Report problems with Grapher, CAS and User Interface

[compsystems]

there is a problem in the graphing, at x=5 you have to paint a dotted line.

[compsystems]

solve([(-4*x-8*y) = 2,(2*x) = (8+4*y)],[x,y]) [enter] returns {[7/4,-9/8]}

There is one bracket too many. {[7/4,-9/8]} => [7/4,-9/8]

[jte]

(05-30-2024 02:34 AM)compsystems Wrote:  solve([(-4*x-8*y) = 2,(2*x) = (8+4*y)],[x,y]) [enter] returns {[7/4,-9/8]}

There is one bracket too many. {[7/4,-9/8]} => [7/4,-9/8]

It would seem to me that this is due to solve returning a set of results. To keep the type of result consistent, a singleton set is returned. (If it were otherwise, and incorporated into a larger routine, that routine would have to check the type of result returned — set or no.)

For example, solve([(y^2) = 2],[y]) returns {[sqrt(2)],[-sqrt(2)]}.

[compsystems]

When plotting, the graphics engine does not calculate the field of view, and the viewing plane is usually empty.
An automatic adjustment is necessary.

[compsystems]

In full screen the simulator makes a bad keyboard adjustment.

Also the virtual key press effect is not displayed very well. I suggest placing a highlighted colored frame to see which key has been pressed.

[compsystems]

The index of a root looks like a factor since the number has a very large character, other characters could be used as superindices. ⁰¹²³⁴⁵⁶⁷⁸⁹

[compsystems]

In the definitions it is not necessary to equal them to zero (=0) or (==0). or at least to be added when you press enter as well as the definition in the history of expressions

[compsystems]

Hello, some suggestions for improving the UI

In CAS Mode almost no ► operator is used to store data, the most common is :=, so I propose for the next update a menu with assignment operators.





Another simple, but significant improvement is that to display the cursor in the expression entry bar you have to press the cursor key up and then press it down again. To go to the expression editing bar first, it would be sufficient to press the cursor up key.

[compsystems]

I think the evaluation priority of the operator (==) follows a strange rule.

Looking at the sequence of expressions from top to bottom, this expression should evaluate the left side, then the right side and then check for equality, because step by step it does it but in one expression it does not.

Moreover, when entering the expression (5==5) == ((e^(π*i)+1)==0) it writes it in a strange way => (5==5) AND (5==(e^(π*i)+1)==0))

[compsystems]

I have never liked the asterisk as a synonym for the multiplication operator, in mathematical notation this symbol is not standardised.

The history of input and output is in the ability to display official signs.

When typed on the input line it is decoded as * and when entered it is shown as × or ·

[compsystems]

Hello
For those of us who use Hp-prime screenshots as documentation, showing the time at the top is not very useful, more important is the date, which can be shown by a flag.

[jte]

(07-09-2024 03:59 AM)compsystems Wrote:  I think the evaluation priority of the operator (==) follows a strange rule.

Looking at the sequence of expressions from top to bottom, this expression should evaluate the left side, then the right side and then check for equality, because step by step it does it but in one expression it does not.

Moreover, when entering the expression (5==5) == ((e^(π*i)+1)==0) it writes it in a strange way => (5==5) AND (5==(e^(π*i)+1)==0))

One possible issue at play here is operator chaining. (https://en.wikipedia.org/wiki/Relational...r_chaining)

For chained relational operators OP1 and OP2, “A OP1 B OP2 C” is interpreted as “A OP1 B and B OP2 C”. Chaining relational operators is fairly typical in a mathematics context, but not so common in a computer programming context.

I'd think the presence of the parentheses rules should rule out chaining in this case.

Some portions of the HP Prime currently chain relational operators whilst other portions do not, or only under certain restrictions. PPL didn't chain relational operators initially, but was later changed to chain relational operators in limited ways. The Advanced Graphing interpreter chains relational operators. Python chains relational operators (and handles parentheses how I'd expect, comparing “1 < 4 > 2” to “(1 < 4) > 2” in a Python interpreter).

[jte]

(05-29-2024 09:06 PM)compsystems Wrote:  there is a problem in the graphing, at x=5 you have to paint a dotted line.

Decades ago, I implemented a graphing engine that would produce this sort of graph (one with a dashed line showing the gap). Hopefully we'll get to the point where something like that could be added to the HP Prime. (This requires a new graphing engine.)

[jte]

(07-15-2024 05:40 AM)compsystems Wrote:  Hello
For those of us who use Hp-prime screenshots as documentation, showing the time at the top is not very useful, more important is the date, which can be shown by a flag.

When I was recently making some changes to the HP Prime Virtual Calculator's screenshot code (addressing a memory leak), I was thinking it could be nice to have some control over screenshot decorations in the UI of the HP Prime Virtual Calculator (not within the calculator, but within the Windows program / macOS application) — in exactly the context you're mentioning, the production of documents that include screenshots . (And, similarly, in the HP Connectivity Kit.) But I wasn't explicitly thinking of date inclusion options. If changes are made there, having some options for datestamping is certainly a good idea.

[jte]

(06-15-2024 04:41 AM)compsystems Wrote:  In the definitions it is not necessary to equal them to zero (=0) or (==0). or at least to be added when you press enter as well as the definition in the history of expressions

When the Advanced Graphing app was first being added to the HP Prime, I first implemented more descriptive error messages, but those were removed due to reported interaction issues with the ARM compiler. It’s something that should be revisited. I’ve created a ticket for this on the bug tracker.

Thinking a bit more about this now, since open sentences are the intent (for V# definitions in the Advanced Graphing app), evaluations outside of plots should also be changed. Flagging errors at the time of entry is possible in some situations. I’ve created tickets for these too.

[Wes Loewer]

(07-06-2024 05:14 AM)compsystems Wrote:  Hello, some suggestions for improving the UI

In CAS Mode almost no ► operator is used to store data, the most common is :=, so I propose for the next update a menu with assignment operators.

I always thought that it was odd that the ► operator was placed on the soft menu. Since the same operator is on the keyboard (Shift-EEX), it seemed a shame to use up a spot on the menu for it.

Better yet, I would love to have built-in MENU and TMENU commands (like the 50g) that allow the user to put whatever they want on the menu. I really miss those.

[compsystems]

Hello

When writing the expression (a<b)+c and then retrieving it from the history line to input, the order of the brackets () is lost, rewriting it to a different and erroneous expression

(a<b)+c [enter] (b+c)>(c+a) // == (a+c) < (b+c) // ok
(a<b)+c [copy] => a<(b+c) bug // (a<b)+c ok

[compsystems]

I have detected another problem
if
(a<b)+c [enter] (b+c)>(c+a) // ok
then why

(-4<-3)+1 [enter] 2 ? is expected (-4+1<-3+1) => (-3 < -2)

does it fail?

is not applying the transformation rule (a<b)+c = (b+c)>(c+a)

Is evaluating Boolean logically first an algebraic error or not?