WP 43S

[Paul Dale]

The 43S project could do with some assistance.

The current simulator runs on Windows and on the DM 42 hardware, it has all the basic functions and menus but there is still a lot to go.

Any assistance would be beneficial.


Pauli

[gomefun2]

Just out of curiosity what level of programming competency goes into coding the WP 43S?

Do you guys all have 5+ years of professional coding in C? or whatever language is being used?

My main concern is that a MISO (multiple inputs single output) is being implemented. It was the key feature that the WP-34s was missing IMO (due to memory limitations I assume).

I am somewhat familiar with numerical methods and programming. If I were to build a MISO solver I'd probably have to read a bunch of books and papers to make it run fast. I am curious how do you guys research the best way to code things? or do you just know from your careers?

[Thomas Okken]

Mathematically speaking, a numerical multiple-input single-output solver, like the ones in the HP-42S and the RPL calculators, is no different than a single-input solver. It just provides a nice user interface for setting the constants and choosing which variable to solve for.

A true multi-variate solver, that finds a zero or extremum in a multi-dimensional search space, would be an entirely different challenge.

[Claudio L.]

(05-01-2019 04:33 PM)Thomas Okken Wrote:  Mathematically speaking, a numerical multiple-input single-output solver, like the ones in the HP-42S and the RPL calculators, is no different than a single-input solver. It just provides a nice user interface for setting the constants and choosing which variable to solve for.

A true multi-variate solver, that finds a zero or extremum in a multi-dimensional search space, would be an entirely different challenge.

Well, newRPL has an implementation of the minimization algorithm by Nelder-Mead for the multiple solver, so it does indeed solve them all at once in multidimensional space. It wasn't as hard to implement as it may seem at first look, so I would encourage him to give it a shot (that one or any other algorithm, I agree with him that the multiple solver is a killer feature on any calc).

[Paul Dale]

(05-01-2019 01:35 PM)gomefun2 Wrote:  Just out of curiosity what level of programming competency goes into coding the WP 43S?

Not a lot. Most code in a calculator doesn't require deep numerical analysis. Most of the code is handling the user's inputs and displaying the results.

Quote:Do you guys all have 5+ years of professional coding in C? or whatever language is being used?

We would expect some experience with C, but again it doesn't have to be large.

Quote:My main concern is that a MISO (multiple inputs single output) is being implemented. It was the key feature that the WP-34s was missing IMO (due to memory limitations I assume).

We didn't think of it and I don't remember it being suggested.

Quote:I am somewhat familiar with numerical methods and programming. If I were to build a MISO solver I'd probably have to read a bunch of books and papers to make it run fast. I am curious how do you guys research the best way to code things? or do you just know from your careers?

Web research was valuable, so was having a number of books to hand, asking questions on these forums helped a lot and there was some experience involved. Really, it was a mixture of sources and a willingness to try and fail.


Pauli

[Dan]

I would have jumped at the chance a few years ago, but I'm in the middle of my own calculator project that I've been working on for over two years. Being the only coder, there's just no time for anything else.

(05-02-2019 01:11 AM)Paul Dale Wrote:  Most code in a calculator doesn't require deep numerical analysis. Most of the code is handling the user's inputs and displaying the results.

Totally agree. As a matter of fact the C math library covers all the basic functions, and they're fast and accurate.

The WP 43S looks like a great project, hope you get some new recruits!

[Paul Dale]

(05-02-2019 04:08 AM)Dan Wrote:  I would have jumped at the chance a few years ago, but I'm in the middle of my own calculator project that I've been working on for over two years. Being the only coder, there's just no time for anything else.

You've got three to five years to go if you're only doing software.

Join the 43S project insead


Pauli

[Dan]

(05-02-2019 07:25 AM)Paul Dale Wrote:  You've got three to five years to go if you're only doing software.

I wish! More like 20, if I live that long!

[David Hayden]

I might be interested in joining. When I look on the sourceforge page it says that specs and manuals are almost complete and you have a simulator, but looking through the page, all I found was the UI.pdf doc. It says the code page hasn't been updated since 2014.
https://sourceforge.net/p/wp43s/code/HEAD/tree/

What am I missing?

Dave

[Paul Dale]

Things have moved to https://gitlab.com/Over_score/wp43s/


Pauli

[David Hayden]

(05-07-2019 01:59 AM)Paul Dale Wrote:  Things have moved to https://gitlab.com/Over_score/wp43s/

Thanks Pauli (and Bill from Smithville who replied privately). I'll take a look.

Dave

[burkhard]

(05-07-2019 01:59 AM)Paul Dale Wrote:  Things have moved to https://gitlab.com/Over_score/wp43s/

I had been led astray by the old dormant Sourceforge site myself. It seems like that should be either shut down or gutted and replaced with a pointer to send people to the new Gitlab one, no? Since many of the primary players are from the WP34S team (which lives in Sourceforge), I had no reason to doubt that what was on Sourceforge for WP43S was current as well.

So it seems that we can download the PC-based simulator and the draft documentation and have some fun? Are UI or documentation suggestions welcomed from non-coding gurus, or is that just an annoying distraction?

You guys are amazing. Looking forward to seeing this develop!

[Paul Dale]

I'm not sure lead astray is the right term

I don't seem to be able to edit the sourceforge page but I've asked for it to be changed to link to the real repository.


Pauli

[Paolo]

Years ago, when the WP 43S project was taking shape, I made some suggestions to the members of WP 34S team on what I would have liked on the new machine.
I repeat those suggestions here, should someone else be interested. Of course, I will rely on the usual good will of the developers :-) I’m in no way able to code them.

- Levi-Civita symbol (at least in 2, 3 and 4 dimensions, better if n-dimensional)
- Kronecker delta
- Dirac delta
- Pauli matrices
- Clebsch–Gordan coefficients for SO(3) / SU(2) (better if expressed as +/-[n/m]^(1/2))
- Lorentz transformation (for any boost)

I would also like some sort of Periodic Table of Elements, could be handy.

By the way, the values of the fundamental physical constants have been updated on May 20th to the 2018 CODATA recommended values (we also have a brand new version of the SI); this should concern both WP 34S and WP 43S.

Thanks to all

Paolo

[Dan]

(05-23-2019 04:11 PM)Paolo Wrote:  Years ago, when the WP 43S project was taking shape, I made some suggestions to the members of WP 34S team on what I would have liked on the new machine...Of course, I will rely on the usual good will of the developers :-) I’m in no way able to code them.

Mmmm...you can't help even a little? I don't think it's realistic (or fair) to make a wish list of what you would like to see implemented and then sit back with your arms crossed and say "I'm in no way able to code them".

[rprosperi]

(05-24-2019 05:05 AM)Dan Wrote:  

(05-23-2019 04:11 PM)Paolo Wrote:  Years ago, when the WP 43S project was taking shape, I made some suggestions to the members of WP 34S team on what I would have liked on the new machine...Of course, I will rely on the usual good will of the developers :-) I’m in no way able to code them.

Mmmm...you can't help even a little? I don't think it's realistic (or fair) to make a wish list of what you would like to see implemented and then sit back with your arms crossed and say "I'm in no way able to code them".

I agree it's not reasonable to simply post a request for such a simple list and then hope/expect it will be included, but for a different reason than Dan, as not everyone that could contribute has the necessary coding skills.

Perhaps some people involved with the project know what these features are intended for, what type of applications could use them, what type of problem they help to solve, etc. but I sure don't and I imagine most folks in the community don't either.

When making such a request, I suggest you describe how such features are useful and try to seek support for some or all of them from the community. Perhaps with some insight as to how such features could be used, more members would join in and support your request and show the team creating the 43S that there is indeed widespread demand for them.

Also, regarding the updated constants, why not include a link to the source of the items, discussion of the changes, etc.

In short, a collaborative approach to requesting features, accompanied by support from lots of potential users is far more likely to be considered for inclusion by the dev. team. For a volunteer project, the request to add some complex totally new features, coming from a single user that doesn't even attempt to justify said inclusion is not likely to be given any realistic priority. Offered as food for thought, not criticism.

[toml_12953]

(05-24-2019 05:05 AM)Dan Wrote:  Mmmm...you can't help even a little? I don't think it's realistic (or fair) to make a wish list of what you would like to see implemented and then sit back with your arms crossed and say "I'm in no way able to code them".

Why? I'm an applications programmer and calculator user and have no C coding skills. Programmers might not know what us "common folk" want or need unless we tell them. If you want a market that's larger than just programmers, you should listen to us. Maybe we non-systems programmer types should start a poll to decide what features we'd like to see. That would let you know what features would be popular.

[Paolo]

I apologize, I didn't mean to be rude but only synthetic. Certainly I didn't want to rise a flame.
I would be honored to contribute updating the fundamental physical constants from primary sources (BIPM, NIST, IERS, USNO, etc).
I can also collect data for a possible periodic table but the rest goes beyond my abilities.
As for the other suggestions, they were "suggestions" and not "requests" but I should have kept them for myself; feel free to ignore them.
However this is a community of very bright and curious people: probably those who already know those objects, maybe used in some circumstances, might think that they are worthwhile. The others will need little time to find out and decide if it is of interest or not.

My best regards
Paolo

[rprosperi]

Personally, I did not find your post rude at all, so no need to apologize.

My comments were meant to motivate you to add to your initial request, by providing those of us that are not familiar with the methods you suggested with information on how they can be used. I agree also with your comment that many of the folks here are bright and curious, and maybe the methods you have requested could be useful to other people that don't happen to be familiar with these tools. Perhaps add just a small description for each one describing what that tool could be used for. If they are important for you (it appears to be physics related) they could be useful for other people too.

Please give it a try, and I expect other people will join in with feedback. And even if these don't get added into the 43S initially, the discussion here may help to educate others and could also lead to info about other resources on these topics you may not be familiar with.

[Leviset]

Pauli,
In the early 90s I was an IT Testing Team Leader for a large IT Shop.
I’d be more than welcome to ‘test’ the WP43S if that’s a help, as there are many more qualified Forum members than me to help with the coding! (I was a Systems Analyst/Programmer from 1979-90 (as a freelance contractor) then an IT Testing Team Leader (mainly on Tandem non/stop Mainframes and IBM 3090) between 1990-1992 and finally the running of Worldwide Computer & Network Live Operations between 1992-2004.
So I’ve not coded commercially since 1990 but have kept my ‘hands in use’ for my own projects although mainly using high level development tools like Microsoft Access, FileMaker Pro & Mathematica with a smattering of Python.
I’m assuming you only have a part built WS43S and are using an emulator of the WS43S to try out your source code?
Whatever system your emulator runs in, I do have Windows 8.1/10, OSX, Linux, iOS and Raspberry Pi environment setups at home.
I’ve also owned the WP34S for a few years.
Dennis