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 »

gomefun2 wrote:
Sat May 16, 2020 11:36 pm
So when is it coming?
Difficult question. As far as I can see now, I'd guess the pilot batch will come before Allschwil - cf. viewtopic.php?f=2&t=2374#p13996. You find many of your other questions answered in that thread.
...many functions are not yet implemented, even though they are picture or explained in the documentation.
Quite natural since the documentation is the spec.

And coding >700 commands etc. takes time. Please also remember it's a spare time project. We've got 5 team members in our project presently and not all of them are coding. If you want to support us in this matter, feel free to apply.
gomefun2 wrote:
Sat May 16, 2020 11:36 pm
This DM43 seems to be a good combination of the WP34 and the DM42.
Also quite natural since its firmware is based on our experience gained with the WP34S (which Pauli and Marcus and me developed) and its hardware will be a slightly modified DM42 (modifications on the outside only).

Thanks for asking.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
Olivier de Nantes
Posts: 84
Joined: Mon Mar 16, 2020 9:37 am
Location: France

Re: 43S News

Post by Olivier de Nantes »

Walter wrote:
Sun May 17, 2020 12:47 am
gomefun2 wrote:
Sat May 16, 2020 11:36 pm
So when is it coming?
Difficult question. As far as I can see now, I'd guess the pilot batch will come before Allschwil - cf. viewtopic.php?f=2&t=2374#p13996. You find many of your other questions answered in that thread.
...many functions are not yet implemented, even though they are picture or explained in the documentation.
Quite natural since the documentation is the spec.

And coding >700 commands etc. takes time. Please also remember it's a spare time project. We've got 5 team members in our project presently and not all of them are coding. If you want to support us in this matter, feel free to apply.
gomefun2 wrote:
Sat May 16, 2020 11:36 pm

This DM43 seems to be a good combination of the WP34 and the DM42.
Also quite natural since its firmware is based on our experience gained with the WP34S (which Pauli and Marcus and me developed) and its hardware will be a slightly modified DM42 (modifications on the outside only).

Thanks for asking.
And that is a very good job on spare time !!!!

Congratulation and thank you for making revival of scientific and programmable calculators.

I really wonder to meet you in Allschwill meeting !

Olivier
Olivier de Nantes (Bretagne)


HP41 (x3 : 2CV / 1CX), HP 42S, HP 48G+, HP 71B, HP 15C LE, HP 35S, HP PRIME

DM41L, DM 41X (Beta - SN: 00078), DM 42 (SN: 1028)
gomefun2
Posts: 35
Joined: Sun May 06, 2018 3:45 am

Re: 43S News

Post by gomefun2 »

I was looking at the WP43 code. It looks like it is written in C.

I was looking at the functions, for example, "cos.c".

I am not a professional programmer by any means, though I am curious how you are doing this. The only thing you #include is a reference to the wp34s file. None of these files contain the c library math.h. I'm guessing this means you are not using it?

Why?

is math.h bloated or slow?

Is it only because this is a microcontroller with limited space? If you had more space would you use the math.h library? does the math.h library not work with the higher precision DM43?

Also, for some of the more complex functions, such as the solver or integrator, how do you choose which algorithm to use?

Just curious.
hoverbeek
Posts: 17
Joined: Sat Sep 21, 2019 3:54 pm

Re: 43S News

Post by hoverbeek »

gomefun2 wrote:
Sun May 17, 2020 6:37 pm
I was looking at the WP43 code. It looks like it is written in C.
Indeed, it is.
gomefun2 wrote:
Sun May 17, 2020 6:37 pm
I was looking at the functions, for example, "cos.c".

I am not a professional programmer by any means, though I am curious how you are doing this. The only thing you #include is a reference to the wp34s file. None of these files contain the c library math.h. I'm guessing this means you are not using it?

Why?

is math.h bloated or slow?

Is it only because this is a microcontroller with limited space? If you had more space would you use the math.h library? does the math.h library not work with the higher precision DM43?
The hardware is quite fast:
The processor is a ST32L476 controller and is a Core: Arm® 32-bit Cortex®-M4 CPU.
The CPU runs at 25MHz when on battery power (1x CR2032 coin cell) or at 80MHz when drawing power from the USB port.
There is also enough space for calculations.
This is not the limiting factor.

The mathematical functions build upon the work done for the WP-34S project:
https://sourceforge.net/projects/wp34s/

We work with 34 and for some calculations 39 digit precision.

Math.h works on doubles, which have 15 digit precision.
So yes, math.h is not precise enough for our goal.

Cos is calculated using Taylor series.
gomefun2 wrote:
Sun May 17, 2020 6:37 pm
Also, for some of the more complex functions, such as the solver or integrator, how do you choose which algorithm to use?
The solver has not been build yet. So we can use any solution possible.
In general, with the solver you enter a function as a program into the calculator.
The solver then tries to find the points where the function is zero.

If you want to solve x^2 +5x = 5, for example, you first rearrange the formula to search for zero, so: x^2 +5x - 5 = 0

You can then enter this as the following program:

[ENTER]
[X^2]
[X><Y]
5 * 5 - +

The solver will generate values for the x register, run the program and tries to find the x where the program returns zero.
DM41X SN: 00070 DM42 SN: 00754
HP35, ..., HP67, HP41(C,CV,CX), ..., HP48GX, ..., HP33S, HP35S, WP34S, WP43S
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 »

hoverbeek wrote:
Mon May 18, 2020 12:47 pm
We work with 34 and for some calculations 39 digit precision.
... and for some with 72, 75, 1002, and even 1065 digits precision. Just for sake of completeness.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
hoverbeek
Posts: 17
Joined: Sat Sep 21, 2019 3:54 pm

Re: 43S News

Post by hoverbeek »

Walter wrote:
Mon May 18, 2020 2:00 pm
... and for some with 72, 75, 1002, and even 1065 digits precision. Just for sake of completeness.
Can you elaborate when we use what precision and why?
Since we are number fetishists, I think the audience would love to hear this.
DM41X SN: 00070 DM42 SN: 00754
HP35, ..., HP67, HP41(C,CV,CX), ..., HP48GX, ..., HP33S, HP35S, WP34S, WP43S
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 »

hoverbeek wrote:
Mon May 18, 2020 8:35 pm
Walter wrote:
Mon May 18, 2020 2:00 pm
... and for some with 72, 75, 1002, and even 1065 digits precision. Just for sake of completeness.
Can you elaborate when we use what precision and why?
Since we are number fetishists, I think the audience would love to hear this.
Oh, the audience is cordially invited to read the Reference Manual, p. 316, and follow the links from there. Whatever questions arise then, I'll try answering. 8-)

(I can write and upload manuals but reading is up to you. Or shorter in German: Selber lesen macht schlau. ;) )
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
jonmoore
Posts: 108
Joined: Mon Apr 13, 2020 4:18 pm

Re: 43S News

Post by jonmoore »

Walter wrote:
Mon May 18, 2020 9:37 pm
Or shorter in German: Selber lesen macht schlau. ;) )
Or shorter still in international "Internet Esperanto".

RTFM, even shorter for those offended by expletives, RTM...

Whether it makes you smarter or not, depends greatly on whether said manual is intelligible!
gomefun2
Posts: 35
Joined: Sun May 06, 2018 3:45 am

Re: 43S News

Post by gomefun2 »

Thanks for answering my questions.

I spent a bit of time looking through the code, I do remember reading a function named Taylor series somewhere in the code. I bet if I could understand everything in this project, I would learn quite a bit about C.

I recently created a C/OpenGL physics simulation, but the way this DM43/WP34 code is organized is much more professional than what I am currently able to do.
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 »

jonmoore wrote:
Tue May 19, 2020 8:29 am
Walter wrote:
Mon May 18, 2020 9:37 pm
Or shorter in German: Selber lesen macht schlau. ;) )
... Whether it makes you smarter or not, depends greatly on whether said manual is intelligible!
Depends on the author and the reader. I've often, however, asked for feedback in this matter (I've read these manuals far too many times to notice my errors etc. anymore). So if and when you come across any errors, typos, wrong terms, false assumptions, missing explanations, etc. you shall feel free to report them. 8-) And don't be afraid: I can stand critics as much as I criticize myself (another ambiguity in English :? - it's all a give and take ;) ).
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
Post Reply