why was the wand made?

[John Ioannidis]

Why was the wand (82153) built? At $125 ($440 in today's money), it was only slightly cheaper than the card reader. Was it worth it as an input-only device? I can't think of any applications where distributing printed barcodes instead of magnetic cards would make a big difference. Were they planning to offer additional software to read other kinds of barcode?

[cruff]

It would be cheaper and faster to print barcode on paper than to reproduce magnetic cards.

[rprosperi]

At the time, printed media was the primary distribution media, for example all the User Library Programs, HP magazines, PPC Journal and more. Other 'competitive' systems used printed assembler and / or BASIC listings, but these had weaknesses such as zeros look like letter o's, 1's look like I's, etc. while the barcode system was relatively error-free and much faster, in addition to being far cheaper than cards.

Today, it seems kinda quaint, but back in the day, these were extensively used by experts and noobs alike, and it was the quickest way to distribute and enter lengthy programs.

[John Ioannidis]

(08-11-2024 01:56 AM)rprosperi Wrote:  At the time, printed media was the primary distribution media, for example all the User Library Programs, HP magazines, PPC Journal and more. Other 'competitive' systems used printed assembler and / or BASIC listings, but these had weaknesses such as zeros look like letter o's, 1's look like I's, etc. while the barcode system was relatively error-free and much faster, in addition to being far cheaper than cards.

Today, it seems kinda quaint, but back in the day, these were extensively used by experts and noobs alike, and it was the quickest way to distribute and enter lengthy programs.

Yes, I remember actually *buying* solutions books that had bar codes in them, and entering them with the wand (which I had gotten as a bonus for submitting a lot of programs to the Users' Program Library - Europe), but I still would found it weird. I also remember Byte magazine having barcode listings in some other format.

I'm looking at a 1978 Key Notes issue; solutions books for the HP67/97 were $10, and application packs, which included cards, were $35. By 1983, solutions books for the hp41 were $20, and packs with modules were still $35, or thereabouts. It's fair to assume that a solutions book with accompanying cards would be $10-$20 more expensive; you'd have to get ten of them to make up for the price of a wand. Was any other software for the hp41 distributed as barcodes? Most user-submitted programs at the users' libraries did not come with barcodes, and when you ordered something, you got it with magnetic cards (submitting programs with lots of cards, then ordering them with bonus coupons was our way of getting free cards ).

What I mostly used the wand for was to create synthetic instruction barcodes (drawing them by hand with a marker!); a lot easier to do than with the byte grabber

Has anyone bothered to disassemble the wand rom? How hard would it have been to write some extra code to read other common barcodes, such as UPC?

[ThomasF]

(08-11-2024 03:23 AM)John Ioannidis Wrote:  Has anyone bothered to disassemble the wand rom? How hard would it have been to write some extra code to read other common barcodes, such as UPC?

Yes, I disassembled the Wand ROM to be able to emulate the Wand with the Pico.
It is all described here.

In short, no it is not possible, all the bar decoding is done in hardware, the start and stop bars are removed, and the decoded bytes are sent to the ROM for checksum calculation and byte decoding.

To use another bar code format (I.e. different bar widths) is impossible without changing the hardware.

You could of course define new types of content, e.g. another checksum type, other kind of data (e.g. MCode data for a flash device), but the physical properties of the bar code must maintain the original format.

Cheers,
Thomas

[John Ioannidis]

(08-11-2024 07:14 AM)ThomasF Wrote:  

(08-11-2024 03:23 AM)John Ioannidis Wrote:  Has anyone bothered to disassemble the wand rom? How hard would it have been to write some extra code to read other common barcodes, such as UPC?

Yes, I disassembled the Wand ROM to be able to emulate the Wand with the Pico.
It is all described here.

In short, no it is not possible, all the bar decoding is done in hardware, the start and stop bars are removed, and the decoded bytes are sent to the ROM for checksum calculation and byte decoding.

To use another bar code format (I.e. different bar widths) is impossible without changing the hardware.

You could of course define new types of content, e.g. another checksum type, other kind of data (e.g. MCode data for a flash device), but the physical properties of the bar code must maintain the original format.

Cheers,
Thomas

I don't see any code, just a description.

[ThomasF]

(08-16-2024 03:07 PM)John Ioannidis Wrote:  I don't see any code, just a description.

Yes, I never bothered to document the ROM, I just disassembled the ROM and used it together with MCode traces from running a real 82153 in my 41 to be able to see how the wand worked while scanning a bar code.
And by looking at the code and the traces, I figured out how the API to the hardware worked, turning on the NUT, the handling of the buffer, checksum calculation etc so I could implement a corresponding API in my Pico code to emulate the Wand successfully.

Maybe one day I try to compile my notes and findings and try to add it to the assembly code to create and publish a VASM listing of the Wand - just as I did with Blinky (82242) - maybe on a rainy day but it is not currently on my to do list ...

Cheers,
Thomas

[vassilisprevelakis]

I believe that the first use of barcodes for program delivery was introduced by BYTE magazine in the 70s. They even trademarked the name PAPERBYTE.

Although the format was described in various BYTE magazines, there is a PAPERBYTE publication by Ken Budnick (https://archive.org/details/Bar-Code-Loa...dnick-1977) which describes the format and provides sample code.

So it was quite logical for HP to use this format (with modifications) for distributing HP-41 programs and data. After all the HP-41 memory was fairly limited, so one could fit barcode for a program in a few pages.

Although offset printing was expensive, I remember in the early 80s I was using the multistrike capability of a Diablo daisywheel printer (with a TITAN 10 metal wheel) to print barcodes.
The Diablo was kept in the machine room and we would submit "print jobs" to be printed on that printer. When printing barcodes, the Diablo would change from the clack clack clack of normal printing to a machine gun rattle that would scare the machine room operators into thinking that it had gone crazy and turn it off. So I would submit sample output with my "print job" to make sure that the operators knew what to expect.

Later, of course, laser printers made barcode production accessible to more people (i.e. ones who could get access to a company or university laser printer).

In the mid-80s, Acorn User, a magazine dedicated to the BBC microcomputer (in the UK) was distributing programs in barcode format. These could be read by a barcode reader they were selling, but I modified my wand to read these programs. I opened the module and disconnected the wand from the module and then connected them together again with DB-9 connectors. This allowed me to connect the wand part to my BBC microcomputer and write a program to read the Acorn User bar codes.

Fun fact: The BBC micro keyboard had a CAPS LOCK led which could be turned on or off under program control. So by pulsing this LED I could trick the wand into loading programs into the HP-41C. I remember using Blu Tack to keep the wand over the LED.

Vassilis
http://www.series80.org

[pascal_meheut]

(08-20-2024 05:20 PM)vassilisprevelakis Wrote:  Fun fact: The BBC micro keyboard had a CAPS LOCK led which could be turned on or off under program control. So by pulsing this LED I could trick the wand into loading programs into the HP-41C. I remember using Blu Tack to keep the wand over the LED.

Nice trick.

[ThomasF]

(08-20-2024 05:20 PM)vassilisprevelakis Wrote:  Fun fact: The BBC micro keyboard had a CAPS LOCK led which could be turned on or off under program control. So by pulsing this LED I could trick the wand into loading programs into the HP-41C. I remember using Blu Tack to keep the wand over the LED.

Nice!
That is similar to what I do with the Pico. The Pico emulates the Wand HW and is connected to a serial port over USB to my PC.

So, I just send the raw BC data to the COM-port, and the Pico (loaded with the Wand ROM in flash) receives the data just as if it had been read by the actual wand and sends it to the 41.

So with just a click on the mouse on a bar code in a document shown on the PC works as my virtual wand.

Cheers,
Thomas

[John Ioannidis]

(08-19-2024 06:46 AM)ThomasF Wrote:  

(08-16-2024 03:07 PM)John Ioannidis Wrote:  I don't see any code, just a description.

Yes, I never bothered to document the ROM, I just disassembled the ROM and used it together with MCode traces from running a real 82153 in my 41 to be able to see how the wand worked while scanning a bar code.
And by looking at the code and the traces, I figured out how the API to the hardware worked, turning on the NUT, the handling of the buffer, checksum calculation etc so I could implement a corresponding API in my Pico code to emulate the Wand successfully.

Maybe one day I try to compile my notes and findings and try to add it to the assembly code to create and publish a VASM listing of the Wand - just as I did with Blinky (82242) - maybe on a rainy day but it is not currently on my to do list ...

Cheers,
Thomas

I meant the code for your emulator.