they got from somewhere beyond my memory now. Oh, and your videos reminded me how on each day I sat down to that machine I'd slide the keyboard out of the way and wonder whyyyy NEC decided to put the power switch there!
The similarities between 8080 assembly-language CP/M programs and 8086 programs on CP/M-86 were vital to making my task possible since I didn't know much about 8086s at the time and I really needed that help. I wish I could remember if the converted program worked well in the end or not, though I *do* remember the APC being a vastly improved computer over whatever 8080/Z80 CP/M machine they were using before!
Back in the day, you could install a "flippy floppy" drive modification, which added two sensors (one for the timing, and one for the write protect), allowing you to use both sides of a 5 1/4" diskette on your single sided drive..
There were also factory 5 1/4" diskettes with two write slots and two timing holes, and a company that offered a punch that punched the write slot and timing hole.
In general, it's not a great idea to use flippy floppies, since the inside of your 5 1/4" diskette has a pad inside to collect dirt. When the diskette is flipped, the dirt can be pulled back off the pad.
Flippy disks, Sierra hole-punched the Kings Quest III disks for CoCo 3, otherwise the 10 disks wouldn't have fit in the Sierra box!
Abacus North, who wrote the Color Program, was a computer store in Anchorage. We price shopped there when buying our kaypro II a few years before they wrote that. Flashback to the smell of new computers and soldering of circuits....
With that high res text screen, the C IDE must have looked amazing!
Ah! The NEC APC series. First PC that I used that could make good-looking plots on the screen that could be easily dumped to the dot matrix printer. Wonderful blast from the past.
This is a genuinely gorgeous and novel machine. I definitely have a soft spot for ambitious hardware releases that never got to see their full potential.
Great video! Man this brings back memories, I started my first computer company in Ohio and we were selling these to all the former engineers from the Tire companies in Akron OH. We had a light fixture company come to us and asked for an inventory control system and these machines came bundled with dBase II Ashton Tate software. I ended up writing the program for the customer in dBase II on the NEC APC. That customer's company stayed and business well through the 90s and they were still using my program and the NEC APC through the mid 90s. One final note, if you logged into the program with the wrong username or password more than three times my program called the POW command, which you talk about in this video, and shut the computer off. 🙂
another great video Adrian.
in part 1, you had mentioned using xmodem or kermit as a possibility for transferring files between kinds of machines. In grad school, I used Kermit-80, Kermit-CMS, and whatever the DOS version was called to move files between a DEC VT-180, the university's IBM mainframe, and the then-rare PCs that two staff engineers were fortunate to have.
maybe part 3 can be a dive into serially transferring files between different machine?
Kermit was/is indeed the protocol of choice for transferring files between semi-obscure systems. There appears to be a version of kermit available for the NEC APC on kermit project website.
From the txt file:
MSKermit on the APC supports both the standard serial port (as port 1) and
the optional (H14) add-on serial port (as port 2). Port selection is
performed using the SET PORT command. Any baud rate up to 38400 is legal
although 38400 has never been tested and may not work well. The port is
always configured as 8 data bits, no parity, and 1 stop bit. Any necessary
parity is supplied by Kermit.
Yeah I wish I had found a copy of KERMIT on the MSDOS disks. It seemed to exist (based on what I can see on the internet) but no copies existed. The non standard Serial cable would be a bit of a problem too. (Centronics type connector.)
@@adriansdigitalbasement Your copy of MS-DOS 2.11 should have ctty, so if you get RS232 working there's the option of pasting a program from your regular PC into a debug session on the APC.
@@adriansdigitalbasement The serial port is connected to the cpu board with a ribbon cable, which seems to be right pinout for D25 ribbon cable connector.
@@adriansdigitalbasement I have the parts (including the clunky CN-36 soldertail connectors), the APC and PC tech manuals, and docs on interconnecting two DTEs together, so I'm gonna work on a null-modem cable so that the APC and an IBM-PC will be able to talk to one another.
Also investigating a native Kermit solution. At first blush, there doesn't seem to be an APC-specific version of MS-Kermit, even though the Kermit docs say one was created. There *IS* source for building Kermit-86 (looking at it now) that has explicit support for the APC, but the chicken-egg challenge persists: how to get the source files ONTO the APC to build them?
with that in mind, I had a thought:
while the IBM-PC + Qume 8in drive didn't seen to work with NFormat or the disk archiver, there's a discussion on Vintage Computer Federation forums titled "NEC APC: RDCPM and a CP/M-86 formatted diskette?" that might have some useful information for accessing 8in disks using 22DISK, listed in your resource list.
Also keep in mind that there _is_ a universal CP/M 8in disk standard (the only disk standard, actually) that might be "more compatible" than the 1.2ish MB format native to the APC's CP/M-86 system. the SSSD format is the same as the IBM 3740 (IIRC, known as "IBM Disk 1") and it stores a whopping 250k or so per disk. It might be possible to use this least-common-denominator to move files between CP/M on the APC and the PC+8inFDD running 22DISK?. Additionally/alternatively, a possibility to to see if DOS and/or CP/M on the APK offer options when disk-fomatting; perhaps a format less-sophisticated than the ones native to APC would let you create a disk that's more directly compatible with ImageDisk/Nformat? Maybe creating an 8in clone of a 360k/720k diskette, for example?
as an adjunct to experimenting with disk formats on the APC, there is MS-DOS utlity on the APC called "RDCPM.COM" which (should) allow its MS-DOS environment to access content on a CPM (I"m giving up on the '/') diskette. maybe in conjunction with a different DOS disk clone (see prior paragraph) that would be easier to deal with on the PC side.
in the meantime, I'll start work on the null-modem cable (who knows, maybe there's a terminal program in those 8in disks somewhere that support zmodem or something) and continue to burn out my grey cells poring over 30-year-old documents for Kermit.
PS: reading now about 8FORMAT ...
Super cool dinosaur, sometimes being the best doesn’t automatically mean success - there are so many examples. Thanks for the video.
Thanks for the breakdown on the index holes. I had often wondered how drives were able to detect different types. Makes sense now, in hindsight.
That reminded me that I have here a couple of Apricot machines (ACT), these ran MSDOS but the disk formatting was very slightly different to make them incompatible with IBM PC's.
Also ACT machines had a character editor and logo creator.
I started in the mid 80´s with an IBM 386 in my dad´s bakery office, and I get such a nerd satisfaction in watching your videos, and I almost learn or remember something, thanks Adrian, keep them coming...
With 1.2MB per-disk, that's pretty good for 1983. I believe the 1.2MB high-density 5.25" floppy format only came out with the IBM AT in 1984, so the 8" disk would have been around with that size for at least a year before IBM went with the 5.25" version.
I vaguely remember reading somewhere that MS-DOS did support 8" floppy disks at some stage, but support was removed pretty early on, so you might be able to read them with an early version (like what was running on the APC, although the IBM compatible version). But I could be remembering incorrectly, so take that with an arbitrary number of NaCl molecules.
When you formatted the disk in NFORMAT!, the format ID was $F9, which is for 1.2MB HD 5.25". $FE is for 1.2MB DSDD 8". It was definitely formatting with 15 sectors of 512B instead of the 8 sectors of 1024B that the APC would be expecting. And that would be the difference in size you saw.
I found this machine a very interesting one for that time. Powerful. The battery backup RAM, the char generator on the RAM, etc. I worked with TRS-80 I and II, Super Brain, Osborne, Commodore PET, TI99 and Apple II at that time and I didn't know about the 8'' disks write protection (maybe because I never really used a TRS-80 model II, just once). Thanks for explaining that.
The DeSmet is a C compiler suite for 8088/86 CPUs. Creates really small code and is just perfect for low ram machines like this one. if you google for desmet C you will find the manual and all.
The maze is also possible on ATARI 8-Bit machines as ATASCII happens to have the same slash/backslash combination as PETSCII, but in another location.
The difficulties of incompatibility is just something of that era, given everyone was competing for people to use their standards, usually claiming better this, faster that, etc., it just seemed to boil down to who could get the most computers out there, and that ended up as being the IBM standard for MS-DOS type computers given companies started cloning their stuff on mass...
It is indeed how much crazy corporate espionage type of moves to wrestle a platform away from company like IBM where Intel and Microsoft became much bigger controllers of the platform is something that companies octotuple check against, No third party Playstations not in any form not even legal emulators and so many other examples of extremely closed off platforms where even Microsoft is trying to make "their machines" more and more of a platform to protect and sell software on, though windows app store and their proprietary UWP format for programs are slightly failures that Microsoft eased on, most new gamepass games being more conventional computer programs, though the strong control from "the store" is there, which is a scary prospect to some like valve making Steam Deck to combat this seemingly one-way future.
@@SumeaBizarro if MS pushes control too much another platform will become more attractive.
* _en masse_
"That's the great thing about standards. There are so many to choose from..."
@@K-o-R there was instead today with the PC including O.S. all are standard because of bad practises of all oems and micro$oft.
The whole aesthetic of the APC and this video gives me Patlabor vibes.
That font is absolutely gorgeous!
If I remember correctly Linux floppy support allows you to customize the disk format. You might be able to read the APC floppies on a PC running Linux. Linux can also write to MSDOS hard disk volumes so it could be used to transfer your files.
That delay lends credence to the lamp being a neon. Sometimes old neon lamps take a while to break past their excitation voltage and light up.
This machine and software bundle is very typical of the early 1980s. If the user wanted to do something, outside of the few programs bundled they had to write the programs themselves. In the mid-1980s, I worked for Visual Technology who made their V-1050 CP/M Plus system. As an employee, I got a system at a steep discount and paid well under the list price of nearly $2500. That system too came bundled with a similar package including WordStar, DR Graph, MultiPlan, C-BASIC, and the Z-80 assembler. The V-1050 came with the built-in capabilities of reading other CP/M system and program disks by running a utility. Among the many the system supported was the DEC Rainbow and the Kaypro systems.
I learned Z-80 assembly on my machine and also wrote numerous programs in C-BASIC. Instead of using the line-editor, I used WordStar in non-document mode. The line editor is similar if not the same as Edlin, making viewing and editing code very frustrating. Once my program was written and compiled in Z-80 Assembler, I'd execute it and view the memory using Z-Sid. Z-Sid is similar to debug on a DOS system.
In many ways, I miss these days. Today, computers are blah, same, similar and not very exciting. This is the same with the operating systems. Linux comes close with its roll your own environment but that too is becoming more of the same from version to version.
John, guess what? Sitting here in my office is a Visual 1050! I haven't fired it up yer, waiting to make a video about it. I think I saw the drives on the 1050 are 96TPI versus the normal 48TPI on other systems, so you get 2x the storage. I figured it was going to have a similar issue to the APC though! Lack of the customized software it needs to really take advantage of the 1050's capabilities.
I got this machine without disks or manuals. I haven't yet looked to see if scans and images exist, but I hope they do or it won't be a very interesting video.
@@adriansdigitalbasement Hi Adrian,
I can't wait to see the system power up and bring back some memories. 🙂
There are disks up on Bit Savers for the system. I recommend downloading both the V1.3 and V.14 contents. Google for the V-1050. There's a couple of sites dedicated to the system and I think one of them has the schematics for the circuit boards. Visual was great that way. The systems came with tech docs as well as a user manual.
Yes, the disks are 96 TPI and the TEAC floppy drives are really noisy and so is the fan inside the case. The 9-inch green or amber CRT is a separate unit. The keyboard to be warned is a Key-Tronic sponge and foil affair. Visual used the same display and keyboard for their V50, V55, V60 and V65 terminals from those components by placing a small circuit board inside the monitor.
I agree, there's not much software to take advantage of the graphics and other capabilities of that system. The system was very powerful for its time. The V-1050 has a 6502 and 32K of RAM dedicated for the graphics portion of the system. The Z80 gets its own bank-switched set. Unfortunately, I can't remember the total amount of RAM on board any longer, but they had plans to come release some add-on cards including some memory boards mounted on some very long-sharp headers which used to gouge my hands whenever I was fixing the board. I tested some early memory boards but they never made it past that point. The system does support a hard drive through the Centronics port in the back.
Yes, the clones really didn't survive but to me these machines are some of the most interesting. For example someone ported DOS 1 to the Commodore B128 (CBM-II) with the 8088 add-on.
Hats off once again. A great visit to an early computer. My computer experience pretty much started with my boss dumping two brand new IBM-XTs on my desk (FDD-only models) and telling me to do something with them. Back then I bought Peter Norton's Guide to IBM PCs and was encouraged to 'play' so that later I could help other users.
I spent a few days with MS-DOS 'debug' just peeking and poking in memory and on the floppies. At the time I was transferring data from a CP/M system via the machines' serial ports and with the help of the Norton book I navigated my way through the directories of several MS-DOS floppies that I'd used. Of course 'delete' doesn't delete necessarily, first it marks files as deleted, then they are overwritten if clusters and/or sectors need to be re-used. Do you think that your blank disks may have erased data and software on them or have they been formatted but not used? There might be something there. I think CP/M has a means by which floppy sectors can be accessed direct, but I can't remember how
Hi Adrian. You can actually use the “one-line maze” program with the regular slash and backslash, you just have to multiply the Rnd(1) result by the difference in ASCII values between the “/“ and “\” characters. So when it gets a “tails” (which I’ll set = 0) it multiplies that times the offset, but 0 times anything is 0, so you just get the ASCII code for the slash. When it gets a “heads” (= 1) it multiplies 1 by the offset, then adds that to the ASCII code for the slash, to get the backslash. It’s not as pretty as the ones that have the “bigger” slash and backslash, but it’ll work. Variations on this should work in other languages, like C or Python, that can be run at a command-line prompt on modern systems.
I came here to make this exact comment. Note also that the rnd function generates a floating point number between 0 and the provided parameter, so it has to be rounded if the difference between characters is greater than one. That single line maze program looks like:
10 ?chr$(47+(cint(rnd(1))*45));:goto 10
(Tested in GWBasic 2.0)
@@michaelcalvin42 Indeed. I came here to see if I needed to comment similarly. I knew someone must have beat me to it, since it's such an obvious point for anyone that knows anything about programming. I assume Adrian just had a brain fart--or perhaps he doesn't know much about programming, even in BASIC?
@@michaelcalvin42 If your BASIC doesn't have a function for rounding (like C64 BASIC doesn't), you can use something like INT(RND(1) + 0.5); essentially add 0.5 and then convert it to an integer; then multiply by the offset and add the base like you already have.
It's amazing how this computer still works fine.
The redefinable character set hypothetically _could_ allow someone to make a really nice PETSCII Robots port for this.
I like the styling of it because it looks so similar to a VT100 terminal. If you could find a terminal emulator to run on it, you could use the computer as a dedicated terminal for other projects (I know you already have dedicated terminals, but why not one more).
Thank you for Archiving the C Compiling Software , Love this NEC , all the Great early 80's Computers . My brain can understand a Lot of Late 70's and Early 80's Computers and it has taken me 39 years to Learn What CPM and DOS that I Know , I sure Hope the collections are found by Young People and Put up on the Net for future Study , I Miss Gary Kildall , I Miss old BBS's , I Collect Modems , Books , Magazines and Computers , Cheres :) QC
This machine is beyond cool. I really hope you do some more videos with it!
Love the 10 print at the end. It's always interesting to see evolutionary dead-ends that die because of market reasons, not technology reasons.
Actually the NEC '765 floppy disc controller (and the Intel 8272A that's pin-compatible) doesn't need the index pulse to read. It only needs it to format a track; the formatting process writes a sector number into each sector on every track, which means it doesn't need the index pulse when reading or even writing- only when formatting.
Interesting! I guess the OS implementation still looks for the pulse anyway. Kind of rude! I know IMD even doesn't work if the index pulse can't be seen (while reading) and it's controlling the disk controller directly.
Sir: I love your videos ! I know essentially nothing about computers, but I learned more about floppy dicks in a few minutes of watching your running comments, than I ever previously knew. You have a great knowledge of the machines from years gone by, a knack for teaching, & excellent articulation. I am in awe & admiration of your work. Cheers!
"floppy dicks" hur hur hur I'm such a child.
I've never used 8 inch drives but I love 5 inch drives. Can't explain why. I think because when I was a child with the family Sinclair Spectrum I thought floppy drives were the coolest things ever. Haha. We are all of our time.
No, Adrian. OLD is the original command to load a program, in Dartmouth BASIC (DTSS),the command is specified in the ANSI BASIC standard, which both TI BASIC, and DRI Personal BASIC implement.
I can fully understand the owner using it well into the 90ties. With wordstar / DB3 and a C compiler, I would have been hard to move to a different setup until the Internet broke out :-)
He said the computer was made in '83. To state the obvious, '93 is "only" ten years later. That'd be like using Windows 8 now, which doesn't seem that absurd.
(Okay, that's not a great analogy.)
The pace of advancement back then was huge compared to now. You can easily use a 10 year old computer today and run Windows 10 on it with all modern software. (Except maybe the most demanding stuff.) Compare that to a 1981 IBM PC 5150 versus a 1991 clone with a 486 chip. The performance improvement in 10 years back then was like 100x faster in some cases. Now they just add more cores but the single thread speed hasn't increased that much.
@@adriansdigitalbasement Yeah, I was sort-of joking. That said, non-GUI DOS applications were still pretty common up to the early '90s, especially in a business context. If whatever company owned this computer had their records on it starting from '83, there had better be a good reason to migrate that data to a new system (and you discovered that it's very hard just to copy bytes, even ignoring potentially translating that data to a new application) or abandoning that data. If I was running a business using this machine for my records in the early '90s, I'd definitely think hard about what advantages existed on a new system that would outweigh the cost of new hardware and data migration. I wouldn't be surprised to find that the stuff you found marked "1993" were the artifacts of a migration effort. Why would a company using a word processor and a database (IIRC) suddenly have a C compiler?
Yes i clearly remember many people in 93 using pure text programs like wordstar, wordperfect and such on pc clones for business use, something graphical like Windows was seriously alienating and made things much slower, so it was a net loss on productivity. I clearly remember one person late 90ies still using her Wordperfect with pure MSDOS, she didn't want Windows or mouse, was in the way. All she needed was to type legal documents and print them, had everything configured margins, fonts, etc. that was it. She was using like a Pentium class computer with that exact same configuration, simply because the older ones died and since MSDOS did still work with later CPUs, it didn't matter. In parallel i have a friend using a Z80 QX10 for almost the exact thing until it broke, in addition to a newer PC for games 🙂
@@wbfaulk early 90s? Longer..I was doing data entry on vt100 terminals hooked to a mini computer early 2000s and my step-dad was doing dbiv programming up until about 2002 at least..lots of DOS stuff hung around longer than it probably should have..heck only a few years ago I saw a business still using what looked like maybe a 386 at best on dos for a bunch of business shit..I offered to trade them for a modern pc no cost because I wanted the retro machine and they said not a chance (was some weird fabric shop)
Even if I’d bought that disk way back then, brand new, I absolutely WOULD have duplicated it! I only use duplicates of purchased stuff and safely store the originals.
Don't copy that floppy! 💾💾
@@mal2ksc copy everything
@@mal2ksc shit is that what they've been saying? I always thought it was copy that floppy. Whoops.
I remember my word processor for my C64 (Starwriter?) said to do this as the first step in the manual. Make a copy, and keep the original disk safe.
@@ax14pz107 Actually when they see you sticking an original disk in the drive they are saying: “Don’t! Copy that floppy!” So the message is definitely saying make a backup.
the font on this machine is beautiful
When I was a kid, I took an after school computer class that had Apple II+ computers. The teacher said he had a black Bell & Howell version of the Apple II with 8" disk drives that had a 1.2MB capacity on their floppies. When I got older and into MS-DOS PCs, I remember hearing the 5-1/4" High Density disk capacity of 1.2MB was to match the capacity of 8" disks. I also remember seeing a 5-1/4 Quad Density disk drive in some PC's in the computer lab, which used disks with a capacity of 720KB.
AFAIK the IBM PC only requires the index hole during formatting. For reading and writing data it is not necessary. So it would be entirely possible to create a flippy disk for the IBM PC, as long as the formatting was specially created on a double-sided drive.
A couple others mentioned this too. I would say a test is in order!
Atari 8-bit slashes are ATASCII code 6 and 7.
That screen looks so pretty, imo a lot prettier than many other computers i've seen here
Don't sound so shocked Adrian... I used my CP/m system well into the early 90s! Wordstar was perfect for me and worked wonderfully... Only purchased an IBM clone in the mid-90s...
You don't need the characters next to each other for the maze one-liner.
chr$(x+y*int(rnd(1)*2))
would jump to either chr$(x) or chr$(x+y).
That's some pretty interesting computer history there! I never knew any of this... I guess this machine has the Y2K bug... Hahaha...
Thanks for the edumication Adrian...
@Adrian's Digital Basement 17:26 Does the disk labeled MSDOS 211 Practice Disk actually have ASM88 on it? or is that just ASM86. If it's ASM88, can you please upload that seperately with MASM, CSTUDIO.S and DEBUG? I couldn't find ASM88.
You can probably get the neon bulb to strike by shining a flashlight on it.
DeSmet compilers have been open-sourced as GPL. Only a few have source+manual though.
Super interesting to have a read through, so simple. Hardly any semantic analysis and no optimization!
It's nothing like the more advanced C compilers from the 80's and 90's such as DigitalMars that is for sure.
One of my favorite videos you’ve done in awhile. I have a pile of disks from my A500 sitting in the basement I haven’t touched in 30 some odd years. Someday I’ll go through them as you did with this computer. I hope.
DeSmet made compilers for several popular languages. They ran on CP/M and later MS-DOS. I think they later changed the name to Utah. I used the MS-DOS version of Utah COBOL and Utah Fortran 66 in college. Very inexpensive.
I have heard, and exoerienced, neon bulbs sometimes seem to need some ambient light to start. Being under the ledge may cause the delay.
When you were randomly pressing keys on the keyboard, it seemed to register the presses of the CAP, ALT and GR keys, though I guess that doesn't matter if you can't press anything else. Just an oddity I noticed.
The APC/PC-98 was certainly a much cleaner, better thought-through design than the IBM PC in many ways. It's a shame that it didn't catch on outside Japan. Speaking of PC-98, I wonder if it's disks are logically compatible with this machine's. IIRC, it used 5.25 inch 1.2 MB floppies but with a non-standard format that sounds a lot like the APC 8 inch layout. Maybe tools for reading/writing PC-98 disks on PC might help for this machine.
Retrobrite! That computer would look awesome if it was its original cream color!
The disk format on the 8" disks seems to be the same format the NEC PC-98 used for its 3.5" disks (which, incidentally, is the third supported format on USB floppy drives in addition to 1.44M and 720K). Maybe you can find software meant for working with PC-98 images or PC-98 disks on DOS/V?
Oh really that's interesting. I didn't know about the 3rd disk format support on USB 3.5" drives. I did some googling for PC-98 DOS support but don't see anything. Just lots of talk about the USB drive thing
I remember a "3 mode floppy support" BIOS option on some socket 7 boards, which is apparently for a 1.2 MB (not 1.44) 3.5" disk format used in Japan. Is that the same thing as the PC-98 format?
The Wikipedia page for the PC-9800 says "Early PC-9801 models supported 1232 KB 8-inch floppy drives and/or 640 KB 5-1⁄4-inch floppy drives. [...] High density 5-1⁄4-inch and 3-1⁄2-inch floppy disks use the same logical format and data rate as 1232 KB 8-inch floppy disks."
@@awd42 Aye that is the same format. PC-98 was the biggest computer platform in Japan before Windows 95 made its differences from PC-compatibles more or less moot.
I found CPM-86 and Concurrent CPM-86 on WinWorld.
Concurrent CPM-86 is a multitasking version of CPM .
In case you need it for archival or restoration purposes.
Adrian , thank you for the video. Wow.. what a beautiful monochrome computer. I liked the monitor. On this monitor, I was taught in school, basic) Even a little sad that almost can't do anything on this computer...But for some reason it seems to me that all the same all is not hopeless )
about the power switch and its Neon lamp; the bulb is failing- and the reason it "sometimes lights up" is based on a few things: 1 the little electrodes are dipped in a slightly radioactive coating which helps the Neon "fire" and this combined with it being exposed to photons (light) also will cause it ti fire at random. a way to "test" this is very simple, simply turn it on in a darkened room and then shine a light on the switch and you'll then see it light up!
Being a programmer and someone who is fluent in ANSI C those C dev disks are super interesting! #ColecoAdam!
Unfortunately the MAME NEC APC emulation seems to not work -- so no way for you to try them out.
@@adriansdigitalbasement Hmm maybe we can reach out to the MAME people? I always have trouble using MAME for microcomputers. It looks like most of the coding for the APC in MAME was done 8-10 years back.
There were about twelve games published by Infocom alone for NEC APC CP/M-86. Including "Zork", of course. So it's strange the one you tried did not work... There was no Zork for IBM PC CP/M-86 as far as I know - maybe that copy of the game is for some other CP/M machine, like DEC Rainbow?
Your year 2000 clock bug had me wondering if anyone had approached a fix for the eventual year 2038 problem. I know DOS will probably keep working, but it would be nice to have the correct dates.
MS-DOS doesn't have a year 2038 problem. That issue is when a signed 32-bit count of seconds past 12:00am, Jan 1, 1970 rolls over. MS-DOS records dates as 5 bits to record the day of the month, 4 bits to record the month, and 7 bits to record the number of years past 1980. That takes us through 2107, though apparently the APIs that access those dates have a problem with dates past 2099.
@@wbfaulk oh, TIL it didn’t just use Unix time
Hey, Dear Adrian, could you please do a video on how you go about creating an archive old floppy disk, particularly ones which belong to systems no longer in use. Among a bunch of 5.25" floppies I bought, here in the UK, I came across an "HH Tigerbyte" "Full CP/M" version 1.1 (or 1.2 if you go by the hand-writing rather than the printed label) for the HH Tiger computer - this was, apparently, a computer developed by Tangerine of Cambridge in the UK. However, they sold the design to HH (Audio) Electronics of Cambridge, and went on the develop the Oric 1, instead. By all accounts, that HH Tiger is an extra-ordinarily rare machine. Sadly, of course, I have absolutely no way of discovering if the original operating system is preserved on this disk, or simply over-written (as quite a lot of the other disks I picked up were). If it does survive, and as far as I can tell from a brief Google-search, this will not be archived and some collectors have these computers and cannot boot them. Maybe, just maybe…? So, how is it done???
Check out the ImageDisk program he uses in part of this video.
It is certainly something people struggle with -- so yeah I'll probably make a new video about it.
Unless you have a machine old enough to run DOS and have a floppy drive controller, you might want to look at getting a flux-level disk interface, such as a Greaseweazle, Kryoflux, Applesauce or FluxEngine (disclaimer: I made that last). This will allow you to archive pretty much anything which is physically compatible using a modern computer.
@@hjalfi That's interesting although I did restore an PC/AT compatible for the purpose.
I remember using that character editor on my dad's APC to make little animations.
loved using my hole punch, first hack
At the ZORK try: the top status line shows that it detected the CTRL, ALT and the GR presses, so at least something went trough...
Once every couple of years, I need to boot up an old PC 8088 to run a program called BoxCar that for some reason only works on an 8088 processor. The software designs Box Culverts. It is the only program available to do some particular calculations for the hunches. Fun DOS stuff.
Dosbox !
Unfortunately disk formats for MSDOS 2.11 on the various machines were not standardised.
So, this NEC has it's own format as do other machines such as the ACT Apricot (which has a non-standard header before the FAT), which has the same software problems.
I guess Microsoft didn't dictate the format and it was left to the manufacturers to decide?
Indeed. I think it was the same as Digital Research did with CP/M.
The FAT filesystem format itself seemed to be the same but the underlying encapsulation could change.
Interesting. Will later versions of MSDOS read and write to earlier disks, or is their some way to copy data back/forth between the formats?
It's more that it's system specific and Microsoft didn't specify the disk format.
In the case of the Apricot the first sector on the drive contains the fully drive parameters including the number heads, sectors and tracks on the medium, the sector offset to the FAT and what would be on a PC the BIOS settings.
Goodness knows what the NEC has there.
@@decidedly_retro What would it take to read the first sector of a floppy? Could it be done in gwbasic, for instance?
Too bad Zork didn't work. I wish I was better in coding so I could try to rewrite that Infocom interpreter to make those games playable on different systems.
By the way, I really like your new way of doing the split screen display while working on the computer connected to your capturing device.
The Infocom interpreter has already been rewritten many times and Infocom games can be played on virtually any computer. "Frotz" is the most notable Z-machine interpreter.
Yeah, just no modern version for APC MSDOS or APC CPU86 :-)
@@adriansdigitalbasement You've got that C compiler. I bet you could compile Frotz on there without too much effort.
Wasted potential having ram limit of 128kb and mentioned in last video highest was 256kb. Game may have entered in to a loop waiting for keyboard bios call to returning scan code. Not trying to be full pc compatible unless only dos functions called, common for clones.
It would be really cool to see what a color version could do.
It may only have a 2 digit year but it had the day of the week correct. October 30, 1922 was a Monday, so I'd wager it is Y2K compatible, at least with CP/M.
Neon bulbs - even newer ones - require a bit of light to "strike"; if this one is old and tired, it might need more light. Try flashing a flashlight into the power switch when it doesn't start right away to verify... There is a "strike in darkness" type, but I doubt that these were ever put into power indicators.
I've seen these neon indicators my whole life but I never knew that tidbit!
I love the shout out for the Coleco Adam!!!
When i was at school in the 80s we used to punch a second hole in the 5 1/4" disks to make them double sided using a small single hole punch. I think there were commercial ones on sale that prompted this called a "banana" disk.
4:24 - "and an Asterix" - that's an asterisk. Asterix is a Gallic warrior from Armorica. 😄
I let that slide, but don't axe me a question. That's battery.
it don't matter :D
Nearly all early PC software ran on non compatible MS-DOS machines. The two notable exceptions were Microsoft Flight Simulator and Lotus 1-2-3, but that was enough to tip the balance towards full clones. In the early days we didn't know if IBM was going to be as aggressive against clones as Apple had been, but once manufacturers saw that it was ok to make clones it didn't make sense to insist on non compatibles (like the TI Professional, the DEC Rainbow or the ones mentioned in the video)
I think early machines recreated the IBM PC architecture in hardware pretty much identically, but didn't have the IBM BIOS. This is what led to some incompatibility, but if the software tried to talk to hardware directly it would work. But this APC is completely different than the IBM PC with regards to the hardware architecture, so the software would have to rely 100% on DOS calls only to work.....
@@adriansdigitalbasement wonder if some modern trickery can be done to make it more compatible.
Something about seeing 2081 on a vintage screen gives me the chills :)
It's a little bit large for a pip-boy.
0:48:
_Adrian:_ "So I wanna see if I can get any MS-DOS software that was designed for the IBM PC actually working on this thing."
_Me, an idiot, but with related experience:_ Not without porting, you can't!
I wonder how hard it would be to port Attack of the Petscii Robots to the NEC APC...
@32:15 There were disk drives for the ADAM? Or is 22 disk just imaging the funky not-quite a normal philips standard compact cassette cassete tapes? I'm genuinely curious now, considering there's a complete (well everything but any old composite monitor.) ADAM up in the attic, next to the //C the c128 and a ][e... with the ADAM having the least sopporting software, other than 2600 cartridge compatability....
When you booted ms dos there was a program called rdcpm , maybe you could read cpm disks from ms dos on the nec apc, and get ms dos programs into it that way?
That charachterset is heavenly beautiful.
There was an oscilloscope card for the PET during the late 70s, a cheap way to replace an expensive Tektronix. If I remember the PET was also used to drive a size A or B x-y plotter about the same time. Back in the day a full blown large size (size E?) flatbed plotter CAD system would cost in the 10s of thousands of dollars.
Love this video, super calc2 for cp/m was one of my first spreadsheets. I recall my cp/m machine had a 3” (not 3 1/2”) drive and a 5 1.4” disk drive and I had to use tdos for cp/m to read 360k pc dos disks.
enjoyed all the videos great ton of information...great stuff very informative and entertaining
Really awsome machine. Maybe if you find a CP/M program that can write MS/DOS disks similarily like the tool for DOS that can write CP/M disks you can make files move with two hoops? Also maybe MAME can emulate this? I do not clearly see it if they have proper emulation, but in case they do, there must be some way to mount images of the disk, add files and write back whole images (or someone could make software for that who is capable enough).
This machine must have been very good for software development because of the added space on the disk drives. The disk that was labelled like "excercise" or what also seem to have some C compilers on it and such - maybe different ones?
I grew up in a college town. My dad was a NEC authorized dealer. When the university opened their own computer store with discounts all the other computer stores went under. That's how I ended up with a NEC APC in my bedroom as a six year old. He eventually found a buyer and I was given a NEC 8201A.
What did you do with the APC when you were that age? Fiddling around with Personal basic?