WP43 Alternative key layout --> C43

This area is for discussion about these families of custom high-end Scientific Calculator applications for SwissMicros devices.
User avatar
alphajuliet
Posts: 9
Joined: Tue Jun 20, 2017 11:37 pm
Location: Sydney, AU
Contact:

Re: 43S Alternative key layout --> WP43C

Post by alphajuliet »

invariant wrote:
Wed Oct 27, 2021 8:48 pm
I ran into the same issue and built the 107 version with the L42 layout. I just put it on Github for your convenience, you can find it here: https://github.com/semi-simple/C43-mirr ... -fixed.pgm
Nice. That’s working fine. Thank you.

A.
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

Good to see it was possible to download and compile C43.
And even better to see help compiling arrive when needed.

I am still flooded with work and will update the tree (with 107+) coming weekend, latest. And I will also include any 43S work that might be updated on Gitlab.
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

invariant wrote:
Wed Oct 27, 2021 8:48 pm
I ran into the same issue and built the 107 version with the L42 layout. I just put it on Github for your convenience, you can find it here: https://github.com/semi-simple/C43-mirr ... -fixed.pgm
I managed to find half an hour to update the hotfix 107+ to 107++.
The ++ indicates two minor updates after 107, including a fix I did a week or so ago on WP43S.

I compiled L1 and L42 for us, https://classic43.com/downloads/C43_107L1++.pgm and https://classic43.com/downloads/C43_107L42++.pgm.
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

C43 progress update:
.
I updated C43 to the latest WP43S code base. Here the progress update: https://classic43.com/downloads/2021-11 ... 107_01.ZIP. The WP43S code is in between releases, so this is what it is, a progress compile.

You can play with the latest equation editor and solvers. WP43S added the equation editor, but there are some design decisions I must still make, based on feedback. The 43S solver works. The issue on lower case or upper case characters for the function names are still to be considered. Currently 43S has sin cos and tan. C43 keys are based on the DM42, which has caps, i.e. SIN COS and TAN. Additionally, I find the typing and changing between numeric/greek/upper case and lower case restrictive. The complete C43 keyboard already works way different from the 43S keyboard, but I do not think it is smooth enough yet.

My latest thinking is that the greek letters (which are all on a softmenu) are in the way on the keyboard. I have already removed the greek letters from the template long ago to remove the keyboard clutter, but left the keys active as hidden shortcuts. But that seems in the way now. There are three sets on the keys, i.e. numeric, caps, lower case, and three shift modes, unshifted, f and g, and two sets of locks: caps on/off, num/on off. You can guess where I am going: as simple as possible but no simpler.

I think the system needs remain more or less as it is now, but with the greek key shifts removed altogether, typing in alpha and equation editor modes may be smoother. The capslock/numlock indicator would remain as is, sans the greek symbol on the indicator.

Some other things are incomplete, such as the catalogue, which I added to the equation entry menu tonight. It will work well, but needs a lot of tweaking first. Rather not use it, it is incomplete in the equation mode. But it is a progress release. I had to trim the DM42 releases a lot to fit into the memory, which is getting fuller and fuller ... Removed for now are: USER keys, test registers 10-20, STAT demos, flag, register and font browsers, Standard extra 43S math: SLVQ, PRIME, BESSEL, ELLIPTIC, ZETA, BETA, ORTHO_POLY. All culled functions are compiled into the Windows version though.



Some notes on the equations and solvers:

1. Type in the formula (with or without the "0 =") like SIN (xx) + SIN (xx^2).
1a. Help with the typing: "SIN f F3 f X f X f F4" first term done. Now "f + SIN f F3 f X f X F3 g 2 f FN4 ENTER"
Clipboard01.png
Clipboard01.png (2.94 KiB) Viewed 3815 times
2. The functions are as per the keyboard, that is on C43 trig is written SIN, hence it is in caps. On 43S it is lower case.
3. Use "xx" as a variable, because my draft solver and plot program expects the variable as xx. That can still change.
4. Once the formula is typed, press ENTER. press "Solver" for the 43S solver.
5. Type in two guesses, say "1.2 [xx] 1.8 [xx]" then press [xx] again. The solver will produce the first real root it finds.
6. Press EXIT.
7. Now try the plotting. Press "[3] CHS ENTER [12] [xxPLT]". The numbers are the left and right x-axis range (-3 to 12) needed for the plot.
20211114-22233600.bmp
20211114-22233600.bmp (12.31 KiB) Viewed 3815 times
8. EXIT will leave you in the C43 plot menu with many graph options. Another EXIT to return to the solver.
9. Try my prototype C43 solver for real/complex roots. Press "[6] [xxSLV]" where 6 is the guess. This function has plenty roots, as can be seen in the graph. I am leaving it in to compare with the 43S solver, but the 43S solver will probably end up a lot faster, sleaker and more efficient than mine, so mine will probably be retired after comparisons have been done.


Another example formula using my solver: I do like the text though. The text is actually strings on the stack, and not a fleeting message which disappears after you press something. Note top right, it quotes the guess provided, i.e. Xo=-1 in this case.
Clipboard02.png
Clipboard02.png (4.46 KiB) Viewed 3815 times
In this example I gave the guess at -1, as I saw on the graph that at about -1 there is a local maximum, hence I gave the guess there. The result was the closest complex root at -1.3-i0.42.
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

I received some user comments based on version 107_01 yesterday, and I realised that I am already working on 107_12! I last released mid-November 2021 and it helps nothing to have an ancient version linked to the post and the latest version on Gitlab only!

I froze the latest test version C43 107_12 https://classic43.com/downloads/2022-01-17_C43_107_12.zip (corrected file), complete with all versions L1, L42 as well as the special WP43S version with my DM42 USER key layout. As was also in the previous release, these are all special compiles with the purpose of leaving alone the standard QSPI file on the DM42. The ‘cost’ of not changing the QSPI file, is I have to reduce the program size to fit the DM42, therefore these versions all have reduced functionality as follows:

The C43 pgm files have the following removed (but all are present in the simulators):
  • KEY PROFILES USER_V43LT, USER_V43, USER_C43, USER_43S; USER_DM42; USER_SHIFTS; USER USER_PRIM00U
  • STAT DEMO DATA SETS
  • Startup register demo values in R10 - R27
  • Complex Solver & graphics & stat graphics
  • Flag-, Register-, Font- Browser functions
  • WP43S programming
  • Math: SLVQ, PRIME, BESSEL, ELLIPTIC, ZETA, BETA, ORTHO_POLY
To fit the WP43S without QSPI, I needed to remove (but again all are in the simulator)
  • math functions Zeta, Slvq, ortho_poly, elliptic, beta, bessel.
Most WP43S work happened in the background since almost all visible user functions have been completed by the 43S team for some time already. Mihail has been pouring a lot of energy into the programming parts and I did the bit of graph plotting of equations over the holidays. The work on programming of the WP43S was imported into C43 as is, but I have not yet considered changes required for the additional and changed C43 function behaviours. I do suspect that some work will be needed in this regard as some key presses on the C43 key layout may not work as advertised in the WP43S manuals. Either way, this can be tested as is.

WP43S added their key assignment method which works well. The ASSIGN function on [f] [ASN] can be used in addition to the C43 canned options [g] [KEYS] menu. An example to map a function from the catalog to a key:

Code: Select all

[f] [ASN] [f] [CAT] [FCNS] [°F->°C] [Σ+] [f] [USER] 
changes the top left key to convert °F to °C. To clear all keys, use [g] [KEYS] [RESET].

The major C43 key changes required for alpha mode editing mentioned in the November post were further worked and tweaked. The C43 g-shift layer for greek letters was removed and replaced with a system shifting only between text and numbers, with a few exceptions in the top two rows. See picture below:

p1.png
p1.png (28.2 KiB) Viewed 2997 times

Excerpt of the new C43 alpha input arrangement. The f-shift is for upper/lower case only, where the g-shift is to access numbers and limited special characters. Except for Σ, ∆ & 𝝿, other greek characters were removed from direct key access and will be selected from a softmenu. Numlock and Capslock are provided as well.

I adapted my graphing module referred to tin the previous November post for use in the WP43S project as an equation visualisation aid. I believe the benefit of visualising equations is essential for both projects, and the proposal was accepted and worked into the 43S. An example using the new code:
p2.png
p2.png (1.59 KiB) Viewed 2997 times

Viewing the graph to determine where to find the roots, specifying [-1.5] [x] [3.5] [x] [Draw]:

p3.png
p3.png (1.97 KiB) Viewed 2997 times
Trying to find the root on the left hand side of the graph, noting that it lies between -5 and -4 minor ticks, type [-1] [x] [-0.8] [x] [x], which finds the root at -0.927 rounded. Trying where there is no real root, at between around +7 and +8 minor ticks, type [1.4] [x] [1.6] [x] [cpxSlv], which results in 1.546 + i.0.111 rounded. The cpxSlv is still work in progress, but the demo in the solver menu already does work.

Edit: Corrected link from 2022-01-14_C43_107_12.zip to 2022-01-17_C43_107_12.zip
Last edited by Jaymos on Mon Jan 17, 2022 6:46 pm, edited 1 time in total.
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
pyridine
Posts: 30
Joined: Fri Sep 01, 2017 11:48 pm

Re: 43S Alternative key layout --> WP43C

Post by pyridine »

Thanks Jaco !
Just to be sure that the right version was published, the C43 Windows simulator shows at startup:
C43L1_105, Apr 2021

(NB no issue when updating the hardware, v107-12 shows up)
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

pyridine wrote:
Sun Jan 16, 2022 2:36 pm
Thanks Jaco !
Just to be sure that the right version was published, the C43 Windows simulator shows at startup:
C43L1_105, Apr 2021

(NB no issue when updating the hardware, v107-12 shows up)
Thx, you are right, I included an old exe file from the folder into the zip file.
C43 Simulator corrected in the zip bundle, see link fixed in the previous post.
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
jwiede
Posts: 44
Joined: Mon Oct 18, 2021 11:17 pm
Location: San Jose, CA USA

Re: 43S Alternative key layout --> WP43C

Post by jwiede »

Any word when we might see a WP43C updated version based on the WP43S 0.22.2 (22-02-22) release?
DM42, DM16L, and a menagerie of HPs thanks to CAS (Calc Acq Syndrome)!
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

WP43C updated version based on the WP43S 0.22.2
jwiede, I should get to update in a week or so.

I can release when the work that I am currently doing for both WP43S and C43 (improving the EQN graph memory etc) is done, or at least stable.

Recently WP43S changed three menu names from plural form to singular form in the cases where it makes sense. In the pic below these changes can be seen in green. I changed my layout to follow the 43S names again in the cases where it makes sense, so I changed the C43 menu names too, also in green.

In C43 we use 4 more plurals, of which REGS (register browser) and FLGS (flag browser) need to stay plural due to the strong connection to 100+ registers and flags they refer to. However, in order to stay closer to the concept, SUMS (WP43S: Σ) could well become SUM, and KEYS (not in WP43S) could well become KEY. KEY would actually work well as it is located next to ASN, forming ASN KEY, which makes more sense in singular from that point of view anyway.

My current thinking is to change these in the upcoming release.




Clipboard87.png
Clipboard87.png (85.46 KiB) Viewed 2288 times
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
User avatar
Jaymos
Posts: 1635
Joined: Sun Nov 04, 2018 7:03 pm
Location: Cape Town

Re: 43S Alternative key layout --> WP43C

Post by Jaymos »

As could be seen in the prior C43 releases, I had to cull significant pieces of code to get the image to fit in the DM42 hardware.

I see no current need for myself to change the QSPI (due to the additional complications in getting it back), so I routinely upload only partially working images onto my two DM42 calculators for testing. I will keep on publishing an image that works without QSPI, at least I use it.

if people need the full compile including a changed QSPI file as well, I can compile it that way and include. But I need a show of hands to know if I must add that to the publication as well. (I do not need to do it, and I will not test it or use it like that, but I can do).
Jaco Mostert
Elec Eng, South Africa
https://47calc.com C47 (s/n 03818 & 06199), WP43 (0015). In box: HP42S, HP32Sii, WP34S&C, HP28C, HP35s, EL-506P, EL-W506, PB700; ex: FX702P, 11C, HP67 & HP85; iOS: 42s Byron, Free42+, WP31S/34S, HCalc.
Post Reply