Are there any bugs in the Classic, Woodstock and Spice calculators besides the famous early HP-35 2.02 bug?

I know of a HP-25 firmware bug, which I discovered already in 1976, and I claim to consider it as a bug, that a HP-29C running program cannot be stopped during PAUSE.

But I do not know any other bug in calculators of this era.

Bernhard

Found

here
"HP-25 Versions

Some HP-25s have several bugs including polar/rectangular conversion errors for angles within +/- 0.000573° of 180° . Also when the last operation caused data in a certain range to be stored in a register and the calculator was at step 00, switching to Program mode would cause a blank display and switching back to Run would display Error."

From:

http://www.finseth.com/hpdata/hp67a.php
[HP-67] Bugs/ROM-Versions::

There exists several specific arguments for which arcsin (and to a

lesser degree arccos) are in error:

x=0.000003000 (0.6% error)

x=0.000004000 (2.5% error)

x=0.000005000 (4.0% error)

x=0.000006000 (7.0% error)

x=0.000007000 (8.0% error)

x=0.000008000 (11.5% error, 5.11078e-4 vs 4.58366e-4!)
The HP-25 bug I found is more spectacular. If you enter a number in RUN mode and switch to PRGM while holding down the key you will see a decimal dot wandering from left to right endlessly when pressing the key repeatedly.

Was this bug known before ?

Bernhard

I'd not heard of it before. Do you have a key by key description?

Pauli

(08-04-2017 11:43 PM)Paul Dale Wrote: [ -> ]I'd not heard of it before. Do you have a key by key description?

Pauli

The sequence I found is the following:

Switch On the calculator in RUN mode.

Press 1 and hold down the key.

Switch to PRGM mode and release the key.

Press 1 again and hold down.

Switch back to RUN mode and release the key.

Then press 1 repeatedly and you will see the bug.

You can use any number key.

Bernhard

(08-05-2017 02:11 AM)PANAMATIK Wrote: [ -> ]The sequence I found is the following:

Switch On the calculator in RUN mode.

Press 1 and hold down the key.

Switch to PRGM mode and release the key.

Press 1 again and hold down.

Switch back to RUN mode and release the key.

Then press 1 repeatedly and you will see the bug.

Confirmed on HP-25 (not 25C) s/n 1605A03202

Interesting find Bernhard.

And it makes me wonder how long you searched before coming up with that reproducible sequence.

(08-05-2017 02:29 AM)rprosperi Wrote: [ -> ]And it makes me wonder how long you searched before coming up with that reproducible sequence.

I found this sequence a few days after I got my HP-25 in 1976. I remember that I tried to perform all possible key combinations mentioned in the manual. And then made some more experiments.

There is another HP-25 bug: The non existent sequence g Sigma Plus can be programmed, shown correctly as 15 25. With help of a PC emulator I found that this step is executed same as NOP no operation. I'm sure this bug has been discovered many times before and perhaps never was mentioned because it is not very spectacular.

Bernhard

At HHC 2010 (aka HHC MMX) I gave a brief talk about bug hunting and presented a short list of "

My Favorite Bugs" in HP calculators. Please note that the list is by no means complete; it's just my favorites. Some are not really bugs, but hidden features or anomalies or simply unexpected weirdness.

Bernhard,

Thanks for the explanation. That sequence produces the bug here too.

I missed the switch back to run mode.

Pauli

(08-05-2017 04:54 AM)Joe Horn Wrote: [ -> ]At HHC 2010 (aka HHC MMX) I gave a brief talk about bug hunting and presented a short list of "My Favorite Bugs" in HP calculators. Please note that the list is by no means complete; it's just my favorites. Some are not really bugs, but hidden features or anomalies or simply unexpected weirdness.

I found both my HP-25 bugs in your list. Thus I'm not a discoverer

Thanks Bernhard

Not a real bug, but all classics and most woodstocks are not able to calculate 2^32 correctly. They tell us 4294967304, correct would be 4294967296. Even more strange, that HP-19C can, HP-25 cannot.

(08-05-2017 08:55 AM)Sadsilence Wrote: [ -> ]Not a real bug, but all classics and most woodstocks are not able to calculate 2^32 correctly. They tell us 4294967304, correct would be 4294967296. Even more strange, that HP-19C can, HP-25 cannot.

HP-25 confirmed! But 2 ENTER ENTER ENTER x x x ..... gives the correct result

Does the HP-34C have any known bug?

Bernhard

(08-05-2017 10:12 AM)PANAMATIK Wrote: [ -> ]Does the HP-34C have any known bug?

There is

this one shared with the 11C and the 15C.

(08-05-2017 10:26 AM)Didier Lachieze Wrote: [ -> ] (08-05-2017 10:12 AM)PANAMATIK Wrote: [ -> ]Does the HP-34C have any known bug?

There is this one shared with the 11C and the 15C.

Confirmed on my HP-34C. What a pity. This could have been my candidate.

Bernhard

(08-05-2017 10:26 AM)Didier Lachieze Wrote: [ -> ]There is this one shared with the 11C and the 15C.

n ENTER CHS 1 +

When performed on HP-25 it always ignores the CHS key and gives n+1 as result instead of 1-n. This could be a bug, because you can see -n in the displayed x register, but x will be overwritten by 1. Also this could be considered normal, because the behaviour is predictable and treats zero like any other number.

Bernhard

My "classic" abacco has No bug, yeahhhh

(08-05-2017 08:55 AM)Sadsilence Wrote: [ -> ]Not a real bug, but all classics and most woodstocks are not able to calculate 2^32 correctly. They tell us 4294967304, correct would be 4294967296.

These early calculators do not feature the extended (13-digit) internal precision routines of the later models (AFAIK since mid 1976). Since y^x is evaluated via e^(y*lnx) the roundoff error of these operations may show up in the last digit(s). This can even happen with newer calculators for very large results, cf. the 15C Advanced Functions Handbook, p. 150 ff.

(08-05-2017 08:55 AM)Sadsilence Wrote: [ -> ]Even more strange, that HP-19C can, HP-25 cannot.

Sure, the 19C is a later model with extended internal precision. But even this is not always sufficient: try 3^201. ;-)

Dieter

Of course you are right and it has to do with internal precision and formula used for

all exponantial calculations. But 2^x values are so common to programmers and so easy to handle for processors of all kinds, that I somehow expected a special treatment.

(08-05-2017 04:36 PM)Dieter Wrote: [ -> ] (08-05-2017 08:55 AM)Sadsilence Wrote: [ -> ]Not a real bug, but all classics and most woodstocks are not able to calculate 2^32 correctly. They tell us 4294967304, correct would be 4294967296.

These early calculators do not feature the extended (13-digit) internal precision routines of the later models (AFAIK since mid 1976). Since y^x is evaluated via e^(y*lnx) the roundoff error of these operations may show up in the last digit(s). This can even happen with newer calculators for very large results, cf. the 15C Advanced Functions Handbook, p. 150 ff.

(08-05-2017 08:55 AM)Sadsilence Wrote: [ -> ]Even more strange, that HP-19C can, HP-25 cannot.

Sure, the 19C is a later model with extended internal precision. But even this is not always sufficient: try 3^201. ;-)

Dieter