The Museum of HP Calculators

HP Forum Archive 14

[ Return to Index | Top of Index ]

Thoughts on OpenRPN - Long and Rambling
Message #1 Posted by rsenzer on 13 Apr 2004, 1:22 p.m.

Some general thoughts about OpenRPN:

I like the concept, however, ...

My thoughts are that an initial step would be to develop an ironclad design specification with a high level of minimal requirements that would receive an "endorsement" as "OpenRPN compliant". Then bid out the design. Let the people with the best resources build it to the desired specifications.

Provide the specs to known calculator companies, computer companies, etc. A small startup electronics company may want to build the products. For that matter, put the specifications on a website along with FAQs, etc. If companies want to add bells and whistles, e.g., algebraic capability, no problem. That doesn't preclude individuals on this board from forming their own company to produce a compliant calculator.

The other advantage is that we could provide different level of specifications for distinct calculator forms. I would like to see three separate specs for 3 different purposes:

1. All encompassing: Graphing calc, etc., with a more HP-41 style programming language. Would like to see some provision for argument passing mechanisms and local variables for procedures and more structured constructs.

2. Two liner: Compact calculator, perhaps the size of a 33S or 32SII but with power of 42S and external interfaces and flash memory.

3. One or Two Line Calculator for examination purposes. This would have battery backed solar power, multiple non-volatile memory registers, have unit conversions and scientific constants built in. Essentially equivalent to a 33S with no programmability and no non-volatile equation storage. This calculator should be really cheap. It would compete with the HP-30S, TI-3X, CASIO fx-XXX models.

All would have an ENTER key in the appropriate place and the appropriate size. All would have full RPN with RollUp, RollDown, LastX, X<>Y, and at least a four register non-volatile stack. All would have the NULL feature that would show the function and cancel the function if held for an extended period of time. All would have a 3 digit exponents and a degree of accuracy comparable to the latest models of HP calculators.

Also, provide an update cycle so that the specification can be appropriately amended from time to time as appropriate.

Issues:

1. Mechanism for creating and updating the document

2. Parties involved in creating and updating the document

3. Mechanism for tracking the progress of the document and companies interested in creating associated products.

4. Marketing: Convincing companies that this is a worthwhile endeavor and a money making proposition.

... As I said, just my thoughts

      
Re: Thoughts on OpenRPN - Long and Rambling
Message #2 Posted by rsenzer on 13 Apr 2004, 1:49 p.m.,
in response to message #1 by rsenzer

The 2 top end models should incorporate financial functionality as well as scientific functionality.

      
Re: Thoughts on OpenRPN
Message #3 Posted by bill platt on 13 Apr 2004, 2:51 p.m.,
in response to message #1 by rsenzer

Hi rsenzer,

Nice write-up.

To me, the idea of "OpenRPN" is all a "mutual compatibility" issue---meaning that all the "models" will be interchangeable from a programming standpoint. This means that there are no "extra" commands in one level---only differences are memory size. I suppose one might allow different commands if they are part of a logical "module" that could be loaded as an option--but then that module better work on all models in the family.

To me, an annoying thing about existing HP's is the not quite compatibility in some cases---(of course this should exist from 25c to 41---but why should the 32sii use a different equation editing than the 17bii for instance). It would be nice to create a robust basis language--and I think RPN is mature enough now to do it!

Regards,

Bill

      
Re: Thoughts on OpenRPN - Long and Rambling
Message #4 Posted by Hugh on 14 Apr 2004, 8:10 a.m.,
in response to message #1 by rsenzer

Great ideas! I hadn't thought of creating a compliance standard, but that is an excellent way to ensure the highest end quality. It's probably inevitable that a company will want to produce at least one of the models devised by the group... A uniform set of quality standards should really help to keep them from being produced with diminished capacities. We'll probably just want to include a clause about compliance in an open license.

            
Re: Thoughts on OpenRPN - Long and Rambling
Message #5 Posted by Nelson M. Sicuro (Brazil) on 15 Apr 2004, 12:16 a.m.,
in response to message #4 by Hugh

I have a suggestion: This project must be hardware-independent, and the specifications must take in account all kinds of input/output (at first display and keyboard). Something like the "core" system querys the hardware ("drivers"?) to know what kind of I/O it has and enable/disable functions regarding it.

To be hardware-independend I think on creating a "virtual CPU", well-defined to satisfy the needs and easy to implement its emulation on any hardware (from PIC microcontroller to Palm and PCs). It'll be great to have the very same calculador on your hand (fisically) and on your palmtop or desktop computer, and it'll be easy to develop programs for it. This way anyone could build a prototype on a breadboard and "play" with it, helping to debug and develop it.

I have the possibility to project custom-made LCD displays, but it'll cost some $$$, but lets define the "core" first! I plan to build a prototype with LED matrix display, by the way... I have something almost done, I'll publish some pictures later.

This can be the "Linux" of calculators, let's keep the idea growing!

Best regards,

Nelson


[ Return to Index | Top of Index ]

Go back to the main exhibit hall