"Zork on my Wangwriter" - that's quite the niche fantasy! Is it perverted that I'm going to watch you do it? Well done all concerned... CP/M..? Awesome!
It was awesome to see for sure! We just gotta figure out the floppy BDOS errors we're getting so I can actually get deeper into the game before it crashes out!
There is a post on a other Wang writer restoration project on GitHub mentioning you and him got the system disk. Congratulations, I'll be waiting to see the next episode.
KeyTronic keyboards like this do not have conductive contacts - they are capacitive. The conductive side of the Mylar fades away from the pcb :) You can actually type on them by touching the PCB! Also, don't judge the feel of all F&F keyboards based on the TexElec pads. The foam they use is too firm, too tall, and sometimes the Mylar has issues too :(
Good catch! You're totally right, it's definitely capacitive and in the back of my head I knew that, but just totally forgot about it when filming. That's interesting that you mention the TexElec pads feeling different, because the keyboard definitely more affirmative to mock type on before I replaced the pads. Maybe after some solid use the TexElec pads will soften up or slim down a bit and the feeling will improve.
As with most 8-bit CPUs, programming the Z80 isn't particularly difficult once you understand assembly concepts. Difficulties more often come from the idiosyncrasies of the specific machine you are working with. Give it a shot. It's fun.
You got right into the point. Is like any microcontroller. You can even pretend you dont need to know the arquitecture because you work on C ... but once you got into the peripherals and the HW you have added ... things start to add up and in some ocasions C is not even good enough, and you can end with asm stubs (NEED FOR SPEED !!!!!!!). LOL
@@AlexanderGrahamSmell More intricate, but not really harder, but more features in the CPU. (For instance, most 8 bit CPUs didnt have multiply instructions, i think the 6809 was the exception to that) Lots of people programmed raw 68000/8086/VAX in the day. When memory limitations were a problem, every byte counted, and even the most efficient compiler wasted memory.
You nailed it, 256Byteram had the BIOS written fairly quickly, but it took us a ton more effort to figure out how on earth the Wangwriter was dealing with the floppy, and we're still not quite there yet!
@@UsagiElectric Amstrad used to manufacture a very popular range of Z80 machines (CPC for home and PCW for office) that used bank switched ram. Maybe the CP/M Plus source is available somewhere? en.wikipedia.org/wiki/Amstrad_CPC#CP/M en.wikipedia.org/wiki/Amstrad_PCW en.wikipedia.org/wiki/CP/M#CP/M_Plus
Did you watch Adrian's digital basement? I saw in his most recent video the CPM transfer program that listed the WangWriter. I thought of this channel when I saw that! He was trying to get Zork running on his NEC X86
@@CommodoreGreg I remember getting it years ago, and paying the shareware fee as well. Must look to see if I still have it on an old hard drive that has not been run in 20 years.....
I hope you can get Wordstar running as it seems an appropriate application for a computer attached to a printer. Back in the day, 1983, I was so proud of myself for mastering our z80 computer, getting it working with a daisy wheel printer, learning Wordstar and putting it all together to submit a junior high assignment. Wordstar even had formatting for things like bold, super and sub-script and possibly also underline. Bold was funny as the printer just hammered on the character several times before advancing the print head.
It is wonderful that 256bytes was willing to invest so much time in this for you. There are some amazingly generous people out there. You are very patient! As a funny side note, I saw a t-shirt that said, " I had my patience tested, I have none." I am fairly patient but I wonder how far I would go with this. I would like to think I would persevere but I don't know. Looking forward to more of the this project. Keep on plugging!
8:55 I had to restore a very similar Key Tronic F&F keyboard from around the same time. The big chip on the right is the keyboard encoder, one of the General Instrument 4592 series. The Interdesign 900c is a custom linear (analog) chip, which essentially amplifies the differences in capacitance when a key is pressed. You can see the likely circuit it contains at the bottom of Figure 1 ("Sample keyboard design rom coded keys") of the AY-3-4592 datasheet.
Oooh, interesting information on the 900C! I didn't think about it being an amplifier, but that makes total sense given that the capacitive key switches will definitely need amplifying and shaping.
@@UsagiElectric And Digikey says thay have them "in stock", though at $310 per each I think this is a Rochester kind of stock, where they will package them in a CERDIP package for you, if you order 64 of them.
Wow. That keyboard gave me flashbacks. Not sure if it was a ADDS Viewpoint 60, or a Liberty Electronics 200, but I know I've seen that PCB before. Nitpick: it's not based on conduction, but rather capacitive coupling. That way a bit of tarnish doesn't affect it.
Adrian Black used the program 22DISK in a recent video to read/write CPM disks on the PC. This program has support for lots of systems, including the WangWriter. Maybe it is of use to you.
Having disassembled, mangled and reassembled the TRS-80 MIII Infocom 'runtime' to make it run on several unusual Z80 based computers this warms my heart. :) The Infocom code seems to be 8080 instructions only, would have to check it more completely, so it'd be likely it could even be made to run on 8080 computers too.
Thank you! I'm still learning Z80 assembly as this is actually my first ever machine with a Z80 processor, but compared to the Centurion assembly, Z80 stuff is far more digestible!
@@UsagiElectric Big fan of the Z-80. Once you get used to it, it is actually quite easy to write assembly on it. I always found the 6502 to be far harder to write for, simply because of the 6502's lack of registers and address modes. Sure, the 6502 had zero page, which you could use as quasi register space, but that was very limited, if you wanted to leave basic and the DOS intact. I may be biased, as the Z-80 was the first processor i learned assembly on. But man i love that chip!
I found it interesting to see that the keyboard PCB and frame seems to be capable of supporting a full number pad and something where we would expect the escape key and a few keys around the nav cluster. Maybe an extended layout was an option? Personally I would find it interesting to expand the keyboard, but don't hack into the original cover, just 3d print an alternate one.
OMG. I first misread it as "Work on the Wangwriter," which I was more than ready to watch. Then I saw what it actually said, and I had to drop everything!
Very cool! I was always on the Apple II side of things so I never really messed about with CPM, but seeing Zork start brings back fond memories of 8 bit computing.
This is actually my first proper Z80 machine and my first times playing around with CPM. Unfortunately, the floppy crashes happen pretty regularly, so we still need to hammer out the floppy read problems we're having before I can get too deep into it, but once we got that squared away, it should be pretty easy to start getting some really interesting things happening!
I worked at a data entry company many years ago and used a Wang. One thing I'd like to say is that they made the BEST keyboards of all time, even better than the legendary IBM Model M "clicky" keyboard. It looked just like the one y0ou have there. It felt like you could drop it off of a skyscraper and it wouldn't even be scratched. Built like a tank.
Good job 256bytes!! I'm going to guess the BDOS error is the floppy doing some kind of time out and CP/M doesn't know how to turn it back on before doing disk access (or not allowing enough time).
I think you may be on the right track! I don't want to spoil too much, but we managed to get something that'll be a perfect reference for reading the floppy, so I'm guessing we'll be able to update our CPM relatively soon with the proper way to access the floppy!
Great episode, congrats to 256byteram to get this running. I think you're right, running Wordstar on a Wangwriter does not feel right. I remember Wordstar very well and found it a great wordprocessor at the time but I am as Curious as Marc to see the Wangwriter running it's very own wordprocessing software 🙂
I love how basically every episode you say you're way of of your debt, no matter the subject, and then proceed to do things people who are out of their debt absolutelly cannot do.
Word Star! It has been SOOO long ago but, I have made CP/M Word Star run on many oddball computers back in the '80s. Word Star does not require much RAM as uses two cache files on disk to keep track of the document. In the '90s I wrote in Borland Turbo Pascal and was happy that the editor used the Word Star control commands. ;)
Brilliant - 256byteram has helped make that Wang a very cool system. Looking forward to seeing the hardware issues sorted out (will likely be a fun piece of work).
Ah, finally someone who recognizes that I am smart. Programmed in assembly on the TRS-80 a.o. to hook it up as a terminal to PR1ME minicomputers at technical college. Those were the days!
I was sure that thing would work as a SBC so it's nice to see that I was right. 😃 I'm sure that the BDOS error is something related to timing for the FDD. Almost certainly, it needs more time to spin back up to speed before attempting to do a read or write operation.
If i remember correctly zork ran on a virtual machine similar to the later sierra games. One of is output options is a printer. Seems to me that the wang writer would make a perfect target to use the printer output to give you a paper trail...
Writing a printer driver will be a much bigger task than porting CP/M. The printer stepper motors and sensors are controlled directly by the main Z80 CPU. (Normally, a printer would have its own microcontroller to handle moving the paper, moving the print head, rotating the daisy wheel, and striking the hammer.)
@@Mueller3D that does complicated things, but if they can figure out the crt and the keyboard by only communicating over the net, I'm sure they can do it. Especially now they seem to have found the system disc. They can reverse engineer the routines needed.
I don't know CP/M but it looks like banked memory support was added with CP/M 3 which can use 3 to 16 banked pages with a common area that can be between ~4K and 32K.
Great videos! Most printers of that era had some sort of built in print test. This was often accomplished by a jumper or dip switch setting, but was sometimes enabled by holding a line feed button while the system was powered on. Not sure if you have looked for any of these to test the printer.
Tektronix computers of the day used the same type of keyboard. The plastic disc on the foam is Mylar with a layer of metal on the foam side. The connection is made capacitively. Feel free to use a DMM and prove me wrong. But that’s what I remember from the old Tek keyboards.
That microcontroller in the top right corner looks like it has a General Instrument logo on it, so it's possibly a precursor to modern day PIC microcontrollers.
I wonder what interpreter is running that Zork. It's not the mainframe version; it's the well-known version that Infocom made for microcomputers. (Well, of course it is. One of the major reasons they did that is the original couldn't fit on disks and in memories of micros of the time.) Specifically, mid-1984, according to the serial number. For portability, they created a special-purpose virtual machine and just wrote interpreters for the different computers they wanted to release to, so they could have the same "story file" per title on all of them, meaning a lot less effort in porting. I guess 256byteram just used one of the existing interpreters for CP/M. Or heck, now that I researched it a little, I guess it's just an existing copy with the interpreter and story file packaged together.
Actually have one of those keyboard boards, the microcontroller is a GI mask programmed part, which you generally need to get the mask code to get a new one, and each mask came with basically a customer code, and the GI logo and base type, and a date code.
My memory of KeyTronic keyboards is you don't feel the foam as you type, it's too soft compared to the normal keypress resistance. Sounds like those replacements are NQR - not quite right. I worked in a telco repair shop and we were forever disassembling them and cleaning out all the dust, that fixed 90% of them. People complain about modern rubber dome keyboards, compared to the stuff we handled back then they are golden, they just keep on working, unlike the junk we were saddled with. If they didn't jam on the way down (or back up), then they didn't register a press at all. Keytronics had good up and down action, shame about the 'contact', others had magnets and glass read switches, so that almost never failed, just the mechanics of the keystem would flog out with key jams all the time (Sagem Telex machines)
Going to guess Wang made CPM run by running the lot just under the paged area, using the paged area as essentially a ramdisk, or as data storage, so that programs could easily be storing data in there, but CPM would flip page when it needed the area, getting 16k of scratchpad RAM for internal operations, and probably using a few pages for disk use as a set of buffers, as 16k does allow a good few sectors to be read in at a time.
That's entirely possible, but I'm not sure how they'd go about loading disk data into the program space when CP/M is selected. Usually the BIOS loads 128 byte sectors to address 80h, but user programm can change that address. If it were changed to a bank that was switched out, it might corrupt CP/M or the software. There's need to be some sort of address check to make sure it doesn't happen. Doable, but tricky.
@@256byteram Well a Z80 can do DMA from memory to memory, so read a sector in, set up DMA to a selected page, and then read a sector in again. depends on just how many sectors the BIOS can transfer per DMA request, if you can do an entire track in one, then the DMA can occur during the step period to the next track transparently. If single track then it will be slow doing a sector per revolution.
I tried to re-house my Spectrum in an old foam and foil keyboard case back in 1985 as the original keyboard flex cables had broken. This did not work, as the key switches do _not_ rely on conduction but are capacitive. (if you check the pads for conduction with a multimeter you will probably find that they have a thin insulating coat on top) This actually marked the point that I gave up professional programming :-) (I had been making some money writing games, but had just started at University and didn't really have the time. The fact than I no longer had a working keyboard was the final decider.)
I had to replace all the foam pads on the SOL-20 I had. At the time, I had managed to purchase what I was told were the last 3 known sets of replacement pads. It's a PITA to clean up the keyboard and replace all those.
Did I see some numbers printed on the backside of the keyboard PCB? Are those by any chance the very scancodes you were looking for in the first place? And: It looks like there are pads on the PCB for a seperate number pad; were there different models of these keyboards, some with, some without a keypad?
The numbers on the back of the keyboard PCB don't appear to correspond to the scan codes. They seem just to be key position numbers. (The 1-4 keys are 20-23 on the back, but are scan codes 6E-6B.)
I just noticed in the short clip of your TI-99/4A keyboard that you have the original function key label strip thingy on there. Mine doesn't have that and I have been unable to find a good high-quality scan to print my own, could you maybe scan that in and publish it? I'd love to have one to make my 99/4A look complete :D
(@0:59) You can see from the memory dump that the string, “Restart POWER UP Procedure” has a terminating null, but no length byte, which would indicate this is a common C-string, while the string, “NOT A SYSTEM DISK” does not (if it had one, it’d be @ 0x0214, which is the 1st character of the “Restart…” string.) The strings are of different length, so they’re not in a table (array) of fixed-length strings either. So how does the printing routine know where the end of the “NOT A SYSTEM DISK” string is, as neither is a Pascal string, either?
This paged memory was common on CP/M machines. I have an emulator for the TRS-80 Model 4 that used paged memory - it put the video display into a "hidden" bank so that the OS had more memory to use. Also, maybe the BDOS error is caused by insufficient time being allowed after starting the drive before trying to read it? Does it only happen after the drive has stopped spinning?
The Keytronic keyboard on my freshly restored Victor 9000 failed in the worst way earlier this year. The mask programmed 8021 died. My first and only Arduino project was to replace the custom 8021 with an Arduino Nano. It works a treat once I worked out the detailed Victor documentation was very wrong in places!
Get your hands on a MEMOTECH MTX512 as it runs a ZILOG Z80A, as it has a package known as Assembler and to access it you just do ASSEM 10 RETURN also has a Dissembler known as PANEL and programs can be written in a mixture of BASIC and Assembler,
That's pretty cool youse got this working to such a degree that it's back to being a useful computer. Although kinda surprised about that foam and foil. I thought that was commonly a budget option, unless this was of an era where even something like this was a "budget" computer. If not, then it's a bit of penny-pinching. :p
Shame the keyboard is foam and foil, by memory I recall the old OIS terminal keyboards were all hall effect based switches... Love the old WANG stuff though!
256byteram... now THAT is a name I never thought I'd hear again. 'Met' him on Hackaday a few years ago, and he sent me a kit for a little CP/M system he'd come up with. My knowledge and resources proved insufficient and Shelby over at TechTangents now owns the hardware I once did. It was a really nice kit, I'm just a blinkin idiot with the iron sometimes! 256, if by some chance you see this, maybe you could help Shelby finally breathe some life into that thing?
@@pc4ad All the traces seem to be there , So i think it would work based on what we can see in the video. Werid they did not add them as the shortcuts with the numkeys was used alot in word software.
![ILLSLICK - THE KING'S LANDING [Freestyle]](http://i.ytimg.com/vi/feIJJnNkTfk/mqdefault.jpg)
Way to engage the community! Congrats to 256bytes, amazing job.
THank you! And 256Byteram absolutely killed it in this one!
An excellent crossover between two 1970s-80s computer nerds. Love both of y’all’s videos!
"Zork on my Wangwriter" - that's quite the niche fantasy!
Is it perverted that I'm going to watch you do it?
Well done all concerned...
CP/M..? Awesome!
Zork on Wang. What a time to be alive.
But will it run Zork II?!
It was awesome to see for sure! We just gotta figure out the floppy BDOS errors we're getting so I can actually get deeper into the game before it crashes out!
I agree. 256ByteRam's effort was outstanding!
He's way smarter than me and this project has been crazy exciting thanks to his efforts!
There is a post on a other Wang writer restoration project on GitHub mentioning you and him got the system disk. Congratulations, I'll be waiting to see the next episode.
I see you found Rich's blog!
But, yup, the next episode is in the editing phase and now and it should be a fun one!
KeyTronic keyboards like this do not have conductive contacts - they are capacitive. The conductive side of the Mylar fades away from the pcb :)
You can actually type on them by touching the PCB!
Also, don't judge the feel of all F&F keyboards based on the TexElec pads. The foam they use is too firm, too tall, and sometimes the Mylar has issues too :(
Good catch! You're totally right, it's definitely capacitive and in the back of my head I knew that, but just totally forgot about it when filming.
That's interesting that you mention the TexElec pads feeling different, because the keyboard definitely more affirmative to mock type on before I replaced the pads. Maybe after some solid use the TexElec pads will soften up or slim down a bit and the feeling will improve.
Great work 256byte ram. You deserve a 512 byte upgrade!
As with most 8-bit CPUs, programming the Z80 isn't particularly difficult once you understand assembly concepts. Difficulties more often come from the idiosyncrasies of the specific machine you are working with. Give it a shot. It's fun.
You got right into the point. Is like any microcontroller. You can even pretend you dont need to know the arquitecture because you work on C ... but once you got into the peripherals and the HW you have added ... things start to add up and in some ocasions C is not even good enough, and you can end with asm stubs (NEED FOR SPEED !!!!!!!). LOL
@@AlexanderGrahamSmell More intricate, but not really harder, but more features in the CPU. (For instance, most 8 bit CPUs didnt have multiply instructions, i think the 6809 was the exception to that)
Lots of people programmed raw 68000/8086/VAX in the day. When memory limitations were a problem, every byte counted, and even the most efficient compiler wasted memory.
You nailed it, 256Byteram had the BIOS written fairly quickly, but it took us a ton more effort to figure out how on earth the Wangwriter was dealing with the floppy, and we're still not quite there yet!
@@UsagiElectric Amstrad used to manufacture a very popular range of Z80 machines (CPC for home and PCW for office) that used bank switched ram. Maybe the CP/M Plus source is available somewhere?
en.wikipedia.org/wiki/Amstrad_CPC#CP/M
en.wikipedia.org/wiki/Amstrad_PCW
en.wikipedia.org/wiki/CP/M#CP/M_Plus
CP/M was originally written for the Intel 8080. With the bundled tools using Intel, rather than Zilog, opcodes.
The new "I'm ready to rock!" is: "I've got my Grease Weasel plugged in!"
This is so amazing, I love these kind of community projects and this is just... magic. Thanks to all!
Thank you so much!
Did you watch Adrian's digital basement? I saw in his most recent video the CPM transfer program that listed the WangWriter. I thought of this channel when I saw that! He was trying to get Zork running on his NEC X86
22disk? That's good to know!
@@HoJoGoGo Yes, 22disk
@@CommodoreGreg I remember getting it years ago, and paying the shareware fee as well. Must look to see if I still have it on an old hard drive that has not been run in 20 years.....
I hope you can get Wordstar running as it seems an appropriate application for a computer attached to a printer. Back in the day, 1983, I was so proud of myself for mastering our z80 computer, getting it working with a daisy wheel printer, learning Wordstar and putting it all together to submit a junior high assignment. Wordstar even had formatting for things like bold, super and sub-script and possibly also underline. Bold was funny as the printer just hammered on the character several times before advancing the print head.
It is wonderful that 256bytes was willing to invest so much time in this for you. There are some amazingly generous people out there. You are very patient! As a funny side note, I saw a t-shirt that said, " I had my patience tested, I have none." I am fairly patient but I wonder how far I would go with this. I would like to think I would persevere but I don't know. Looking forward to more of the this project. Keep on plugging!
8:55 I had to restore a very similar Key Tronic F&F keyboard from around the same time. The big chip on the right is the keyboard encoder, one of the General Instrument 4592 series.
The Interdesign 900c is a custom linear (analog) chip, which essentially amplifies the differences in capacitance when a key is pressed. You can see the likely circuit it contains at the bottom of Figure 1 ("Sample keyboard design rom coded keys") of the AY-3-4592 datasheet.
Oooh, interesting information on the 900C! I didn't think about it being an amplifier, but that makes total sense given that the capacitive key switches will definitely need amplifying and shaping.
@@UsagiElectric And Digikey says thay have them "in stock", though at $310 per each I think this is a Rochester kind of stock, where they will package them in a CERDIP package for you, if you order 64 of them.
Wow. That keyboard gave me flashbacks. Not sure if it was a ADDS Viewpoint 60, or a Liberty Electronics 200, but I know I've seen that PCB before. Nitpick: it's not based on conduction, but rather capacitive coupling. That way a bit of tarnish doesn't affect it.
That's an astounding feat of technical wizardry!
Technical Zorkery
@@hicknopunk "What sort of Zorkery is this?!"
Thank you! 256Byteram is properly brilliant, I'm just pushing buttons and flipping switches, he did all the proper hard work!
That's so cool, Nice work 256bytes. :)
Hey kids, tell your mother you want a Grease Weasel for Christmas!
Adrian Black used the program 22DISK in a recent video to read/write CPM disks on the PC. This program has support for lots of systems, including the WangWriter. Maybe it is of use to you.
Having disassembled, mangled and reassembled the TRS-80 MIII Infocom 'runtime' to make it run on several unusual Z80 based computers this warms my heart. :) The Infocom code seems to be 8080 instructions only, would have to check it more completely, so it'd be likely it could even be made to run on 8080 computers too.
Thank you!
I'm still learning Z80 assembly as this is actually my first ever machine with a Z80 processor, but compared to the Centurion assembly, Z80 stuff is far more digestible!
@@UsagiElectric Big fan of the Z-80. Once you get used to it, it is actually quite easy to write assembly on it. I always found the 6502 to be far harder to write for, simply because of the 6502's lack of registers and address modes. Sure, the 6502 had zero page, which you could use as quasi register space, but that was very limited, if you wanted to leave basic and the DOS intact.
I may be biased, as the Z-80 was the first processor i learned assembly on. But man i love that chip!
"A lot of dead keys on this keyboard..." "BAH GAWD THAT'S ADRIAN BLACK'S MUSIC!"
I found it interesting to see that the keyboard PCB and frame seems to be capable of supporting a full number pad and something where we would expect the escape key and a few keys around the nav cluster. Maybe an extended layout was an option?
Personally I would find it interesting to expand the keyboard, but don't hack into the original cover, just 3d print an alternate one.
OMG. I first misread it as "Work on the Wangwriter," which I was more than ready to watch. Then I saw what it actually said, and I had to drop everything!
Very cool! I was always on the Apple II side of things so I never really messed about with CPM, but seeing Zork start brings back fond memories of 8 bit computing.
CP/M works great on an Apple ][ with a Z80 Softcard. It was my first dual CPU computer, lol.
This is actually my first proper Z80 machine and my first times playing around with CPM. Unfortunately, the floppy crashes happen pretty regularly, so we still need to hammer out the floppy read problems we're having before I can get too deep into it, but once we got that squared away, it should be pretty easy to start getting some really interesting things happening!
I worked at a data entry company many years ago and used a Wang. One thing I'd like to say is that they made the BEST keyboards of all time, even better than the legendary IBM Model M "clicky" keyboard. It looked just like the one y0ou have there. It felt like you could drop it off of a skyscraper and it wouldn't even be scratched. Built like a tank.
Good job 256bytes!! I'm going to guess the BDOS error is the floppy doing some kind of time out and CP/M doesn't know how to turn it back on before doing disk access (or not allowing enough time).
I think you may be on the right track! I don't want to spoil too much, but we managed to get something that'll be a perfect reference for reading the floppy, so I'm guessing we'll be able to update our CPM relatively soon with the proper way to access the floppy!
hmmmm.... "BDOS Error".... that takes me back to my youth!
I dunno... on many CP/M systems 47K is a MASSIVE amount of memory.
Good work Mr. 256
Great episode, congrats to 256byteram to get this running. I think you're right, running Wordstar on a Wangwriter does not feel right. I remember Wordstar very well and found it a great wordprocessor at the time but I am as Curious as Marc to see the Wangwriter running it's very own wordprocessing software 🙂
I love how basically every episode you say you're way of of your debt, no matter the subject, and then proceed to do things people who are out of their debt absolutelly cannot do.
* depth
@@andrewdunbar828 I stand corrected. I'm dutch, so English is not my native language.
@@janklas7079 No worries mate. My Dutch is a lot worse.
The power of collaboration - amazing progress. Hellorld is wearing a big thin though- it was amusing once or twice 😊
Word Star! It has been SOOO long ago but, I have made CP/M Word Star run on many oddball computers back in the '80s. Word Star does not require much RAM as uses two cache files on disk to keep track of the document. In the '90s I wrote in Borland Turbo Pascal and was happy that the editor used the Word Star control commands. ;)
Brilliant - 256byteram has helped make that Wang a very cool system. Looking forward to seeing the hardware issues sorted out (will likely be a fun piece of work).
Sublime stuff going on here.. love it.
Ah, finally someone who recognizes that I am smart. Programmed in assembly on the TRS-80 a.o. to hook it up as a terminal to PR1ME minicomputers at technical college. Those were the days!
Pr1me was a Honeywell spinoff, actually some decent minis in their day.
Was l33t speak already a thing back then lol.
@@belstar1128 Pr1me invented it
Tremendous work!
It's good to see your 'Wang" is working better. ; ]
I remember Zork on my Atari 800XL - I would have taken in the moments much more had I have realised how special those days were.
I was sure that thing would work as a SBC so it's nice to see that I was right. 😃 I'm sure that the BDOS error is something related to timing for the FDD. Almost certainly, it needs more time to spin back up to speed before attempting to do a read or write operation.
First!
Your videos are such a inspiration!
If i remember correctly zork ran on a virtual machine similar to the later sierra games. One of is output options is a printer. Seems to me that the wang writer would make a perfect target to use the printer output to give you a paper trail...
I first played Planetfall on a printer.
It is a virtual machine, yes. Z-Machine version 3.
Writing a printer driver will be a much bigger task than porting CP/M. The printer stepper motors and sensors are controlled directly by the main Z80 CPU. (Normally, a printer would have its own microcontroller to handle moving the paper, moving the print head, rotating the daisy wheel, and striking the hammer.)
@@Mueller3D that does complicated things, but if they can figure out the crt and the keyboard by only communicating over the net, I'm sure they can do it. Especially now they seem to have found the system disc. They can reverse engineer the routines needed.
I don't know CP/M but it looks like banked memory support was added with CP/M 3 which can use 3 to 16 banked pages with a common area that can be between ~4K and 32K.
Awesome progress!! amazing!
You have to go to CP/M version 3 or greater to make easy use to bank-switched RAM, I believe.
Hey! Don't poke the bunny! 😄
Great videos! Most printers of that era had some sort of built in print test. This was often accomplished by a jumper or dip switch setting, but was sometimes enabled by holding a line feed button while the system was powered on. Not sure if you have looked for any of these to test the printer.
Tektronix computers of the day used the same type of keyboard. The plastic disc on the foam is Mylar with a layer of metal on the foam side. The connection is made capacitively.
Feel free to use a DMM and prove me wrong. But that’s what I remember from the old Tek keyboards.
That microcontroller in the top right corner looks like it has a General Instrument logo on it, so it's possibly a precursor to modern day PIC microcontrollers.
I wonder what interpreter is running that Zork. It's not the mainframe version; it's the well-known version that Infocom made for microcomputers. (Well, of course it is. One of the major reasons they did that is the original couldn't fit on disks and in memories of micros of the time.) Specifically, mid-1984, according to the serial number. For portability, they created a special-purpose virtual machine and just wrote interpreters for the different computers they wanted to release to, so they could have the same "story file" per title on all of them, meaning a lot less effort in porting. I guess 256byteram just used one of the existing interpreters for CP/M. Or heck, now that I researched it a little, I guess it's just an existing copy with the interpreter and story file packaged together.
Actually have one of those keyboard boards, the microcontroller is a GI mask programmed part, which you generally need to get the mask code to get a new one, and each mask came with basically a customer code, and the GI logo and base type, and a date code.
My memory of KeyTronic keyboards is you don't feel the foam as you type, it's too soft compared to the normal keypress resistance. Sounds like those replacements are NQR - not quite right.
I worked in a telco repair shop and we were forever disassembling them and cleaning out all the dust, that fixed 90% of them.
People complain about modern rubber dome keyboards, compared to the stuff we handled back then they are golden, they just keep on working, unlike the junk we were saddled with. If they didn't jam on the way down (or back up), then they didn't register a press at all. Keytronics had good up and down action, shame about the 'contact', others had magnets and glass read switches, so that almost never failed, just the mechanics of the keystem would flog out with key jams all the time (Sagem Telex machines)
This video is riveting and scintillating!
Z80 assembly is pretty easy itself but interacting with hardware can be quite tricky and require more arcane knowledge.
WOW .... Superb job !!!!!!!.
Im slightly creeped out by the dog sitting behind the Centurion wearing french fries on its head
Going to guess Wang made CPM run by running the lot just under the paged area, using the paged area as essentially a ramdisk, or as data storage, so that programs could easily be storing data in there, but CPM would flip page when it needed the area, getting 16k of scratchpad RAM for internal operations, and probably using a few pages for disk use as a set of buffers, as 16k does allow a good few sectors to be read in at a time.
That's entirely possible, but I'm not sure how they'd go about loading disk data into the program space when CP/M is selected. Usually the BIOS loads 128 byte sectors to address 80h, but user programm can change that address. If it were changed to a bank that was switched out, it might corrupt CP/M or the software. There's need to be some sort of address check to make sure it doesn't happen. Doable, but tricky.
@@256byteram Well a Z80 can do DMA from memory to memory, so read a sector in, set up DMA to a selected page, and then read a sector in again. depends on just how many sectors the BIOS can transfer per DMA request, if you can do an entire track in one, then the DMA can occur during the step period to the next track transparently. If single track then it will be slow doing a sector per revolution.
I tried to re-house my Spectrum in an old foam and foil keyboard case back in 1985 as the original keyboard flex cables had broken.
This did not work, as the key switches do _not_ rely on conduction but are capacitive. (if you check the pads for conduction with a multimeter you will probably find that they have a thin insulating coat on top)
This actually marked the point that I gave up professional programming :-)
(I had been making some money writing games, but had just started at University and didn't really have the time. The fact than I no longer had a working keyboard was the final decider.)
I heard that Early Computers Project has the system disk for the wang writer.
You may want to ask them if they can lend it to you.
I had to replace all the foam pads on the SOL-20 I had. At the time, I had managed to purchase what I was told were the last 3 known sets of replacement pads. It's a PITA to clean up the keyboard and replace all those.
Did I see some numbers printed on the backside of the keyboard PCB? Are those by any chance the very scancodes you were looking for in the first place? And: It looks like there are pads on the PCB for a seperate number pad; were there different models of these keyboards, some with, some without a keypad?
The numbers on the back of the keyboard PCB don't appear to correspond to the scan codes. They seem just to be key position numbers.
(The 1-4 keys are 20-23 on the back, but are scan codes 6E-6B.)
@@Mueller3D Just add 110 (dec)! 😁 ...ok, now I'm really just grasping at straws. 😂
That Keytronics keyboard looks familiar. I bought a similar one a long time ago and I did some reverse engineering on it.
Am I wrong or are the scan codes of the keyboard printed in base10 on the back of the PCB? You can see the numbers when he disassembled the keyboard.
I just noticed in the short clip of your TI-99/4A keyboard that you have the original function key label strip thingy on there. Mine doesn't have that and I have been unable to find a good high-quality scan to print my own, could you maybe scan that in and publish it? I'd love to have one to make my 99/4A look complete :D
interdesign 900C may be some sort of gate array. they were later built by ferranti (which did gate arrays)
So the keyboard's PCB has circuitry for a missing numeric keypad. Cost saving for this model?
(@0:59) You can see from the memory dump that the string, “Restart POWER UP Procedure” has a terminating null, but no length byte, which would indicate this is a common C-string, while the string, “NOT A SYSTEM DISK” does not (if it had one, it’d be @ 0x0214, which is the 1st character of the “Restart…” string.) The strings are of different length, so they’re not in a table (array) of fixed-length strings either. So how does the printing routine know where the end of the “NOT A SYSTEM DISK” string is, as neither is a Pascal string, either?
This paged memory was common on CP/M machines. I have an emulator for the TRS-80 Model 4 that used paged memory - it put the video display into a "hidden" bank so that the OS had more memory to use. Also, maybe the BDOS error is caused by insufficient time being allowed after starting the drive before trying to read it? Does it only happen after the drive has stopped spinning?
Awesome, now do Doom!
Or Bad Apple!!
The Keytronic keyboard on my freshly restored Victor 9000 failed in the worst way earlier this year. The mask programmed 8021 died. My first and only Arduino project was to replace the custom 8021 with an Arduino Nano. It works a treat once I worked out the detailed Victor documentation was very wrong in places!
I wonder if he's the same 256byteram who built the 8080 computer which plays "Still Alive"? (Also on TH-cam)
Get your hands on a MEMOTECH MTX512 as it runs a ZILOG Z80A, as it has a package known as Assembler and to access it you just do ASSEM 10 RETURN also has a Dissembler known as PANEL and programs can be written in a mixture of BASIC and Assembler,
What is a creased weasel?
That's pretty cool youse got this working to such a degree that it's back to being a useful computer. Although kinda surprised about that foam and foil. I thought that was commonly a budget option, unless this was of an era where even something like this was a "budget" computer. If not, then it's a bit of penny-pinching. :p
Awesome work, but can it run Snek?
I had a Wang word processor which had the Adventure text game. So Colossal Cave, 350 point version.
Legend!
Shame the keyboard is foam and foil, by memory I recall the old OIS terminal keyboards were all hall effect based switches... Love the old WANG stuff though!
This looks very, very like the keyboard in the Compaq Portable Plus I'm trying to restore. Texelec's kit made short work out of this job.
Is this episode early? I usually find the new episode upload some time during the night (after midnight) monday morning, yes I live in CET timezone.
Yup, it was a few hours earlier than normal! I figured I'd kick it up early since there wasn't an episode last week.
très bon travail
14:39 but there's still a cable unpluged...it's for the printer board yes but still
I'm amazed by the community. And big thanks to 256byteram!
so, are those numpad contacts wired up to the MCU?
That was pretty
Respect
Would it be considered bad form to modify the hardware to allow selectable ram at c000?
Bunny!
256byteram hasn't posted to his channel in a while.
looks like the 900c might be some kind of clock gen ic from what I found online
could also be some sort of gate array.
素晴らしい!
+5 (age) Years for the Beard! :D
Amazing work!
But holy duck, how are you still alive using Discord light mode?!?
AFAIK CP/M cannot deal with paging and is generally limited to 64k of RAM.
Please search for Compucorp 675 too. I have it and no disks.
256byteram... now THAT is a name I never thought I'd hear again.
'Met' him on Hackaday a few years ago, and he sent me a kit for a little CP/M system he'd come up with. My knowledge and resources proved insufficient and Shelby over at TechTangents now owns the hardware I once did. It was a really nice kit, I'm just a blinkin idiot with the iron sometimes!
256, if by some chance you see this, maybe you could help Shelby finally breathe some life into that thing?
Why do you say “et” instead of “at”? Pretty different. ;)
The keyboard is capacitive the pads are isolated.
Wangwriter. Wasn't that Lauren Boebert's hubby's tattoo artist?
Its because a z80 its a z80 everywhere. Know what the biggest installed base is? Alarm systems
Could you 3dprint a new layout for the keyboard , then add more keys to the keyboard such as numpad?
Saw that as well, and there seems to be space for an 'escape' key as well
@@pc4ad All the traces seem to be there , So i think it would work based on what we can see in the video. Werid they did not add them as the shortcuts with the numkeys was used alot in word software.
But does it run Doom?