# HP Forums

You're currently viewing a stripped down version of our content. View the full version with proper formatting.

Hello

Spn: =( Como solucionar la ambigüedad

Si se define una matriz dentro de las variables M0, .... M9 se debe respetar el álgebra lineal o no?

internamente el autor del xCAS puede tratar listas, vectores, matrices, polinomios, conjuntos y otros como un solo tipo de dato, pero externamente debe ser transparente para el usuario, la matematica no es ambigua, el xCAS si =(

Eng: =(
if define a matrix variable M0, .... M9, Respect the linear algebra ...

one solution is to use prefixes as does MAPLE

Edited: 27 Aug 2013, 11:25 p.m. after one or more responses were posted

New BUGs:

must also solve polynomial ambiguity

poly1[1,2]+poly1[3] // OK

but

[1,2]+[3] /=/ poly1[1,2]+poly1[3]

[HP-Prime] x+y == y+x return FALSE

x^2+x == x+x^2 return FALSE

[TI-CAS] x+y == y+x return TRUE

x^2+x == x+x^2 return TRUE

request: rename conversion commands, between polynomial coefficients and polynomial, because are confusing

Poly->Coef (on catalog) => symb2Poly?? (on entry line) // SymPolyToCoef or SymbPoly2Coef or Poly2Coeff OK

Coef->Poly (on catalog) => poly2Symb?? (on entry line) => poly2Coeff // OK

Edited: 28 Aug 2013, 9:14 a.m. after one or more responses were posted

Hello all.

The real question is "even though HP ignored the bugs and bug reports on the HP-33s and 35s, will they remedy HP-Prime bugs?"

Bernard Parisse says:

giac is certainly not a nspire clone, it's designed to be a competitor to major CAS like maple

---------------

First step is to resolve the ambiguity when handling a single type of data, maple can do it, because the HP-Prime no? What hardware or software limitation exist?

Edited: 28 Aug 2013, 9:48 a.m.

Quote:
The real question is "even though HP ignored the bugs and bug reports on the HP-33s and 35s, will they remedy HP-Prime bugs?"

I suspect yes, for two reasons: (1) Fixing bugs in the upgradeable-firmware Prime will be much less expensive for HP than fixing bugs in ROMs (33s, 35s), AND (2) They have already stated that they will support firmware upgrades for Prime; they never said that about the 33s or 35s.

On the other hand, I would put good money on HP ignoring "bug reports" which are either incomprehensible (e.g. Google Translate refuses to translate postings into English if the non-English is interspersed with mathematical gibberish) or are merely lists of behaviors that are intentionally different from other calculators' behaviors.

Disclaimer: My opinion only. I don't work for HP or anybody else.

Hi there.

If it can be revealed, how would the firmware upgrades be performed? More specifically, would the upgrades be WiFi based or do they require a Windows PC and MacOS users are out of the loop?

The firmware upgrades can already be performed through a Windows tool provided inside the ZIP file which contains the firmware.

For firmware upgrades, the Prime is behaving as a standard Mess Storage Device, so the firmware upgrade tool could be reimplemented on MacOS X, Linux, and pretty much any host platform with USB support. The USB packets going through the wire have already been partially analyzed by third parties.

Quote:
[...] the Prime is behaving as a standard Mess Storage Device [...]
Nice one ;-).

Hi joe, I totaly agree with you.

We can't say that SOLVE (eq,[x,y,z]) is buggy because it returns :

{ sol1, sol2, sol3 } and you would have prefered :
x=sol1 and y=sol2 and z=sol3

or that {[1,2], [3,4]} as result of 'solve' is a bug because you prefer
(x=1 and y=2) or (x=3 and y=4)

However I hope improvements with the CAS in the way to manage matrices with adequate warnings like in HOME. imho it's not a big problem but it's disturbing

>> Gilles Carpentier: We can't say that SOLVE (eq,[x,y,z]) is buggy because it returns :

{ sol1, sol2, sol3 } and you would have prefered : x=sol1 and y=sol2 and z=sol3

---------

I never said that if solve returns the right part of the solution is an error or BUG, I said it was ambiguous

{7, 8, 9} is ambiguous may mean coefficients of a polynomial, a vector, a set, a matrix etc.

{ var1 = expr1, var2 = expr2, ... } ~ MAPLE is a better way to give the answer, this feature must clone because the HP-Prime have commands to separate the left part or right part of an expression as object1 = object2

example

if you want in a program to obtain just the right from

{ x = 7, y = 8, z = 9 }

right( {x = 7, y = 8, z = 9}({1}) ) return 7

right( {x = 7, y = 8, z = 9}({2}) ) return 8

right( {x = 7, y = 8, z = 9}({3}) ) return 9

request for HP-Prime development group

1: rename right to RightPart

2: RightPart operate in parallel on the list (one step)

rightPart( { x = 7, y = 8, z = 9} ) return { 7, 8, 9 } = )

3: requires a new command of expression to list

Expr2list({ x = 7, y = 8, z = 9} ) return { 7, 8, 9 } = )

other example

solve( x²-10*x+y² = -8 and x*y²+x-10*y = -8,{x,y} ) =>

{
[ x = 1, y= 1],

[x = 2.19, y = 3.020]

}

Expr2list(
{
[ x = 1, y= 1],

[x = 2.19, y = 3.020]
}) =>

{
[1, 1],
[2.19, 3.020]
}

>>

However I hope improvements with the CAS in the way to manage matrices with adequate warnings like in HOME. imho it's not a big problem but it's disturbing

----

I think that all CAS check arguments, this means that you lose a bit of processing time, but nothing extraordinary

-----

Bernard Parisse Say:

but I don't believe it would be a good idea to clone the HP49/50 or TI nspire CAS behavior vs lists/vectors because that would probably require a lot of work to add strict checks (and would require a lot more conversions commandcalls by the users after).

No quiero que clonen la HP49/50 or TI nspire CAS, quiero que clonen MAPLE, maple chequea argumentos esto no va a ralentizar la calculadora (de ahora en adelante haré las comparaciones mas con MAPLE y otros CAS de PC), y si se requiere conversión de de tipos de datos para pasar entre vectores, listas, matrices, conjuntos, pues hay que hacerlo, nada del otro mundo asi se simple

I: LIST TO MATRIX/VECTOR/POLY

0: list2mat( {{ 1, 2 , 3 }} ) => [[ 1, 2, 3 ]] or

mat{{ 1, 2 , 3 }} => [[ 1, 2, 3 ]]

1: list2vect( { 1, 2 , 3} ) => [ 1, 2, 3 ] or

vect{ 1, 2 , 3 } => [ 1, 2, 3 ]

2: list2poly( { 1, 2 , 3 } ) => x^2+2*x+3 or

poly{ 1, 2 , 3} => x^2+2*x+3

II: MATRIX/VECTOR/POLY to LIST

0: mat2list( [[ 1, 2 , 3 ]] ) => {{ 1, 2, 3 }} or

list[[ 1, 2 , 3 ]] => {{ 1, 2, 3 }}

1: vect2list( [ 1, 2 , 3 ] ) => { 1, 2, 3 } or

list[ 1, 2 , 3 ] => { 1, 2, 3 }

2: poly2list( x^2+2*x+3 ) => { 1, 2, 3 } or

list( x^2+2*x+3 ) => { 1, 2, 3 }

Edited: 28 Aug 2013, 10:58 a.m.

How long before we hear of the first Prime virus?

a BUG is a Virus inside a program

new BUG

inv(M1) => [1/440, 1/528]

see image up

If the Prime platform makes enough (and the right kind of) people interested in it, arbitrary code execution on the Prime will be a matter of weeks or months. From that point, anything is possible, the (legitimate !) Linux use, which would open the Prime to being much more than a calculator, being among the first things which could appear.

The Prime uses a common, publicly documented S3C2416 processor (already largely supported by Linux; also, some old, out-of-tree QEMU targets support other related S3C2410 and S3C2440 chips), its firmware upgrades are clear-binary. The TI-Nspire series lacked those two aspects, delaying arbitrary code execution on the Nspire by more than two years compared to the usual amount of time.

So far, HP has been doing quite a good job raising user / tinkerer interest in the platform :)

That said, nowadays, tinkerers tend to be more interested by the likes of Raspberry Pi Model B, BeagleBone Black, CubieBoard and others, which are less portable than calculators (no built-in battery, color screen and keyboard), but are far cheaper, more powerful and versatile. The Prime's announced price tag is in the league of powerful quad-core boards such as Wanboard Quad, ODROID-X2, UDOO quad, and even fake 8-core (big.LITTLE) boards such as ODROID-XU...