HP Forums
newRPL - build 1255 released! [updated to 1299] - Printable Version

+- HP Forums (https://www.hpmuseum.org/forum)
+-- Forum: Not HP Calculators (/forum-7.html)
+--- Forum: Not quite HP Calculators - but related (/forum-8.html)
+--- Thread: newRPL - build 1255 released! [updated to 1299] (/thread-9700.html)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33


RE: newRPL - build 1089 released! [update:build 1111] - okkama - 10-11-2018 03:36 PM

Code:
(0,-1)
-1
^
(-4,855790.E-32,1)
Must be (0,1).


RE: newRPL - build 1089 released! [update:build 1111] - Claudio L. - 10-12-2018 02:26 PM

(10-11-2018 03:36 PM)okkama Wrote:  
Code:
(0,-1)
-1
^
(-4,855790.E-32,1)
Must be (0,1).

The result is not wrong considering is numeric, not symbolic. Unless the case i,-i is treated as a special case, you won't get an exact result.
If you want Z^-1 it's best to use INV(Z) which uses multiplication by the conjugate, and can give you an exact result for exact arguments.
The general case Z^a is used when you use ^, which has no choice but to use transcendental functions:

Z^a = exp(ln(Z^a)) = exp(a * ln(Z)) = exp( a * (1-i*π/2) )
** I replaced already Z=(0,-1) above as r*exp(i*Arg) = 1*exp(i*-π/2)
= exp( a - a*π/2 i ) = exp(a) * exp(i*(-a*π/2))
= exp(a) * [ cos(-a*π/2) + i * sin(-a*π/2) ]

Now we can see the problem: cos(π/2) can't ever be zero unless you can exactly represent π/2, which is numerically impossible, so it returns what you saw: a tiny number in the order of 10^-32 (because you used the default 32 digits, the result will vary at different precisions).

Unlike many other calculators, newRPL will "by design" NEVER try to cheat or manipulate results in any way.


RE: newRPL - build 1089 released! [update:build 1111] - acoto - 10-13-2018 08:00 PM

(09-27-2018 09:28 PM)Claudio L. Wrote:  Then there's the "responsive mode" that came from our friend acoto's concern about speed. When system flag -88 is set, newRPL only waits 30 ms before running at full speed. With flag -88 cleared, the normal is 500 ms before the faster clock kicks in.

A couple weeks away, and this magical things happens. Awesome Claudio!

Downloading, I can't wait to test.


RE: newRPL - build 1089 released! [update:build 1111] - acoto - 10-13-2018 08:47 PM

A little test

<< TICKS 0 1 64000 FOR X X + NEXT SWAP TICKS SWAP ->>
T STO
-88 CF
T
2048032000
1286550
-88 SF
T
2048032000
847550


Works like a charm!


RE: newRPL - build 1089 released! [update:build 1111] - Claudio L. - 10-15-2018 01:22 PM

(10-13-2018 08:47 PM)acoto Wrote:  A little test

<< TICKS 0 1 64000 FOR X X + NEXT SWAP TICKS SWAP ->>
T STO
-88 CF
T
2048032000
1286550
-88 SF
T
2048032000
847550


Works like a charm!

There are side effects, however. They are minor annoyances, for example in responsive mode, the editor might take more than 30 ms to redraw the command line, causing the system to hit the turbo. Try to edit a long list and hold the cursor right, you'll see a little jerkiness when the high speed kicks in, the cursor "catches up" by moving faster then back to normal speed.
Also, clock switching requires adjusting the display clock, which in turn resets the grayscale dithering generator. If you edit a large multiline text, select the entire text (so there's large gray areas), then hold cursor up or down to scroll line by line and you'll notice how on every clock change the gray color "freezes" showing a dithered pattern. It's a flash, but can be annoying when repeated.
It does feel much more responsive, though, but after a short while I ended up back at the original setting on my calc because I edit a lot of test programs.

PS: Are you aware that TEVAL is implemented already?


RE: newRPL - build 1089 released! [update:build 1111] - orthocube - 10-21-2018 12:30 PM

I printed new labels for my HP 39gs, attached it to a double-sided tape, and painstakingly put them on my calculator. With this, I will be able to use newRPL with correct labels on my HP 39gs.

Please don't change the key mappings for the 39gs, or at least if you did, please put an option to use the original 50g keymapping, or else my effort will be wasted Smile

Thanks!


RE: newRPL - build 1089 released! [update:build 1111] - Claudio L. - 10-21-2018 07:43 PM

(10-21-2018 12:30 PM)orthocube Wrote:  I printed new labels for my HP 39gs, attached it to a double-sided tape, and painstakingly put them on my calculator. With this, I will be able to use newRPL with correct labels on my HP 39gs.

Please don't change the key mappings for the 39gs, or at least if you did, please put an option to use the original 50g keymapping, or else my effort will be wasted Smile

Thanks!

I don't have a 39 or 40, and never used one so I don't know where a 39 user would expect things to be. Unless somebody gives me complete new key assignments in source code form, your stickers will be the way to go. And even in that event, it would be optional, not enforced.


RE: newRPL - build 1089 released! [update:build 1127] - Claudio L. - 10-24-2018 09:32 PM

All ROMs as well as the Android port have been updated to build 1127.

See first post for changes.

Other unrelated changes:
I added a list of all units in the Units menu to the wiki, together with their definitions in newRPL.

Before testing the new ROM:
Make sure you backup (SDARCHIVE/USBARCHIVE) your calculator before the update. There were some changes under the hood that might cause problems.
It did on mine once, but I could never reproduce it afterwards on my calc or x49gp or desktop, so it's reasonable to assume my calc had some data corruption from unstable builds during my multiple tests. If I'm wrong and your calc crashes, restoring from backup should resolve all issues, as it did on mine. The new SDRESTORE/USBRESTORE is by design capable of reading the old backup format correctly.
After the update, make sure you create a new backup of your calc using the new format with a different name (just in case you want to go back to an old ROM version for any reason).

New versions starting from build 1127 on will read old and new version archives, but write only new version archives.
Old builds won't be able to tell the difference between the new and old version, and will fail to restore a newer backup (though I haven't tested this scenario).


RE: newRPL - build 1089 released! [update:build 1127] - orthocube - 10-25-2018 12:27 PM

Awesome! Thanks!


RE: newRPL - build 1089 released! [update:build 1127] - Vtile - 10-27-2018 10:45 AM

Really awesome that this journey of NewRPL continues. I feel a bit guilty not to contribute anything, but the new job have taken all the (nerdy)free time for too long.

That said I have installed the android version to my phone... Looks pretty nice so far.

I have a suggestion while trying to wrap my head around how everything works..

*Menus: I noticed that the LS will enable the "nxt..." softbutton to turn to "Prev". This is logical and do work as in stock 50g hardbutton. However would it be good that when you press the LS the text in the softmenu will change to ".Prev"? This naturally takes some memory resources out of more important things.
Another thing I do wonder is if there is go back to previous menu level. If not would it work if you press RS and "NXT.." Again maybe with changing text in softmenu.

Thumbs up Claudio and for others who work with this project.


RE: newRPL - build 1089 released! [update:build 1127] - Vtile - 10-27-2018 10:51 AM

(10-21-2018 12:30 PM)orthocube Wrote:  I printed new labels for my HP 39gs, attached it to a double-sided tape, and painstakingly put them on my calculator. With this, I will be able to use newRPL with correct labels on my HP 39gs.

Please don't change the key mappings for the 39gs, or at least if you did, please put an option to use the original 50g keymapping, or else my effort will be wasted Smile

Thanks!

For those who are looking a solution for keyboard labels. One way to go is to use these thermal printed "Dymo" etc. industrial / office label makers to print the texts. Then cut them to right size (smaller than the key surface, see further). Glue them in place with their own glue. To finalize apply a two layers of clear nail polish lacquer on top of them. I have found that it is best applied as a droplet above the sticker ones or twice. The good place to practice and test the materials is inside of the battery compartment.


RE: newRPL - build 1089 released! [update:build 1127] - Claudio L. - 10-27-2018 11:32 AM

(10-27-2018 10:45 AM)Vtile Wrote:  Really awesome that this journey of NewRPL continues. I feel a bit guilty not to contribute anything, but the new job have taken all the (nerdy)free time for too long.

That said I have installed the android version to my phone... Looks pretty nice so far.

I have a suggestion while trying to wrap my head around how everything works..

*Menus: I noticed that the LS will enable the "nxt..." softbutton to turn to "Prev". This is logical and do work as in stock 50g hardbutton. However would it be good that when you press the LS the text in the softmenu will change to ".Prev"? This naturally takes some memory resources out of more important things.
Another thing I do wonder is if there is go back to previous menu level. If not would it work if you press RS and "NXT.." Again maybe with changing text in softmenu.

Thumbs up Claudio and for others who work with this project.

I've always wanted to have the menus be completely dynamic when you press the shifts. Not just NXT but all menu labels updated when the shift state changes. The only reason I decided against it is that it not only needs redrawing, I need to call RPL programs for each label. Menu definitions in newRPL allow a program to be provided instead of text as a label. It will call the program to get the label every time it needs to be displayed. Calling all those programs every time the user presses a shift could slow down typing, when you have a large text it's already hard to keep up with scrolling and redrawing the text itself. In most cases menus don't use text as a label but the command itself, therefore the decompiler needs to be called every time we update the menus, and the decompiler is quite slow as well.
In the end I decided to prioritize quick response while typing over visual effects.
Perhaps I could do it as a long press on the shifts only, which would mean the user stopped typing and is waiting for the menus to show something. I'll give it some more thought.


RE: newRPL - build 1089 released! [update:build 1127] - The Shadow - 11-25-2018 03:56 PM

I'm not sure how big of an effect it will have on the Units menu, but the CGPM voted recently to redefine the International System of units, the changes go into effect on May 20 of next year.

Just as the speed of light in vacuum has an exact value now (since the meter is defined in terms of it), at that point so will Planck's constant, Boltzmann's constant, the charge of the electron, and Avogadro's number. (As well as a few more obscure numbers.) Since you'll never have to worry about them changing in the future, they may be worth adding to the menu.


RE: newRPL - build 1089 released! [update:build 1127] - Claudio L. - 11-26-2018 04:19 PM

(11-25-2018 03:56 PM)The Shadow Wrote:  I'm not sure how big of an effect it will have on the Units menu, but the CGPM voted recently to redefine the International System of units, the changes go into effect on May 20 of next year.

Just as the speed of light in vacuum has an exact value now (since the meter is defined in terms of it), at that point so will Planck's constant, Boltzmann's constant, the charge of the electron, and Avogadro's number. (As well as a few more obscure numbers.) Since you'll never have to worry about them changing in the future, they may be worth adding to the menu.

I'll review the entire set of units in newRPL for correctness before the May 20 date, there might be little tweaks needed here and there in terms of conversions. I didn't research it too well yet, but for example if the kilogram is redefined does it mean the pound will be redefined to match? if so the current conversion kg/lb stays the same, otherwise the lb stays the same then we need to update the factors. For example, the ft was redefined in 1959 as a constant in terms of the metre (0.3048), so redefining the metre would automatically redefine the ft, but I'm not sure the same happened with the pound.


RE: newRPL - build 1089 released! [update:build 1127] - The Shadow - 11-26-2018 04:54 PM

I just looked at the 1959 document from the National Bureau of Standards, and in addition to defining the foot in terms of the meter, it also defined the pound in terms of the kilogram:

NBS Wrote:1 pound (avoirdupois)= 0.453 592 37 kilogram

The source document can be found here as a PDF.


RE: newRPL - build 1089 released! [update:build 1127] - ijabbott - 11-26-2018 07:11 PM

(11-26-2018 04:54 PM)The Shadow Wrote:  I just looked at the 1959 document from the National Bureau of Standards, and in addition to defining the foot in terms of the meter, it also defined the pound in terms of the kilogram:

NBS Wrote:1 pound (avoirdupois)= 0.453 592 37 kilogram

The source document can be found here as a PDF.

Also, in the UK:

[UK] Weights and Measures Act, 1963 Wrote:The yard or the metre shall be the unit of measurement of length and the pound or the kilogram shall be the unit of measurement of mass by reference to which any measurement involving a measurement of length or mass shall be made in the United Kingdom; and- (a) the yard shall be 0.9144 metre exactly; (b) the pound shall be 0.45359237 kilogram exactly.

— Weights and Measures Act, 1963, Section 1(1)



RE: newRPL - build 1089 released! [update:build 1127] - Claudio L. - 11-29-2018 09:57 PM

(11-26-2018 07:11 PM)ijabbott Wrote:  
(11-26-2018 04:54 PM)The Shadow Wrote:  I just looked at the 1959 document from the National Bureau of Standards, and in addition to defining the foot in terms of the meter, it also defined the pound in terms of the kilogram:


The source document can be found here as a PDF.

Also, in the UK:

[UK] Weights and Measures Act, 1963 Wrote:The yard or the metre shall be the unit of measurement of length and the pound or the kilogram shall be the unit of measurement of mass by reference to which any measurement involving a measurement of length or mass shall be made in the United Kingdom; and- (a) the yard shall be 0.9144 metre exactly; (b) the pound shall be 0.45359237 kilogram exactly.

— Weights and Measures Act, 1963, Section 1(1)

Thank you! That means there will be relatively few if any changes to be made. I should add all the constants but not as units, as actual constants that can be used in formulas. For the time being I'm working on a new symbolic rules engine from scratch. So far it applies basic rules, it's up to the level of the original rules engine I wrote 3 years ago, but this one is much cleaner code so I should be able to extend it to do the more advanced stuff with a lot less headaches. The old code became so convoluted I couldn't go on.


RE: newRPL - build 1089 released! [update:build 1127] - compsystems - 12-07-2018 11:45 AM

Hi, I think that the hp48 and hp50 series are the only calculators that allow you to enter menus and symbols as a second option for each change key, (holding the change keys). Is this option contemplated in NewRPL?

for example hp50
[right shift] + [n] = char view
[hold right shift] + [n] = char menu

[right shift] + [o] = Equation Writer
[hold right shift] + [o] = ` `

[right shift] + [7] = Numerical Solver View
[hold right shift] + [7] = Numerical Solver Menu

[right shift] + [9] = Time View
[hold right shift] + [9] = Time Menu

Idea for NewRPL
for example if there are variables in the menu
[a] [b]

[F1] reads the contents of a
[<¬] + [F1] executes Y=
[hold <¬] + [F1] places the variable 'a'


RE: newRPL - build 1089 released! [update:build 1127] - Claudio L. - 12-07-2018 10:03 PM

(12-07-2018 11:45 AM)compsystems Wrote:  Hi, I think that the hp48 and hp50 series are the only calculators that allow you to enter menus and symbols as a second option for each change key, (holding the change keys). Is this option contemplated in NewRPL?

for example hp50
[right shift] + [n] = char view
[hold right shift] + [n] = char menu

[right shift] + [o] = Equation Writer
[hold right shift] + [o] = ` `

[right shift] + [7] = Numerical Solver View
[hold right shift] + [7] = Numerical Solver Menu

[right shift] + [9] = Time View
[hold right shift] + [9] = Time Menu

Idea for NewRPL
for example if there are variables in the menu
[a] [b]

[F1] reads the contents of a
[<¬] + [F1] executes Y=
[hold <¬] + [F1] places the variable 'a'

And one more time you are proving you never bothered to even try newRPL... should I give up on you or keep insisting? I think you have some good ideas, but to cooperate with the project you need to be familiar with it. The idea above has been part of newRPL from the very first ROM that was released (I forgot how many years ago).
Yes, the keyboard has hold-shift planes and yes, the keyboard is fully configurable by the user, more than the 50g ever could.
And before you suggest it: so are the menus!


RE: newRPL - build 1089 released! [update:build 1127] - compsystems - 12-07-2018 11:52 PM

The problem is that I have not read the entire manual, soo had tried some combinations and did not get any output, I deduced that it had not been incorporated.

Looking for other combinations. I only find the options between [M] and [Enter]

RS_hold + [spc] = ;
RS + [spc] = ,

LS_hold + [0] = INF ;
LS + [spc] = ¯inf,