82242A simulated

[mike-stgt]

Short: I proudly present a simulation of the HP 82242A Infrared Printer Module.
That was Jan/2021.
Now: the IR printer module simulation is now included in NutEm/PC filed in Warren Furlows library.

/M.

Screen shot of ooNUreFIP with IR printer simulation

[dlachieze]

Impressive work as always! Unfortunately I'm not familiar with REXX and can just look with amazement to what you're achieving.

[mike-stgt]

This append is obsolete since the IR printer module simulation is now included in NutEm/PC filed in Warren Furlows library.

/M.

[TwoWeims]

Now if we could only get this into the DM41X...

[mike-stgt]

HP-41 infrared module does not support barcode printing

Sylvain said lately in another thread.
This leads to the question, what he meant by 'support'. Is the 82242A Infrared Printer Module too filigree, too fragile or too poor to take the burden of barcode printing? That would be 'support' in the sense of je ne supporte pas du tout les voyages. Or is it just the lack of a "Bar-Code Input" option as it offers the 82162A? So the show stopper is rather the 82240A/B printer not the IR module?

In short: virtually it works but I never tried with real devices (some just don't exist in haptic reality).
For quite some time Christoph's 82240B IR printer simulation serves as target system for my "IL-to-IR bridge" what together acts as an HP-82162A, graphics and barcode included. This is why I conclude a real IR printer may print barcodes the same way or at least similar as an 82162A printer. For a (virtual) test I replaced in PRBC program (see Plotter ROM manual section 7) the BCO function since it does not cooperate with the IR printer module. Summary: it works as supposed, alas much slower than expected.

The program (using Plotter ROM, Extended IO, IR Printer, IR Printer Module):

Code: Select all

 01▶LBL "PBCIR"
CF 23  "NAME?"  AON 
STOP  AOFF  FC?C 23  RTN
ASTO 12  VIEW 12  RCL 12
1,1  SF 25  PINIT  FIX 0
CF 29 

 17▶LBL 01
"ROW: "  ARCL X  ACA 
ADV  BCP  STO 11  R^ 
CHS  SF 20  0  X<>FIO 
SIGN  SF 19  XEQ 03 


 32▶LBL 02
ENTER^  ATOXX  X<>FIO 
CLX  7 

 38▶LBL 03
15  FS?C IND Y  GTO 04
16  FS? 20  SIGN  * 
GTO 06 

 47▶LBL 04
FS?C 20  GTO 05  SF 20
240  XEQ 11  GTO 06 


 54▶LBL 11
SIGN  CLX  68  X<> L 


 59▶LBL 12
ACCOL  DSE L  GTO 12 
RDN  ADV  RTN 

 66▶LBL 05
17  * 

 69▶LBL 06
XEQ 11  DSE X  GTO 03 
X=0?  GTO 03  RDN 
FS?C 19  RTN  ISG X 
X<0?  GTO 02  SF 01 
SIGN  SF 19  XEQ 03  ADV
ADV  RCL 12  RCL 11 
X≠0?  GTO 01  BEEP 
.END. 

The result (reduced by half -- exactly 50%, resized, not resampled so white and narrow black bars are 2 pixels in width and big black bars 4 pixels -- rearranged):

Simple example

PBCIR.png (275 Bytes) Viewed 320842 times

This output works with either the "Software Decoder for HP-41 Barcodes" of M. Hepperle (set threshold to 150), or the "Software Defined Wand", an inclosure of my NutEm/PC.

Edit: supplemented information about NutEm/PC that meanwhile comes with a SDW ("Software Defined Wand") which works directly with the emulated HP41 w/o intermediate processing.

[pcscote]

HP-41 infrared module does not support barcode printing

Sylvain said lately in another thread.
This leads to the question, what he meant by 'support'. Is the 82242A Infrared Printer Module too filigree, too fragile or too poor to take the burden of barcode printing? That would be 'support' in the sense of je ne supporte pas du tout les voyages. Or is it just the lack of a "Bar-Code Input" option as it offers the 82162A? So the show stopper is rather the 82240A/B printer not the IR module?

Hello Mike,

My statement was taken from the 82242A Infrared Printer Manual, page 82 → Features Not Available → Bar Code Printing

The Infrared Printer Module does not support bar-code printing functions.

I have entered your program into one of my 41CX half-nut with the needed physical modules.
The printer is currently printing the bars and I will come back with a picture of the result when completed.
I will also read the barcodes with a wand to validate that it works.

Great program!

Best regards,

Sylvain

• Edit #1: 2022-06-01 12h00 EDT (GMT-4h)
  Printing the PBCIR program, a row takes between 15 to 20 minutes to print and its length is between 10 to 11 inches.
• Edit #2: 2022-06-01 20h15 EDT (GMT-4h)
  Still printing (ROW 19) ... LOL
• Edit #3: 2022-06-01 22h18 EDT (GMT-4h)
  Printing has just finished, it took 10h18m to produce 24 strips of barcode paper for a 165 bytes program.
• Edit #4: 2022-06-01 22h57 EDT (GMT-4h)
  With the 41CL running in turbo 50 mode, the internal code run faster but the printing speed stay basically the same.
  I have not analyzed the code but it sure looks like the final speed is limited by the I/O transmit speed.

[mike-stgt]

My statement was taken from the 82242A Infrared Printer Manual, page 82 → Features Not Available → Bar Code Printing

The Infrared Printer Module does not support bar-code printing functions.

Thank you for clarification. Highly likely "HP" perceived that this option of the 82162A was not worth to copy to the IR printer and its module.

I have entered your program into one of my 41CX half-nut with the needed physical modules.
The printer is currently printing the bars and I will come back with a picture of the result when completed.

OK then next week...

I will also read the barcodes with a wand to validate that it works.

Thank you very much to conduct a test with real machinery. But I'm afraid it's waaaay too slow. When I see the CPU count on my NutEm/PC it' is not realistic to run it on real HW.

Great program!

Thank you, but it's just a quick-n-dirty to show it would work, I did neither a speed-up by replacing the ACCOL by something faster nor compact the code, nor check if all GTO/XEQ are in compilable distance. In fact I just migrated the method I simulate the barcode feature in my oo82162A from REXX to the HP41. There are many much more effective ways to print bar code.

Frankly -- who still does have the need to print bar code today? That belongs to the "once upon a time"-bin -- The other way round yes, there are still addends in manuals and elsewhere with example programs shown as bar code. Once I wanted to scan all and prepare "supplement ROMs" with it (and gave up after I've done those few of my interest). The highest hurdle is the performance of the "software-definded wand".

Amicalement,
M.

[pcscote]

Mike,

I have entered your program into one of my 41CX half-nut with the needed physical modules.
The printer is currently printing the bars and I will come back with a picture of the result when completed.

OK then next week...

LOL ... it is slow but I expect the printing to finish before the end of the day.

I will also read the barcodes with a wand to validate that it works.

Thank you very much to conduct a test with real machinery. But I'm afraid it's waaaay too slow. When I see the CPU count on my NutEm/PC it' is not realistic to run it on real HW.

Six stripe of barcode was printed, so it was time to test them.
It took several tries to load some rows, but I was able to successfully load the six stripe for a total of 17 program lines. (all goods)

Great program!

Thank you, but it's just a quick-n-dirty to show it would work, I did neither a speed-up by replacing the ACCOL by something faster nor compact the code, nor check if all GTO/XEQ are in compilable distance. In fact I just migrated the method I simulate the barcode feature in my oo82162A from REXX to the HP41. There are many much more effective ways to print bar code.

Still it works and you have proved that we can print barcode over IR, so BRAVO!

Frankly -- who still does have the need to print bar code today? That belongs to the "once upon a time"-bin -- The other way round yes, there are still addends in manuals and elsewhere with example programs shown as bar code. Once I wanted to scan all and prepare "supplement ROMs" with it (and gave up after I've done those few of my interest). The highest hurdle is the performance of the "software-definded wand".

I do it all the time, I save all my programs on magnetic card and print the barcode on pdf files in case the mag cards fails, it also ack as an offline archive.

Amicalement,

Sylvain

[mike-stgt]

Bonsoir Sylvain !

Still it works and you have proved that we can print barcode over IR, so BRAVO!

And you proved that what I did virtually also works with concrete/tangible equipment. Thank you.

In fact, the method is known for quite a while now, it was Christoph who first published an "IL-to-IR bridge" sending printer output from virtual IL to his IR printer simulation. BTW, what he calls 'a simulation' is pretty close to the real thing. I copied his IL2IR idea but enhanced it with graphic and bar code ability.The resulting oo82162A is partly a (close to perfect) remake of the well known HP-IL printer and same time kind of a virtual 'Red Eye' sender (at least for the simulated IR printer). Thus it's only a little step to drop the IL2IR bridge and use my 82242A simulation directly. Theoretically

If the reasoning is "feasible but not useful" it's a result as well.

/M.

[mike-stgt]

Hoped SACA of "THE HP-IL Enhancement module" (sic), SKWID INC 1986, could help as replacement of ACCOL, alas --
NO IL PR. ROM
For a speed-up I have in mind something like this, or a mod of the BCO function making it more tolerant about the printer.

[mike-stgt]

• Edit #4: 2022-06-01 22h57 EDT (GMT-4h)
  With the 41CL running in turbo 50 mode, the internal code run faster but the printing speed stay basically the same.
  I have not analyzed the code but it sure looks like the final speed is limited by the I/O transmit speed.

Sure, the IR printer module must not send more than the printer may print in given time. So its transmision is throttled, see DELAY function.
To print the the three rows shown above my NutEm/PC counts nearly 25E6 CPU cycles, what takes (on real HW 64 word times last about 1/100 seconds) few minutes more than one hour. I assume (guesswork I don't like, but my IR printer simulation does not pinpoint such details, in fact the delay is part of the simulation) that the throttler does not discern if a byte sent will print as a character or as a column of pixels only what is 7 times faster.
/M.