Let's vote for suggestions and bugs

[compsystems]

index:=0; find("ab","abracadabrant") [↵] returns -1 // bug. GIAC => [0, 7]
index:=0:; F_:=[H_,E_,L_,L_,O_]; find(L_,F_) [↵] returns 2 // bug. GIAC => [2, 3]

[compsystems]

all statements when executed in the history, if they contain the SET data type become LISTS, a list does not always have the same mathematical properties of the sets

D_ := set[ "H", "E", "L", "L", "O"]; [↵] set["H","E","L","O"]
E_ := [ "H", "E", "L", "L", "O" ]; [↵] ["H","E","L","L","O"]


D_ == E_ false

set[ "H", "E", "L", "L", "O"] minus set[ "H", "E" ]; [↵]
set["L", "O" ]

set[ "H", "E", "L", "L", "O"] union set[ "H", "E" ]; [↵]
set[ "H", "E", "L", "O"]

[compsystems]

diference cmd does not work on a set
DIFFERENCE(set["b","a","c","a"],set["a","d"]) [enter]
{"b","c","d"}

set[ "b", "a", "c", "a" ] minus set[ "a", "d" ] [↵] set["b","c"]

[toml_12953]

I'd like the option of using a fixed-width font in the Terminal. The way it is now is that a space character is narrower than a letter and that makes it harder to line up output.

[John Keith]

I strongly second the above.

[Oulan]

Here is not very useful program

Code:

#pragma mode(separator(.,;) integer(h32))
LOCAL mat:=[[1,2],[3,4]];
EXPORT test() BEGIN
 LOCAL li:={1,2,3,4};
 LOCAL ii;
 FOR ii FROM 1 TO SIZE(li) DO
  PRINT(STRING(ii)+" "+STRING(li(ii)));
  li:=append(li,ii);
  IF ii>10 THEN BREAK;END;
 END;
 PRINT(li);
 //ii:=SIZE(mat)(1);       // is not parsed
 ii:=SIZE(mat); ii:=ii(1); // is ok ...  
END;

On the calc, it works as planned, SIZE(li) seems to be evaluated once at start of loop,

Code:

1 1
2 2
3 3
4 4
{1,2,3,4,1,2,3,4}

but on the virtual prime there is some bug ...

Code:

1 1
2 112
3 112
4 101
append(li,ii)
 Error: Bad Argument Value

Btw a second annoying behaviour, see the two last lines ... not a bug but a limitation
Other troubles: k is a 16x16 matrix (full rank), f is a vector (not a {1,16} matrix)

Code:

  M1:=k;
  M2:=TRN(f);
  M3:=simult(M1,M2);
  M4:=simult(k,TRN(f));

Third line is correctly evaluated, fourth one leads to an error (Bad argument type)

Btw, (M1^-1)*f give a full zero solution, which is an error, simult gives the right solution

P.S. calc and virtual one at the last revision 2.1.14181

Olivier

[toml_12953]

It'll probably never happen but I'd like to see the following (in no particular order):

1) The ability to print to a physical printer directly without the use of the CK.

2) The option of being able to pass parameters by reference as well as by value.
Currently you can only pass by value and so have to resort to using global variables.
This goes against the rules of encapsulation.

3) Structures built into HPPL rather than having to resort to tricks to simulate them.

4) The option for a fixed space font in Terminal. The proportional font makes it impossible
(or at least inordinately difficult) to line up output neatly.

5) Create a second version of the Prime without CAS or provide a way to totally disable
it. TI has CAS and non-CAS versions of its Nspire calculators and it works very well
for them. Too often I have errors caused by conflicts with the CAS which I (as a
minority here, I'm sure) never use.

This is my current list but I'm sure it will grow in time. I'm not complaining about the Prime. I consider it to be the best calculator on the market. I would just like to tweak it to be even better!

[tom234]

It would be nice to have a soft key menu on the screen lower as in all the other HP since HP27s. You can program this calculator 3 different ways to Sunday but you can't have a fast solve it doesn't take that much memory. The HP27s can convert BASE number systems at the touch of a button Oct, Bin, Hex etc
do polar coordinates. In fact, the Solve is a better entry-level display.
A lot of horsepowers but no innovation.

Aftermarket libraries, really I thank G someone was ingenious enough to be kind.

all the subroutines you can ask for but they are not in a function menu system that works intuitively like the HP's of the past.

So its more of a programming computer. That's it. Can I get some more powerful subroutine calls please Hew!

[Oulan]

After a while on TI and Casio calcs (graph90+E python is rather speedy)
I tried again to make a program on the Prime and bingo, BREAK
breaks again the calc.

Code:

EXPORT Queens(N)
BEGIN
 MAKELIST(0,X,1,N)▶L1;
 0▶S;0▶T;1▶I;1▶J;0▶C;
 WHILE I>0 DO
  IF C<N THEN C+1▶C;T+1▶T;I-1▶J;
   WHILE J>0 DO
    IF C==L1(J) THEN BREAK;END;
    IF ABS(C-L1(J))==(I-J) THEN BREAK;END;
    J-1▶J;
   END;
   IF J≤0 THEN C▶L1(I);I+1▶I;
    IF I>N THEN S+1▶S;I-1▶I;
    ELSE 0▶C;END;
   END;
  ELSE I-1▶I;L1(I)▶C;END; 
 END;
 PRINT(T);
 PRINT(S);
END;

On the virtual calc, no problem, but regularly on the calc this lead to the blue screen of the reboot ...

[FabVir]

The HP Prime is very close to be a great platform, very powerful but with some small / annoying issues still to be corrected and improved; e.g. in this blog there is an old list that is still valid:
http://www.numericana.com/answer/hp-prime.htm#opinion

would it be nice at least to correct the 0^0 issue and to fix the list of conversion factors and the list of constants, correcting those that are wrong and adding those missing.
This is a very short effort, but hugely beneficial to greatly improve the already high quality of the calculator.

[Wes Loewer]

(11-11-2019 10:25 PM)FabVir Wrote:  http://www.numericana.com/answer/hp-prime.htm#opinion
would it be nice at least to correct the 0^0 issue

There are times that I wished 0^0 were defined as 1, but I don't think I would describe the issue as needing "correcting" as the Prime's behavior is in line with generally accepted practices.

The linked article describes the 0^0 issue as a deal-breaker for the HP Prime. This makes me wonder what calculator he would be willing to buy. Are there any current calculators that use 0^0=1 ?

The hp 48/49/50 made 0.^0.=1. but 0^0 was undefined.

On the ti 89, OS 2.09 used 0.^0.=1. and 0^0=1 but for both displayed: "Warning: 0^0 replaced by 1". However, on OS 3.10, these were changed to undefined. The nspire series gives undefined.

Numworks gives undefined. The Casio fx-9860G gives an error, but I'm not familiar with the other Casio models.

So, are there any current hand held calculators for which 0^0=1 ?

[DrD]

(11-12-2019 03:28 PM)Wes Loewer Wrote:  

(11-11-2019 10:25 PM)FabVir Wrote:  http://www.numericana.com/answer/hp-prime.htm#opinion
would it be nice at least to correct the 0^0 issue

So, are there any current hand held calculators for which 0^0=1 ?

Just for fun, I tried wxMaxima, XCAS, and Octave. (I use these on my PC). Only Octave had 0^0 = 1, of these three programs.

Here is another opinion on the matter: https://en.wikipedia.org/wiki/Zero_to_the_power_of_zero

[attachment=7807]

[Joe Horn]

(11-12-2019 03:28 PM)Wes Loewer Wrote:  So, are there any current hand held calculators for which 0^0=1 ?

These don't qualify as "current" but the following HP models return 1 for 0^0 (if the inputs are reals, not RPL integer-type objects):

20b
30b
28C/S
40gs (maybe other 39/40 models?)
48S/SX/G/GX/G+
49G/g+
50g

I imagine that the 48gii does too, but I don't have one to check.

The HP-71B returns either 1 or an error for 0^0 depending on system flag settings.

[CyberAngel]

(11-12-2019 07:17 PM)Joe Horn Wrote:  

(11-12-2019 03:28 PM)Wes Loewer Wrote:  So, are there any current hand held calculators for which 0^0=1 ?

These don't qualify as "current" but the following HP models return 1 for 0^0 (if the inputs are reals, not RPL integer-type objects):
X
The HP-71B returns either 1 or an error for 0^0 depending on system flag settings.

DEFAULT OFF
Error number 6
Does not set an exception flag.

[rprosperi]

(11-12-2019 07:17 PM)Joe Horn Wrote:  The HP-71B returns either 1 or an error for 0^0 depending on system flag settings.

Like so many of these kinds of things, the 71B seems to have the right answer. An option lets both camps have the 'right' answer they need, with the flexibility to use the wrong answer, when it's right.

[FabVir]

(11-12-2019 04:07 PM)DrD Wrote:  

(11-12-2019 03:28 PM)Wes Loewer Wrote:  So, are there any current hand held calculators for which 0^0=1 ?

Just for fun, I tried wxMaxima, XCAS, and Octave. (I use these on my PC). Only Octave had 0^0 = 1, of these three programs.

Here is another opinion on the matter: https://en.wikipedia.org/wiki/Zero_to_the_power_of_zero

The link to wikipedia suggests that the higher consensus is on 0^0=1, also for practical reason in simplifying many cases.
Even if I am absolutely not an expert on the subject, I have never found a convincing demonstration of why it should be left undefined or why the result 0^0=1 should be wrong.
In my opinion, it should either be implemented as 0^0=1 as in many other HP calculators, or at least as in the HP71 with a flag that allows to choose between 1 and undefined.

[CyberAngel]

(11-12-2019 03:28 PM)Wes Loewer Wrote:  

(11-11-2019 10:25 PM)FabVir Wrote:  http://www.numericana.com/answer/hp-prime.htm#opinion
would it be nice at least to correct the 0^0 issue

There are times that I wished 0^0 were defined as 1, but I don't think I would describe the issue as needing "correcting" as the Prime's behavior is in line with generally accepted practices.

The linked article describes the 0^0 issue as a deal-breaker for the HP Prime. This makes me wonder what calculator he would be willing to buy. Are there any current calculators that use 0^0=1 ?

The hp 48/49/50 made 0.^0.=1. but 0^0 was undefined.
X

I like the 48-series way!
HP Prime CAS has on the CAS Settings page 1
Exact [ ]
but it seems have no effect on the 0.^0. handling
Perhaps the HOME Settings page 1 could benefit on a separate

Default [ ] Complex [ (a,b) ] [ ]
Setting which is similar to DEFAULT OFF in the 71B & undef (or NaN or ? or ...)

Default [V] Complex [ (a,b) ] [ ]
would then give for this problem the desired numeric value of 1.

Note that currently on the Prime both approx(oo - oo) and approx(oo^0) give undef
VPN

[cdmackay]

Unsurprisingly, perhaps, the WP 34S gives 1.

[CyberAngel]

(11-13-2019 07:18 PM)cdmackay Wrote:  Unsurprisingly, perhaps, the WP 34S gives 1.

(11-13-2019 04:53 PM)CyberAngel Wrote:  Note that currently on the Prime both approx(oo - oo) and approx(oo^0) give undef
VPN

What does WP 34S do for the above?

[CyberAngel]

FP doesn't work in CAS using HP Prime ver A Software Version 2.1.14341 BETA

Looks easy to fix

MS Windows Emulator works

MS Store Production version does not work

Android Free 2.1.14346 CAS 1.5.0 does not work
___________________________________________________________
The same problems is with IP