WP43 News

This area is for discussion about these families of custom high-end Scientific Calculator applications for SwissMicros devices.
RNav
Posts: 15
Joined: Mon Apr 22, 2019 2:01 pm
Location: Metro Detroit, MI USA

Re: 43S News

Post by RNav »

Have fun boys and girls! I am a big fan.
28c, 28s, 48sx, 17bii, 32s, 33s, 35s, Prime, WP-34s, DM42 (daily user)
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 »

RNav wrote:
Sun Oct 27, 2019 2:50 am
Have fun boys and girls! I am a big fan.
Thanks - though we'll leave the fun boys aside. ;) :P

EDIT: This was a meeting of eight HP aficionados and some 30 calculators. Most of the calcs were modified HP products: Low Power Woodstocks, WP34S, WP31S, and some very early samples of WP43S based on SM DM42. Enough to play with and admire. :D

The entire WP34S/31S development team signed some of the respective calcs. Now we'll wait until one of them will appear on eBay. ;)
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
User avatar
wiljea
Posts: 14
Joined: Tue Sep 04, 2018 6:20 pm

Re: 43S News

Post by wiljea »

Hi to you, participating members of this meeting of the "WP43s" development community. I hope you will have, or already had a good reunion. I just wanted to encourage your collaborative work toward this truly ultimate RPN calculator. Aready appreciating the "WP34s" technological tour de force, I am also a big fan of your work who dreams of this future unprecedented evolution. Let's hope that this goal of yours would also appeal to some other future collaborators willing to contribute and accelerate its achievement. Good luck !
DM42 SN: 03090
HP-29C, HP-41C, HP-15C, HP-33s, HP-35s, WP 34S, Free42, DM42
(Fortran, C++, Mathematica, Maple, Python, …) Physics aficionado
User avatar
wiljea
Posts: 14
Joined: Tue Sep 04, 2018 6:20 pm

Re: 43S News

Post by wiljea »

43S Perspectives about emulator and calculator integration …
(I left to moderators’ judgment, to see if this theme should be addressed in a separate forum entry. Maybe it should, due to the diverse and too numerous questions…)

There is a clear trend in many recent developments of handheld calculators to provide an emulator running in a computer environment. These communicating emulator platforms are supplied to the user for different purposes. This approach can indeed be a powerful marketing tool in the education world as it allows teachers to demonstrate various usage and mastering techniques (as seen for instance in the case of the HP Prime, also in the NumWorks project and some other handheld calculators). Otherwise, in the case of the 43S/43C, the emulator allows a natural test bench for software implementation & development. Anyway, these emulator integrations may turn out to be an interesting or fruitful paradigm which need to be examine more closely for merits and potentialities. I want here to envisage specific integration goals requiring either once at a time data exchanges, or, why not, even real-time requested data transfers between physical and virtual implementations, and also with external entities (like a target file or a computer interface). Let me try to explain some of these visions.

Concerning more precisely the "43S" development and its emulator integration, for discussion purposes, let’s envisage the following suggestions, some being straightforward, easily or already attainable, and some others, more or less wildly imaginative (maybe laughable, excuse some of my overwhelming exploration that I didn’t censure) and perhaps useful if some merit can be envisioned.

1) Punctual data exchanges between 43S calculator and its emulator.
(An easy suggestion).

1 a. These exchanges seem a priori possible by design of the future 43S (isn’t it?) and they should ideally include both states and programs. This differs from the DM42 case whose programs and states can be exchanged between different calculator units but where only programs remain compatible with Free42 (which is not a DM42 emulator, as we all know: if you try to share states you may encounter a “corrupted file” message). This could be useful to continue a numerically intensive tasks on the emulator taking advantage of the computer faster implementation.

1 b. (Hypothetic preamble: let’s suppose that the 43S calculator and DM42 do eventually share the same hardware platform as intended.) Would it be conceivable that they (i.e. the 43S & DM42) may share some level of compatibility between their respective states? (Is it desirable? I am not sure …).

1 c. (With the same 1b preamble, but on a more restrictive point of view) … Or, to feature a partial compatibility (between 43S & DM42) between some part of their respective very basic program architecture (I mean in term of the basic functionalities, excluding some more advance features already implemented in the WP34S, like : very specialized functions, local variable handling, 8 level stacks, etc.)?

1 d. (With the same 1b preamble, but on a more restrictive point of view) … Or, to feature a partial compatibility between some part of their respective very basic data structures (of 43S & DM42) in term of the direct register content (including possibly defined variables containing either, complex, matrix or alphanumeric entries)?

2) Punctual data exchanges between external virtual entities (clipboard or target files) and either the emulator or the 43S calculator.
(Another simple suggestion).

2 a. Providing that 1a is fulfilled, at minimum this would concern only the 43S emulator (which would provide the indirect link from the calculator to reach the computer files through emulator exchanges). Note that the emulator aspect of this approach is already implemented with real success in the Free42 environment (where the clipboard allows to import or export data to a spreadsheet, for instance). But, as far as I know, this remains impossible with the DM42 (or is there any intent from SwissMicros to eventually provide such possibility?).

2 b. Through the USB connection, would it be conceivable for the 43S calculator to write directly into or to exchange data with a computer file (at least to the clipboard)?

2 c. Indirectly (through the emulator, as in 2a) or directly (as in 2b) would it be possible to foresee data transfer to some other and quite different computer files? (And here, you can let your imagination run away…). For instance, to overcome the actual screen or printing limitations, let’s say that your program draws directly to a dimension adjustable 2D computer wider screen (and, why not, your program would be able to write into a virtual 3D object.) … Or to virtually access anything else in the same (out of your mind) futuristic spirit?

3) Real-time data transfers (or requested on-line data exchanges) between external virtual entities (active computer interfaces) and either the emulator or the 43S calculator.

3 a. In the hardwired case, I imagine the 43S calculator interacting directly with third party hardware interfaces (like data acquisition or robotics hardware). I doubt that the DM42 communication hardware (through its proprietary USB implementation) would be able to allocate interruptions and execute live data transfer. Simply because this was not intended at first. But some future hardware versions of the handheld 43S calculator might very well do so. The competition is already there (HP Prime and some others, I think). A very basic instance of on-line requested data transfer would be to use the calculator as a simple USB external numeric keypad (a bit expensive gadget, I agree, but is it possible that this was once considered by HP?) or the converse, the computer keyboard controlling directly the calculator (might be useful for classroom demonstration).

3 b. The software case seems more easily achievable as the 43S emulator would communicate through one of the computer’s I/O accesses to control some external hardware (and pilot a robot for instance).

4) Real-time data transfers (or active data exchanges) between virtual running entities, such entities being different running programs and even many instances of the emulator (including, why not, the 43S calculator) running all together in parallel.
(Some very far-fetched suggestions, I guess, and I laugh with you, but…).

4 a. Between different emulator instances running simultaneously. For instance, it is actually quite amazing to see the work load distribution of the windows environment running in parallel 4 instances of Free42 testing intensively for a huge prime number (you should try this!). Imagine, a possible implementation of quasi parallel computations exchanging their data (even if these tasks are not well suited for interpreter codes). For instance, as one already running instance is about to terminate the calculation of some 3D field evolution, a second family of running instances is already performing ray-tracing representations of each instantaneous field images and a last family would produce and show the final video-rendering animation on the computer screen or on your future holographic projector (this is more like a kind of collaborative sequential parallelism, I know).

4 b. The 43S calculator might be participating in this collective task even if it runs at a slower speed, why not? As would also other (and surely faster) programs, written in some entirely different language (compiled or interpreted like Python, for instance).

POST SCRIPTUM.
1) I hope you had fun dreaming with me in these number crunching futuristic fantasies, aimed to reinvigorate, if not reinvent, the old paradigm of handheld RPN calculators with these perspectives of future emulator intregration! You may have some doubt about the pertinence of these advanced suggestions. But think about it, in: hydrodynamics*, thermo-hydraulics*, neutronics*, meteorology*, aerodynamics, real-time simulators* (including games and virtual reality renderers), numerical relativity, etc. … in mostly all these technologic and scientific fields, any simulation attempt relies on the stability and convergence of various numerical methods. And the fine tuning of the latter can benefit when they are introduced, tested (in small implementation) and exemplified using matrix algebra. But isn’t that precisely one of the fortes of the DM42/Free42 environments ? I undersdtand that 43S & DM42 are a priori quite different beasts, but the 43S developpment may profit from the future inclusion of even more powerfull desirable matrix functions (exponentiation, Padé approximant, etc.). In that sense, playing with matrices on your pocket calculator is already like playing with a downsized version but numerically quite significant (due to high numerical precision) handheld version of Matlab. Imagine what the future 43S may reserve. Thank you, four your attention.
2) After proofreading (still a good idea), I made minor corrections/clarifications. Do not hesitate to react on either of these ideas (if you find some interesting), since it could probably help revive this forum, judged by some as a little inactive. Note (*), I had various work experiences in those fields.
DM42 SN: 03090
HP-29C, HP-41C, HP-15C, HP-33s, HP-35s, WP 34S, Free42, DM42
(Fortran, C++, Mathematica, Maple, Python, …) Physics aficionado
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 »

@wiljea: Good grief :shock: , we'd be happy when the 43S would do what it should to full extent... I think we'll come back to your ideas :idea: :idea: :idea: :idea: when we approach this milestone... but don't hold your breath.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
Thomas Okken
Posts: 1100
Joined: Tue May 02, 2017 5:48 pm
Location: Netherlands
Contact:

Re: 43S News

Post by Thomas Okken »

wiljea wrote:
Mon Oct 28, 2019 6:57 am
[...] the DM42 case whose programs and states can be exchanged between different calculator units but where only programs remain compatible with Free42 (which is not a DM42 emulator, as we all know: if you try to share states you may encounter a “corrupted file” message).
That could be remedied, if SwissMicros choose to do so. Since version 2.5, Free42 has an architecture-neutral state file format, allowing states to be exchanged between Free42 instances running on different platforms. If the DM42 were to use that new state file logic as well, and write its own extension data at the end of the state file, its state file format could be made compatible with Free42. I don't know how useful that would be, but it should be doable.

Free42-like Copy and Paste is another matter. The DM42 is a single-app platform, unlike a PC or smartphone, so there's nothing to share that clipboard data with.
User avatar
wiljea
Posts: 14
Joined: Tue Sep 04, 2018 6:20 pm

Re: 43S News

Post by wiljea »

@okken: Thank you Mr. Okken for your clear indication about the fact that since version 2.5, you already paved the way for true Free42-states compatibility over many different platforms including iOS and Android (as indicated on your website), which is more than pertinent for the young scientific generations (… be sure that the word will spread). Together with the Free42 copy-and-paste capability, your integration work is very impressive if not unprecedented!

If other members (DM42 owners) show their interest in such possibility to share Free42-states with their DM42 calculator (going in both directions), I think, we should officially ask SwissMicros if they may want to implement this compatibility (providing it is possible or feasible for them. But I am not sure this is the right forum to do so). As I mentioned earlier, it would be useful (at least for me) to take advantage of the Free42 lightning speed when this emulator runs on a faster CPU some complex numerical scheme (potentially converging too slowly on DM42).

@walter: OK, understood (I really tried)! And I thank you also Mr. Walter for your reply. I hope that “ good grief ” was an euphemism of not a too bad perception. Anyway, be sure that I was not trying to discourage anyone by some too enthusiastic (perhaps awkward) suggestions. Nor did I try to obtain any new engagement whatsoever. I certainly respect the seriousness of your actual commitment and the many efforts required to successfully complete the 43S project as it is currently defined.

At least, may I ask if options 1a) and 2a) are probable? As interim closure and to promote peace of mind, I shall postpone sine die any further suggestion and live with more tangible expectations. “ Qui vivra verra! ” In the mean time and if so, let again this forum go back in Morpheus arms... where androids dream... of electric sheep! (... In celebration of Philip K. Dick, nothing to do with fractal flames or cheap calculators).

“Androidly” yours, with all my respect to both of you (and without euphemism)!
DM42 SN: 03090
HP-29C, HP-41C, HP-15C, HP-33s, HP-35s, WP 34S, Free42, DM42
(Fortran, C++, Mathematica, Maple, Python, …) Physics aficionado
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 »

wiljea wrote:
Wed Oct 30, 2019 9:44 am
At least, may I ask if options 1a) and 2a) are probable?
Yes, you may. But I admit your text was simply too much for my little mind - so I'll read it again when time will come. Thus, I can't answer you now.
wiljea wrote:
Wed Oct 30, 2019 9:44 am
As interim closure and to promote peace of mind, I shall postpone sine die any further suggestion and live with more tangible expectations.
Feel free to suggest whatever you think being necessary at whatever time you think being appropriate. No need to step aside. If viable, however, please make your proposals a bit smaller and easier for us pedestrians - but this is just my personal opinion now.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
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 »

There's fun work to do right now: Filling the catalog of functions and quickly sorting character strings, dealing with complex infinities and constants of various precisions, etc. ;) You can watch most of it on GitLab. 8-)

And if you want, you can use our simulator for testing. Please look into the Reference Manual, App. D, for how to set up and start it. :D

As usual, remarks, comments, found bugs, questions, etc. are welcome. 8-)
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
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 »

There is an old mathematical claim that all natural numbers from 1 to 100 can be expressed as sums of three cubes of integers, i.e. n=i^3+j^3+k^3. Until 2018, this was proven for all n but 33 and 42 (sic!). Earlier this year, the proof for 33 was achieved at Bristol University (UK) using a supercomputer for 3 weeks. Only 42 was missing still. Using a huge computer network, two mathematicians (of Bristol University and MIT) now found i, j, k are
12 602 123 297 335 631,
80 435 758 145 817 515, and
−80 538 738 812 075 974
to solve this problem. Using the WP43S simulator, you can verify this recent mathematical discovery in a minute.
WP43 SN00000, 34S, and 31S for obvious reasons; HP-35, 45, ..., 35S, 15CE, DM16L S/N# 00093, DM42β SN:00041
Post Reply