One more letter in the display - Comments on V20N5

What a wonderful issue V20N5 was! Definitely, that’s what the HP-12C deserved - I don’t know any other electronic device on the market for as long as 20 years.

As to Juergen Rodenkirchen’s question in V20N5p4, I think he will enjoy the database that Craig Finseth maintains (http://www.finseth.com/~fin/hpdata.html).

It has detailed descriptions of all HP calculators. That’s the first resource to be consulted, every time I can add a new calculator to my collection. It would be great to read a review on Craig’s project in a future Datafile, especially a few words about those mysterious HP-12BII, HP-56 and HP-66.

I have another answer, though somewhat delayed. In V19N6p34 José Chinchilla asked who wrote the HP-41C manual. According to an HP Journal article (March 1980, V31N3) the author was Ray Tanner.

More on hex digits

I’ve observed that the undocumented keyboard combination ON+n is equivalent to ON+PMT on the HP-12C. This operation is deserving of thorough study, since I suspect that what I described in my previous article (V20N5p26) is just the tip of the iceberg. Tony Duell’s remark about the abnormality of that 0.1 (V20N5p24) made me think of potential dangers when using non-naturalised numbers. For example, this sequence: 3 ON+PMT 1/x seems to hang the machine - ON+n or ON+PMT must be pressed to bring it back to life. Sometimes, misleading results are obtained: 1001 ON+PMT g-LN returns 8.19 after 10 seconds.

How do other Voyagers behave? I have no access to them at the moment; hopefully, someone will explore further. With reference to earlier calcs, here’s an excerpt of an article in the November 1975 issue of the HP Journal, describing the way the HP-21 family displayed messages:

"An accessory ROM composed of sixteen words of seven bits each. The words 0 to 9 are used to generate the digits. One word, 15, is a blank. Three of the remaining words generate the letters E, r, and o [...] to spell ‘Error’ [...] One remaining word generates the letter F. This is used by the HP-25 to spell ‘OF’ when a storage register overflows."

Let me cite a few cases where non-naturalised numbers are used.

• Someone recently posted to the forum of the Museum of HP Calculators (http://www.hpmuseum.org) saying that, with his newly acquired HP-67, he had got a card that displayed a funny text. This unexpected ability is due to the fact that, on the HP-67, certain altered numbers are displayed as letters. They are created synthetically (corrupting the RAM by turning the machine off briefly). I hasten to add that there was a warning in that thread not to try to print such numbers on the HP-97, because that could burn out its printhead.
• In the May 1983 HP Journal (page 27), it is stated that matrix descriptors in the HP-15C (displayed as A, b, C, d and E) are internally represented by a 1 in the mantissa sign digit and a hex digit corresponding to the matrix name in the most significant digit of the mantissa field. For instance, 1C000000000000 represents the matrix name C.
• One more case. Let’s interrupt the HP49G while computing a long CAS operation: Run 3493799781771879973943 NEXTPRIME and press ON after 3 seconds. On version 1.19-5 (I couldn’t check the last one), this gem is returned: 5>;31;2349=918821? The symbols : ; < = > ? which follow the numbers in the character set, represent the hex digits A to F.

Our Editor, Bruce Horrocks, kindly communicated to me that I have been declared the winner of Bill Butler’s challenge. Well, except for last-minute surprises... Many thanks for the prize, Bill! It has been a thought-provoking challenge, with interesting results. It’ll be funny to explain why my renewal only cost 7 cents!

All the best wishes you can imagine.