WP43 News

This area is for discussion about these families of custom high-end Scientific Calculator applications for SwissMicros devices.
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: 43S News

Post by Walter »

Walter wrote:
Sat May 29, 2021 2:20 pm
PierreMengisen wrote:
Sat May 29, 2021 12:31 pm
As for the OWNER'S MANUAL I think it's a bit too early to make definitive corrections. In the meantime I have listed the layout references that are currently incorrect. They are the references on pages 83, 131, 132, 142 (2*), 146, 150, 270, 276 and 281. Control made on the WP 43S U v0.20 version.
This is a known problem: either the graphic layout is ahead or the software when you change the keyboard. So, after each keyboard change, you need some time to reach equilibrium ... ummh ... equality again. The OM is of 2021-05-20, i.e. really old. Ongoing revision is looking at (hopefully) all the layout references and exchanging them. Thanks in advance for a little patience.
Here you go: an updated OM was uploaded just minutes ago. Enjoy!
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
User avatar
ManuMa
Posts: 24
Joined: Sun Mar 15, 2020 10:14 pm
Contact:

Re: 43S News

Post by ManuMa »

May I suggest a couple of modifications to the layout?

First of all, I think the long press of a key feature is very powerful, and it could be used more.
I suggest using a long press of the ENTER key to perform the FILL function. I think it is natural to enter the "X" register all the way up to the upper one, either the "T" or the "D", with a long press of the ENTER key. This also serves to free up a shifted function of ENTER. More on that later.
I also sugest to use the long press on the BACKSPACE key to perform a CLSTK. Instead of entering four (eight) ceros manually, or even using FILL, having CLSTK on a primary key after a long press I think would be very nice.
The free space left on the ENTER key is used now by the SHOW function. I think it naturally belongs to the ENTER key, as already done with the MANT function on the historic models.
The empty space on the "-" key left by the SHOW function, is now used by the "#" function, that matches perfectly there, just near INTS.
The space that "#" occupied on the EXP key is now used by the TONE function. I miss the possibility to use sounds to communicate, and no interface would be complete without taking care of one of the most important senses of the human being.
Two more minor changes deal with something that is concerning me since the beginning, and are the keys containing the logarithms. I have rearranged them so one is for Ln and e^x, and the other is for Lg and 10^2. Seems more tidy to me.
Opinions?

Regards,
Manuel.
Attachments
WP43s MOD.JPG
WP43s MOD.JPG (150.62 KiB) Viewed 2404 times
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: 43S News

Post by Walter »

ManuMa wrote:
Sun May 30, 2021 1:52 pm
May I suggest a couple of modifications to the layout?
...
Opinions?
Yes, you may, in principle. But you have to live with our feedback, too. Having said this, please hold on:
  • As long as I'm responsible for the UI of the 43S there will be no 'long press' (no HP calculator needed this so far, and I don't see any need here either). KISS.
  • Currently, the shortest way to clear the stack is 0 FILL. Easy enough.
  • About the position of #: please check p. 28 of the OM and then think why # is located where it is. Got it? On the other hand, I can't see a link of TONE to BITS.
  • There are many possible positions for SHOW. I don't see a special link between SHOW and the stack, though.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
User avatar
ManuMa
Posts: 24
Joined: Sun Mar 15, 2020 10:14 pm
Contact:

Re: 43S News

Post by ManuMa »

Hi Walter,
Walter wrote:
Sun May 30, 2021 6:36 pm
...
Yes, you may, in principle. But you have to live with our feedback, too...
Thank you, I wasn't expecting otherwise.
Walter wrote:
Sun May 30, 2021 6:36 pm
As long as I'm responsible for the UI of the 43S there will be no 'long press' (no HP calculator needed this so far, and I don't see any need here either). KISS.
I wasn't aware you were a long press hater. However, long press has nothing to do with complication. In fact it is more simple a long press than a three stroke instruction (CLSTK), so I don't understand why you think that what I suggest is more complicated than the existing solution.
BTW, it would be nice if the user asignments could be done also to a long press key.
Walter wrote:
Sun May 30, 2021 6:36 pm
Currently, the shortest way to clear the stack is 0 FILL. Easy enough.
That's why I sugested a much more easy way. (1 stroke instead of 3).
Walter wrote:
Sun May 30, 2021 6:36 pm
About the position of #: please check p. 28 of the OM and then think why # is located where it is. Got it? On the other hand, I can't see a link of TONE to BITS.
Tone has the same relationship to BITS as "a b/c" to "a.FN", or ".d" to "Lg", or "h.ms" to "10^x", that is, none.
Walter wrote:
Sun May 30, 2021 6:36 pm
There are many possible positions for SHOW. I don't see a special link between SHOW and the stack, though.
AFAIK, SHOW is for showing all digits or characters stored in X, that is part of the Stack, so the link between SHOW and the Stack is more than evident to me.

Anyway, thank you very much for this wonderful calculator. I wasn't so excited by a calculator since a very long time.

Regards,
Manuel.
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: 43S News

Post by Walter »

ManuMa wrote:
Sun May 30, 2021 9:35 pm
Walter wrote:
Sun May 30, 2021 6:36 pm
As long as I'm responsible for the UI of the 43S there will be no 'long press' (no HP calculator needed this so far, and I don't see any need here either). KISS.
I wasn't aware you were a long press hater. However, long press has nothing to do with complication. In fact it is more simple a long press than a three stroke instruction (CLSTK), so I don't understand why you think that what I suggest is more complicated than the existing solution.
Look: On all usual calculators, a keystroke is a keystroke. You press a key, it registers, and the assigned function will be executed. This is what I call simple. - With your suggestion, OTOH, you press a key for a certain time, it registers, and if you press it for less than a limit time, the assigned function A will be executed else function B. This looks more complicated since you have to watch registration and the limit time - press either clearly below the limit or unambiguously above.
ManuMa wrote:
Sun May 30, 2021 9:35 pm
Walter wrote:
Sun May 30, 2021 6:36 pm
About the position of #: please check p. 28 of the OM and then think why # is located where it is. Got it? On the other hand, I can't see a link of TONE to BITS.
Tone has the same relationship to BITS as "a b/c" to "a.FN", or ".d" to "Lg", or "h.ms" to "10^x", that is, none.
But # has a relation to BITS. So breaking this is the opposite of an improvement. And # specifies short integer input like d.ms sexagesimal angles, .d dates, and h.ms times - all in the same row of keys and with the same color.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
vgoudreault
Posts: 19
Joined: Sat Oct 17, 2020 2:32 am
Location: Greater Montreal, Canada

Re: 43S News

Post by vgoudreault »

Not wanting to sound out of line, but wasn't there at least on some HP calculators, the feature that a 'long press' would bring out a message on the display (indicating what function was about to be launched) and if the press was kept longer, the function call was cancelled?
This cancellation on long press is even emulated on the Free42 programs, by the way.
This behaviour implies that the function has to be launched on the key release, not on the key press.

Having different functions assigned to the same key, and launched with long presses, would be counter intuitive, and if someone is slow on the keypad, dangerous as one would have to establish the transition time between the end of function one mapping and the beginning of function two mapping.
If there is ever a vote on this and that my opinion would matter, I would be against alternate functions on long press.

The other point brought up, the one about the TONE, intrigues me. Not about the need to have it as a function taking actual real estate on the face of the calculator -- being in a menu is good enough since it is likely going to be called only from within a program -- but because of its comparative restriction. Tone takes only one argument, from 0 to 9, and there is no published relation between the tone argument and the actual frequency.
Why isn't there also an advanced tone function, similar to the BEEP of the HP-48 and successors, that takes a frequency in Hertz and a duration in second? Is this "10 tones only" a limitation of the hardware?
User avatar
PierreMengisen
Posts: 305
Joined: Wed Nov 29, 2017 1:38 pm
Location: Neuchâtel CH

Re: 43S News

Post by PierreMengisen »

Over_score wrote:
Sat May 29, 2021 6:47 pm
Try INTS 2 2 2 D # H then # 1 0 or # D (the D on the ln key)
OK, I'm getting to what I wanted.
But:
(a) unable to enter a number in direct form of a power of 10. Obligation to pass by the decimal number then to pass in BASE10 mode with f# 10. (e.g. 1 E 18 f# is not possible, it is necessary to make 1 E 18 ENTER f# 10 and then erase the Y register)

b) erratic behavior for large numbers (e.g. 1 E 18 f# 10 10 f# 10 * gives an erroneous result without owerflow indication)

Whatever the base you are working in, large numbers end up being erratic (e.g. by repeatedly ENTER x end up with 0 and sometimes 1 in the chosen base).
Pierre
[TI59 with PC100C; TI-84 Plus CE-T; HP41CV with HP IL loop & 2*82161A DCD & 82162 TP; HP15C; HP28S; DM41; DM41L; DM42; DM41X]
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: 43S News

Post by Walter »

Please check: 2^64 = 1.845x10^19 is the upper limit for short integers. If you want to calculate with greater integers, we sincerely recommend using long integers. The OM is your friend, look up pp. 141ff.
1 E 18 f# 10 10 f# 10 * gives an erroneous result without owerflow indication
This input sequence doesn't work at all. Please try to reproduce. 8-)

EDIT: If, however, you have meant (but not written) 1 E 18 EXIT f# 10 10 f#10 x instead, then you get this:
.
Overflow.png
Overflow.png (8.36 KiB) Viewed 2239 times
with a nice little overflow indicator as advertized in the OM (which is your friend as mentioned). 8-) 8-)

EDIT 2: Integers on the WP43S are not designed to work with exponents. So you observed a feature, not an error. Did I say the OM is your friend? :geek:
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
User avatar
PierreMengisen
Posts: 305
Joined: Wed Nov 29, 2017 1:38 pm
Location: Neuchâtel CH

Re: 43S News

Post by PierreMengisen »

Walter wrote:
Mon May 31, 2021 10:48 am
Integers on the WP43S are not designed to work with exponents. So you observed a feature, not an error. Did I say the OM is your friend? :geek:
I understand. It would be interesting to have an OWERFLOW error in this case as for the real numbers.

Did you find the reason for the fCC and gCC exchange?
Pierre
[TI59 with PC100C; TI-84 Plus CE-T; HP41CV with HP IL loop & 2*82161A DCD & 82162 TP; HP15C; HP28S; DM41; DM41L; DM42; DM41X]
User avatar
Walter
Posts: 3070
Joined: Tue May 02, 2017 11:13 am
Location: On a mission close to DRS, Germany

Re: 43S News

Post by Walter »

PierreMengisen wrote:
Mon May 31, 2021 11:43 am
Walter wrote:
Mon May 31, 2021 10:48 am
Integers on the WP43S are not designed to work with exponents. So you observed a feature, not an error. Did I say the OM is your friend? :geek:
I understand. It would be interesting to have an OWERFLOW error in this case as for the real numbers.

Did you find the reason for the fCC and gCC exchange?
Do we have to throw an error if we light an indicator? Overflow is an inherent feature of finite integer arithmetic.

Regarding |x| & Cie.: Did you follow Jaco's recommendation (delete backup.bin)? Or did you follow mine (install in a separate folder)? What did you observe?
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
Post Reply