... maybe two or even four:

Looking at the top two rows of keys on our hypothetical calculator, I did show you a layout earlier as displayed in the left picture below. Alternatively, these rows may look like the right picture - BUT there's one vacant location there. Any suggestions what shall go there?







You may even suggest more since 'I am not married to | |' (as people say here). I'd be also willing to kick the cube root and cube if you have a really good idea requiring four locations.

Now the small print: THIS IS NO POLL. Call it brainstorming if you like - or 'let 100 flowers blossom' as Mao Ze Dong once declared (don't be afraid, you won't experience the same fate as those 100 flowers then). And remember the top row keys will be reassigned with each and every menu call. So don't suggest putting anything there you want to see and use all the time.

d:-)

Edited: 13 Dec 2012, 7:20 a.m. after one or more responses were posted

Whatever you do, please do *not* lose the "i" - so not four spaces available but three at-most, yes?

A few suggestions if I may:

1. Not sure if y^x is somewhere else but that'll have priority over the x-th root of y

2. Consider a twist of math in the layout, say x!, or nPr / nCr. Yes sure they are in their dedicated menus but...

3. I know this is controversial but a Hyperbolics prefix will also come handy.

now let the flames begin! (oops, sorry "games") :-)

Edited: ok I saw y^x - so never mind #1 above - or maybe do: it's a secondary function now, I'd say it can be made primary instead.

Edited: 13 Dec 2012, 7:19 a.m.

Olá Ángel,

Do you wear your fire-proof suit? Ok, so let me count really slowly: One vacant location plus | | gives two vacant locations. Ok so far? Now take these two and add cube and cube root - what do you get? Four! Great, Ángel :-) Now sit down and enjoy the rest of the class ;-)

H.MS+ please. The only funcion I really need a pocket calculator for...

Quote:

---

3. I know this is controversial but a Hyperbolics prefix will also come handy.

---

Where to put it and what to lose in favour of it. It also increases the complexity of the keyboard input engine which is something I'd like to avoid -- the 34S went majorly overboard in this respect and a significant portion of the code is dealing with this rather than implementing functions.

Plus, the keyboard will be user definable so you can place them where ever you want.

- Pauli

The obvious place for this and H.MS- would be on the shifted + and - keys instead of %+ and %-.

- Pauli

Appreciate the arithmetic lesson Walter, now I know why I need calculators. It's always good to know where our limits are.

So the vacant position was not shown, uh? Interestingly enough I made the assumption the picture was describing the situation, but it appears not. Or maybe the text wasn't appropriate for the picture? Who cares.

Übrigens, auf Spanisch man schreibt "hola" mit "h". Ich glaube ohne "h" ist Portugiesisch.

Hola Ángel,

Appreciate the Spanish lesson :-) But we are not through yet with arithmetics, I'm afraid. But maybe it's a vision problem? Come on, what do you see below | | ? Nothing? Me too! And that's what I call a vacant location :-) Ah, I knew you'll need your glasses today ...

The only glasses I need are those filled with good Rioja - or Pinotage, Shiraz... I recommend you try it too, specially before posting your hall-mark pleasantries.

Quote:

---

The obvious place for this and H.MS- would be on the shifted + and - keys instead of %+ and %-.

---

Agreed. But I really would prefer it unshifted... (my most often used app on the iPad is a calculator that does nothing but add time values)

I don'n like the idea at all. The one on the left hand site is perfect. Please do not clutter the keyboard with exotic functions. Keep it simple.

How about adopting elementary modular arithmetics 'y-1(mod x)' as well as 'y(mod x)' ?

Quote:

---

It's always good to know where our limits are.

---

Let's return to calculators.

d:-)

Edited: 13 Dec 2012, 10:25 a.m.

Quote:

---

I don'n like the idea at all. The one on the left hand site is perfect. Please do not clutter the keyboard with exotic functions. Keep it simple.

---

Since the keys are assignable everybody can build his own layout!

Franz

I would even suggest a much more extreme layout change:

Remove even the log and trig functions from the keyboard and put them into menus!

For example instead of [sin] [cos] [tan] use only one prefix key [TRIG] which shows [SIN][COS][TAN][ASIN][ACOS][ATAN] in the menu line (called with [F1]...[F6]).

The same method for a [HYP] prefix and a [LOG] prefix.

This would make keys free for things that are really needed on the keyboard, like [i] for complex entry or [alpha] for alpha entries.

Franz

Quote:

---

Plus, the keyboard will be user definable so you can place them where ever you want.

---

Yes, but will it be easily re-labelable? I guess my assumption has been that we are discussing the keyboard function layout that will be hard-coded, so to speak, on the keys and keyplate. So re-assignment of functions will require some means to re-label the keys and keyplate, ala the wp34s overlays produced by Eric Rechlin. Is this the plan? If so, I think we need to develop additional infrastructure to produce the overlays, as it might be too heavy a burden to place entirely on Eric. (I won't claim to speak for Eric, so he may be ready, willing and able to take on the task.)

FYI, alpha is already a primary function and [i] is ... ummh ... only required if you need complex input - which is a feature definitely not requested by everyone here.

Thus I put [i] on the top row, to show it to the knights of complex domain (BTW, [i] is also in the unshifted row of menu CPLEX). Same motivation for [Sigma+] for statisticians (which is also in the unshifted row of menu STAT) and LG for ... ummh ... those who aren't aware of LN yet. Also x-ROOT(y) and LOG[sub_x](y) are asked for frequently. The remaining stuff is more or less arbitrary as I mentioned in the OP.

d:-)

Edited: 13 Dec 2012, 10:22 a.m.

Quote:

---

FYI, alpha is already a primary function and [i] is ... ummh ... only required if you need complex input - which is a feature definitely not requested by everyone here.

---

"A feature not requested by everyone"? Well, NO feature is requested by EVERYone, so with this argument you won't have to put ANY function on a primary key. ;-)

But as I already said: everyone want his own layout, so it would be the best to make the calculator with a naked keyboard.

Franz

Edited: 13 Dec 2012, 10:37 a.m.

As much as I like the || function, I think replacing that key's dedicated functions with Mean and Standard Deviation would group better with the dedicated Sigma+ and Sigma-. Move the Mean and Standard Deviation key to the second column and shift the others right one position.

Mark Hardman

OK, I rephrase: [i] is required by far less people than SIN.

d:-)

Quote:

---

OK, I rephrase: [i] is required by far less people than SIN.

---

You see what I mean? ;-)

Franz

Well, that's nothing else than IF you do statistics then [Sigma+] is indeed needed for EVERY number input, so it's not just a function but more like a number input (or decimal point or +/- or EEX).

You see what I mean? ;-)

d:-)

I want USER as a primary (unshifed) key, but that should be on the third row. RDN could move to the second row.

It would be nice to have y^x as a primary key. xth root of y, if it's on the keyboard at all, could be a shift of y^x. I never use it, but perhaps % could be a shift of 1/x.

If the top row is not dedicated to soft menus (as in the right illustration), I'd prefer Sigma+ to be in the second row rather than the first row.

I'd certainly rather see something else more useful replace x^3 and cube root, but I'm not sure what.

Overall, I think what I'd like to see is something like:

1st row unshifted: 1/x      ln    log    sqrt   y^x     ? 

1st row f shift:   %        e^x   10^x   x^2    xroot   ?

2nd row unshifted: Sigma+   RDN   i      SIN    COS     TAN

2nd row f shift:   Sigma-   RUP   angle  ASIN   ACOS    ATAN

3rd row unshifted: STO      RCL   f      g      ALPHA   USER

That still leaves one top row key unassigned; I'm not sure what I'd want to put there.

Edited: 13 Dec 2012, 12:53 p.m.

Quote:

---

I want USER as a primary (unshifed) key, but that should be on the third row.

---

d:-)

USER isn't necessarily a toggle, though it can be. The behavior is more like USER mode of the HP-48.

The default behavior is that it shows the USER1 annunciator, affects the next key or shifted key sequence, and turns off. If you press it twice in a row, it locks, and shows the USER annunciator, until at a later time you press USER again to turn it off.

By setting a flag, you can disable USER1 and have it act as a pure toggle. On the HP-48, flag -61 ("User Lock") is clear for USER1 mode, and set for USER as a toggle.

It is for the USER1 behavior that it is very desirable for USER to be a primary key.

It might be useful to have a similar ALPHA1 feature, but for ALPHA, the default should be to lock, and the user should have to set a flag to get the ALPHA1 behavior. For the HP-48, that is flag -60.

Edited: 13 Dec 2012, 2:18 p.m.

Quote:

---

So re-assignment of functions will require some means to re-label the keys and keyplate, ala the wp34s overlays produced by Eric Rechlin. Is this the plan?

---

I would be in favour of a "keyboard rearrangement kit" containing a new front plate (so breaking the heat steaks on the original one is not an issue) and a few extra keys with different primary labels. This would allow the user to shuffle the keys around as he pleases and then re-assemble the calculator.
The secondary key labels could be done "HP-41-style" using replaceable overlays.

Who needs trigonometric functions if you have complex numbers? ;)

No, but seriously, for me [i] would be a very important key. Proper complex number support is the only thing I miss in the WP34s. It does handle complex numbers, but (at least for me) using them doesn't come natural. This is one of the big improvements I would like to see in the 43s

The [i] key could be one of those optional keys in the "keyboard rearrangement kit" described a bit further up this thread.

Cheers,
Harald

I'm pretty sure (based on the WP 34S) that I'll like whatever you finally decide on!

Myself, I'd like:

• unshifted x^2, sqrt(x), 1/x, y^x; I use these keys so often.
  
• Although I would use "parallel", it wouldn't be often enough to justify a primary key place. So:
  
  • move "parallel" under sqrt(x);
    
  • move the ln/e^x key up next to the 1g/10^x key, shifting the third and forth keys in the top row to the right;
    
  • have x^2 as the primary key function of the key now vacant to the left of sqrt(x).
    
  • I'd also swap y^x and y^(1/x) so that y^x is primary.
    
  • In the empty parking space (lot?) under x^2 I'd have ... abs(x), perhaps? This goes nicely with x^2 and would be useful with complex numbers.
    
• Having said all this I'd prefer the keyboard layout with the blank top row. I feel a soft menu is easier to use when the keys referring to it were blank. The top row of the WP 34S keyboard always feels a little uncomfortable with its dual usage; I often forget whether labels A-D are defined or not. Still, I accept that it can work either way, especially here where a soft menu will be displayed.
  

Anyway, to repeat: you have my total confidence!

Nigel (UK)

Quote:

---

Proper complex number support is the only thing I miss in the WP34s. It does handle complex numbers, but (at least for me) using them doesn't come natural.

---

d:-)

I'm not up to copy - I'm thinking what will make sense. "ALPHA1" in an environment where we allow for multi-character variables doesn't seem too useful IMHO, "ALPHA" applies better. "USER1" ... hmmm ... I've to think about it. "USER" applies for sure.

d:-)

Eric,

did you read the small print in message #1?

d:-)

Oooh, I like those top two rows a lot! I agree that y^x is a better primary key choice than x-root(y).

Sigma+/Sigma- or R>P/P>R seem like good candidates for the last key-- a tough choice, and some would likely prefer >H.MS/>H.

-Bill

I very much doubt they'll be heat stakes on this hardware.

Templates liek the 41 series have been discussed and seem like the path to goodness.

- Pauli

USER mode? I don't really see how unshifted H.MS+ can be justified as a top level function. Still will be considered nonetheless.

- Pauli

I use the alpha1 mode on the RPL machines a lot. It additionally works press and hold and type as much alpha as you want, release to drop alpha mode.

I don't use USER mode much, but USER1 makes sense in a similar way.

- Pauli

Could shifter function keys be used to give another 6 user definable functions ..?

Mike T.

I very much expect that there will be heat stakes on this hardware. The alternatives are more expensive.

Richard Ottosen disagrees with me, but I spent time researching it, and the difference is significant.

Edited: 13 Dec 2012, 6:47 p.m.

Yes, I did. Why?

I use USER1 all the time on the 48, and miss it when I use the 41, 42, and WP-34s.

USER1 effectively acts as a meta-shift, doubling the number of functions directly accessible with only one extra key press.

Sometimes I use USER lock, but not as much as I use USER1.

I think this is one of the best UI improvements that was made on the 48 with regard to the 41.

I like R->P and P->R for the top row.

Eric said: "I'd certainly rather see something else more useful replace x^3 and cube root, but I'm not sure what."

Gene: I'd like to second that. x^3 and cube root are not what I would ever suggest as keyboard dedicated functions...IMO of course.

Just too easy to do otherwise, even IF you lose a stack level.

So my vote is for nearly anything else in place of those two. :-)

Gamma and digamma then :-)

- Pauli

R>P and P>R would be my choice, too. However, despite the hue and cry when they weren't included on the 35S, I'm sure there are some who have never used them. In that case, sigma +/- wouldn't be a terrible choice. Better than cube and cube root, anyway.

I don't use the % functions either. I've tried now and then on the 15C but I always spend more time looking at the cheat sheet on the back than it would have taken me to do the math. Would parallel get more use? It wouldn't be out of place as the f-shifted function of 1/x.

-Bill

See how we are all different. I never use ALPHA1 or USER1 - these are the first two flags I change to opposite of default setting, after system reset.

Since you put some basic functions in first row - 1/x, e^x, ln, sqrt. These won't be accessible whenever a menu is called :-?

d:-)

Seems the small print of my OP was a bit too small. If I see e.g. 1/x or SQRT in the top row I guess you may have overlooked this statement:

Remember the top row keys will be reassigned with each and every menu call. So don't suggest putting anything there you want to see and use all the time.

This means functions for permanent use will start in row two. Of course, each and every location may be reassigned in USER mode. Maybe a lot of freedom, but that's what we are planning.

d:-)

Now we're talking !

Quote:

---

I hate to repeat, but feel free to do a better complex number support within the HW constraints set - the code is all open for you. Looking forward to your results.

d:-)

---

Hi Walter,

sorry if it came across wrong. I agree that within the hardware constraints the solution you have found for the WP34s is as good as it gets. And my comment wasn't meant as a complaint at all. All I was trying to say is that the WP43S won't have those constraints and therefore complex numbers can be supported much better. And (in my opinion) it would be a shame to miss that opportunity. The [i] key would make complex number entering much easier and would be a big improvement over the 42s, which has very nice complex number support otherwise.

I understand however, that complex numbers might not be everyone's priority. Hence the suggestion, given the re-assignable keyboard, to make provisions for the user to re-arrange the physical keyboard as well.

Cheers,
Harald

Good morning!

Quote:

---

... that complex numbers might not be everyone's priority ...

---

I think "priorities" should really be considered the heart of the project. Ideally they will be polled and evaluated before fine tuning the placement of individual functions onto keys! What percentage of users need complex numbers? Hyperbolics? Financial functions? Time and date calculations? Matrices? Probability? Real-time stuff? Mathematical and physical constants?

Everybody would have to count (or estimate) how often he presses which key and communicate his top 30 (or so). A statistical anaysis will then show which 30 functions should get the most prominent locations on the keyboard. Just like Mr. Sholes did when he designed the QWERTY typewriter keyboard!

Regards
max

0 - 9 . EEX CHS ENTER + - * / six function keys, two shifts, alpha, x<>y, backspace, exit, STO, RCL, Rv and we're up to 34. Some disagreement about a few of these is inevitable but on the whole they are the top ones.

- Pauli

Quote:

---

0 - 9 . EEX CHS ENTER + - * / six function keys, two shifts, alpha, x<>y, backspace, exit, STO, RCL, Rv and we're up to 34. Some disagreement about a few of these is inevitable but on the whole they are the top ones.

---

OK. But the keyboard will have more than 34 keys, maybe 39 like the voyagers? So five more top-level functions to assign. Logs and trigs? Complex? Matrices? Finance? Programming (on my never-to-be-forgotten-all-time-forever-truly-madly-deeply beloved Ti59 seven out of 45 keys are unshifted programming keys which I found very useful at the time when I needed it)?

And has anybody thought about a double (or triple) action shift key? One press "f" second press "g" third press "alpha", and so on. That would free another key or two. R/S on the ON key: When its already on, On becomes R/S.

Edited: 14 Dec 2012, 4:16 a.m.

The keyboard features 43 keys. Thereof, six are in the top row. Remain 37 for permanent assignments. Please look at the last picture published some days ago showing the full keyboard:

Quote:

---

And has anybody thought about a double (or triple) action shift key? One press "f" second press "g" third press "alpha", and so on.

---

d:-)

The more times I see this picture, the more I like the clear top row.

- Pauli

Then how about this one:

?

d:-)

Quote:

---

Then how about this one:...

---

Personally, I prefer the other one! Menus (or catalog as on the HP41 or Ti Voyage) would be the last resort for me, really only for very special functions that the majority of users will never need...

Quote:

---

Menus (or catalog as on the HP41 or Ti Voyage) would be the last resort for me, really only for very special functions that the majority of users will never need...

---

Which is neither the case of logs nor the case of trigs...

I quite like this one. I'd even go as far as putting LN and e^x in the LOG menu. Two key access to these functions is no worse than many existing calculators. Plus they can be assigned if single key access really is important to someone.

The cleared LN key would become USER (or better move alpha there and put USER where alpha is.

In fact HYP needn't be there either -- these function can live in LOG or TRIG or both quite happily for three key access.

- Pauli

I think the fundamental problem we're trying to address is that there are too many functions which are deemed by some to be "too important to not be on the keyboard".

Some people use complex, some use logs, some trig, some H.MS, some TVM, some integer mode, some the advanced mathematics functions, ... Pleasing everyone is clearly impossible. USER mode was supposed to be a way around this by allowing anyone to put their most dear functions in easily to access locations, yet nobody seems willing to compromise and admit that their favourites might actually deserve such relegation.

I personally find it difficult to believe that there is really is a sufficient demand for very many top level functions to be exposed at all. Stack manipulations, numeric entry, enter, sto and rcl are all good. At this point I'm running out of steam. Add in sqrt, maybe x^2, 1/x and y^x and we're pretty much done for the universally useful functions.

This leaves space for a top level USER, top level alpha, two shift keys, LOG and TRIG as menus (in addition to being assignable).

- Pauli

I quite like that one!
The most common functions are there, the others are easily accessible by pressing two buttons rather then one. And most importantly entering numbers (even complex ones) doesn't require shift keys or menus :)

There's no point in putting functions that are never used on the first row.

Take a look at what functions the 42S puts on the first row. I don't necessarily agree with all their choices (e.g., Sigma+), but they had fewer keys available than the 43s.

I want Sigma+ on the *second* row, because some of the time I use it in conjunction with stuff on menus other than the STAT menu.

Has it been decided that the top row will be dedicated menu keys (i.e., no other function markings)? I agree with Maximilian and Gerson that menus are a last resort (and not for trigs and logs).

I disagree with Paul Dale that there are too many too-important functions for the keyboard. Below, Eric Smith presented a fairly workable approach that got all the fundamentals by assigning functions to the top row keys. Walter correctly points out that these functions won't be available when a menu is called, so some tweaking would be justified. One approach would be to put the more "controversial" functions that some say they never use (trig, logs, i/theta) up there and put the less controversial functions (1/x, sqrt, y^x) safely on the second row with sigma and roll.

Essentially, the keys on the top row would act like a "default" menu and would be re-assigned through soft keys. Sure, when re-assigned, F1...F6 might be aesthetically nicer, but having "default" functions on these keys wouldn't prevent their use as soft keys. The reverse wouldn't be true. I can see the reasoning, though, that the top row wouldn't get blue-shifted keys since they already serve a third function as soft keys.

If the constraint of no blue-shifted functions on the second row is relieved, there would even be plenty of space for the rectangular/polar, H.MS, radians/degrees conversions and other nice-to-haves like parallel, combinations/permutations, etc.

-Bill

Quote:

---

Has it been decided that the top row will be dedicated menu keys (i.e., no other function markings)?

---

Quote:

---

I agree with Maximilian and Gerson that menus are a last resort (and not for trigs and logs).

---

Quote:

---

One approach would be to put the more "controversial" functions that some say they never use (trig, logs, i/theta) up there and put the less controversial functions (1/x, sqrt, y^x) safely on the second row with sigma and roll.

---

Quote:

---

Essentially, the keys on the top row would act like a "default" menu and would be re-assigned through soft keys.

---

Quote:

---

Sure, when re-assigned, F1...F6 might be aesthetically nicer, but having "default" functions on these keys wouldn't prevent their use as soft keys. The reverse wouldn't be true. I can see the reasoning, though, that the top row wouldn't get blue-shifted keys since they already serve a third function as soft keys.

---

Quote:

---

If the constraint of no blue-shifted functions on the second row is relieved, there would even be plenty of space for the rectangular/polar, H.MS, radians/degrees conversions and other nice-to-haves like parallel, combinations/permutations, etc.

---

Exception: The blue-shifted user assignments in row *two* shall become directly accessible also out of USER mode.

Warning: Being able to reassign each and every key would allow for overwriting the toggle USER and a few other essential functions. We may think about saving these but that's not decided yet.

Hope this explains a bit. Else please continue asking.

d:-)

Quote:

---

If the constraint of no blue-shifted functions on the second row is relieved, there would even be plenty of space for the rectangular/polar, H.MS, radians/degrees conversions and other nice-to-haves like parallel, combinations/permutations, etc.

---

What is the point of having two shift keys and not fully using them? Just to make the keyboard less cluttered or because of something else I am missing? 20+ more useful functions available right from the keyboard outweighs a so called cluttered keyboard, IMHO.

Gerson.

I concur. Please read above.

d:-)

Quote:

---

Please read above.

---

I confess I haven't read everything. Before doing so, let me suggest eliminating the digits 2 through 9, thus saving eight primary function positions for more useful things. After all, there are 10 kinds of people, those who don't know binary and us :-)

Quote:

---

After all, there are 10 kinds of people, those who don't know binary and us :-)

---

I thought is was "10 kinds of people: those who know ternary, those who don't, and those that don't care" :-)

Quote:

---

Plus, the keyboard will be user definable so you can place them where ever you want.

---

:-D Sounds like you want to help Eric and Richard making the hardware.

d:-)

I like it, with the exception of LN which belongs with LOG menu. Maybe another menu key instead? STAT?

Now I saw what Pauli suggested and fully agree - USER is probably the best bet.

Edited: 14 Dec 2012, 2:46 p.m.

Quote:

---

I disagree with Paul Dale that there are too many too-important functions for the keyboard. Below, Eric Smith presented a fairly workable approach that got all the fundamentals by assigning functions to the top row keys.

---

I'm not objecting to Eric's proposal. It seems reasonably well thought out. It misses out several requested functions: H.MS+ H.MS- D>R R>D Phi Phi[super]-1[super], parallel and probably others I've forgotten (gamma, comb, perm?). It also includes the not universally wanted i and angle. Even logs and trig functions aren't universally wanted -- there isn't a perfect set that doesn't include everything.

This is a balancing act. There are too many desirable function to fit. If we had enough key positions for everything everyone wants, we simply wouldn't be having these discussions.

- Pauli

Walter-

Thank you, that does explain a bit... I think. Let me know if I'm understanding correctly.

You're focusing on the primary and f-shifted keys because these will be difficult to change (heat stakes, key caps, etc.). The g-shifted key labels are printed on the face and would be easier to change.

There may or may not be *default* key assignments for the top row based on opinion. (In case I miss the poll, my vote would be for yes ;). If there are not default assignments for the top row, there will only be 12 spaces for default functions: the primary and f-shifted keys on the second row. (Hence the poll.) Still, a user would relatively easily be able to create an overlay that assigns their own default functions to the top two rows of g-shifted keys giving them 12 more directly accessible functions (although the 6 in the top row might be impacted by active menus?).

If I'm understanding correctly, that's very exciting, although it would change my response to your later "most important functions" poll. (My apologies for responding to that poll before reading your post here.) I'm still a bit confused by that poll, though: some of the functions on your list appear to already appear on or below the third row in the Spee Dee Design keyboard. I'm thinking specifically of USER/ALPHA and the percent functions, but there may be others. Has the thinking changed since the Spee Dee post to move >all< functions to the top two rows? I think that would be unfortunate. Keyboard space is too valuable to waste opportunities to tuck functions away wherever they might fit.

Thanks again for explaining.

-Bill

Bill,

Thanks for asking. There's at least one who reads thoroughly :-)

Quote:

---

You're focusing on the primary and f-shifted keys because these will be difficult to change (heat stakes, key caps, etc.). The g-shifted key labels are printed on the face and would be easier to change.

---

Quote:

---

There may or may not be *default* key assignments for the top row based on opinion. ... If there are not default assignments for the top row, there will only be 12 spaces for default functions: the primary and f-shifted keys on the second row. ... Still, a user would relatively easily be able to create an overlay that assigns their own default functions to the top two rows of g-shifted keys giving them 12 more directly accessible functions (although the 6 in the top row might be impacted by active menus?).

---

Quote:

---

If I'm understanding correctly, that's very exciting, although it would change my response to your later "most important functions" poll. ... I'm still a bit confused by that poll, though: some of the functions on your list appear to already appear on or below the third row in the Spee Dee Design keyboard. I'm thinking specifically of USER/ALPHA and the percent functions, but there may be others.

---

Hope this explains a bit more. Else ...

d:-)

What about putting menus on the blue-shifted top row and leave them blank (or labeled F1 to F6) otherwise. This would allow for just two functions (plain and gold-shifted) on any menu but would keep the menu selection close to the menu execution. Keep in mind that menus should be sticky (the display is large enough). Functions which are used frequently (and therefore shouldn't be shifted) but not by everybody, can be safely put in sticky menus. Once displayed, they are just one key press away. I would qualify statistical functions (including SUM+) as such but also trigs or complex functions (including <, re, im, re<>im, maybe except the i key).

The free blue positions can be used for other functions. User mode should always include a user menu so the preferred way of using user mode would be to assign F1 to F6 and their secondary function first. This would automatically display the assigned functions in the menu area. Other keys can be assigned, too, of course. Outside user mode, a default menu should appear. We'll need a key position to select the default/user menu while any other menu is selected.

On a graphing calculator, I've been toying with the idea of a "virtual keyboard overlay":

• Operator presses USER key
  
• The screen shows a map of the keyboard with the user key assignments. Unassigned keys display their default function. In other words, the map shows what each key does.
  
• Optionally the operator presses the shift key(s) and the map changes to show the shifted key assigments
  
• Operator presses one of the keys
  
• Calculator reverts the display to it's previous value and executes the user-assigned function.

Honestly, I believe that this functionality will completely eliminate the endless discussions that we have over which functions should go where. It would make the calculator much more useful because it could be easily tailored to whatever problem domain is most interesting to the individual user.

Dave

Dave,

I've read about such an overlay in another post recently, but can't find it anymore. Maybe it was yours, too? Anyway, I doubt such a mapping of 43 keys at once will work in real life. The distance from each physical key to its assigned label is just too far IMO.

What we plan to do is a smaller range of it: whenever a menu is called, up to 18 labels are presented - 6 unshifted, 6 f-shifted, and 6 g-shifted functions of the top 6 keys. I think that's viable, but wouldn't strain it further. Your idea of switching the display by pressing a prefix may make it even easier:

• Operator calls a menu.
  
• The screen shows the 6 unshifted top row key assignments.
  
• Optionally the operator presses a shift key and the screen changes to show the shifted key assigments
  
• Operator presses one of the keys
  
• Calculator reverts the display to it's previous value and executes the user-assigned function.

d:-)

The HP15C has F-shifted functions on the number key. Could be done here as well?

Generally speaking, each and every user shall be free to assign each and every function/operation to each and every level (unshifted, f- or g-shifted) of each and every key. In the default startup layout, we left the f-shifted digit keys free so far for less clutter. But that may change like many other things.

d:-)

I think it's essential to show the function of a shifted key before you press the respective shift. Especially when using menus, you'll find yourself often searching for a function by selecting the menu and looking at the displayed labels.

BTW, I don't like to have 18 labels displayed, because it's not obvious which row belongs to which shift. What about my proposal of putting menus on the blue shifted positions and use only the gold shift on the F keys? (see here!)

Walter,

Yes, that was probably my post that you saw. I've been experimenting with the concept on a 50g and I think I can provide enough context clues in the map to make it easy to find the right keys. The cursor keys help for the first 3 rows and using reverse video for the keys that are black should be enough for the rest.

I think the distinction between 18 and 12 soft keys will have to wait for a prototyping stage.

Just how hard is it to viably navigate eighteen? Is is possible to visibly indicate which row corresponds to which shift? We simply don't know.

You're not the only one concerned about potential over clutter in menu navigation :-)

- Pauli

Quote:

---

I think the distinction between 18 and 12 soft keys will have to wait for a prototyping stage.

Just how hard is it to viably navigate eighteen? Is is possible to visibly indicate which row corresponds to which shift? We simply don't know.

---

d:-)

I realize I seem having reached the stage of Karl Schneider in a way meanwhile - I can subsist on my older contributions to some extent at least. Hope he's well. Long time no read.

This is still quite different to actually using such a menu system :-)

With fixed meuns (that don't go away on selection), another row of soft keys will clutter the display.

- Pauli