New Features and Improvements WANTED!

[-TJ]

Dear Group,


Have you ever felt that the TI-Nspire or the TI-Nspire CAS lacked
something you wanted? Was the TI-Nspire and the TI-Nspire CAS
v1.7.2471 update unsatisfying? Yes, I know that improvements can not
be made immediately, but the point is that the number of improvements
simply doesn't add up. For the time Texas Instruments took, there
hasn't been major improvements. And as much as we argue about this, it
won't help at all. Instead of talking about what improvements have
been made, lets discuss about what improvements should be made!

Why not dump a list of all of the things that you might want to
consider the TI-Nspire calculator family to have. Any improvements
could be suggested; here are just a couple things I would like to see
in the next update to v1.8:

---

001: Faster boot time - If only it could start faster... Sometimes, I
don't have any patience. It loads all of the documents from the NAND
Flash Rom to the SDRAM. Why not just have an index of the files on the
calculator available, and open a document only when necessary? The
calculator doesn't take that long to load documents from its NAND
Flash Rom. When exporting documents to a computer via a usb cord, file
transfers are pretty fast; simply opening one file would be faster.

002: Developer UI - Every time the calculator boots, it loads
seemingly useless "help" information and things that are unnecessary.
There should be a more streamlined version of the operating system,
exclusively for "expert" users. Since the operating system takes a
significant portion of the memory, when the calculator is low on
SDRAM, there aren't many options left than to reboot. These are the
times when I feel like operating system has a lot of "junk". When it
adding the suggestion 001 onto this one, speed memory availability
will significantly increase!

003: Emulation Font Size- When we use TI-Nspire or TI-Nspire CAS
Computer Software Teacher Edition's SmartView Emulator, we don't have
the ability to change the font! Why not?!

004: Emulation Button Feedback - It would be more realistic if the
emulator would mimic the look of a button being pressed rather than
turning the button red. It feels uncomfortable leaving it red, after
pressing it.

005: Emulation Touch Capability - Since I am using a pen, touch, and
multi-touch input device built on my computer, it would be nice for me
to simply press multiple buttons at a time, on the software. Pressing
[ctrl][up] would be much easier than having to press them one at a
time. In addition, it seems difficult to press any alphabetical key on
it too.

---

Once again, the following were just some couple suggestions made.
Continue adding your own wanted features, improvements, bug fixes, and
anything else. Together, we can see the progress at which Texas
Instruments is developing and improving. =]

IT WOULD BE NICE TO HAVE A SUGGESTION NEATLY FORMATTED WITH NUMBERS;
REFERRING TO PREVIOUS SUGGESTIONS COULD THEN BE EASILY MADE!

[CONTINUE ON WITH NUMBER 006!]


Sincerely,
-TJ

[Eric Findlay]

006: Custom Errors - With the new programming capabilities, they have
added the Try...Else blocks, but the errors are set, and sometimes you
want one that's not listed. Thus, I propose the "throw" command, like
in C++, where you can define an error and then pass the error message to
the user like a system-produced error message (with abilities like go to
and view, etc.)

Responses:

001 - we've discussed this off the thread. First of all, it doesn't
load all the documents, and it DOES keep an index to the files. This is
called the File Allocation Table. When you are looking at the list of
files on the calc, then select one, that is when it loads. Changing the
way the OS boots or files are loaded will almost assuredly not happen
because it would require a near rewrite of the entire OS.

If you're referring to the load time of a document or loading all the
pages/problems at once, this goes back to our out-of-thread discussion.
It may also be because you work with abnormally large files - larger
than the average person.

Computer transfer time is faster than opening because it does not need
to translate the data into its visual representation and display it, it
merely sends it off down the wire.

002 - If you think it loads too slow, then adjust the power down options
so that it never loses the OS. Then you never have to wait for it to
load, but the batteries won't last as long - it's a trade off. Again,
this is also because you use abnormally large files. It will not be a
problem for 90%+ of the users.

003-005 - are you talking about the computer software? Is this
discussion about the computer software or calc OS?

004 - it stays red for purposes of demonstrations, so your students can
see which button you pressed last. I don't see them changing this.

> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 8.5.409 / Virus Database: 270.14.3/2410 - Release Date: 10/02/09 18:24:00
>

--
For I am convinced that neither death nor life, neither angels nor
demons, neither the present nor the future, nor any powers, neither
height nor depth, nor anything else in all creation, will be able to
separate us from the love of God that is in Christ Jesus our Lord.
- Romans 8:38-39 (NIV)
--
Eric Findlay
AKA Eagle-Man

[Michael Ball]

006: Larger Font Size (On Device). It'd be nicer if the font were
larger for notes and for Calculations pages. It's slightly better
than the 89, but I want a bigger font. Large should be Larger. Maybe
pt size options.

007: Easier link to a Function Table from G&G. I know you can do a
half page one, but I'd really like an easy link to a full page Fn
Table. Or maybe a user preference setting would work too.

008: more conversion options, especially for velocity

009: Subfolders for documents

[Eric Findlay]

006M (for Michael): Even bigger than the Large font? Wouldn't you then
lose too much screen space to very little text? (especially with split
screen)

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 8.5.420 / Virus Database: 270.14.3/2412 - Release Date: 10/03/09 18:34:00

[Susan Reeve]

I agree with needing larger font size, especially for the notes page.  It would be nice to have a title on the page that is bigger than the other print. --- On Sat, 10/3/09, Eric Findlay <eagl...@duetsoftware.net> wrote:

[Joe]

What boot time do you experience and is it for the nspire or nspire
cas handheld? I'd like to compare it to the boot time for my netbook
which boots up and is ready to use in about 28 seconds.

On Oct 3, 5:51 pm, -TJ <tjra...@gmail.com> wrote:

[-TJ]

Dear Micheal,

The whole discussion includes both, the calculator and the software.

Hey Micheal Ball, I saw your TI-Nspire video on YouTube! =P
Hehehe...

Sincerely,
-TJ

---

Dear Eric,

Replies to comments on the following suggestions:

006E: Doesn't the TI-Nspire have a function to identify error? It is
"errorCode". Although it isn't in the catalog, type it in and it won't
be in italics because it is a reserved variable for the calculator.
When you use the Try command, for example...

Define abc(x)=
Prgm
:Try
: Disp x&"bc"
:ElseIf
: Disp errCode
:EndTry
EndPrgm

and you run abc("a") then it will display "abc", but if you run abc
(a), it will display 210, which is the errCode. The reference guide
says that when errCode=210, then "Invalid Data type An argument is of
the wrong data type". You can see this on page 124 of the TI-Nspire
(Non-CAS) Reference Guide.

Sincerely,
-TJ

---

Dear Group,

From now on, Eric's Suggestion 006 will be referred as 006E and
Micheal's suggestion 006 will be referred as 006M.

Sincerely,
-TJ

[-TJ]

Dear Group,

I apologize for sending multiple messages. Please excuse me for this
inconvenience.

Sincerely,
-TJ

---

Dear Micheal,

Wow, your netbook boots up fast!!!
My tablet boots up Windows 7 Ultimate 64-Bit w/ Windows Tablet PC
features in 32 seconds. That's still close. =P

The boot time for the calculator varies on the amount of data stored
on your NAND Flash ROM.
The more space your documents take up, the long it takes for it to
load.

TI-Nspire CAS v1.7.2741
00:00:00:00 - 00:00:03:76 >> Loading File System and Operating
System
00:00:03:76 - 00:00:16:93 >> Loading UI and Settings (w/ clock/
waiting cursor)
00:00:16:93 - 00:00:22:22 >> Loading Documents (w/o clock/waiting
cursor)
00:00:22:22 - 00:00:48:62 >> Finished

My computer boots faster than that! >=[

I don't have a TI-Nspire (Non-CAS), so I can't time that. Sorry. =/

Sincerely,
-TJ

---

Dear Eric,

The following are replies/comments to your responses of my
suggestions:

001: I thought the NAND Flash ROM contents were copied onto the SDRAM
at every boot up! Is that wrong? I always explained it like that. And
I'm hesitant to mess around with the boot loader, for my calculator
might become permanently disabled. It does have 2 boot loaders,
though.
Well loading the display contents sort of relates to our out-of-group
discussion on how some of the content could be loaded to display,
which the rest remains in untranslated. If it takes almost a second to
load the entire document and consuming large amounts of memory, why
not wait a fraction of a second, only loading the first page, and
saving memory?

002: But then again, a lot of the "help" files are useless to us!
Those are only for new users. It takes up memory, and that is
inefficient. That is why I want to get rid of it. =/

003-005: I am talking about the TI-Nspire CAS TE Computer Software
with its built-in TI-SmartView Emulator.

004: Well, there should be an option for turning the red thing off, so
it seems like a virtual calculator on a screen (intuitive to multi-
touch tablet inputs).

Sincerely,
-TJ

---

Dear Susan,

006M: I also agree with Micheal's suggestion 006M! I would like have a
font size option too! That would make directions for tutorials and/or
demos much more comfortable to look at.

Sincerly,
-TJ

---

Dear Group,

010: I want the text to be in any combination of bold, italics,
underlined, crossed-out, subscript, superscript! Usually, I want the
title of the document to be bold and underlined! Unfortunately, Texas
Instruments totally forgot about this!

011: I want to get rid off the "Hints" menu thing in all of the pages
(for calculator, g&g, l&s, notes, and d&s)! It seems useless for
experts, and a waste of time to those who want more SPEED...
Apparently, this calculator sometimes goes up to 3 times slower than
the TI-89, despite TI-Nspire's increadible speed (up to at least 10
times the speed of TI-89).

Sincerely,
-TJ

[Eric Findlay]

Again, TJ, if this is really such an issue for you, disable the
Hibernate setting and this boot up time will NEVER happen again (ok, it
will happen when you switch face plates, reset or update the OS).

Home>8:System Info>2:System Settings>Hibernate = Never.

All it will cost you is battery life.

And to the people who want a larger font - Have you tried switching the
System Font to Large? It's quite big on the Notes page. I wouldn't
want anything much bigger than that or you can't fit much on the screen.
The Large font allows for about 35 characters on one line of the Notes
page. Medium allows for about 40 characters, and Small allows for about
50 characters.

The font size can be found at
Home>8:System Info>2:System Settings>Font Size

--Eric

[Ross]

The features I'm hoping to look forward to in 1.8 is:
1. Built in function for finding the inverse of a function. If the 84
can do it, then why shouldn't the Nspire as well?
2. The function "input". We need some better way to have user input
in programs
3. Functions for programs to directly communicate with graphs and
geometry screen to do things like create new elements, modify existing
elements, ect.

[Nelson Sousa]

programs can modify existing elements. How do you think I get away
with most of the stuff I create? ;)

just create any object and link it to a variable. You can have a
program modify that variable and the object will also be modified. If
there are many objects you can have them drawn over a scatter plot and
have a program modify the lists that define that scatter plot.

Almost every document I've created uses functions (or spreadsheet
formulas in most cases) and in some cases programs to modify the way
things are displayed.

What you can't do is create new G&G objects using programs and/or
functions. It surely requires some planning to get things going but
there are some amazing things being acomplished even with the limited
input/output functionality.

As for boot time: if the unit isn't in hibernation mode, it's
instantaneous; if not it should take roughly about 1 minute, maybe
less (I don't have my Nspires here with me to try that)


Nelson

[Daniel Dudley]

I would like a simple way to "unsplit" the screen. For example, I may want to show a relationship between something in a spreadsheet on a graph just for a couple of instances where it would be inconvenient to keep flipping back and forth between screens. I just want to see a cause and effect, and then return it to either the spreadsheet or graph in the full screen. Does this make sense?

Dan

________________________________

[-TJ]

Dear Group,

I know I write in great detail, but when others reply to what I write,
they include what others have written as quoted text. This is just
making the discussion extremely long, even though the messages are
quite short.

IF POSSIBLE, I WOULD ADVISE NOT TO INCLUDE QUOTED TEXT IN YOUR
MESSAGE!

The only reason I quote text is to refer what they say as a number!
Because they forget to add a number, I have to go back and assign the
number as well as to mention what they said. =/
Could everyone just label your suggestion by number(s)? It would just
make life easier... =/

PLEASE REFER/LABEL YOUR SUGGESTIONS NEATLY FORMATTED WITH NUMBERS!


Sincerely,
-TJ


--- THE FOLLOWING IS TO CLARIFY AND ORGANIZE CONTENT MADE FROM OTHERS
AS WELL AS TO REPLY TO COMMENTS ---

Dear Joe,

When I said:
"Dear Micheal,

Wow, your netbook boots up fast!!!
My tablet boots up Windows 7 Ultimate 64-Bit w/ Windows Tablet PC
features in 32 seconds. That's still close. =P

The boot time for the calculator varies on the amount ..."

I meant to say "Dear Joe," I apologize for this confusion.

Sincerely,
-TJ

---

Dear Eric,

001: Thanks for your comments, truly I am not concerned about the boot-
up speed. I may have not stated my opinions clearly; let me rephrase
them. What does concern me is the "useless" content; I judged this by
how slow the calculator boots up.

006M and 010: I wanted to add the functionality of changing the font
size. What I mean is changing the font size of particular text, not
the system font size. Changing the system font size will make all font
bigger; this may be a user preference, but I am talking about making a
more dynamic type of text. As of suggestion 010, it would be nice to
have text to be bold and italics, italics and underlined, or bold and
underlined, etc. With a combination of settings that intertwine
together, the results will become more unique and custom made/suited
for the application/function/program. Sometimes, people want LARGE
titles and smaller sized text underneath, emphasizing certain things
and not others.

Sincerely,
-TJ

---

Dear Group and Ross,

012, 013, and 014: Ross wrote:
"The features I'm hoping to look forward to in 1.8 is:
1. Built in function for finding the inverse of a function. If the 84
can do it, then why shouldn't the Nspire as well?
2. The function "input". We need some better way to have user input
in programs
3. Functions for programs to directly communicate with graphs and
geometry screen to do things like create new elements, modify existing
elements, ect. "

The features he mentions/wants will be referred as 012, 013, and 014
respectively.
Just to repeat with emphasis ...

THE FEATURES HE MENTIONS WILL BE REFERRED AS 012, 013, AND 014
RESPECTIVELY.

Responses:
013 and 014: What do you mean? Are you trying to emphasize custom
interactive programs that do not require the base pages (instead of
using calculator, G&G, L&S, notes, and/or D&S, we should use a custom
made window specifically for the program, such as assigning the keys
to do different things, similar to the programming capabilities to the
ti89)? Sorry, that was a run-on sentence; I hope that made sense.

Sincerely,
-TJ

---

Dear Group and Micheal Ball,

015: Micheal Ball wrote:
"Ok, I think this would be cool for both the software and the
calculator:

I'd like the ability to control the computer software with a
calculator. It's much faster this way, rather than clicking on screen
buttons. Or, if I could simply view the calculator screen on the
computer in real-time—not just screenshots (which are useful too)."

015: As a response, I wrote:
"I feel EXACTLY the same way! I had the same feeling when I first
started using the calculator! This will absolutely make it easier to
use and demonstrate! As you said, we could use the buttons on the
calculator to control whatever comes up on the screen, and the buttons
would be similar to how we use a keyboard! Instead of using the
computer keyboard, we can use the calculator's keypad! Instead of
having to type of click buttons with the mouse, we use the calculator,
and then the software could be projected onto a projection screen! In
this case, the calculator is doing minimal work, just sending the keys
and buttons that are being pressed, and the rest is taken care of by a
computer! In this case, since we could use the computer to do all of
the work, everything will run much more smoothly and faster!"

MICHAEL BALL'S SUGGESTION WILL BE REFERRED AS NUMBER 015!

Sincerely,
-TJ

---

Dear Nelson,

001: What do you mean when you said "it's instantaneous"? I didn't
understand.

Sincerely,
-TJ

---

Dear Group,

016: Dan wrote:
"I would like a simple way to "unsplit" the screen. For example, I
may want to show a relationship between something in a spreadsheet on
a graph just for a couple of instances where it would be inconvenient
to keep flipping back and forth between screens. I just want to see a
cause and effect, and then return it to either the spreadsheet or
graph in the full screen. Does this make sense?"

HIS SUGGESTION WILL BE REFERRED AS 016!

Responses:
016: I think this is a great idea, but I don't exactly know what you
mean when "return it to either the spreadsheet or graph in the full
screen". Do you mean to have the page in two screens, but the screens
can hide in a corner, so that you can open and pull one out to see it
in full screen and hide the other one in the corner?

Sincerely,
-TJ

[-TJ]

Dear Group,

Once again, just to get this in everyone's head, just PLEASE use
numbers to any suggestion you add.
Also use the suggestion numbers when referring/commenting on them. =/

I really don't want to stress over this simple direction, and its
already getting harder to track the suggestions and who is commenting
what; implying everything will soon fail.

Sincerely,
-TJ

[-TJ]

Dear Group,


Once again, the following were just some couple suggestions made, but
we need to continue adding more.
Continue adding your own wanted Features, Improvements, Bug Fixes, and
ANYTHING ELSE.

Together, we can see the progress at which Texas Instruments is
developing and improving. =]

Sincerely,
-TJ

[Steve Phelps]

perhaps not new features, but things I do NOT like...

I really dislike how when you hover over a graph, the function
equation show. Really do not care for that.

[Wayne]

017: I would like an option for the DelVar command that deletes all
variables in a problem, say DelVarAll.

Wayne

[Daniel Dudley]

Isn't this the Clear a-z under menu->4? Or do you need something in a
program?

Dan

-----Original Message-----
From: tins...@googlegroups.com [mailto:tins...@googlegroups.com] On
Behalf Of Wayne
Sent: Tuesday, October 27, 2009 9:26 PM
To: tinspire

[Susan Reeve]

How about scientific notation?  Anyone know where that is?  --- On Wed, 10/28/09, Daniel Dudley <Dan...@vvsaz.org> wrote:

[Andy Kemp]

Do you mean like 5.2x10^11?  If so you want the EE key and enter 5.2E11

If you want answers returned in that form then the setting in under document settings - exponential form - Scientific, if you prefer it to round to nearest majour unit e.g micro, nano ect then choose engineering instead of scientific

[Susan Reeve]

I want 5 x 10^5 times 6 x 10^4 to come out as 3 x 10^10.  I can do it on my TI-34, but even when I put the document setting in scientific, it still comes out as 30000000000.  I think this is a feature that we need to have if we want to put it in the hands of middle schoolers.  Thanks for any help you can give. --- On Wed, 10/28/09, Andy Kemp <an...@1kemp.co.uk> wrote:

[Andy Kemp]

I just did 5E5 * 6E4 in Scientific mode and got 3E10 which is what you wanted...  Try starting a new document to make sure there isn't something odd happening...

[Nelson Sousa]

5x10^6 is an integer, an exact number; if you multiply by another integer you get an exact answer, unless you specify Approximate mode.

However, using 5e6 you're specifically saying you want scientific notation, which uses approximate numbers. 

Answers to approximate questions are approximate, answers to exact questions are exact.

Does this help?

Cheers,

Nelson

[Wayne]

The clear a-z applies only to one-letter variables. I want to clear
all variables, whether they are one- or multi-letter names.

> Wayne- Hide quoted text -
>
> - Show quoted text -

[Sean Bird]

I imagine HOME > New Document
wouldn't satisfy for the situation you have?

Or CTRL HOME (Tools) > Insert > Problem

[Daniel Dudley]

With “pretty print” what’s up with still using that darn “E” for scientific notation.  It still confuses my students (though I don’t know why).  I guess that would be 018 for suggestion.  Write it in proper scientific notation.

 

Dan

 

---

[Susan Reeve]

I agree!  Make scientific notation look like scientific notation.  Even a lot of the scientific calculators show it as a power of 10.   This is what I'm looking for too.  The trouble is students are so used to thinking the calculator is always right, that they write down exactly what is on the screen.  The TI-36 doesn't even show a 10, it just assumes the student knows it's there, and you get students writing 3.5^12 on their paper, and you wonder where they've been all week, until you find out that was what was on their calculator screen.  Thanks for everyones help on this, knowing the E trick at least gets us in the right form.

[-TJ]

Dear Everyone,

017: I like your idea, about DelVarAll, to delete all variables in the
problem without having to delete the problem variables one by one!
Although, as someone mentioned, there is a thing called "clear a-z" or
something like that, DelVarAll can be a function that deletes all
variables, not only the ones that are in the alphabet! Sean mentioned
how you can just make a new document, but that is not really
convenient. It would be easier if there is a function. This is a great
idea, Wayne!!!!

018: Sean's suggestion reminds me about something. I wish there was a
shortcut to make a new problem, just as you can by inserting a new
page with [ctrl][i].

Sincerely,
-TJ

[Nelson Sousa]

I don't like the idea of a "delvarall". It's dangerous and it doesn't
really solve anything.

For me, on most situations, I can have up to 20 variables that I want
to delete, but still a few that need to be kept.

It's be better to allow wildcards (* to specify any set of characters
and ? to specify any character) or to specify the type of variable to
delete (delvar type=real or delvar type=list, for example).

I don't like the idea of having to delete variables writing the names
of all of them (ok, I use the var menu to paste the names, but still,
to delete 20...), but having a "delvarall" wouldn't do me any good
either.

A mass deletion feature could be easily programmed if we had a
function to list all variables on a problem. Example, varlist() could
return a list with the names of all defined variables; then, a few
simple loops could allow some simple yet powerful mass deletion.

Cheers,
Nelson

[Wayne]

I agree with the approach of allowing wildcards in the delvar command
rather than a global delete of all variables. The varlist approach is
also a reasonable way to provide some added functionality. Either one
would make my life a little easier. I just don't like to have to
paste all the names of the variables when I want to delete a list of
variables. That is particularly annoying when I want to delete
variables that have been named in a group (like tvm. for example).

Wayne

> > -TJ- Hide quoted text -

[lee kucera]

I just want it to clearly and easily show all relevant data points on
a boxplot--like the 84 does. It only displays Q1, median and Q3. The
endpoints are problematic depending on whether you have any outliers.
If there are no outliers, you can hover over the endpoint and display
it. However, it is way too easy to change a data value while trying
to get it to show when there are outliers. URGH!
lk

lee kucera
a.p. statistics
leek...@gmail.com

[Sheila]

006 Larger font size on device-definitely font too small, sometimes
hard to distinguish between multiply signs, decimal points, and
negative signs.

007 I definitely think there should be an easier way to see a function
table than a spllit screen that is not easily read and only shows one
function at a time. This list and spreadsheet way is a nice look but
it is too much trouble...there should be a cntrl graph which goes to
tables of the graphs easily.


0017 I want to graph inequalites, including vertical lines. Again if
the 84 can do it, the nspire needs to do it better. Also it is too
difficult to delete an inequality after graphing, can be very
frustrating as a student.

> > all variables, whether they are one- or multi-letter names.- Hide quoted text -

[pam]

Sheila,

To see a function table as a whole page, add a L&S page and then "View
Function Table." My students find this to be a fairly simple way to
see the numerical values of up to three functions at a time.

Pam

[Mrs. Heidi Rudolph]

Did anyone mention rechargeable batteries? (I was scrolling through
rather quickly...)

And what about turning a Key Press history feature on to alleviate the
RED button feature that some object to?

> > all variables, whether they are one- or multi-letter names.- Hide quoted text -

[Eric Findlay]

Hello, Heidi. May I ask you to expand on your battery suggestion? Do
you mean a special built-in rechargeable battery?

I, for one, would be opposed to that for a few reasons:
1) You would need a special charger that you would have to drag around
with you everywhere. It limits portability.

2) If the batteries die, you have to wait for it to recharge before you
can continue using it.

3) When the battery eventually stops holding a charge, it would be
either expensive or impossible to replace, so you'd have to buy another
calculator.

4) Rechargeable AAA batteries are cheap and readily available, so you
can have a spare set ready to go when the current set dies.

--
For I am convinced that neither death nor life, neither angels nor
demons, neither the present nor the future, nor any powers, neither
height nor depth, nor anything else in all creation, will be able to
separate us from the love of God that is in Christ Jesus our Lord.
- Romans 8:38-39 (NIV)
--
Eric Findlay
AKA Eagle-Man

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 8.5.423 / Virus Database: 270.14.42/2473 - Release Date: 10/31/09 21:14:00
>

[Nelson Sousa]

I use rechargable AAAs on my Nspires and I'm happy with them. I bought
the new generation of batteries that have a much longer shelf life,
they usually last a few months if I'm not using the handheld much
(thank goodness for hibernate!), a few weeks if I use them a lot.

Cheers,
Nelson

[Eric Findlay]

I think this is #19

019a: Search function for notes pages. So if there's a long notes page,
you can search for a word and jump to that portion.

and/or

019b: Hypertext/bookmarks that could click to jump to a different page
or another part of the notes page.

I would like for both of these to be programmable, too. For instance,
you could do "search(problem, page, search term)" or something like that.

--
For I am convinced that neither death nor life, neither angels nor
demons, neither the present nor the future, nor any powers, neither
height nor depth, nor anything else in all creation, will be able to
separate us from the love of God that is in Christ Jesus our Lord.
- Romans 8:38-39 (NIV)
--
Eric Findlay
AKA Eagle-Man

> anything else. Together, we can see the progress at which Texas

> Instruments is developing and improving. =]
>

> IT WOULD BE NICE TO HAVE A SUGGESTION NEATLY FORMATTED WITH NUMBERS;
> REFERRING TO PREVIOUS SUGGESTIONS COULD THEN BE EASILY MADE!
>
> [CONTINUE ON WITH NUMBER 006!]
>
>
> Sincerely,
> -TJ
> >
>
>
>
>

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 8.5.409 / Virus Database: 270.14.3/2410 - Release Date: 10/02/09 18:24:00
>

[DrBob]

Hi Gang,

Just joined the group. I'll put in my 2 cents worth in relation to my
background as research physicist/engineer and teacher (secondary/
tertiary), amongst other things. I'll make some comparisons to
features of similar technologies I use, such as the HP 50g, Casio
fx9860g AU, and recent high-end models by SHARP.

I held off on purchasing the nSpire CAS until recently in the hope
that any bugs would be ironed out. Sadly, I was disappointed (and have
informed TI of such). They have taken software previously used on the
PC ("Derive", I believe, which is what I used to know it as) and tried
to port it to a slower platform. By slower I actually mean the method
of input used. Keyboard and mouse are much more flexible than
'cramped' keyboard and trackpad (or whatever you call it). This
hampers operations straight away, wasting precious time for students
during exams, etc.

So, how to fix this? (I think we're starting from #20 now :-):

020 Need to look at how to speed up moving around objects on screen.
Either speed up pointer movements or use valuable and easily
accessible keys like <TAB> to flick between screen objects, especially
on the GRAPHS or STAT plots screens (compare to the reliable and
intuitive way the old Casio 9860g does it).

021 Adjust the KEYBOUNCE setting (such as on the HP calculators) such
that the first key pressed is the only one registered, ignoring any
other key accidently pressed for the next few milliseconds. This will
stop the ludicrous situation of trying to enter numbers (which is most
affected by this) and finding interspersed letters and pronumerals
accidently entered in your calculations (again, a huge time-waster,
especially for those with, shall we say, stout fingers :-).

Here's a problem that should have been fixed over the last 20 years,
but hasn't.

022 All calculators, as far as I can see, still don't put the
negation operator (-) at the top of the order of precedence. If you
try typing "-2" squared on any machine you are still given the answer
"-4", which is of course totally wrong. If you try typing "-2" squared
squared, etc, on some models they even put in the brackets for you,
and you can see that the calculator interprets this as '-((2)^2)
^2' (with the negative sign ignored till the end), again giving the
wrong answer. In the 'olden days' of line-by-line calculations we
always got around this by remembering the venerable brackets; however,
with machines such as the nSpire CAS, where you can select a previous
answer and press <ENTER> to place it into the present calculation this
presents somewhat of a problem.
Here's a typical sequence of events:
I first calculate 2 x -4 = -8 (correct ..... one tick! :-)
Then, for fun, 3B x C = 3BC (cool)
Now, however, I find I need to calculate 5^2 + (previous answer of
'-8')^2 (ie. 5 squared plus minus 8 squared),
so, I type "5^2 + ", then go up 2 lines to select the "-8" from
before, and press <ENTER> to give me what I thought I wanted, ie. 5^2
+ (-)8^2
I really wanted 5^2 + (-8)^2 = 89 , but instead get 5^2 + -(8)^2 =
-39, NOT GOOD!
So, to fix the problem we need to make the negative sign 'stick', and
be considered part of, the number. ie. Treat it like a variable, as in
t = -8, therefore t^2 = 64
Note that the situation is slightly different with pronumerals. If I
have y = -t^2 (ie. minus t-squared), then t^2 takes precedence (which
is good if 't' is negative, as this gets evaluated first), then apply
the negation operator. Quite a simple fix though -- treat numbers as a
'sign + an integer/real), etc. Could TI be the first to correct this?!

This leads to the next series of problems to fix on the nSpire CAS, to
do with the handling of units, which is quite bad:

023 A number, say, with a unit (such as metres, '_m') should be
treated as a single object, such as on the HP 50g calculator, where
this combination is stored as a string. If I do the following simple
operation on the CAS, which should be intuitive, you get the wrong
answer:
We all know that 1 divided by 4 metres should give 0.25 m^(-1).
The CAS however insists on treating the unit as multiplied onto the
number; hence, 1 divided by 4 metres. ie. "1/4._m" comes out as
0.25._m (not 0.25 metres to the minus one). Confusing for students.
The HP does 1/'4_m' = 0.25_m^-1, as expected (with the '4_m' stored as
a single string). Again this problem is exacerbated by copying answers
from previous lines containing units.

024 Speaking as a physicist now, the units under the heading
"Velocity" list knots, kilometres per hour and miles per hour. The
base SI units of metres per second have been forgotten. NOT GOOD! :-)

025 By the way, these should be under the heading of "Speed", NOT
"Velocity". A velocity has both magnitude and direction (and TI
developers should know this already :-)

026 The fact that the conversion of temperatures can't be handled by
the usual conversion operator approach is laughable. Any child could
program a calculator routine to handle this. The problem, it appears,
is due to the way the nSpire handles the combination of the degree
symbol and the following 'C', 'F' or 'K' (Kelvin).

027 Please give us the option of whether to use derived units in the
answers or not. Not all answers with a unit of 's^-1' should be quoted
as 'Hz' (Hertz). Often an answer should be left as 'kg m s^-2', say,
rather than 'N' (Newtons), especially if we want the students to work
out the unit conversions themselves.

028 The Data & Statistics application is great. Pity that you can't
use the fitted equations for anything there and then. You can't use
any fitted curves for quick prediction of X from Y values, or Y from
X, in that application. You can't even edit a TRACE point on the curve
manually to put in a desired X or Y value as you can do in the Graphs
and Geometry application. The only way at present is to recall the
fitted equation in the calculator application (through STAT
variables), cut out the relevant section to the right of the equals
sign, paste it into a Graphs and Geometry screen, and then play with
the curve that way. This is very poor design! All other machines let
you fit a curve, then directly put in a value of X, say, and get a
corresponding value of Y. The one other way to cut and paste fitted
equations is to fit the data in the spreadsheet (and even then you
can't predict any values within that application). How many times do
you have to perform the fitting before you can both visualise
(Australian spelling! :-) the data and predict values.

029 This highlights the inconsistent cutting and pasting options. You
can't cut something in the STATS application and paste it into the
CALCULATOR page, for instance. What a pain!

030 It seems to be a gimmick to have the ability to move points in
the STATS (Data & Statistics) window. I don't think you can even enter
a specific (x, y) value.
More useful is the ability to be able to enable and disable (include/
exclude) points from a fitted dataset to see the effect on the fitted
curve. This is actually extremely useful, especially when explaining
the effect of outliers. Very useful in research applications as well
(and is a good selling point for some of the top statistics and data
analysis packages).

I'll finish on the next couple of points (for today -- phew!):

031 The CREATE NEW MATRIX menu item <MENU> + 7 + 6 + 1 is totally
useless and just wastes time. The CAS has such great matrix
functionality it's hard to believe this 'clunky' function slipped
through. It produces a matrix filled with zeros (ie. '0'), which you
then have to laboriously edit. TERRIBLE! This menu should just call up
the beautiful matrix template option in the CATALOG ('book' key + 5).
Much more useful!

032 Has anyone seen a working 'wizard' yet?? I've had the "Use
Wizard" box ticked in the CATALOG ('book' + 5 option) section of the
template application all along. Haven't seen a working wizard yet!

The following relate to the Graphs and Geometry application:

033 Need to improve the usability and speed of the Graphs and
Geometry application. It's too hard to toggle quickly between which
graphs you want displayed or hidden.

034 Need to be able to tab between functions to select them, as well
as tabbing between function zeros (like the Casio calculators). This
having to select an area for each zero, max or minimum is laughable.
Good for young students, but far too cumbersome. Occasionally the max
is off the screen and all you want is a value, NOT to have the screen
scroll so you can see the point.

035 Want to be able to start a TRACE anywhere on a function plot.

036 Tracing is too slow (but don't want to increase space between
points). Just speed it up or let us start the trace from anywhere near
the value we are seeking.

037 TheGraphs Application points out that there is a y-intercept, but
it doesn't give you a value (as when it works out other intercepts
between curve, or curves and the x-axis, ie. zeros).

038 The Conics library has a few problems also, but I won't get into
that now!

Phew, that's enough for now!............. Sorry everyone!

Any comments?


You know, with the sluggish speed of the trackpad a small mouse to
link into the USB port wouldn't be a bad idea :-)

Cheers from sunny 'DownUnder'!

[Marc Garneau]

Welcome to the group!

You've raised some good points (and to be honest, I didn't read every detail), but I did want to comment on one that caught my attention:

Regarding #022, there's a reason why it hasn't been fixed - it's not broken!
I'm curious as to why you think it's "of course totally wrong". Why would you distinguish between -t^2 and -4^2? The result of -t^2 will always be negative in real numbers. This is true whether t is positive or negative. t represents the set of real numbers (unless otherwise specified). The case of -4^2 can be seen as a substitution of 4 for t. The answer should be negative. With the way it is (and has been), the mathematics is consistent. Algebraic terms don't change the mathematical meaning of operations. No doubt it raises some confusion, but when there's confusion I'd rather have a system which is consistent than one which treats numbers and variables differently - that would ultimately lead to more confusion, in my opinion.

Marc Garneau

[Nelson Sousa]

There are valid points on what you said, but there are also quite a
few with which I totally disagree. Some remarks:


- TAB is used to move around various objects / action areas. If you
tab from the graphs entry line you get the cursor moving from one
clickable button to the next and then to the graph window itself. It
works for all "active areas" on a given application; tab also allows
us to select an object lying under another one (a circle that crosses
a line, for example, you tab to select the line or the circle. On the
L&S app tab moves from the xlist select box to the ylist select box
and to the graph window. Shift+Tab moves the other way around.

- -2^2 _is_ -4. If we mean the square of -2 we write down (-2)^2. We
use parenthesis in real life notation. If your notation would be used
then to specify the symmetric of the square of 2 one would have to
write -(2^2), which is quite an awkward way of writting. Likewise,
when we want to add a positive number and a negative number we write
down 2+(-2), not 2+-2.

- m/s is the standard unit the Nspire uses. But, not having a standard
name (we write only m*s^-1) it doesn't show on the units' selection.
However, speed is presented as "_m/_s" or (_m*_s^-1).

- temperature conversions: I want to convert 0ºC to ºF. Is 0ºC an
increment in temperature (which is the case if we're computing the
equilibrium temperature of a system), or means the conversion between
temperature values? The answer differs because the temperature scales
have different origins.

- Matrices: you can create matrices quite easily without going
throught the meny: either using the math template (there are two
templates, one for 2x2 matrices and another where the user specifies
the dimensions) or pressing Ctrl + (which is the [ ] symbol). The
latter will create a 1x1 matrix with an empty entry; you can then
press the "new line" key (next to Enter) to insert a new line and
Ctrl+new line to create a new column; I almost never use the menu to
enter matrices, it's too time consuming;

- Wizards: stat regressions use wizards. There are other functions
that use them too. Most of the functions, however, have no wizards.
Maybe we can rephrase this and say that the "use wizard" option should
be displayed only when a wizard is available

- As for any G&G object you can use the context menu (Ctrl+Menu) to
see available options to that object. Hide/Show is always one of
those;

- Multiple zeros / area selection: no algorithm can guarantee that all
zeros are found for any function. Having a calculator determine all
zeroes at once has two disadvantages in my opinion: first the user
trusts the calculator and may miss quite a few zeros that are too far
to the left or right of the screen. Given that the Casio calc argues
that the function determines "all" zeroes on may actually believe it.
But the truth is that all numeric algorithms are applied to intervals,
not to ]-infinity, +infinity[, so any numerical calculation should
also specify an interval. As for the point of interest being off the
screen, you can use Zoom Fit. I actually prefer that the calculator
doesn't adjust my screen automatically and leave that option to me.


Nelson

[DrBob]

Thanks for the comments.

Let me rephrase things. As it stands at the moment (and for many
years) the system used by calculators IS in fact inconsistent. Why do
we have a special negation key, ie. (-), on the keyboard? We expect
that when we press it instead of the normal minus operator that the
object following it should have its sign changed, and should therefore
be considered as a negative number (in the case of negation of a
positive integer, say). Yes, on a piece of paper we need to enclose,
say, (-2), by brackets as we don't have two obvious different types of
minus sign. Yet in our heads we store it as if it were a single
variable (and operate on it as such), ie. "minus two".

Let me first distinguish between the negation operator, (-) -- or
'small minus', and the subtraction operator [-] -- or 'large minus'.
You can see that these are represented differently on the calculator's
screen. If you play with both of these on most calculators you can see
that they're actually treated the same, which actually defeats the
purpose. The nSpire CAS even assumes you've made an error and converts
[-] to a small (-) if you accidently type ([-]2)^2 (just use the large
minus sign here). Try it and you see that the CAS converts it to the
small (-) negation operator.

The whole idea of a lot of these WYSIWYG calculators, those with
algebraic interpretation systems, is to reduce the need for the
operator to enter brackets.

Here's where I think the inconsistency lies at the moment.

You assign 't:=(-)2', ie. minus 2 is stored in variable 't'.
Then t^2 is 4 (as expected) and (-)t^2 is -4 (again, as expected).

If the calculator stored (-)2 in its memory as a single variable
(without a specific instruction by the operator), then (-)2 squared
would come out as +4 (as we expect if we were to square minus two 'in
our heads', so to speak). The following operations would then be
either valid, or invalid (and give an error) on the CAS systems:

a) 3+(-)2 -- Valid (with answer = 1, as this is just
interpreted as three plus negative two)
b) 3+[-]2 -- Invalid (error as can't chain two algebraic
operators together. Would expect CAS system to give syntax error as
with the following operation, which I hope you would all find invalid
--> 3x/2; try it! Three TIMES DIVIDED by 2 is not valid; just as 3
add 'take away' 2 is invalid)
c) 3(-)2 -- Invalid (as all you've written are 2 numbers,
three and negative two, without specifying an algebraic operator
between them).
d) 3[-]2 -- Valid (as it's just 3 take away positive two).
e) [-]t^2 -- Invalid (as there's nothing to the left of the
t^2 to subtract it from)
f) (-)t^2 -- Perfectly valid, as you've pointed out for all
't' values (positive or negative), so you end up with a negative
answer.
g) g^2(-)t^2 -- Invalid (as for point (c), above)
h) g^2[-]t^2 -- Valid (as in point (d), above)

Hopefully, you can see that this system is self-consistent. What we
have used up until now from the calculator companies has been
inconsistent. This system would do away with errors (on mechanical
calculating devices anyway) to do with remembering to open a bracket
before copying a negative number from a previous answer and then
needing to enter a close bracket afterwards in order for calculations
involving negative numbers to work.

You appear to favour the inconsistency of entering some numbers in
brackets and others without. I agree with that when working on paper
(as a human), but certainly not on computer or calculator.

Easiest thing for the CAS system is to store any number (temporarily)
as a string containing 'sign + number + unit' together, then parsing
any calculation or equation this is put (or pasted) into to work out
the context and perform the desired calculation.

Treat the small negation sign, (-), as if it were a totally separate
object to the large [-] sign. We would not be having this argument if
the signs didn't look so similar :-)

[DrBob]

Thanks Nelson,

In trying to counter your counter-arguments I found out a few
interesting things. I'll try to be brief as I haven't slept much in
the 40 degree Spring (yes Spring!) weather here :-)

I'll address each of your points in turn.

I know that TAB works to jump between major sections of the screen,
but I'd still rather use another combination of keys for this, so that
TAB could be used to jump around within each screen element, such as
selecting the various functions plotted on the graph window, points
that you've placed, zeros marked and text regions, etc. At present you
TAB into a major screen region and are then forced to use the slow-
moving cursor to select specific things. This, I find, is just a time
waster.

I've already answered the problem with the negation operator in a
previous posting this morning. The one things I would like to tack
onto the end of that is that I would expect that the TI could insert a
pair of brackets around any negative number I enter as (-)2, say, as
one would on paper, so that it would literally appear as "(-2)" on
screen, thereby answering other peoples' concerns about my advocating
leaving out the brackets during equation/number entry.

"kph" and "mph" are just human acronyms for km/hr and miles/hr, so I
don't see why we can't have another for metres/sec. Note that the HP
50g lists the following units under "Speed":
1) m/s 6) knot
2) cm/s 7) c
3) ft/s 8) ga
4) kph
5) mph
Bit more there than the nSpire.

One would normally consider a temperature 'conversion' (and that's the
important word here), a conversion between 0ºC to ºF, as being between
zero degrees on the Celcius scale to the corresponding value on the
Fahrenheit scale (using something like C = (5/9)(F-32). The conversion
here would be absolute. Increments in temp should be handled
separately.
I don't see a need for a separate function to be used by the nSpire.
TI just mentioned it in passing in their manual (function "tmpCnv()").
Is the other function 'Deltasymbol'tmpCnv() one that deals with
increments? I haven't even got that one to work properly yet.

I'm glad we agree about the matrix creation issue. The menu system
method is far too slow; hence, I would change its implementation.

Instead of what you call wizards I'd like to see the function name
plus a skeleton framework of parameters displayed (such that you could
type over the parameters). For instance, if I insert the temperature
conversion function, that instead of just seeing "tmpCnv()" on my
entry line (in CALC screen) I'd be presented by three placeholders to
fill in (like with the templates TI provides for matricies, say, with
empty placeholders), or something like in the function menu, ie.
"tmpCnv(Expr_unit1, _unit2). It's a pain to have to go back and check
how many parameters to enter. And yes, I do realise some functions
take multiple parameters numbers :-). You could always make the "?"
button call up the relevant help details for the function currently in
focus.

I agree that one could use the (Ctrl+Menu) Hide/Show command to hide a
certain function from your plot; however, what you forget is that once
you've hidden it there's no quick way to show it again other than to
TAB to the function entry window, use up-arrow, somehow get to the
left of the entry line, then select the "eye" icon to show the
function again. So... there we have it.... another waste of time. Far
easier to have a quick way of calling up the list of plottable
functions and quickly select/deselect functions :-)

In trying to look at a better way of determining the zeros of plotted
functions I've found a few interesting things. The problem with
missing some of the zeros is in the determination methodology. It
appears that on most of the calculators I tested on this problem
(nSpire CAS, Casio fx9860g AU, TI 89, TI 83 and HP 50g) suffer from
trying to zero in on the solutions using an engineering brute-force
approach rather than an algebraic one. They zero in on the answer
using some variant of some Newtonian stepping type approach, which
suffers when the increments used are too small. Even so, I found that
the TI's suffered the most (timewise) in having to specify the
interval over which to look. In all cases (for TI) if the interval
encompassed (accidently) 2 or more zeros then the TI would only report
the left-most. The Casio (as long as the total interval on screen
wasn't too huge) quickly found the zeros and cycled through them. I
thought I'd got away from making up high-school-like examples but I
decided to whip up a few.
Enter the following into your nSpire CAS G&G function entry screen:
f1(x) = x^4 + 0.02x^2 - 1e-4x - 2e-6
and try to find the zeros quickly. Ask your students. The good ones
will realise that they're looking for 4 roots.
Unless you zoom in very closely (indeed) the CAS system will take
ages. The Casio find all of the roots in rapid succession (say, less
than 30 sec).
It's a lot easier if you look at the way I made up the function:
ie. f1(x) = (x+0.01)(x+0.02)(x-0.01)(x-0.02)
Here the 4 roots are obvious, and it's plain to see how I've tried to
make it a problem for the CAS system by placing the roots close
together.
The next example caused problems for all machines. The Casio initially
only found 3. Again, the TI's had problems.
I designed it such that you would lose some of the zeros if you tried
to zoom in (as you might advocate when using the CAS way of needing to
specify a small range around each zero).
So, here's the function:
f2(x) = x^7 + 3x^6 + 2.9995x^5 - 0.00149996x^3 - 4.9988e-4x^2 +1.2e-7x
+ 4e-8
or, written more obviously,
f2(x) = (x+0.01)(x+0.02)(x-0.01)(x-0.02)(x+1)^3
ie. it's just f1(x) with an extra (x+1)^3 tacked on the end (to make
it interesting).
In this case it was necessary, on all systems, to zoom in somewhat to
work out the zeros; yet, the Casio still made the job easier once the
range in x-values was reduced.
To me, when I ask for zeros in a graphical window, that means find the
zeros in the x-range defined by the boundaries of my screen. From what
I can recall my SHARP models allowed for 3 methods of zero (and curve
intersection) estimation. It's a very tricky business, I agree,
Nelson. But again, I must stress that my concern is about the time it
takes for students to do this sort of thing in a exam. In the 'real
world' it's not such a big problem.

As for manual screen scrolling -- what I've found to be a real pain
with the CAS is if I want to TRACE a particular part of a function
that I may have zoomed in on. Since TRACE doesn't allow me an
arbitrary start point on the curve I often find that the screen starts
scrolling all over the place while I wait the ages it takes to get to
my area of interest -- and wo-betied I accidently press ESC during
that time (and have to start again! :-). I'd like the option to lock
the WINDOW, and to start the TRACE at any point I select.

[DrBob]

Just a quick one:

Ignoring problems with the Conics library for the moment, I'll start
at #38

After playing around with the Geometry app for a while I was quite
impressed. The one thing I would like to have is the ability to enter
an exact angle figure. Dragging the arms of the angle around doesn't
give you enough control. Much easier to type in a value. Then you
could actually solve some real-world problems rather than just play
around.

See -- I can write short ones! :-)

Cheers!

PS: Let's make it #39 Who'd like to see easier formula entry in the
spreadsheet. Once you start writing an equation its hard to select the
cells you want to operate on (to put in the equation). You either need
to remember the relevant cell addresses or waste time going through
the rigmarole of (CTRL MENU) 5: RANGE SELECTION every time (yawn!).

PPS: #040 Would like to be able to create a sequence (10, 20, 30, etc)
at any point in a data column.

PPSS: #041 Would like to be able to plot parts of the data in various
columns (NOT whole columns).

OK ---- Enough!! :-)

[Eric Findlay]

A few comments, Bob.

> Let me first distinguish between the negation operator, (-) -- or
> 'small minus', and the subtraction operator [-] -- or 'large minus'.
> You can see that these are represented differently on the calculator's
> screen. If you play with both of these on most calculators you can see
> that they're actually treated the same, which actually defeats the
> purpose. The nSpire CAS even assumes you've made an error and converts
> [-] to a small (-) if you accidently type ([-]2)^2 (just use the large
> minus sign here). Try it and you see that the CAS converts it to the
> small (-) negation operator.

The reason for the replacement and assumption of user error stems from
the fact that [-]2 is an invalid use of the [-] operator, which requires
2 operands (as you mention below), so the user must mean the (-) operator.

> Here's where I think the inconsistency lies at the moment.
>
> You assign 't:=(-)2', ie. minus 2 is stored in variable 't'.
> Then t^2 is 4 (as expected) and (-)t^2 is -4 (again, as expected).
>
> If the calculator stored (-)2 in its memory as a single variable
> (without a specific instruction by the operator), then (-)2 squared
> would come out as +4 (as we expect if we were to square minus two 'in
> our heads', so to speak). The following operations would then be
> either valid, or invalid (and give an error) on the CAS systems:

This is, as you say, expected, because t is replaced by ((-)2), thus,
t^2 is substituted with ((-)2)^2. What you suggested earlier, the
example of -2^2 = +4

> --> 3x/2; try it! Three TIMES DIVIDED by 2 is not valid; just as 3
> add 'take away' 2 is invalid)

What you have typed here is 3 ecks (the variable x) divided by 2, which
is quite a valid statement. The Nspire uses the dot representation of
the multiplication operator. If you try 3 (dot) divided by 2, you get a
syntax error.

> Hopefully, you can see that this system is self-consistent. What we
> have used up until now from the calculator companies has been
> inconsistent. This system would do away with errors (on mechanical
> calculating devices anyway) to do with remembering to open a bracket
> before copying a negative number from a previous answer and then
> needing to enter a close bracket afterwards in order for calculations
> involving negative numbers to work.
>
> You appear to favour the inconsistency of entering some numbers in
> brackets and others without. I agree with that when working on paper
> (as a human), but certainly not on computer or calculator.
>

If the system would remove the need for brackets around negative numbers
when copying a previous answer, then that would be making an assumption
about what the user means to type. You state earlier that this kind of
assuption should be wrong, so why would it be acceptable here?

Also, due to the order of operations, it makes sense to have (-)2^2=(-)4
because using the standard definition of order of operations, negation
has a similar priority to multiplication/division, which is lower than
exponents. So, when you evaluate (-)2^2, you first evaluate 2^2
(resulting in +4), then you evaluate (-)4, resulting in -4. These CAS
systems are just following proper order of operations procedures.

However, during a bit of research, I did find mention to a few systems
that give negation a higher priority, including the C programming
language (which has no exponential operator), and Excel.

> Easiest thing for the CAS system is to store any number (temporarily)
> as a string containing 'sign + number + unit' together, then parsing
> any calculation or equation this is put (or pasted) into to work out
> the context and perform the desired calculation.

Actually, in terms of programming, dealing with strings is not only
difficult, but takes a lot more processor time and memory than dealing
with numbers. To remove the sign from the string, you have to compare
the character to +/-, and that's only if it HAS a sign. Then you have
to compare the next characters to all possible digits (0-9) until they
do not match. Then you have to perform calculations to recreate the
actual number. Then you have to compare the next characters to all
possible first letters of units, second letters of units, etc. It's
much more work. (not including turning the number into the string in the
first place).

--Eric

> Treat the small negation sign, (-), as if it were a totally separate
> object to the large [-] sign. We would not be having this argument if
> the signs didn't look so similar :-)
>
> On Nov 11, 6:17 pm, Marc Garneau<pi...@telus.net> wrote:
>> Welcome to the group!
>>
>> You've raised some good points (and to be honest, I didn't read every detail), but I did want to comment on one that caught my attention:
>>
>> Regarding #022, there's a reason why it hasn't been fixed - it's not broken!
>> I'm curious as to why you think it's "of course totally wrong". Why would you distinguish between -t^2 and -4^2? The result of -t^2 will always be negative in real numbers. This is true whether t is positive or negative. t represents the set of real numbers (unless otherwise specified). The case of -4^2 can be seen as a substitution of 4 for t. The answer should be negative. With the way it is (and has been), the mathematics is consistent. Algebraic terms don't change the mathematical meaning of operations. No doubt it raises some confusion, but when there's confusion I'd rather have a system which is consistent than one which treats numbers and variables differently - that would ultimately lead to more confusion, in my opinion.
>>
>> Marc Garneau
>
> >
>
>
>
>

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 8.5.425 / Virus Database: 270.14.61/2497 - Release Date: 11/11/09 19:41:00
>

[Eric Findlay]

A few more comments:

> I agree that one could use the (Ctrl+Menu) Hide/Show command to hide a
> certain function from your plot; however, what you forget is that once
> you've hidden it there's no quick way to show it again other than to
> TAB to the function entry window, use up-arrow, somehow get to the
> left of the entry line, then select the "eye" icon to show the
> function again. So... there we have it.... another waste of time. Far
> easier to have a quick way of calling up the list of plottable
> functions and quickly select/deselect functions :-)

You can easily get to that view by typing Menu, 1, 3.

> 38: After playing around with the Geometry app for a while I was quite

> impressed. The one thing I would like to have is the ability to enter
> an exact angle figure. Dragging the arms of the angle around doesn't
> give you enough control. Much easier to type in a value. Then you
> could actually solve some real-world problems rather than just play
> around.

I agree. I'd like to be able to determine the angle of the lines by
typing it in.

--Eric

[Nelson Sousa]

I'll try to answer shortly to all the points you've raised, but I
won't enter into a discussion about which design convention is better.
When it's a matter of fact that a given feature is missing or that one
way is canonical whereas Nspire's way isn't, I'll try to comment on
that.

On the Nspire the (-) sign can be replaced by the - sign and it'll be
interpreted correctly as a sign and not an operation. Even if you
write 2+-1, using the "big minus" instead of the "small one" the
Nspire recognizes it as 2+(-1). The two minus signs were kept,
presumably, for compatibility purposes, being that users are used to
having a key for subtraction and a different key to enter negative
numbers. But yes, the (-) is (almost) redundant. About the writting,
given that the sign is different, parenthesis are not necessary and
are left out (probably) for simplicity. Either way, it's a design
choice. Whether I like it or not, it's not wrong. Questionable, yes,
but not wrong per se.

As for units: all possible speed units are included on the Nspire; you
can convert 120 kph > m/s and it'll display the unit adequately. It's
simply not listed on the unit list, beucase it's a quotient between
two units, so it's not a different item. It's only a matter of naming
convention, not a limitation.

Temperature: as it's ambiguous, it's always better to force the user
to explicitly choose an operation. tmpcnv converts between temperature
values and deltatmpcnv converts temperature intervals (so,
deltatmpcnv(1ºC,ºk) returns 1ºK whereas tmpcnv(0ºC,ºk) returns 273.15
ºK). Again, it's not a limitation but a matter of design choice. Who
developed unit conversion (back on the early 90s for the TI-92)
prefered to force the user to make the conversion explicit. Either way
is valid as long as it's documented, both have their problems: with
tmpcnv you'll have a lot of users complaining that temperature
conversions don't work, if you have a standard conversion to assume
it's a value conversion you'd have users complaining that when
calculating equilibrium temperatures in a physics problem the
calculator returned a wrong answer).

Wizards / list of parameters: the parameters used by (almost) every
function are listed at the bottom of the catalog window; there's a
small "expand" button that you can click to open a description of the
function usage (and the argument names usually let you know what types
are accepted; So, for example, when det is selected you have at the
bottom of the catalog window "det(squareMatrix[,tol])". Whenever I
have doubts about any function I use the catalog and so far it hasn't
let me down.

Help: Ctrl+? already provides some level of help. Not very detailed,
sure, but some useful tips for the current application. One of the
tips refers to entering matrices. Perhaps you could read the various
documentation pages on the examples folder and the tips under Ctrl+?,
I'm sure you'll find a lot of useful information there.

Hide/Show: Press Menu, 1:Actions and 3:Hide/Show. It allows
hiding/showing various objects at the same time and displays hidden
objects in gray.


Zeros of functions: All graphing calculators use the same family of
algorithms: Newton's method with some improvments for performance. The
reason the Casios determine "all" zeros is because they start with the
interval [1e-99,1e+99] and use the first zero as the first endpoint of
the second iteration of the zeros algorithm. Flaws: f(x)=sin(1/x).
There are an infinite number of zeros as x approaches 0. Neither
Casio, nor anything else can determine "all" zeros. But their zero
functions claims that it can. Newton's method can determine one root
or any finite number of roots, it cannot, ever, determine all roots if
there are infinite roots.

Trace: it has changed the behaviour over time according to user's
requests. It's not perfect yet, maybe it'll be changed on the next
version? I think currently (OS 1.7) it starts on the middle of the
screen.

Adjusting angles on a discret set of possible values: there are
workarounds. You can either use scatter plots (see ppt presentation
available at http://nelsonsousa.pt/index.php?lang=en&cat=2&subcat=9&article=57)
or using a trick that involves degenerate polygons (see tutorial at
http://nelsonsousa.pt/index.php?lang=en&cat=2&subcat=5&article=42)

#40: see Generate sequence under the L&S menu when you're on a
columns' definition. You can also use the seq command (see the ppt
presentation given above).

#41: see Left, Right and Mid functions.


As you can see, quite a few of your desired features are already
implemented. Maybe you can first explore an app and ask how to do
something specific before claiming it's missing?

Cheers,
Nelson

[Ray Fox]

Why is it you can not graph f1(x)=sin(theta)??

 

Sorry, when I copied the greek symbol it came up incorrect.

[Nelson Sousa]

because sin(theta) is a function of variable theta, not x.

If theta is defined, then that's simply a constant. If it's not
defined you get an error. You must use variable x for graphs of
functions, t for parametric graphs, n for sequence plots and theta for
polar plots.

Nelson

[Ray Fox]

Thanks,

 

It would help if the textbooks and Nspire would use the same conventions.

 

Ray

[DrBob]

Interesting that certain individuals choose to interpret things
written to suit themselves.

Thanks for pointing out to me that the CTRL + MENU is used to hide a
function, yet only the MENU + 1 + 3 is useful to get it back. Another
great leap of logic for TI! I am glad, however, that the grayed
(hidden) functions are shown when this is chosen, which is somewhat
better. Thanks for the advice.

Eric, when I wrote the following,

> --> 3x/2; try it! Three TIMES DIVIDED by 2 is not valid; just as 3
> add 'take away' 2 is invalid)

you can see that I meant 3 times divided by 2. I even stated it in the
following sentence. So no need for the comment you made below:

>>What you have typed here is 3 ecks (the variable x) divided by 2, which
is quite a valid statement. The Nspire uses the dot representation of
the multiplication operator. If you try 3 (dot) divided by 2, you get
a
syntax error.

I admit the use of the 'x' was somewhat misleading. Next time I'll use
'*' as the dot '.' would also confuse some people :-) My point, as you
so rightly point out is that there should be a syntax error. TI have
chosen (which I think is good in this case) to interpret 3(-)2 as 3
times 'minus 2' (which is fine). But if I wrote 3(-)2^2 I'd rather it
be interpreted as 3 times 'minus 2' squared. Would be nice to see "3.
(-2)^2 = 12" on the display in this case! :-)

These electronic devices should be programmed as 'expert systems'.
Thus, to get around any confusion, let users enter bracket, -2,
bracket if they want (as for the last decade), or recognise '-2' as
the value of an internal variable and display it specifically as
"(-2)" (with brackets) when it's cut and pasted into an equation on
screen. Simple! And it allows everyone to get what they want. I won't
go on about it further as it's to do with the programming languages
used and the fact that people have gotten so used to the problem that
they choose to ignore simple solutions.

> As for units: all possible speed units are included on the Nspire; you
> can convert 120 kph > m/s and it'll display the unit adequately. It's
> simply not listed on the unit list, beucase it's a quotient between
> two units, so it's not a different item. It's only a matter of naming
> convention, not a limitation.

Derived units such as 'kph' should also then be worked out as
_km*_hr^-1. If you consider that kph and mph are valid then metres-per-
sec should be considered equally so. Better to write as km/h, mi/h and
m/s to distinguish the lot :-) (noting, of course, that m/s is the
only valid SI derived unit here).

This brings up a further problem I noticed yesterday. In trying to
work out the amount of time it would take for the equivalent of all
the water in the Antarctic to flow down a river (with dimensions 1000m
by 20m at a flow rate of 5 kph) I got partway through when I realised
I wanted to convert the flow rate to m/s. Trying to use the conversion
operator (which I shall note as -> ) in mid-flight resulted in an
error. I could have started again, or used another calculator window,
but thought it would be OK to do it this way. I tried it on the HP
50g, which uses a conversion function (which is long-winded in its own
right), but it worked:

HP 50g version:

(1000_m) *( 20_m) * convert(5_kph,1_m/s) = 27777.7777778_m^3/s

Which is right for the volume flowing per second. The problem here --
like with the nSpire -- was that to do any useful calculations by
'chaining' unit calculation together I had to put brackets around
anything that had units, which is a real pain.

nSire CAS version:

(1000_m)(20_m)(5_kph -> _m*_s^-1) = ERROR

Gives a 'Missing bracket' error. Pity! Seems to want to interpret
everything to the left of the conversion operator (everything!) as
what you want converted. Here there needs to be a change in the
precedence of this operator and the brackets, at least.
Funnily enough, if I enter the following:

(1000_m)(20_m)(5_kph) = 27777.7777778_m^3/s

it gives the perfect answer (with the 'kmh' already converted). This
is good, but certainly a feature which helps kids cheat. They get out
of having to learn how to do the conversion from km/h to m/s (not
good!). Watch out for this feature! :-)

> Temperature: as it's ambiguous, it's always better to force the user
> to explicitly choose an operation. tmpcnv converts between temperature
> values and deltatmpcnv converts temperature intervals (so,
> deltatmpcnv(1ºC,ºk) returns 1ºK whereas tmpcnv(0ºC,ºk) returns 273.15
> ºK). Again, it's not a limitation but a matter of design choice. Who
> developed unit conversion (back on the early 90s for the TI-92)
> prefered to force the user to make the conversion explicit. Either way
> is valid as long as it's documented, both have their problems: with
> tmpcnv you'll have a lot of users complaining that temperature
> conversions don't work, if you have a standard conversion to assume
> it's a value conversion you'd have users complaining that when
> calculating equilibrium temperatures in a physics problem the
> calculator returned a wrong answer).

Nelson, as for the deltatmpcnv() function. Surely you would want your
students to do the conversion by hand. I would place the standard
tmpCnv() functionality under the unit conversion menu and have a
separate deltatmpcnv() function for specialist applications such as
you mention. This would make the whole conversion functionality of the
nSpire more consistent. If I didn't have a manual to look at I would
have been left scratching my head. Thanks for pointing out the use of
the deltatmpCnv() function, by the way, as it's not covered by the
reference manual (unless I missed that too :-). Looks to have been
added in haste.

> Wizards / list of parameters: the parameters used by (almost) every
> function are listed at the bottom of the catalog window; there's a
> small "expand" button that you can click to open a description of the
> function usage (and the argument names usually let you know what types
> are accepted; So, for example, when det is selected you have at the
> bottom of the catalog window "det(squareMatrix[,tol])". Whenever I
> have doubts about any function I use the catalog and so far it hasn't
> let me down.
>
> Help: Ctrl+? already provides some level of help. Not very detailed,
> sure, but some useful tips for the current application. One of the
> tips refers to entering matrices. Perhaps you could read the various
> documentation pages on the examples folder and the tips under Ctrl+?,
> I'm sure you'll find a lot of useful information there.

Nelson, any idiot -- including this idiot -- knows about the CTRL - ?
help listing. I've read them (yes I have). When someone suggests a
better way, one that doesn't require flipping backwards and forwards
between the app and the CATALOG listing, it might be an idea to
consider it, instead of making derogatory comments (even though I'm
probably making one now :-) What I was suggesting was context
sensitive help specific to the present task. Say I remember that there
is a temp conversion function, but I forgot the parameters required. I
type "tmpCnv(" on the keyboard. I then hit '?' and the context
sensitive help brings up the function description listing. Much faster
than going to the catalog, scrolling through it, etc. That's all I was
suggesting. Or else, get the machine to bring up a 'skeleton' listing
of the function on the entry line for you to fill in. I am suggesting
this after experience on many machines where many prgrammers have
chosen to write excellent functions, but forgotten to include help
listings or parameter details. And this will happen, I believe, with
people writing the add-on library functions. Anyway, it was just a
suggestion to save time.

As for the zeros of a function. Let's get off the nSpire vs. Casio
debate. Did you even try to input the functions I made up? And if so,
how long did it take you or your students to work out the roots. The
point being is that it takes too long using the present graphical
approach and that it's way too prone to errors in the case of clusters
of roots that are close to each other, or multiple clusters of roots
which may themselves be separated from each other by some distance in
x, say. Just try it out (please :-).

I'll have a look at those powerpoint presentation soon. Thanks!

Hey Nelson. Quick question. You don't work for TI, do you? :-)

Cheers!

[Nelson Sousa]

See attached file for some unit conversions regarding that example and
how you should have done it in the first place (I hate to make
derrogatory comments as well, but how to use unit conversion is in the
Reference guide, including temperature conversions - see pages 92, 93
and 116 on my version of the reference guide, which is probably
outdated, so pages may have shifted). Nspire CAS converts what you
wanted, quite well, actually, albeit using different syntax rules than
those of the HP.

As for the rest of the points you make: I've stated my opinion, you've
stated yours, I have no particular interest in continuing the
reply-over-reply this thread will become.

But your last question deserves an answer: no, I don't work for TI. I
give professional development courses for teachers on how to best use
technology in teaching (including TI's) and I work as a consultant in
educational technology.

Nelson

[pam]

About finding zeros:

If you use the intersection of the graph and the x-axis, then all zero
points will be identified. After that, just show the coordinates of
each point. I find this to be a much simpler procedure than tracing
the graph. Plus, it drives home the idea that zeros of the function
are the x-intercepts of the graph.

Pam

[John Hanna]

'all zeros'? some, perhaps. When you use the intersect tool with a curve and
the x-axis. It shows the zeros in the current viewing window. When you
change the window settings intersection points tend to disappear.

Sail Upwind,
John Hanna
T3 . Teachers Teaching with Technology
973.398.3815
jeh...@optonline.net
www.johnhanna.us
"the contact between the acute angle of a triangle and a circle has no less
effect than that of God's finger touching Adam's in Michelangelo." - Wassily
Kandinsky

[Nelson Sousa]

After getting a little carried away building a tutorial (should be
published sometime today) I've discovered, the hard way, that the
number of pages in a problem is limite to 50 (ok, anyone in his right
mind will never create over 50 pages in a single problem, but
apparently I'm not in my right mind).

Because of thatI had to move a bunch of pages from one problem to
another, which can only be done one by one.

So,

#n+1 (replace n with the last number on this list): Allow for multiple
page select, copy, paste and move.

#n+2: I was playing with angle measurements both in degrees and in
radians and it's somewhat annoying having to go all the way to the
Document settings dialog to change the display units. I'd rather have
angle units as an attribute of angle measurements, the same way we can
define the number of decimal places on the attributes of a number. We
can keep the default settings defined on the Document / System
settings dialog boxes, but each angle measurement could be switched
from Deg to Rad on the fly.


Nelson

[DrBob]

Correction to what I wrote above.

deltmpCnv() is written up in the manual. Sorry about that.

Still believe tmpCnv() should be integrated into unit conversion app.
though.

MENU 1 3 (with grayed hidden graphs able to be picked so that they can
be seen again) doesn't work if par of your 'hidden' graph doesn't fall
in the x-y range shown on the screen. Pity!

Nelson, you're good with these things: -- I want to declare a global
variable that I can access in any of my documents. Any suggestions?

Cheers!

PS: I'd like to hear some of your suggestions as to what could be
improved on the nSpire CAS, seeing as you've had a lot more time to
play with these units.. What do you reckon?

Thanks again :-)

[Nelson Sousa]

You can't define a global variable. But you can have a document on
MyLib folder and access any function and / or program from it.

So you can work around it: suppose your global variable is x=2. Create
a new document and call it "global"; place it on MyLib. On that
document define a function

x( ):=return 2

Set this function as LibPub (so that it can be accessed from the catalog.

Update the libraries and from any document type in "global\x( )".
it'll return 2. I cant' think of any other way

As for CAS: probably the thing I missed most while in college (I had a
TI-92 Plus and then a Voyage 200) was the ability to solve equations
where variables were matrices and functions. So just like we can
define a variable as a matrix, the same way we cna add _ to the end of
a variable name to specify it as a complex number. The goal would be
solving exercises like given matrix A and diagonal matrix B, which
matrix S is the solution of A=S^-1*B*S? As for funcitons as variables,
the main application would be solving equations involving differential
operators (curl, div, grad).

Cheers,
Nelson

[MattG]

Great idea!

006 - Area approximations built into the OS... I think the
functionality of doing rectangle approximations and trapezoid
approximations is sorely missed and would help thousands of students.
The programs out there are great but why not centralize?

Matt Gironda
Science Park HS, Newark, NJ

[DrBob]

Sorry Nelson,
I've been preoccupied somewhat at this end.

Thanks for the advice. Using the library function approach worked
well. The only problem now is that it started me playing around with
writing programs on the unit, only to be shocked at finding that the
Input/Output options are limited to output only (ie. "Disp"), unless
you want to edit parameters during the initiation of a program. Weird.
No 'Input' or 'Prompt' type options, as well as a lack of graphics
options. These must be on the drawing boards for the next OS version
(surely). Hmm.....

> As for CAS: probably the thing I missed most while in college (I had a
> TI-92 Plus and then a Voyage 200) was the ability to solve equations
> where variables were matrices and functions. So just like we can
> define a variable as a matrix, the same way we cna add _ to the end of
> a variable name to specify it as a complex number. The goal would be
> solving exercises like given matrix A and diagonal matrix B, which
> matrix S is the solution of A=S^-1*B*S? As for funcitons as variables,
> the main application would be solving equations involving differential
> operators (curl, div, grad).

Good idea. Listen, do you know the best place to write to (in the case
of TI) so that we can get them to look at some of these suggestions
such that they might actually have a chance at being included in the
next release?

[DrBob]

Sorry Nelson,
I've been preoccupied somewhat at this end.

Thanks for the advice. Using the library function approach worked
well. The only problem now is that it started me playing around with
writing programs on the unit, only to be shocked at finding that the
Input/Output options are limited to output only (ie. "Disp"), unless
you want to edit parameters during the initiation of a program. Weird.
No 'Input' or 'Prompt' type options, as well as a lack of graphics
options. These must be on the drawing boards for the next OS version
(surely). Hmm.....

> As for CAS: probably the thing I missed most while in college (I had a
> TI-92 Plus and then a Voyage 200) was the ability to solve equations
> where variables were matrices and functions. So just like we can
> define a variable as a matrix, the same way we cna add _ to the end of
> a variable name to specify it as a complex number. The goal would be
> solving exercises like given matrix A and diagonal matrix B, which
> matrix S is the solution of A=S^-1*B*S? As for funcitons as variables,
> the main application would be solving equations involving differential
> operators (curl, div, grad).

[Nelson Sousa]

To overcome the limitations of programming here's what I usually do:
1. Build a geometry construction with the desired properties and some
variable storing;
2. On a spreadsheet invoke functions that will return the results;
store them in variables as well.
3. Back to G&G that new set or variables will be used to build the
"output" objects.

To get an idea of the power one can use, take a look at the Organizer
or the Nspir3D programs from my site (www.nelsonsousa.pt). Although
there are no output functions I managed to get some pretty neat stuff
working.

As for sending suggestions to TI: you can make your "demands" here,
there are TI people out there reading this (right guys? ;) ). From
what I've seen in recent OS updates, the improvements come from user
feedback (although the order in which they're implemented probably
depends on a lot of factors)

Cheers,
Nelson

[Rex Boggs]

From: "Nelson Sousa" <nso...@gmail.com>

> As for sending suggestions to TI: you can make your "demands" here,
> there are TI people out there reading this (right guys? ;) ). From
> what I've seen in recent OS updates, the improvements come from user
> feedback (although the order in which they're implemented probably
> depends on a lot of factors)

As the manager/moderator of the tinspire discussion list, I can assure you that I have approved many
list members who have .ti in their email address. So I think all suggestions made via this list
about changes to the tinspire firmware are being heard by the right people.

Having said that, I hope such emails remain a minor part of the list activity. Most of us here are
trying to find better ways to use the tinspire to improve learning in our classrooms. Which was
why the list was created in the first place.

Cheers

Rex Boggs
list moderator

[DrBob]

Thanks everybody for getting back to me on this. I suspected as much.

[DrBob]

I'm just going to add one thing to my previous posts about CAS systems
such as the nSpire handling of negative numbers and units.

Those who stated that these systems were working correctly and that
storing a number as (sign) + (numeric) + (unit) would be too hard on
the processors when they had to parse an equation should note that the
HP 50g (which you might consider as inferior to the TI) does this
quite easily, especially in RPN (Reverse Polish Notation) mode. Store
the whole lot as one object on the stack. No problems there.

As for the disagreement about the handling of the negating operator
and its precedence in operations, note again that RPN calculators,
which were of course developed in the 1960's (before CAS systems, and
by none other than HP, I believe) have handled this correctly for
decades, since their inception. I should also credit HP with the stuff-
up in the precedence under the CAS system, as they were the first to
program these as well. If you look at Maple, Mathematica, Sage, etc
packages, not how they handle negatives.

When you enter a negative No. and try to operate on it the TI, say,
should perform the old twos complement, tack on any relevant unit,
store it as a single object on the stack, display it on the screen
surrounded by brackets, then all will be well :-)

In RPN (on HP 50g):
===============
Note: Specific keys are highlighted with brackets.

eg. Square 'minus 2 km'
------------------------------------
2_km (+/-)
(ENTER)
(x^2)
(ENTER)

Answer = +4_km^2 (correct)

eg. Store 'minus 2 seconds' in variable 'T', then work out 'minus (T
squared)'
----------------------------------------------------------------------------------------------------------------
2_s (+/-)
(ENTER)
T
(STO->)
Now tried storing this as a string (note single quotes), as it
works as well: ---> (') (+/-) T (x^2) (')
Screen displays '-T^2' (as expected)
(--> NUM) ie. please interpret this and display in numeric form.

Answer (after parsing the string above) = -4_s^2 (as expected)


So, please don't say that these machines can't do the right thing
guys :-) And they have plenty of processor power. This method of
handling numbers as well as units works both with explicitly entered
negative numbers as well as with algebraic representation where
pronumerals (or variables) represent the values.

Last thing about this is that I don't hold the HP up as the be-all and
end-all of calculators. It does have (as with the TI, Casio, SHARP,
etc) problems when in algebraic interpretation (CAS) mode, both with
negatives, and (at times) with variables; and this is a problem of its
own making as they were the first to bring out the CAS systems in the
70's or 80's.

So, the answer and techniques have been around since the 60's. It's
not rocket science (which is stuff I deal with on most days :-) All of
the computer-based packages do it well. So now it's time for handheld
calculators to catch up. Who will be the first company to do it
correctly? (hint, hint) :-) ANd.... you can still enter brackets if
that's what you really really really want, for those who don't like
change! :-)

Phew.... enough already you (and I) say!

Cheers

Please don't be angry that I've raised this again! Let's here some
other suggestions from people (just to shut me up, at least :-)!

Have a good weekend!

[DrBob]

Gotta love this crazy english language. Drop one letter and it changes
everything.

I wrote:
>If you look at Maple, Mathematica, Sage, etc
packages, not how they handle negatives.

Of course, read this as:
"NOTE how they handle negatives."

Cheers.

[Joe]

I guess I should get my two cents in before this topic stalls out.

First let me say that a similiar set of posts was generated by the TI
nspire discussion group that TI killed, and that list was forwarded to
TI during the nspire development phase. As I recall, exactly none of
the suggestions were incorporated into the design. So I have no hope
that this latest list will have any impact on TI's thinking and
development spending but here goes anyway.

My interest is in teaching undergraduate college mathematics and using
the TI cas. With out the ability to plot several surfaces, rotate,
and perhaps zoom and move them, the TI cas is simply incomplete. It's
like buying a car and finding out that you have to buy and install
after market steering gear for it. In any case, because of the lack
of 3d graphing in an age where it is readily available through many
other options, I cannot and do not recommend the TI handheld. It is
simply to limited, and there are simply better options, such as having
the TI cas windows software in an inexpensive netbook computer along
with at least one other math program that does have good 3d graphing
capability. Having said that, I recognize that most high school
teachers never acquaint their students with surfaces so it is not
likely that TI will fix this glaring deficiency unless they wish to
continue enjoying the college student market. I say this because with
the small inexpensive and very portable netbook computers, along with
the low cost math software that is now available, software that does
3D graphing among other things, I find a definite trend away from
handhelds and toward computer math software. So, if TI were to invest
in modernizing their cas, adding a decent 3d graphing capability along
with fixing some other competitive deficiencies such as the lack of
div, grad, and curl functions along with laplace and inverse laplace
capability should be addressed. Of course it is also necessary to fix
the lack of input and output progamming statements.

[elkar]

Only to adhere the input Command

http://www.univers-ti-nspire.fr/forum/viewtopic.php?f=10&t=288

[Joe]

Mr. Elkar,
Please be so kind as to translate the statement in French at the link
you give, or elaberate on what you said. I don't understand what you
are trying to say by "Only to adhere the input Command."
Thank you.

[Sean Bird]

Respectfully, I believe he did try to translate the French.

I'm fairly certain he was requesting the ability to have an input command after a program has started.

There are certain places in the world were this must be important to a math curriculum. One of the things I've been delighted to say to fellow teachers is that students will not be playing 'video' games. Although this absences has been to the chagrin of some students. 
I'm sure I'll be amazed at what Nelson, Steve, and others from around the world will come up with if there was input, but I'm quite pleased with the 'programming' methods we are currently using that do not necessitate much of the Program Editor (which has some capabilities that are superior to any previous device). I appreciate the efforts of those who are considering how to make advanced authoring techniques more time efficient and less labor demanding. 

I feel bad for someone trying to use a translator engine to interpret/understand what I just wrote.

To showcase a bit of what can be done to promote inquiry and understanding through pattern recognition with the TI-Nspire I'll make another post with a different subject. (This thread is getting quite long and I'd like to honor Rex's polite request to discuss how the Nspire can be currently used in the classroom.)

- Sean

[Eric Findlay]

Wow. I don't understand all of those posts, but from what I do
understand, it's pretty much just a summary of the argument we've had in
this group.

The first person annoyed because they want an input function, and the
only way of passing values is as a parameter to a function. They then
go on to say that they think the programming capability on the Nspire is
very limited, not adaptable enough for the course their teaching (I
don't know what that course is) and not as good as the previous version,
considering how long it's been since the last new calc.

The responder basically states that the NSpire is a complete revamp of
the calculator system, not building off the old calculators, and that
they're still making modifications. Then they point out all the great
new stuff on the calculator that proves that it is better than the older
models, they mention libraries and how most people (including students)
use them to work around these "shortcomings."

They end of by saying that there are two ways to work around the "no
input" problem. I believe they suggest recursive functions (probably
related to the course) and using a list as one of the arguments.

Like I said, it's basically a summary of the discussion we've already
had. So, people on both sides, be assured that you're not the only one
who thinks like you.

--
For I am convinced that neither death nor life, neither angels nor
demons, neither the present nor the future, nor any powers, neither
height nor depth, nor anything else in all creation, will be able to
separate us from the love of God that is in Christ Jesus our Lord.
- Romans 8:38-39 (NIV)
--
Eric Findlay
AKA Eagle-Man

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 9.0.709 / Virus Database: 270.14.76/2518 - Release Date: 11/21/09 11:41:00
>

[-TJ]

Dear Eric,

The fact that there are posts suggests that there is productivity,
which is always beneficial. Let everyone dump their thoughts and views
into this discussion and lets see what we can get out of it. After
all, that is the main purpose of this Google Group. Can you quit
arguing, there is nothing we can do to stop others from sharing
stating their opinions.

Sincerely,
-TJ

[DrBob]

Hi Gang,

Good to see I'm not the only one stirring things up!

Nobbling the nSpire so as to stop students from playing games in class
is no excuse. Being able to interact with a running program on a
'personal' level through keyboard input, say, is a very powerful
learning and motivational tool. Anyway... I'll assume direct input and
prompts will be added later. 3D-graphing would, of course, be handy.

Now, just to take the educator hat off and don the analyst's one. Just
a couple of issues to help with usability again. There appears to be a
problem with scaling of the special characters. Hopefully it's already
been commented upon.

Not sure which number to make this:

#3?

There is so much screen 'real estate' (ie. masses of pixels) to play
with yet readability suffers due to the size of some key characters. I
shall just comment on the 'negate', ie. or small negative symbol (-),
and the multiplication symbol, both of which are undersized and just
make things difficult to read (with all the decimal points, dots, etc
thrown around the screen).

The negative symbol suffers from a scaling error, I believe. It should
be made somewhat bigger to aid readability (at least half the size of
the 'minus' symbol). In fact, you can actually see that the nSpire
does this, but not at the base level of operations. Try the following
and you will see that when raising something to the power 3 or 4 times
the 'negate' symbol actually gets bigger (and more readable) while the
numbers reduce in font size. Interesting!

Try this (or a variant):

-2^(-2)^(-2)^(-2)^(-2)

ie. Minus 2 raised to the power of minus 2 a few times.

After about 3 power raises you can see that the negate symbol
increases to a more readable size.

Should be an easy programming fix.

Now, TI correctly (I feel) used the asterisk (*) symbol for
multiplication. Try it by selecting it from 'chars' menu. The trouble
is that the symbol comes out so small on the screen that it just
appears as a dot, which is a bit messy as there are enough decimal
points, etc, floating around. The symbol hardly scales at all as you
increase the font size.

Lets fix these scaling issues so that the display becomes more
readable.

Hopefully this isn't too controversial! :-)

Cheers!

[Nelson Sousa]

There are mitigating factors to the input/output problems, as well as
3D graphing. Not a definitive solution (and one may argue that it
should be part of the OS and not extra documents, but still, they're
mitigating factors):

- 3D plotting: check out
http://www.univers-ti-nspire.fr/resources.php; this file allows
plotting up to 2 functions z(x,y) with rotation.
- 3D geometry: limited functionality due to performance issues, but
still allows up to 3 planes, 6 lines and 8 points, draws perpendicular
and parallel objects. Downloadable from
http://nelsonsousa.pt/index.php?lang=en&cat=2&subcat=3&article=52
- Input/Output: as long as your program can be done by functions (and
ALL programs that return numbers can be done with functions), you may
use sliders and/or geometric objects to serve as your input and have
the results from those functions to manipulate/generate geometric
objects in real time. You can find some examples of such functionality
on the section TI-Nspire/Programs of my website and some techniques
explaind on TI-Nspire/Tips and Tricks. URL: www.nelsonsousa.pt


Cheers,
Nelson

[Eric Findlay]

TJ, I was not arguing, but merely giving a rough translation of the
link provided. Furthermore, I was noting in a positive way that others
had the same thoughts for both side of the argument, to show that both
sides of the issue are not limited to this group.

I'm not sure where you got the opinion that I was arguing further.
Sorry for the confusion.

--
For I am convinced that neither death nor life, neither angels nor
demons, neither the present nor the future, nor any powers, neither
height nor depth, nor anything else in all creation, will be able to
separate us from the love of God that is in Christ Jesus our Lord.
- Romans 8:38-39 (NIV)
--
Eric Findlay
AKA Eagle-Man

> No virus found in this incoming message.
> Checked by AVG - www.avg.com

> Version: 9.0.709 / Virus Database: 270.14.77/2520 - Release Date: 11/22/09 11:40:00
>

[-TJ]

Dear Eric,

Ah, well on that note, you make an interesting observation, yet I am
guessing they repeat their views in a single coherent discussion
rather than a scatter of multiple discussions. Maybe, it is too, the
name of the discussion that invites others to state their opinions.

Sincerely,
-TJ

[DrBob]

Just a quick one I should have mentioned earlier. I'll start the
numbering system off again:

042 Would be good to have an AUTO SAVE function. There have been
quite a few times when I've been on the road, done some lengthy work,
only to find that I've forgotten to hit save at some critical point
only to find that the CAS hangs on some calculation or program and
reboots itself (Microsoft would be proud! The 'grey' screen of
death! :-). So, it would be nice to have the option to have your work
saved every 10 minutes or so.

Occasionally the machine has just 'hung' while running a faulty
program. I've tried the usual way to break out of this, but it
occasionally doesn't work. Is there another key combination to kill
the current process without rebooting when this happens?

Cheers!

PS: Nelson, I've looked through quite a lot of your stuff. Thanks -
learned quite a bit. Keep up the good work :-)

[Nelson Sousa]

I find it somewhat odd that you encounter so many crashes (not that an
auto-save feature wouldn't be welcome, I'd love that too). Make sure
you have the latest OS, some older versions were more prone to
crashes...

As for interrupting a calculation: pressing ON for 1 second or two
should be enough.

There's also the maintenance menu but it only allows for reboot,
delete OS and complete format, so it doesn't help saving the work in
progress..

Cheers,
Nelson

PS: glad you liked it.

[DrBob]

Hmmm.... maybe I try to push the technology a little too hard at
times :-)

Am running 1.7.2741 Would be happy to get an update (even beta)
before the scheduled April 2010 one.

Actually, it would probably be a good idea for a few of us to be
allowed to take beta versions 'for a run' before the next release.
Would give us something to do over the holiday season! :-)

Cheers!

[Nelson Sousa]

I think a few of us are. It would be very surprising if there were no
beta testers...

Nelson

PS: I'm not one of those few.

[DrBob]

OK, next one!

043 Please, please give us an easier way to pick ranges while
building equations in the spreadsheet app. This needing to call up a
menu option every time you want to add a cell reference (range) is way
too slow. Take a leaf from all the other spreadsheet apps out there
(even on handhelds.

Next small problem to fix:

Noticed this fault (again somehow linked to the negate (-) operator
within the spreadsheet while trying to write a simple formula
involving exponentials to the negative power of the value stored in
another cell):

044 Need to fix small inconsistency in menu system when creating a
spreadsheet formula involving the negate operator, (-). Found this
one while trying to use the exponential function (ie. e^x) in an
equation. Try this:

- Enter a value in spreadsheet cell a(1).
- Try to calculate e ^ 'minus a(1)' value, but using the range
selection option mentioned above.
- Ideally I would hope to use the following keystrokes: <=> <e^x>
(-) <CTRL> <MENU> <5> < left arrow > <ENTER>, to get "=e^-a(1)".

What you find, instead, is that the RANGE SELECTION option on the menu
is not selectable.

If you try to put <e^x>, but with a positive power, the RANGE
SELECTION option stays selectable.

Note that <MENU> <1: ACTIONS> <3: SELECT> <3: SELECT RANGE> still
stays selectable in either case, even though it comes up with the
following error:
"Cannot do a range selection in a formula not beginning with an
equals", if you try to use (-) before the range select.

The only way around again is to use brackets just prior to selecting
the range. So <=> <e^x> <(-)> <(> <CTRL> <MENU> <5> < left arrow > <)
> <ENTER> works; but what a hassle.

The CAS system should interpret any Range Selected as an object, thus
freeing us from this continual bracketing nonsense whenever we need to
use, or calculate, the negative of some quantity. Let's streamline and
minimise the keystrokes required (with all operations and in all
apps).

The simplest example is just trying to find the negative of the value
stored in cell a(1).
Only way to do it is <=> <(-)> <(> <CTRL> <MENU> <5> a(1) <)> <ENTER>,
with brackets.

Much too long-winded, and gets worse the more complex your equation
becomes.

Now that I know the problem I can work around it; but, please, please
there needs to be a quicker way to select ranges so that we can use
the spreadsheet 'on the fly', for quick calculations.

Cheers again!

[John Hanna]

[JEH>] See below...

043 Please, please give us an easier way to pick ranges while
building equations in the spreadsheet app. This needing to call up a
menu option every time you want to add a cell reference (range) is way
too slow. Take a leaf from all the other spreadsheet apps out there
(even on handhelds.

[JEH>] To avoid the 'Cell Reference/Variable Reference' dialog box, don't
name variables with column names (like C or X) or with cell names (like D1
or X1) or you can precede the variable name with a ' (single quote) such as
'X1. Did you know that when you are building a cell formula that you can
address other cells in the software (and TE emulator) by mouse a
point-n-click? Not sure if you can select a range this way. Perhaps the
handheld will have a mouse cursor in every app ...someday...

[Sean Bird]

Let's say for example that you want to find the average from cell b4 to b25, just type

=mean(b4:b25)

You don't need to use a range selection or select variable - Although thanks for pointing out that that exists.

If you want to do e^(-a1), just type that in.
If you want to recall a variable, you could use the var button, but for cells it is just as easy to use b5, etc.

I like what John just posted.
Have a great day.
Sean Bird

[Sean Bird]

Oh, and have you tried CTRL Click on L&S it is a quick way to fill down or copy/paste a selection. See the Tips & Tricks of http://education.ti.com/sites/US/downloads/pdf/TI-NspiringTimes3.pdf for more info about that.

[DrBob]

Thanks guys.

John, I know what you're saying, but I'm not talking about named
variables. I'm actually just talking about cell references.

Remembering all of the cell references, b5, c17, etc, is fine for very
small spreadsheets and mini-equations. I am talking about using large
sections of the spreadsheet, utilising individual cells (not just
whole columns). Once you get to that level you don't want to be
planning your equations on paper so that you don't stuff up cell
references. You want to actually put your cursor (quickly) over the
cell/cell-range you want and enter it into your equation.

The problem is that the data/formula entry line needs to utilise the
up-down-left-right movements due to its method of equation entry. If I
want to skip around the spreadsheet while editing some formula I'd
need to be able to hit TAB or something to free up the cursor for
range selection, then hit ENTER (for example) again to take me back to
my point in the formula entry window. That might work. An external
mouse interface to the USB port on the nSpire might add extra
functionality too. Or even free access to the mouse cursor while the
text cursor stays in the formula window.

One can only wish! :-)

Oh yeh, and Sean, thanks for the comments. My whole idea is to
streamline things, making the units actions a bit more consistent and
thereby speeding things up and simplifying things for the user; hence,
my prompting for ways to get rid of redundant things like bracket
entry where none should really be required (as the software should
interpret what the user wants). eg. Interpret negative 2 squared to
give positive 4, ie. (-)2^2 = 4, instead of having to specify the
extra brackets around the negated 2, ie. ( (-)2 )^2 = 4 It's quite
simple to do. But I won't go on about it as I've talked about this in
many previous posts.

I'll mention a few more wishes in the next post (below)...

Cheers!

[DrBob]

Am I blind or are there no options for rescaling axes? So here's the
next wish:

045 Other options, especially LOG-LOG or SEMI-LOG axes (for a
start). I know I can rescale to linearise plotted values, but students
probably wouldn't understand it too well.

Next, I never thought I'd have so many files in so many folders on the
nSpire. So let's see a small change in file handling and navigation of
the directory tree:

046 Improve navigation in directory tree. When within a folder,
press left arrow to go up to the next level (closing the present
folder), alternately press right arrow to enter a folder or launch (or
tab) a file (also allowing the ENTER key to do what it normally does).
This saves all the trouble of scrolling up to the folder name and
hitting ENTER to close it, then running back down the directory tree
to open another folder.

047 Allow selection of multiple files for deletion, say (ie. tagging).

[DrBob]

Hopefully the last one for today.

047 The ability to open 2 (at least) documents at once. How often
have you been in the middle of playing around with a file when you
have wanted to check a value you had saved in another document. You
may not want to save the changes you've made in the present file (but
you want to keep your temporary stuff for the moment).

At present the only way is to save to a temporary file, open the
second one, then close this again and open the temporary one (which
you then have to remember to delete after you finish as well).
What a pain! :-)

[DrBob]

Whoops, that should have been 048 !

Sorry! :-)

[DrBob]

OK, one more...

People are going all out and putting so many sliders on the screen
that the screen is so cluttered it's almost unreadable (with these
things overlapping).

049 Have a combined slider for all variables that are to be
controlled in this way. The variables, plus their present value would
be displayed, while clicking on them would bring up the single slider
with endpoint values shown for the particular variable.

Sliders are much better than the up down arrow interfaces as the
arrows force you to look away from the graph, etc, while you're
adjusting them. I'd still rather be able to TAB around inside a box
containing all the variables that could control -- TAB or use arrow
till I get the one I want, the CLICK to select it, then right/left
arrow to adjust. Easy!

[Nelson Sousa]

That one is quite interesting, I subscribe that one as well!

Instead of having several sliders, having only one with multiple
variables would be great. I've created two (very lousy) mockups, one
for expanded slides another for minimized ones.

Cheers,
Nelson

[DrBob]

Thanks Nelson,
Yes, I was thinking about something like your top panel (sliders.jpg),
but with only a single slider (horizontal or vertical) that would have
differing ranges displayed depending on the variable selected. This
would save a lot of screen 'real estate' (ie. one slider instead of
many).

The problem with the minimised ones, as you've shown in
"sliders2.jpg", is that the operator, or student, needs to shift their
attention from the plot to the arrows on the slider controls whenever
they want to change direction (ie. decreasing to increasing the
variable's value), whilst trying to note the effect on the plot. We
had this problem years ago with adapting Mac software to run huge mass
spectrometer systems. You had to keep your eye on a needle being
deflected left or right while adjusting a variable (accelerating
voltage, say) up or down using the arrows on a scroll bar. Very
inefficient system (programmers' fault :-). Better to put in a manual
dial to change the variable (which could be moved without looking at
it). This dial approach was much like the present slider system, where
you 'grab' the control (once), then change the value of the variable
with your thumb movements (left, right, up or down, and without
distraction) while concentrating on the output (graph, etc). Nice
system -- except for the large screen area required.

I can see the value of students seeing all the sliders with their
differing limits; but maybe we could have an option to either show all
sliders (as in your first diagram) or use a single combined, or
compound, slider (sort of an expanded versus summary view).

Cheers again!

[Michael Ball]

As far as:

> The negative symbol suffers from a scaling error, I believe. It should
> be made somewhat bigger to aid readability (at least half the size of
> the 'minus' symbol). In fact, you can actually see that the nSpire
> does this, but not at the base level of operations. Try the following
> and you will see that when raising something to the power 3 or 4 times
> the 'negate' symbol actually gets bigger (and more readable) while the
> numbers reduce in font size. Interesting!
>
> Try this (or a variant):
>
> -2^(-2)^(-2)^(-2)^(-2)
>
> ie. Minus 2 raised to the power of minus 2 a few times.
>
> After about 3 power raises you can see that the negate symbol
> increases to a more readable size.
>
> Should be an easy programming fix.

It took me 5 power raises for it to change size, either that or it's
an illusion. Anyway, I think this is odd, but I couldn't see it if it
got much smaller.

050: (Wow) I'd like the option to remove the down-scaling for
exponents and nested functions. This would make me quite happy
combined with 006M. :)

As far as the whole minus substitution issue: I've avoided any
problems as a student by using parentheses where I'm unsure. I've
never actually used 3(-)2 as a means of returning -6 either, and I
normally would have expected an error which is how the 83/84 work.
While I think the goal of the calculator is simplicity for students, I
think (as a HS student) that any extra work like that gets more
confusing. I like the distinction in the two symbols and I think it
can help avoid errors. We've been taught that -x^2 = -(x^2), and I've
had this discussion with a friend of mine who has his PhD and agreed
with me, though not everyone does. But, this is the way (all?)
textbooks seem to teach it.
My one gripe is that 2 - 4 = -2 (correct) and that Ans^2 = 4 is
correct, but that inserting -2 into a problem will return -(2^2) and
not treat it as the answer (a variable). I guess my solution would be
to have the Nspire automatically inset parentheses around the answer
when you insert it, which is what I do-if nothing else, because I've
learned order of operations and I know how I want to do the math. I
believe I'll continue to do this until there's brain nodes connecting
to the USB port. :)

[-TJ]

Dear Everyone,

Wow, my discussion started small, and now it is HUGE!
This must be very important and might as well be the most influential
discussion in the entire TI-Nspire Google Group.

Sincerely,
-TJ

[tmb]

018: How about sending a picture/jpeg/etc.? Ti-Connect could send a
bitmap from the computer to the 84. With better resolution this could
be fun.
If you had a photo of a suspension bridge send to Nspire, then try to
graph the curve (catenary I think) [now I just take a transparency and
trace the image from the monitor (may have to re-size it) or magazine,
then lay it atop the screen and look through it]. This will be handy
for my Calc-Art project.

Post your photo, or group shot from a conference, personalize a
document, maybe get a Calvin/Hobbes comic appropriate to the
Activity. Collect photos that exhibit geometric shapes, measure the
flag pole problem presentation.

[yes, this feature could be inappropriately used; much like cameras on
cell phones. just gotta teach appropriate use]

On Oct 28, 7:29 am, Sean Bird <covenantb...@gmail.com> wrote:
> I imagine HOME > New Document
> wouldn't satisfy for the situation you have?
>
> Or CTRL HOME (Tools) > Insert > Problem
>
>
>
> On Wed, Oct 28, 2009 at 9:37 AM, Wayne <waynep...@bellsouth.net> wrote:
>
> > The clear a-z applies only to one-letter variables.  I want to clear
> > all variables, whether they are one- or multi-letter names.

[Joe]

A rationalize command. Since auto simplify doesn't rationalize a
denominator when it contains a variable, for example 1/(sqrt(x)+1) the
only way to force the cas to do that fis to have a rationalize
command. Perhaps this problem has been dealt with years ago in some
manner?

> > > all variables, whether they are one- or multi-letter names.- Hide quoted text -
>
> - Show quoted text -

[-TJ]

That's a good idea, Joe... Unfortunately, I don't have the expertise
to program the calculator to do such without any problems or errors.

[Wayne]

051 I would like to have the system display the name of the current
document on a page of the document. The document name could perhaps
be displayed on the bottom line of a calculator page together with the
display of the number of lines in the calculator history or on the top
line together with the display of document settings. It might also
be displayed in some dedicated line when a Notes page is opened.

Wayne

[Sean Bird]

Yes, it has happened often that a student asks, "Which file are we opening up again?"
I have enjoyed the visual aid provided by the TE software that has the name of the file written across the top.
But that would help. Of course many file names are longer than what may be readily visible on the top of the handheld.

I wonder if other users would find the file name more helpful than the settings, e.g. "RAD AUTO REAL", if it had to be an either/or option?



Sean Bird
Indianapolis, IN
http://covenantchristian.org/bird/Nspire.html