Dazzler: Gdemo switches to black and white and plot writing random chars to the s-132 terminal

[ronaldlw]

Any one else seeing this?

Issue 1 Related to Dazzler Switching to Black and White

When I run gdemo the S-132 display output looks normal at first, showing color until the demo switches to black and white mode and then back to color. After going into black and white mode it never comes back to showing color and any program writing to the dazzler stays in black and white as well. A hard restart (with the reset switch) of the imsai seems to return it to normal. I can run kscope for hours and it never sticks in black and white (I'm guessing this is because it never switches the dazzler into black and white mode). Here's a video showing this happen in real time:
https://photos.app.goo.gl/LpJkgUwmxQ1ojyWF9

Here are a couple of screen grabs from the video.

Here it is showing color:

Then the next stage where it switched to black and white (by design I believe):

Then this should be color, but from here on out it's all black and white:

The Web-UI Dazzler display works as expected and never looses color output.

Issue 2 Related to The Plot Program Running and Outputting to the Tektronix Web UI Display

Also, if I run the plot program a bunch of random chars fill the S-132 TTY terminal and, usually, it requires a restart of the S-132 (with it's reset button) to get normal operation again. Is that expected? i.e. the tektronix display is not compatible with the S-132, or do I just have something misconfigured?
Here's a picture of the S-132 CRT screen:

Same for the S-132 TTY screen:

[ronaldlw]

See the same thing with the VIO display as well.

[Retro-Tech-Reboot]

Sorry if this is all obvious.
It looks like the output meant to go to the Dazzler isn't routing to it which is why it's ending up in the user console.
Just so it's said the Dazzler doesn't display on the console device. The Dazzler was implemented as an S-100 card that required a composite monitor to display on.
The Dazzler is implemented via the "Dazzler either on the S-132" and in the "Desktop GUI".
On the S-132 the Dazzler screen needs to be selected so it will output on the VGA monitor. I forget which function keys do this? (F5?)
In the "Desktop GUI" it needs to be connected to the system. There is a "websocket" (power cord icon) button on the Dazzler frame that needs to be enabled (green). You will need to open the Dazzler using the desktop icon (looks like a TV).
I susp[ect you don't have the "websocket" connected on the Dazzler device on in the "Desktop GUI"

[ronaldlw]

Thanks for your reply @Retro-Tech-Reboot. I think maybe I didn't make it clear, but there are two different issues related to the S-132 add on card, so not related to the web interface at all.

Sorry if this is all obvious. It looks like the output meant to go to the Dazzler isn't routing to it which is why it's ending up in the user console.

The issue with garbled output to the tty and crt output on the S-132 isn't related to the Dazzler, that happens when using the tektronix display in the web-ui while the S-132 is connected. The garbled output doesn't show up on the web-ui TTY or CRT screens, just the output from the S-132 card for the TTY and CRT when running the plot program to show the demo drawings on the tektronix display. I'm guessing the tektronix interface is some how in conflict with the S-132 device. A added some more pictures of the output to the original post.

Just so it's said the Dazzler doesn't display on the console device. The Dazzler was implemented as an S-100 card that required a composite monitor to display on. The Dazzler is implemented via the "Dazzler either on the S-132" and in the "Desktop GUI". On the S-132 the Dazzler screen needs to be selected so it will output on the VGA monitor. I forget which function keys do this? (F5?)

The dazzler issue is with the display I'm seeing on the output of the S-132 card that is connected to my monitor.

For that issue I linked a video showing what happens, in the video you see the gdemo output starting out with color and then switching to black and white mid way through it. So you can see that I have selected the CRT output of the S-132 since the gdemo info does display on the monitor.

In the "Desktop GUI" it needs to be connected to the system. There is a "websocket" (power cord icon) button on the Dazzler frame that needs to be enabled (green). You will need to open the Dazzler using the desktop icon (looks like a TV). I susp[ect you don't have the "websocket" connected on the Dazzler device on in the "Desktop GUI"

This doesn't apply since I'm talking about the output from the S-132.

[Retro-Tech-Reboot]

I take it you have a VGA monitor in the VT 100 cabinet. :) Pretty cool.

I'd guess the PLOT.COM program for the Tek terminal you're using is the one I wrote. Can it do a 'slide show' of all the plot files it finds? If so, then it probably is. It's been ages since I wrote this. I don't 100% remember how I was using it or if I saw the same thing.

I did some coding against the Dazzler as well, wrote some graphics primitives for the low-res mode. In my memory GDEMO switches to greyscale mode as part of the demo. I'll need to verify this once I have my emulator set back up.

I need to get my system set back up. It's in the stack of about 1000 things I need to get too. :)

[Retro-Tech-Reboot]

Yep, that's my code. I did have an S-132 when I wrote PLOT.COM and validated both the DZLR: and S-132 implementations. I've worked 1:1 with the High Nibble on several of his projects to test, provide feedback, etc.

[Retro-Tech-Reboot]

The 3D printed VT100 looks great. It fooled me. I spent a lot of years coding on a VT100, mostly PASCAL and C. I moved to a VT340 and eventually a VAXstation. Those were fun times, at least they were in my memory. :)

[ronaldlw]

Yep, that's my code. I did have an S-132 when I wrote PLOT.COM and validated both the DZLR: and S-132 implementations. I've worked 1:1 with the High Nibble on several of his projects to test, provide feedback, etc.

That's awesome! What did you write that in ?

[ronaldlw]

The 3D printed VT100 looks great. It fooled me. I spent a lot of years coding on a VT100, mostly PASCAL and C. I moved to a VT340 and eventually a VAXstation. Those were fun times, at least they were in my memory. :)

I only worked on terminals some in college, but not in my main job. I did work with a tektronix display some too in college for CAD classes. I didn't realize those were persistent phosphor screens (or even what that was) at the time. Pretty cool tech!

My plan is that the IMSAI will have a place on a shelf at some point and I figured having a more time period accurate display would be cool. Of course the real thing is not practical lol.. Plus I had a lot of fun redesigning the project. I obviously cut off the back, but I also made it so it can go together without needing glue which was an interesting exercise. I'd like to re-print it on a larger bed printer at some point to make it so it doesn't have to be 4 large parts (plus several smaller parts), but for now it's working out well.

[ronaldlw]

Yep, that's my code. I did have an S-132 when I wrote PLOT.COM and validated both the DZLR: and S-132 implementations. I've worked 1:1 with the High Nibble on several of his projects to test, provide feedback, etc.

That's awesome! What did you write that in ?
I'm blind, i now see the .c file on the disk lol. Very cool

[Retro-Tech-Reboot]

A note on the Tek terminal emulator.

A VT100 terminal has two serial ports. The primary one that is connected to the host computer and an "aux" port that is software controlled. An sequence is used to control the aux port.

The PLOT program enables the AUX port which results in printf() output strings being sent to the Tek terminal instead of the VT100 display.

This mode is entered in plot.c with this code.
auxOn(); /* output to the VT100 is redirected to the VT100 AUX output */

When the plot program exits, it disables the aux port with this code:
auxOff(); /* will now print to the TTY: */

PLOT.COM opens each file to be displayed and simply printf()'s the contents, with AUX set the program output goes to the TEK terminal.

If you -C out of PLOT.COM the function to disable the AUX port is never called. This results in the TEK terminal becoming the TTY:

While PLOT is running the strings sent to the Tek terminal are commands. The Tek terminal processes these sequences and takes the appropriate action. If the Tek terminal doesn't recognize a specific command sequence, it's sent back to the VT100 over the aux port. This is where the garbage on the VT100 comes from. In the end some of the .plt files I collected look to have invalid sequences in them.

In my library for VT100 control (vt100.c) this is how auxOn() ;and auxOff(); are implemented:

/* Tell VT100 to change to AUX output. /
/ Output to the VT100 is still parsed /
/ by the VT100 but is sent to the aux /
/ port on the VT100 instead of too /
/ the screen. /
auxOn(delay)
int delay;
{
printf("%c[?38h",27); / [?38h */
if (delay > 0) sleep(delay);
}

/* Tell VT100 to change to AUX off /
auxOff(delay)
int delay;
{
printf("%c%c",27,03); / */
if (delay > 0) sleep(delay);
}

[ronaldlw]

Thanks for the info! What's interesting is that the output seems to go to the s-132 tty but not the tty in the web browser.

[ronaldlw]

@Retro-Tech-Reboot would you mind sharing what c compiler and any other ancillary things you used to build the plot program? I tried to build it with a couple of c disks I found on the z80pack site, but it complains about missing header files and whatnot.
Thanks!