[M100] Homemade 8085

[Kevin Becker]

This kid made a CPU on breadboards which he says is based on the 8085.

https://www.youtube.com/watch?v=g_ZaioqF1B0

Re: [M100] DVI help

[Randall Kindig]

Brian, you are absolutely correct.  You sent me everything I possibly needed to 
understand how this should work and I managed to ignore some of it.

I really, really appreciate all your help.

I’m just pumped that both DVI’s work when for a while I thought neither did.

I will check the disk to see if has the 200 files on it, although I don’t have 
plans to get a T200.  I had one a few years ago and sold it.

Thanks again!

Randy
> On Feb 19, 2018, at 2:42 AM, Brian White  wrote:
> 
> The manual, which is on-line btw, does mention somewhere in there that the 
> first step loads software from the disk to the dvi. Then the 2nd step loads 
> from the dvi to the M100. It also shows the backup procedure and all kinds of 
> other stuff.
> 
> Also, 3 observable things do happen when the M100 is loaded successfully, 
> it's just that none of them obviously look like "Everything worked, we're 
> ready."
> 
> 1) The M100 menu screen blanks and redraws.
> 
> 2) The free memory display on the M100 menu reduces by about 5K.
> 
> 3) The TV screen clears. The the Microsoft/Tandy boilerplate goes away 
> leaving only a cursor.
> 
> I should have thought to ask if you saw these in the first place. I did say 
> "nothing happens" but that still isn't good enough because how could you know 
> what counts as "something" when it's just things like that?
> 
> Though, I believe this is all in the video I posted. Did I ever post a link 
> here? I don't remember, maybe I only posted that to facebook...
> https://youtu.be/Mt6AmyLKkXQ 
> Not wait, I did post it here because didn't you point out that it looked like 
> I was using the serial port? So you saw the entire boot process. So you have 
> no excuse after all :)
> 
> The service manual is also on-line which describes even more about how the 
> dvi works.
> 
> http://www.classiccmp.org/cini/pdf/Tandy/Disk%20Video%20Interface%20Manual.pdf
>  
> 
> 
> http://www.classiccmp.org/cini/pdf/Tandy/Disk%20Video%20Interface%20Service%20Manual.pdf
>  
> 
> 
> It went similar with me. I thought I had 2 dead DVIs and 3 bad original 
> system disks, Then it turned out I had 2 good dvis, 3 good original system 
> disks, and merely one bad drive in one dvi. But I also had an unconfirmed 
> cable I made myself too, not one anyone else claimed was correct and tested. 
> I had no reason to believe I had it wired right. Now after finding some 360K 
> drives on ebay and finding just enough documentation on-line to set the 
> jumpers right, I have 2 fully working dvis, with 2 drives each no less.
> 
> Now the last thing is, does your system disk contain files with 200 in the 
> name, like FMT200 and BAK200 ? And if not, do you have or plan to have a 200? 
> There are two versions of the system disk. One is for 100 & 102 only, the 
> other is for 100, 102, & 200. If you don't have the 100/200 system disk I can 
> make a copy.
> 
> -- 
> bkw
> 
> On Feb 18, 2018 6:31 PM, "Randall Kindig"  > wrote:
> Wow, I hooked up the original DVI box now that I know how all this works, and 
> since I verified the cable/M100/boot disk combination that works.
> 
> Guess what, it works too!!
> 
> I think what was throwing me was that when you reset the M100, the disk light 
> on the DVI doesn’t even come on most times.  You just hear a very brief disk 
> noise from it, a second or two, and that must be all it takes to load the 
> DISK BASIC.  You then go into BASIC on the M100, type SCREEN 1,1 and voila!
> 
> I had originally assumed the disk drive light would come on and that it would 
> take several seconds to load DISK BASIC, so I was looking for that behavior.
> 
> I now have a preponderance of riches.  Two working DVI boxes.  I’ll probably 
> sell a fully tested system on eBay with cable, boot disk, M100, DVI and 
> recover some of the funds I spent on all this.
> 
> Thanks to everyone who helped with this. Especially you, Brian.  I owe you 
> that beer :)
> 
> Randy
> 
> 
> 
>> On Feb 18, 2018, at 5:45 PM, Randall Kindig > > wrote:
>> 
>> This worked perfectly, Brian!  I now have a backup system disk and both have 
>> the write-protect tab on them.
>> 
>> thanks!
>> 
>> By the way, the system is consistently booting now.  Just had to shake the 
>> dust off of it, I guess.
>> 
>> Still would like to see if I can get the other DVI box working.
>> 
>> Randy
>> 
>> 
>>> On Feb 17, 2018, at 10:24 PM, Brian White >> > wrote:
>>> 
>>> The easiest way is using the backup program on the system disk.
>>> 
>>> It's possible but I think it's probably more difficult with a modern pc. I 
>>> don't think just any modern floppy controller chip 

Re: [M100] TDock

[Joan Leach]

I remember reading in Puppy's Barry blog when he was building out his Pi 
systems that he used 5-way expansion to his various RAM cards and 
devices...that should help?
I have a netbook, laptop and xeon server that cannot use standard hard drives 
for various reasons, so I use a read-only boot device and save out to another.
Joan in Reno

  From: John R. Hogerhuis 
 To: m...@bitchin100.com 
 Sent: Monday, February 19, 2018 11:37 AM
 Subject: Re: [M100] TDock
   


On Mon, Feb 19, 2018 at 11:26 AM, Kevin Becker  wrote:

Running the Pi from a read-only filesystem would prevent shutdown problems.
https://www.raspberrypi.org/ blog/adafruits-read-only/




True, but then you can't use it for storage, right?

If you're just using it as a display and a bridge to cloud storage that would 
be sufficient.

I guess root filesystem were r/o, set noatime, and all logs went to ramdisk 
with strict growth /cleanup limits, but store all modifiable files to a 
separate partition or usb key.

Then you can store files, but at least you stay bootable in the face of any 
corruption.
-- John. 

   

Re: [M100] modern cross-assemblers for M100?

[John R. Hogerhuis]

On Mon, Feb 19, 2018 at 11:15 AM, Ken Pettit  wrote:

> On 2/19/18 10:29 AM, John R. Hogerhuis wrote:
>
>
>
> On Mon, Feb 19, 2018 at 10:10 AM, megarat  wrote:
>
>>
>> So I'm curious about what cross-assemblers people would use today for
>> developing such software on the M100.  Is there a standard "go-to" package
>> that is used by the Model T community?  I noticed GNUSim8085, which seems
>> intriguing, but I haven't looked at it too closely.
>>
>>
>>
> VirtualT. There's a rudimentary debugger built in, not source level but
> that's less important for assembly code since it does show the instructions.
>
> As to assembler, as (as8085) is freely available.
>
>
>
> VirtualT also has a built-in assembler.  I use it exclusively for writing
> 8085 assembly for the Model T.  It is what I used to write AsciiPixels,
> TDock and several other assembly packages that I have yet to complete or
> release (but that have A LOT of code written).
>
> In my Personal Libraries section at Club100, there is a simple example
> project for the VirtualT Assembler / IDE under "VirtualT IDE".
>
> Ken
>


Ah, sorry I intended to mention that. I think I finally got a working
project going at some point.

-- John.

Re: [M100] TDock

[John R. Hogerhuis]

On Mon, Feb 19, 2018 at 11:26 AM, Kevin Becker 
wrote:

> Running the Pi from a read-only filesystem would prevent shutdown problems.
>
> https://www.raspberrypi.org/blog/adafruits-read-only/
>
>
>
True, but then you can't use it for storage, right?

If you're just using it as a display and a bridge to cloud storage that
would be sufficient.

I guess root filesystem were r/o, set noatime, and all logs went to ramdisk
with strict growth /cleanup limits, but store all modifiable files to a
separate partition or usb key.

Then you can store files, but at least you stay bootable in the face of any
corruption.

-- John.

Re: [M100] TDock

[Kevin Becker]

Running the Pi from a read-only filesystem would prevent shutdown problems.

https://www.raspberrypi.org/blog/adafruits-read-only/



On Mon, Feb 19, 2018 at 1:22 PM, John R. Hogerhuis  wrote:

>
>
> On Mon, Feb 19, 2018 at 10:04 AM, megarat  wrote:
>
>>
>> A good counter-argument to this is that if using a computer as a
>> foundation for building your new device is the easiest, cheapest, and/or
>> quickest way to do it, then those benefits might be worth the cost of
>> having the computer involved.
>>
>>
> I'd agree with that.
>
> The Pi is also small and a popular platform. Lots of support and
> peripherals. It kind of future proofs us. It makes it possible to extend
> the system yourself or with applications written by others.
>
> But for that same reason there's a lot of stuff to "turn off" or configure
> for our purpose and possibly some hardware beyond just the I/O to make it
> stable.
>
> Persistent linux filesystems are not compatible with being hard power
> cycled. Journaling filesystems help but even those seem to get corrupted in
> my experience.
>
> https://github.com/wware/rpi-shutdown
> https://hackaday.io/project/25107-single-supercapacitor-
> ups-for-raspberry-pi
> https://www.pi-supply.com/product/pi-ups-uninterrupted-
> power-supply-raspberry-pi/
> https://www.modmypi.com/raspberry-pi/power-1051/ups-boards-1051/ups-pico
>
> -- John.
>

Re: [M100] modern cross-assemblers for M100?

[Ken Pettit]

On 2/19/18 10:29 AM, John R. Hogerhuis wrote:



On Mon, Feb 19, 2018 at 10:10 AM, megarat > wrote:



So I'm curious about what cross-assemblers people would use today
for developing such software on the M100.  Is there a standard
"go-to" package that is used by the Model T community?  I noticed
GNUSim8085, which seems intriguing, but I haven't looked at it too
closely.



VirtualT. There's a rudimentary debugger built in, not source level 
but that's less important for assembly code since it does show the 
instructions.


As to assembler, as (as8085) is freely available.



VirtualT also has a built-in assembler.  I use it exclusively for 
writing 8085 assembly for the Model T.  It is what I used to write 
AsciiPixels, TDock and several other assembly packages that I have yet 
to complete or release (but that have A LOT of code written).


In my Personal Libraries section at Club100, there is a simple example 
project for the VirtualT Assembler / IDE under "VirtualT IDE".


Ken

Re: [M100] modern cross-assemblers for M100?

[John R. Hogerhuis]

On Mon, Feb 19, 2018 at 10:10 AM, megarat  wrote:

>
> So I'm curious about what cross-assemblers people would use today for
> developing such software on the M100.  Is there a standard "go-to" package
> that is used by the Model T community?  I noticed GNUSim8085, which seems
> intriguing, but I haven't looked at it too closely.
>
>
>
VirtualT. There's a rudimentary debugger built in, not source level but
that's less important for assembly code since it does show the instructions.

As to assembler, as (as8085) is freely available.

I used Telemark Assembler for HTERM but it is no longer available
apparently. It has doesn't have any weird syntax or unused features /
complexity to work around.

And you'll need rom entry / address definitions. I hacked mine as I went
but Willard Goosey supports a more complete out of the box def.

-- John.

[M100] modern cross-assemblers for M100?

[megarat]

Greetings again, folks.  So my primary reason for getting back into the M100 is 
to sharpen my skills in 8085 assembly programming.  I have ROM2/Cleuseau, both 
as a standalone option ROM and by way of REX, and while this is a satisfying 
way to play around an learn, I have a long-term aspiration of creating some 
software for the M100 that would be implemented by way of option ROM, and I 
don't believe that I can do this entirely within the M100.
So I'm curious about what cross-assemblers people would use today for 
developing such software on the M100.  Is there a standard "go-to" package that 
is used by the Model T community?  I noticed GNUSim8085, which seems 
intriguing, but I haven't looked at it too closely.

Any pointers would be appreciated.

All the best,
CAM

Re: [M100] TDock

[megarat]

My feelings about this are pretty similar, but they don't have to do with 
"modern vs. retro", but rather, "how much life support is required to get the 
system working"?  Accomplishing a single technical objective with a 
full-fledged computer with a modern operating system ,when it could be 
accomplished by something much more simpler, just strikes me as overkill and 
inelegant, with a huge surplus of moving parts.  You inherit all of the 
problems and maintenance issues of the computer itself, just to accomplish that 
one task.  This is especially the case if the OS is Windows, as Windows can be 
so much more fussy, insecure, and unstable.  
A good counter-argument to this is that if using a computer as a foundation for 
building your new device is the easiest, cheapest, and/or quickest way to do 
it, then those benefits might be worth the cost of having the computer 
involved. 

CAM



> All of the discussion about getting DVI to work and a modern DVI solution
> made me start thinking about TDock again and what it was that made me stop
> working on it.  Largely is was simply lack of time, but also I believe I
> also felt that while using RaspberryPi was maybe the cheapest solution,
> somehow it never sat quite right with me to use a 1GHz processor with 1G
> RAM as a display terminal for a 2.4MHz / 32K processor.
>
> Like it has already been stated, if you are throwing a Linux box into the
> mix, then why not just run VirtualT on the thing and be done.  Especially
> if you are bound to a dispay anyway.  Somehow using a lower-end
> microcontroller with a VGA output from a CPLD / FPGA always seemed like a
> more "retro" solution.
>
> Am I crazy for thinking this way?  Or do the majority of people not really
> care about "retro" value and only care that it is dirt cheap (or as close
> as possible)?