A Serious Scientific Instrument



#15

Thanks for your numerous reactions and comments after my last proposal (the equivalent amount of thanks goes to you). Based on these, I got many ideas to ponder: Why is there more response for some minimal cosmetic changes than for efforts to improve an instrument in total? Why does one create more excitement by changing 3 colors or (not and!) reshaping 4 keys? What kind of keyboards do people have allowing them to delete by moving a cursor to the left? How big is the fraction of melapetrophils in this honorable forum? What happened to all those members striving for the optimum? Etc.

Nevertheless, I won't give up so easily, but try again with a more "serious" and cleaner outfit:

Don't be afraid, that's it. No repelling explanations nor frightening requests for comments and the like. Just a picture. Don't bother. Everything is calm and peaceful and quiet. Good night :-)


#16

Walter,

I'll try to rise to meet your challenge :)

On your handsome keyboard, and while I haven't scrutinized it, I'd point out that your UP key is the only alpha key. This could be frustrating when modifying programs or editing equations (the 48 series has this problem).

ECL


#17

The 48/50 have letters on STO/RCL. I do not see an issue with that given that this new keyboard has an alpha key. That would push Z up to 3. You could add a comma. Since symbols are used infrequently consider a few on the arithmetic operators. I would also consider the darker (or lighter) bar under menus unless all CAPS is how you distinguish them.


#18

I suppose you're consciously divorcing your speculation from reality, and with that in mind, it's a fine exercise. (I do like the overall result!)

The alpha key is, I think, the proposed feature least likely to be realized in a 35s. Apparently built upon the 33s, the 35s will no doubt impose very strict limitations on the nature of any alpha entry, relying upon keystroke context for automatically entering alpha mode.

Explicit manual control of alpha mode implies more of a 42s- or 48g-style alpha capability, with lengthy variable and program names, etc. Fundamentally different, and not gonna happen. (If only to insure the test-use certifications, let alone the software redevelopment and testing costs.)

But in its layout, legend placements, and color choices, your re-imagining of the 35s is beautiful!

H-P: there is much wisdom here, worthy of your consideration!

#19

Hello,

I think this keyboard is better with the cursor keys à la HP48.

I see 2 flow in this keyboard.

First, I don't see how you swith between normal and user keyboard, my proposal is to put it on the alpha/assign key.

Second, I think that the logic/binary keys are specific to the logic/binary mode and have nothing to do there, many floating point keys are meaningless in that case. With the HP41C, this kind of problem is solved with a keyboard overlay. On the HP41C, when you enter the stopwatch mode, you have a specific keyboard and an overlay.

In binary mode, an overlay should bring us something like the key board of an HP16C with shifts, rotates, word size ...

Last thing, I know it is guessing with capabilities, but I see nothing related to time/stopwatch/alarm

Happy flaming :-)
Patrice

#20

Walter,

Very nice work, as always. My only major suggestions are to put the alpha labels on the same level as the lower yellow labels. In addition, I think the yellow labels should ALL be placed under the keys rather than next to the blue labels occupying the same space. Not only would that improve clarity, it would also make the use of overlays feasible. It also appears that there is ample space for a set of soft keys beneath the LCD so a nice 42s style menu system could be implemented.

Keep up the good work!

-Hugh

#21

Hi, Walter --

Well, that's an earnest graphic effort for which you may be commended. The "goal" also seems ambitious, beyond what the HP-35s is intended to be -- perhaps more like an "HP-45s".

It looks elegant, but there's a certain amount of, er, inconsistency in the layout -- sometimes, the yellow legends are below the key, and sometimes they are on the sloped face. Sometimes they are in capitals, and sometimes in title case. Pi, x2, ex, and |x| are barely visible -- somewhat of a carryover from HP's own graphic.

(Remember that the original concept was that the yellow legend was to be centered above the key and the blue legend was to be centered on the sloped lower part of the key, leaving the large main legend centered between them.)

How practical is modern I/O (hardwire or IR) on a calculator that uses low-charging-capacity 1.5V cells? The three 13/44/76/357 button cells of the Pioneer (and Voyager) lines were poorly-suited for such use in 1988; this, and limited open space on the front, is probably why bidirectional hardwire I/O was not provided on the HP-17B/27S/42S. The thin CR-2032 cells used by the HP-33S seem to be even worse in that respect.

There's a "MATRIX" legend, but have you given thought to how it should work? (I'm still pondering it.)

Is the display unit well-suited for a full ALPHA mode?


As for me, I'd like to learn more details about how the HP-35s will actually work, as well as experiment with one, before I make detailed suggestions about how to improve it. I guess that we'll just have to wait...

-- KS

Edited: 6 June 2007, 11:36 p.m.


#22

Actually, a great idea would be to keep those gold labels above the keys and center the green alpha labels below them. One way to make overlays even more versatile would be to put the gold labels above the keys, the blue labels below the keys, and green alpha labels on the lower sloped face of the keys. I think that would be a pretty slick way of handling it.

As far as the batteries are concerned using coin-cells in a calculator thats ~0.7" thick is just plain dumb. The 33s isn't in mass production yet, so I hope they will be smart enough to use 3 or 4 AAA's to power the thing and get 10 years of battery life.

I think Walter's current layout is much closer to something along the lines of a 41sx... Although for that purpose I hope they go with a 2sx


#23

Hugh; i am guessing the batteries will be set up like the 33s. you can change one at a time and not have to risk loosing the memory like in a 32, 42, or the time/date settings on the 41. there's no reason to worry about how long the batteries last if changing them doesn't affect the memory. it will probably be a good design.

#24

I don't recall this being discussed in all that's appeared about alternate keyboards:

It seems to me that it makes more sense to put log and 10^x together on one key and ln and e^x on a different key, rather than having the two log functions on the same key and the two exponentiations on a different key. All (most?) other keys, where it's logical, put inverse operations together, such as rad <-> degrees; hms <-> h.hh; Rec <-> Pol; sqrt <-> x^2; etc.

My '11C, '41CX, and '48GX have logs and exponentiation the way I like them (the "old" way, let's say). My 32Sii has them the "new" way (i.e. ln and log together). Anybody know why this changed between models? Which makes more sense to you?

I'd also prefer pi as an unshifted key. As an astronomer using angles all the time, I use pi a WHOLE lot more than many other functions on the keys. Perhaps pi could be the primary function of what is now the () and [] key (makes sense to have them with MATRIX, though)

Other functions not so necessary(?): roll up (almost, but not quite, of course, the same as ENTER); complex roots (I think the way somebody else put it: take the reciprocal of the root and then do y^x; RAN # (what do you really need this for?!); % and delta% (again, just about any user of a calculator like this can probably do % calculations faster by keying them in than he/she can remember the details of how the % key works! - see recent discussion about this matter).

Glad to see that ON is a single push. I think it was (yellow?) shifted in the original version (at least it was in one of the cached versions). OFF should then be a yellow shifted (rather than blue shifted) ON (i.e. immediately above ON). I suspect most of us turn our calcs off more often than we clear the summation registers!


#25

Quote:
My '11C, '41CX, and '48GX have logs and exponentiation the way I like them (the "old" way, let's say). My 32Sii has them the "new" way (i.e. ln and log together). Anybody know why this changed between models? Which makes more sense to you?

I like the 32Sii way (unshifted e^x and ln side by side)
because, as a math teacher, I use the natural log much
more often than log base 10.

I wish my calculator had "log_b" and "b^x" keys instead
of "log_10" and "10^x", with the value of b settable by
the user in a special register. I'd set mine to base 2,
since I use it more than base 10. This is admittedly a
minor point, since base-2 logs are easy to compute with
3 extra keypresses.


#26

Quote:
I wish my calculator had "log_b" and "b^x" keys instead of "log_10" and "10^x", with the value of b settable by the user in a special register.

I think that is by far the coolest idea I've heard in a LONG time! Excellent thinking!

I'll be implementing that as standard routines on my 33s. It would indeed be very cool if such was available on the 35s.

Great idea!

#27

Quote:
I wish my calculator had "log_b"

It's a pity x|/y (xroot) will be a shifted function on the HP-35S. On the HP-33S logxy can be obtained with only three keystrokes:

[LN] [x|/y] [LN]

For instance:

32 ENTER 2 LN x|/y LN   ->   5   (log232)

#28

Quote:


It's a pity x|/y (xroot) will be a shifted function on the HP-35S. On the HP-33S logxy can be obtained also by:

32 LN 2 LN /    ->   5   (log232)

Dennis


#29

Quote:
On the HP-33S logxy can be obtained also by:

32 LN 2 LN /    ->   5   (log232)

This is not logxy, in the sense x and y refer to stack registers, as on the key labeled yx.

Actually, the sequence I presented simulates a hypothetical logxy key. This might be more interesting programatically though. For instance, assuming a in stack-register Y and b in stack-register X try to write a shorter RPN program than the following to compute logba:

L0001 LBL L
L0002 LN
L0003 x|/y
L0004 LN
L0005 RTN

(15 bytes on the HP-33S; 7.5 bytes on the HP-32SII)

Regards,

Gerson.

#30

Good,

but, even if someone pointed out that they're useless, to perform ASINH (hyperbolic arcsin) you've got to hit

[yellow] HYP [blue] SIN

or am I wrong?

-- Antonio


#31

Thanks for the feedback! So you shall get some, too:

  • yellow print on the keyplate is for menus - this explains most of your remarks
  • HYP is a menu containing 6 items
  • there are 3 blue menus on the keys 1, 2, 3
  • everything else is a function
  • Patrice, you are right with USER mode - I'll think about it
  • Dave S., I agree on your remark about LOG/LN. The reason I distributed it this way was simply the space available. A solution may be to abbreviate the decimal log with "LG". OK?
I'll come to your other remarks later. Promised!
#32

Well, talking about a serious instrument, you should consider my 65 modification:

(google for "all your base" to experience what has been a running gag since 2001...)


#33

Nice HP-65 redux!
B^)

It made me wonder, (pardon my not knowing this) did HP's have a F-1 button?

Ren

dona nobis pacem


#34

Quote:
did HP's have a F-1 button?

AFAIK only one: HP65.
#35

Walter, it looks great. Only two things I would like changed. I'm old school and would like to see the F & G key back, and I would also like to see Hewlett * Packard spelled out on the very lower lip of the calculator.


#36

John, what is "F & G" ??


#37

Sadly, that's what 98% of everyone who would buy a machine with "f" and "g" keys these days would say.

WE all know what they are, but the majority of purchasers would wonder if the "f" key was broken...imagine the calls to tech support "What is this "f" key and why doesn't it do anything?"


#38

Gene, and others who miss the f 'n g keys...

Maybe the f and g could be printed on the shift keys along with the curvy arrow, maybe on the lower bevel, the same color as the function.


Hmmm, maybe they could make the key tops holographic, depending on the viewing angle, it would show the curvy arrow or f 'n g.

Ren

dona nobis pacem

#39

Shame on me, I should have known this! :-(

But in CAPITALS ... was looking like one single operation ... simply couldn't get an idea :-/


#40

Quote:
Shame on me, I should have known this! :-(

But in CAPITALS ... was looking like one single operation ... simply couldn't get an idea :-/


Sorry about that Walter....

#41

Thanks a lot for your kind responses and valuable critics. As Luiz rightfully suggested, there shall be one proposal at the end. So let me sum up the open points and wishes:

  • "f" & "g" will appear along with the curved arrows as Ren proposed, Nelson.
  • USER mode will be selectable, Patrice. I'd put it on "g" "UP".
  • MATRIX menu shall work as in 42s, Karl.
  • I/O: I thought of an USB socket.
  • SOFT KEYS like in 27s, 42s, 48 would be very nice to have, but would require a different LCD. I didn't want to redesign that much, but keep as much as possible of the 35s HP showed us. Well, I failed with the cursors already... ;-)
  • ALPHA keys: I must admit I had an 48sx on the desk. Nevertheless, I'll follow ECL's arguments, and will have alpha-free cursors. To make up for this, I'll put letters on STO/RCL as Egan suggested. Paul (B.), you're right, I'm longing for a 42sx more than for a 33sII - IMHO the 42s was (and still is) far superior to the 32s, 32sII, and 33s. And if there will be a 35s for NCEES tests very soon, HP is free to put something serious on top of it :-) It's still a very limited alpha feature.
  • GENERAL FUNCTION SET & KEYS: there were different wishes, so we have a list of candidates and a drop list below.
  • GENERAL LABELLING: Going way back in history, an HP21 looks far cleaner than a 31e or a 10c. The 32sII put it to the extreme negative. Thus, labelling itself will stay as proposed for sake of clarity. Please also see my explanations posted earlier in this thread.
  • COLORS OF KEYS: Jürgen, I agree this may give it a kick, but unless there is a broad majority wanting this and agreeing on the colors, I won't do it.
  • OVERLAYS: For user- or application-specific reassignments, overlays with flaps may be used (the flaps covering the slanted front of keys, but may be cut away). I vote against overlays for parts of the standard function set. Instead, there shall be as much as possible directly accessible from the keyboard.
Candidates: Pi unshifted (Dave S.), EQN unshifted (Bill), e^x & LN unshifted, LOG_b & b^x (Paul G.), stopwatch (assume >1 function, Patrice), ALL YOUR BASE ARE BELONG TO US (sorry, Frank, I see low chances for this important operation ;-), %T (Carlos).

Drops: Binary & logic functions (Patrice), x_SQRT_y, R_UP, RAN#, %, %CHG (Dave S.). (If there isn't a space problem, I won't drop anything.)

Anyway, right now the keyboard is full. So, whoever wants a candidate included must say which function shall be replaced (i.e. deleted from the keyboard or moved to another location). Paul G. already told us he wants LOG_b & b^x replacing LOG_10 (= LOG = LG) & 10^x.

Hope I didn't forget anybody's contribution. Else please remind me. In either case, it's your turn again :-)

Addendum: Most probably the 35s will be released as published. So whoever is completely satisfied with the 35s as is, you'll get it. I'll buy one for sure, too. And I'll continue "tuning" (or dreaming ;-). Just imagine. "Maybe I'm a dreamer, but I'm not the only one!"

Edited to update the candidates and drops. Added the last paragraph.

Edited: 9 June 2007, 3:33 p.m. after one or more responses were posted


#42

Hello!

Quote:
Hope I didn't forget anybody's contribution. Else please remind me. In either case, it's your turn again :-)

I havent contributed yet, but i will do now, with a reminder :-) :

This thing is supposed to be a reborn HP-35. A very simple (by todays standards) scientific calculator. With no shift key. Simple and clear.


For me, the HP-35S should be just the same: Simple and clear. No need to read the manual, no need to install communication software on my Mac (that HP will never supply anyway, other than Ti :-( ), no need to press "XEQ Alpha Letter Letter Letter Alpha Enter" to invoke a command like on the '41 (horrible, terrible, unergonomic nightmare of a user interface!), no need to hassle with menus and cursor keys. Sine of 31 Degrees? "3 1 SIN" is all I ever want to have to type into any pocket calculator to do that.


A simple machine for quick and simple calculations is all it should be. For the not-so-simple calculations absolutely nobody (apart us nutters!) has used a pocket calculator in the last 30 years anyway...


Therefore I would suggest to wait with all these ideas and improvements for the 35-years-jubilee-version of the '41....


Greetings, Max

Edited: 8 June 2007, 3:11 a.m.


#43

I kinda like Maximilian's suggestion! (I might even get my unshifted pi key back!)

Isn't "retro" in these days?


#44

Quote:
I might even get my unshifted pi key back!

Dave, you will get it, just you tell me what I shall drop for it. Referring to "retro" I suppose you'll be not satisfied with the original function set of HP35A, will you?

Edited: 8 June 2007, 2:24 p.m.


#45

Hello!

Quote:
Referring to "retro" I suppose you'll be not satisfied with the original function set of HP35A, will you?

Well, I at least would be completeley satisfied with the original function set. For evereything that requires more functions than that, I wouldn't use a pocket calculator anyway. I live in the year 2007 and get paid for 2007 work, so I am not able (nor are my employers!) to afford the luxury of doing this work with a toy, whatever sophisticated this toy might be, when the competion uses proper tools to do the same work.


Come to think of it, the calculator that gave me the biggest satisfaction ever was the Privileg basic scientific machine (D-883) that I got for christmas when I was 13 or 14. Somehow, I am still after the same degree of satisfaction, and I dont feel, that matrix operations and triple-nested menus will bring that back to me...


Greetings, Max

#46

"Referring to "retro" I suppose you'll be not satisfied with the original function set of HP35A, will you?"

Well, like Maximilian, that would meet most of what I use a calculator for. I just dragged out both my original '35 and my '41CX. Other than the summation+ key, which I've used on a few occasions (mostly when teaching physics labs) for linear least squares fitting, there aren't any other new keys on the 41 that I use (except for those few times when I actually programmed the 41 - which I did in the field back before the days of laptop PCs, which I would probably take now). Similarly for my 32Sii and 42S - there isn't much new there that I need.

As has been mentioned elsewhere in the Forum, one could imagine a split between simple calcs (the 35S) used for simple purposes, and the heavy iron (a 50G) used for more complex problems.


#47

Quote:
one could imagine a split between simple calcs (the 35S) used for simple purposes, and the heavy iron (a 50G) used for more complex problems.

Agreed for the 35s. With 50g, the one and only problem for me is it is RPL instead of RPN. As others in this forum, I see a gap left by the late 42s and even 41c. I hope for a midrange calc with some (limited) possibility to exchange data with the rest of the world (e.g. with a PC), especially for program loading and backup. Else, several kilobytes of memory are of very limited use. If we call it "45s", we may believe we won't have to wait for the 41 anniversary ;-)

Edited: 9 June 2007, 3:02 p.m.

#48

Although our wishes may only help for future developments or just to fantasize, I would never drop the "x_SQRT_y, %, %CHG" keys in a scientific calculator. What is more, I would even consider a "%T" key for recurrent calculations.
Just my 2 cents.


#49

Buenas tardes, Carlos, I put your suggestion into the summary. Though you did not tell us what shall be dropped for %T.


#50

Thanks Walter for your reply. I should have chosen to drop a shifted key like <-ENG-> or the blue shifted computer terms like XOR, etc. but I wouldn't like to be murdered :) by my dear fellow colleagues who claim them for their own reasons, because I preferred a more comfortable way of dividing numbers (for prorating).
That said, let me point out that I found the delta percent and percent functions, sharing the same key place, mostly interesting, as well as the elegant proposal for logs in any base.

#51

Here we go:

Please see the keys K, L, M, where the logs were rearranged (thanks to Dave S.), and a log to base "b" is introduced (thanks to Paul G.). The base may be stored and recalled via "STO b"/"RCL b" using the SQRT-key. The function xSQRTy was deleted because I needed the space for "b" and the operation may be reached as fast via "1/x" "y^x". The shift keys now show f & g (thanks to John Nelson). Besides, I relocated some keys for better grouping (especially for EQN, thanks to Bill P.).

In alpha mode, the cursor keys are "alpha-free" now (thanks to ECL). Some symbols are on the arithmetic operators (thanks to Egan). The keytop functions of those keys attributed with a green character may be accessed via "g" + the respective key (thus, the normal g-shifted operations shall be not accessible in alpha mode, except GTO and LASTx; f-shift may be used, however).

Thanks to everyone who contributed! Comments, critics etc. are most welcome 8)

Edited: 10 June 2007, 2:31 a.m.


#52

The Backspace key is *still* there :-(
It is simply not needed when you have a left arrow key that can be used as a backspace key in normal RPN entry mode.

Get rid of the backspace key forever, move the 1/X key down, and give LOG it's own key.

Then I'll be able to sleep at night!

Dave.


#53

Dave, I'm simply too stupid to understand :-/ Assume you are editing a matrix. You want to go e.g. from element a33 to a12 to clear one digit of it. But you must not delete a33 nor a22 on your way. So how do you want to do this without a left cursor plus a backspace key?? And why do 48 and 42s have cursors and backspace??Please explain to me like to a 4-year-old toddler. You are not allowed to sleep before ;-)

Edited: 10 June 2007, 6:46 a.m. after one or more responses were posted


#54

Quote:
Dave, I'm simply too stupid to understand :-/ Assume you are editing a matrix. You want to go e.g. from element a33 to a12 to clear one digit of it. But you must not delete a33 nor a22 on your way. So how do you want to do this without a left cursor plus a backspace key?? Please explain to me like a 4-year-old toddler. You are not allowed to sleep before ;-)

Simple, you have backspace (or delete digit) as a shift function.
Guess it comes down to whether or not you want the keyboard optimised for matrix operations or basic scientific functions. For me, I'd use basic scientific functions 99.9% of the time so would rather have the extra keys for that purpose.
I just think it's smarter to combine the left arrow and backspace functions for normal operation, gaining an exra valuable key. And in my opinion matrices are not "normal" operations, so a shift function would suffice. But of course, YMMV.

I could complain about the other "wasted" keys, but I'll let that slide! :->

Dave.


#55

And why do 48 and 42s have cursors and backspace?? (you see I'm trying hard to act like a 4-year old)

And what are the other "wasted" keys? Why?


#56

Quote:
And why do 48 and 42s have cursors and backspace?? (you see I'm trying hard to act like a 4-year old)

Because they are more complicated calculators.
I agree with Maximilion above when he said the 35S should be a *simple* scientific calculator. That means lots of keys dedicated to simple everyday functions.

If I want to edit matrices I'll do it on a nice big 4 line screen thanks, not a pokey one or two line display.

Quote:
And what are the other "wasted" keys? Why?

A simple calc shouldn't even have dedicated cursor keys, so there is 4 wasted keys for starters.
I won't go into the others!

OMG, I just notice the ENG and <ENG shifted function are now gone!
Every basic (or even complex) scientific calc should have a dedicated ENG button. I was even happy with the shifted versions on the orginal 35S, at least they were in a nice prominent position.

These changes make this calc almost useless for me for everyday use. Just like big graphing calcs are useless to me for everyday use.

Dave.


#57

Thanks for your opinion. Let's wait for more feedback.

#58

Hi, Dave --

Quote:
If I want to edit matrices I'll do it on a nice big 4 line screen thanks, not a pokey one or two line display.

I'd prefer a PC screen with Matlab or similar, but for a calculator, I actually prefer the "single element plus menu" display methods of the HP-42S over the methods of the RPL-based 48/49/50, with their four or five lines. I think it's best to display one complete datum -- real or complex -- than to try to cram as many elements as possible into a small display.

Quote:
A simple calc shouldn't even have dedicated cursor keys, so there is 4 wasted keys for starters.

Including the abominable "raised silver bar", the HP-33S has five more keys than the proposed HP-35s -- 48 versus 43. You're right that including four cursor keys will eliminate some desired primary-function keys. However, the cursor keys could be put to very effective use:

  • Program navigation: Up/down scrolls by line; left/right could jump to adjacent labels.
  • Equation navigation: Up/down jumps to adjacent equation in memory; left/right could scroll cursor within an equation (a feature missing from the HP-22S/32SII/33S, but present in the HP-17B/BII/27S).
  • Menu navigation: Up/down jumps to next screen or level; left/right moves to different selections
  • Matrix navigation: If matrix capabilities were present, the cursor keys would allow the easy navigation provided for the HP-42S and RPL-based HP-28/48/49/50.

Quote:
Every basic (or even complex) scientific calc should have a dedicated ENG button.

I like Casio's way of doing things, which may have been adopted for the HP-35s. Use "<-ENG" and "ENG->" to place the display in ENG mode and to move the decimal point for scaling the value as desired. There is no need to include ENG by itself as a separate mode setting under the DISP menu, because it is equivalent to SCI except that the exponent is an integer multiple of three. "ENG n" versus "SCI n" also makes no difference if used as an integrand-uncertainty setting.

-- KS


Edited: 11 June 2007, 2:26 a.m.


#59

Quote:
Including the abominable "raised silver bar", the HP-33S has five more keys than the proposed HP-35s -- 48 versus 43. You're right that including four cursor keys will eliminate some desired primary-function keys. However, the cursor keys could be put to very effective use:

* Program navigation: Up/down scrolls by line; left/right could jump to adjacent labels.
* Equation navigation: Up/down jumps to adjacent equation in memory; left/right could scroll cursor within an equation (a feature missing from the HP-22S/32SII/33S, but present in the HP-17B/BII/27S).
* Menu navigation: Up/down jumps to next screen or level; left/right moves to different selections
* Matrix navigation: If matrix capabilities were present, the cursor keys would allow the easy navigation provided for the HP-42S and RPL-based HP-28/48/49/50.


Agreed they can be put to good use, but of course a basic scientific calculator doesn't need any of those functions now does it :->

Quote:
I like Casio's way of doing things, which may have been adopted for the HP-35s. Use "<-ENG" and "ENG->" to place the display in ENG mode and to move the decimal point for scaling the value as desired. There is no need to include ENG by itself as a separate mode setting under the DISP menu, because it is equivalent to SCI except that the exponent is an integer multiple of three. "ENG n" versus "SCI n" also makes no difference if used as an integrand-uncertainty setting.

Indeed. As an electronics engineer the ENG key is my most used key after the 4 functions and EXP, as I am always dealing in M,k,m,u,n,p etc,
I can't imagine any scientific calculator without this key. I'm happy with <ENG as a shift function, but ENG simply must be a dedicated key on any calculator that claims to be suitable for engineering use.
On more complex calcs like the original proposed 35S, shift functions for both are a reasonable compromise. HP had the good sense to not only add them, but make them prominent.

And before anyone mentions it, ENG mode is *not* the same thing as an ENG key. Having everything always display in ENG mode can be incredibly annoying, and I always want the option of pressing the ENG key when needed. Simple and effective, the way it should be.

Dave.


#60

No one of you seems aware of the following fact:

the xROOTy key is essential when the operand is negative; I have a HP-32II in front of me:

27
ENTER
+/-
3
xROOTy

yields (correclty) -3.0000, because the function x^3 is reversible. The result is obviously correct because -3*-3*-3 is -27. Try with, say -27.3 as operand and you'll get -3.0111, which is correct.

Now:

27
ENTER
+/-
3
1/x
y^x

returns "INVALID y^X", because the function a^x (with x real) admits only positive values for "a", and its graphic (similar to the exp case) shows this.

So it's simply not true that the [1/x] [y^x] sequence is equivalent to the [xROOTy] key.

-- Antonio


#61

Antonio, you will get the same result on a 42s unless you allow complex results. Is there a way to do this with a 33s?

BTW, I count on y^(1/x) being identical to xROOT(y) - else you force me to read my old math books again ;-)

Edited: 11 June 2007, 12:32 p.m.

#62

Hi, Antonio and Walter --

Antonio seems to have replied to the wrong post, but I'll address it anyway.

This topic was discussed more than a year ago, but I haven't made the effort to look up the thread.

"x_root_y" is not absolutely essential, but is somewhat useful, as it allows the real-valued odd-integer root of a negative real value to be calculated without cumbersome special code to ascertain odd integer x and negative y. (Checking for MOD(x, 2) = 1 is probably the best way to verify an odd-valued integer.)

"x_root_y" is also sometimes a bit more accurate than [x< >y][+/-][x< >y][1/x][y^x][+/-], probably because full internal precision is maintained throughout, rather than rounding the result of [1/x] to 12 digits.

In the strictest mathematical sense, the "correct" cube root of -27 is 1.50(1 + i*sqrt(3)) = (1.50 + i2.59808) = 3.00 @ 60 degrees. This is the "primary" root, which has the lowest angle on the complex plane, measured counterclockwise from the positive real axis.

This can be obtained on the HP-33S (or HP-32S/SII) by the following procedure:

0
ENTER
-27
ENTER
0
ENTER
3
CMPLX1/x
CMPLXyx

CMPLX"x_root_y" is not supported.

-- KS

#63

Quote:
Agreed (that cursor keys) can be put to good use, but of course a basic scientific calculator doesn't need any of those functions, now, does it? :->

But the HP-35s would be more than a basic scientific calculator -- more like an intermediate model. The TI-30X and the original HP-35 are basic scientific calculators, but in the case of the 35, basic is all there was...

Quote:
And before anyone mentions it, ENG mode is *not* the same thing as an ENG key. Having everything always display in ENG mode can be incredibly annoying, and I always want the option of pressing the ENG key when needed. Simple and effective, the way it should be.

I think what you mean is that ENG should function as implemented by Casio for years, allowing a value to be displayed to the nearest ENG exponent using a single keystroke, and to have that exponent shifted using one or two keystrokes. The "=" key cancels ENG mode.

My 1981 Casio fx-3600P and 2005 fx-115MS operate in exactly that fashion, which I also find convenient. However, both are algebraic and non-programmable, and perhaps "therein lies the rub." IF ENG is treated only as a temporary display mode (like SHOW), it could be nonprogrammable. Maybe SHOW (in addition to FIX/SCI/ALL) could cancel ENG, and the left and right cursor arrows could provide ENG-> and <-ENG (if not used for matrix navigation)...

-- KS


#64

Quote:
But the HP-35s would be more than a basic scientific calculator -- more like an intermediate model. The TI-30X and the original HP-35 are basic scientific calculators, but in the case of the 35, basic is all there was...

What I mean by "basic scientific" is non-programmable.
Non-programmable scientifics range from the basic early ones like the HP35 through to say the Casio FX-991ES which is about as advanced as you can get without having programmability.
I call them all "basic scientific" calc, although some are obviously a lot more advanced than others and can have a solver etc.

But once that programability option goes in then you have to start dedicating keys and having menus etc. That to me becomes an "advanced" scientific calc.

How do others classify and categorise various calcs like this?

Quote:
I think what you mean is that ENG should function as implemented by Casio for years, allowing a value to be displayed to the nearest ENG exponent using a single keystroke, and to have that exponent shifted using one or two keystrokes. The "=" key cancels ENG mode.

Yes, I'm talking about the Casio's, I love the way they implement the ENG mode, to me it's just perfect. I've never liked the ENG "mode" used on the HP's.

Dave.

#65

The cursor keys in addition to the backspace are extremely useful when you need to edit an equation or if you are in (gasp) algebraic mode. This has always been one of the most serious shortcomings of the 33s.

#66

Quote:
...and a log to base "b" is introduced

Instead of b, bx and LOGb I would suggest the most used LOG and 10x functions are kept and
logxy
is introduced. The latter will require two arguments but will be used only occasionally by just a few users.

Regards,

Gerson.

Edited: 10 June 2007, 8:15 a.m.


#67

So these 3 keys may look like this then:

With b = 10 as startup default you will reach the same with the previous function set, and only experienced users will change b at all. I leave it to the forum to decide.

Edited: 11 June 2007, 1:54 a.m.

#68

Much better.

But I have one suggestion:

"i" is still wasting a primary, and the equation mode is still buried. Do the following:

the "i" keycomes "EQN" as primary, and the --> key becomes SOLVE as primary. The --> and the "i" go to secondary positions on the 0 and the comma key. Additionally, why is the comma primary?

And what is the --> key for, anyway?

Edited: 10 June 2007, 10:19 a.m.


#69

Hi, Bill --

Quote:
"i" is still wasting a primary, and the equation mode is still buried.

I disagree, and believe that HP's concept should be retained. Those of us who use complex numbers regularly would appreciate having i unshifted, with the angle symbol blue-shifted on the same key. I also believe that a menu of functions germane to complex numbers (not ARG, which I believe is like ATAN2) should be the yellow-shifted function for the key. Please see the following post from 2004:

"User-friendly complex numbers" at
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv014.cgi?read=63415#63415

Quote:
And what is the --> key for, anyway?

My sentiments exactly!

-- KS


#70

Bill & Karl,

Quote:
And what is the --> key for, anyway?

Well, you could have read in my earlier post:
Quote:
an unshifted -> is introduced to allow double use of mode keys also for conversions (e.g. blue-shift 6 sets hexadecimal mode, while -> blue-shift 6 converts to hex display); this makes the menu BASE obsolete

This shall work for the keys "4" through "9".

@Karl, I second you keeping "i" as primary key. BTW, I'm slowly approaching the state where I can quote my older posts, too ;-)

@Bill

Quote:
why is the comma primary?

So far, the decimal separator has been primary always :-) Of course you may also set the calc to use a point as decimal separator. However, the layout chosen has an advantage in alpha mode: there is a point available already via "f" "*" (the function DOT outside alpha), so we save one key by using the comma as default decimal separator. And AFAIK it's the default for the majority anyway.

Edited: 10 June 2007, 6:31 p.m.

#71

I think OFF should be shifted yellow. Swap the shift colors for OFF and clear sigma.


#72

Why? (remember the 4-year-old ;-)


#73

well, if I was a 4-year old too, I would just pout and say I want it that way ....

But, in a more measured tone, I would say that's the way it is (not necessarily the color code, but the functionality of the key immediately above the ON button) on at least the 32, 33S, 42S, and 48 calcs. Plus, it seems most obvious and convenient to turn it OFF by pressing the button immediately above the ON, rather than the button two rows up.


#74

:-D OK, Dave, you can have it! FYI, I made it blue because it shall invert the primary function of this key (like blue SIGMA-, x^2, R_UP, ASIN etc.). I did not have any yellow inversions of primary functions so far. But, for sake of convenience, and to prevent you from pouting... ;-)


#75

Thanks. Now if HP would only make it your/our/this way!

Interesting point about inverse functionality, though.


#76

So this part of the calc may look like this now: also following an advise of Brian (see here. Hugh, this shall be interesting for you!).


Edited: 12 June 2007, 8:53 a.m.

#77

... there is an alternative design:

Everything else is the sum of all the contributions above, wherever there were good arguments or a clear majority.


#78

Walter,

Either version is fine with me!

Just remember to send the idea(s) to your HP insiders

;^)

Ren

dona nobis pacem

#79

Actually, I don't see the need of wasting a key position just to allow for logs of other bases, since this easily accomplished, either manually or by means of a very short program. Anyway, if this is really needed, the proposed solution might lead the user to errors in case the base (b) is accidentally changed. On the HP-33S it's quite common /c is involuntarily changed from its default value (4096).

Regards,

Gerson.


#80

Gerson, I count your post as a vote for this:

Looks a bit cluttered. However, I can make an exception (I hate exceptions ;-) and put the golden LOGxy function below of the key. BTW, for sake of consistency, this function shall be LOGy x instead of LOGx y.

The respective line of keys may now look like this:

Your votes, please 8)

Edited: 14 June 2007, 2:46 a.m.


#81

Hello Walter,

Quote:
BTW, for sake of consistency, this function shall be LOGy x instead of LOGx y.

For the sake of consistency, I would keep LOGx y. In yx, for instance, first we enter the number and then the power to which we want it raised. Likewise, entering the number and base, in this order, appears to be more natural to me. Another matter for voting :-)

In order to avoid cluttering and exceptions you dislike, what about using lower case letters for those functions? I mean log, ln and logx y.

Best regards,

Gerson.


#82

Ola, Gerson,

Quote:
I would keep LOGx y. In yx, for instance, first we enter the number and then the power to which we want it raised. Likewise, entering the number and base, in this order, appears to be more natural to me.
I can argue for LOGy x the same way: first enter the base and then the number which should be analyzed versus this base.
Quote:
Another matter for voting :-)

Agreed.
Quote:
... what about using lower case letters for those functions?

Fine idea. You can see it above already. However, many people seem to be inclined (?) to CAPITALS. Therefore, I confined lower case to variables and to modes so far (needing the space in Rec Pol and Deg Rad, and writing Cx Re consistently). And I started to use it for functions on the keyplate to indicate the difference to menus. And even lower case logx y will not fit nicely on the key.
Quote:
I mean log, ln and logx y.

IF the universal logarithm shall be called logx y or logy x (which is perfectly OK for me) THEN the decadic logarithm can't be named LOG again but needs a different name, e.g. LG as proposed above.

Com os melhores cumprimentos,

Walter

Edited: 16 June 2007, 11:05 a.m.

#83

"I feel, I feel, I feel ...
I feel like a morning star"

Sorry -- it just reminded me of that. (Those who don't get the reference can Google it.)

Seriously, I've decided to wait until I have my own HP-35s before expressing detailed suggestions for improvement. I probably won't depend upon a "surprise in San Diego"; that way, I'll arrive at HHC with well-formulated constructive input, both suggestions and criticisms.

In addition to the implementation of complex numbers, the main mystery to me is how the "800+ registers" would be addressed, and whether and how two-character variables are supported without an ALPHA mode. Any major improvements would have to wait for an HP-35sii or HP-45s to be developed. Take note: The "morning star" would be great for matrix editing.

-- KS

Edited: 16 June 2007, 2:09 a.m.

#84

As many of you already suggested, such a calculator will be more useful and really fun with a slightly bigger display. Keeping the given outer dimensions it may look like this:

With the coarse resolution of the 33s/35s this would allow for 4 lines. Using something a bit more advanced, 6 lines are possible with a font still perfectly readable for the true old fans of HP. And their (hopefully many) coming successors could go up to 8 lines easily. With a good display resolution the number of lines may be user-settable within a given interval. Either way the cursors make more sense than before.

For the logs: I think this version keeping b, bx, LOGb is still better than the alternative with the universal LOG, 10x, LG for at least 2 reasons:

a) easier error correction,

b) I expect users to stay with a specific base for some calculations, so it's more efficient setting the base once instead of having to enter it for each and every operation.

And for those folks who will never need any other base than 10? They will just stay with the startup default of b = 10 and never change it :-)

As usual, do not spare with comments and critics! 8)

(Edited to update the picture: LCD + cursor print)


Edited: 21 June 2007, 3:27 a.m. after one or more responses were posted


#85

Quote:
As usual, do not spare with comments and critics! 8)
Well, I was happy with the versions without the weird cursor keys, and have been completely uninterested since you put them back in.

#86

Well, I did not find your vote in this thread above. Anyway, you may take the top 2 lines of keys as they were in Version 1.1 and regain your interest :-) And I recommend you find some more votes for this, else I fear you will not end as majority leader ;-)


#87

Quote:
Well, I did not find your vote in this thread above.

Firstly, I'm just stating a preference, not "voting." I don't see the point of a vote, since I highly doubt the result will have any influence on what HP does. Secondly, I'd already stated my preference earlier, and saw no need to repeat it. The rest of the discussion has centered on details that don't matter much to me. But for the record, all of your versions that have "normal" cursor keys are acceptable to me, and none of the versions with the "diamond" pattern keys are.

Quote:
And I recommend you find some more votes for this, else I fear you will not end as majority leader ;-)

As I said, I don't see the point of voting, because I expect that even 100% agreement here will make no difference in the outcome.


#88

So there's no mercy even for black diamonds ;-) So far I assumed you had a color problem with the 35s's diamond, but now I know. Thanks for your explanation.

Ref. to "voting": I know HP is free to (and will presumably not) take anything from these proposals, regardless how many forum members stand behind them (free of charge, anyway). Nevertheless, I'm interested in the opinion of the educated majority of these members. And I make it visible, so we can talk about a concrete model instead of pondering wishlists which's combined effects blast any design. IMO we had too many easy wishlists in the past, and very few people made a serious effort to check them at least by modelling. As people say here: "The devil hides in detail!"

Thanks again for making your opinion clear.


#89

Quote:
So far I assumed you had a color problem with the 35s's diamond, but now I know.

Actually, it was both color and shape.

#90

I love the big display.

Now give it graphing features and a CAS, plus a thoroughly revised and evolved RPN programming environment, and we'll have the flagship RPN calculator that HP would have produced by now if it had never taken the left turn into RPL. 8)

Regards,
Howard


#91

Thanks, Howard!

Quote:
Now give it graphing features and a CAS

Please remember this keyboard is fully occupied already. So, if you want any additional functions, you have to kick out as many others (see above). Please specify your wishes, I cannot do it for you. IMO, for a real flagship a third shift key may be inevitable.

Edited to correct an ambiguity.

Edited: 20 June 2007, 5:00 p.m.

#92

FYI, the picture in my earlier post is updated to show how a 7 line display may look like. Luiz's 42s-font was used (thanks!).

IMO such a display is still clearly readable in full natural size, even above 50 :-) Principally, as long as you can read the yellow and blue print, you shall be able to read this display as well. And you may set the LCD to 6, 5, and 4 lines for bigger letters, but I cannot give you a bigger print ;-)


Edited: 21 June 2007, 5:11 p.m.

#93

Hallo Walter!
It looks great.

But what do you think about colouring the Number-Keys and the Arithmetic-Keys in a other colour like older HP (Spice, Woodstock, Champion and so on).

I think this will give it a kick, doesn’t it?

Thanks Jürgen

#94

Hi;

all 'n all we want either to present our best shot about what we call the best calculator ever or we are getting back to rebuilt an old HP successful calculator. Many attempts to get such combinations failled so far, and HP (the ACTUAL HP) offered a bunch of solutions that worked in many areas.

From time to time in this forum we experience the chance to answer to surveys about our favourite HP calculator (In some surveys, our favourite calculator despite brand). Well, if there is such unique model that satisfies all needs, no survey is needed, right?

When the HP28C was introduced, I felt as if HP was ahead of time, showing us what we WOULD need and never considered possible. The whole 48/49/50 series was a consequence of such vision. At that time, I wrote a book about such changes ('Da HP41 para a HP48', or 'From the HP41 to the HP48') and I intended to guide former RPN users into the new RPL usage/programming paradigm.

I feel as if the message has been lost, though. Not that I disagree with the RPN-driven models, I actually believe RPN helps students to understand calculus subjects. It is something like the abacus users when they actually see it (as a visual reference) instead of touching the device itself; reasoning over the RPN stack and operating structure seems to force the brain to understand the problem and find a solution in a different perspective. I remember that some classmates of mine actually got the main ideas in the many subjects after trying them out with an RPN calculator. Creating a program to solve the problems was a passage to understanding both problem and solution. I actually experienced this.

Did not have the chance to go ahead with RPL, because I was no longer a student when the HP48SX was introduced. The HP28 was not so often seen, but I remember joining a group of about ten students with HP28C´s that wanted to know how to use it the best way possible in algebraic environment. My only and last HP28C class...

Time has gone and now HP seems to listen to us, users, prior to introduce a new product. Interesting, we all had to wait till HP decided to show us what it was hiding that was astonishing, new and desirable. Now we are trying to get from HP what we believe is mostly important, and there are many close, yet different proposals. If I'm there I'd be confused, because it seems to me that HP would need to offer some different versions of the HP35S to satisfy some of us. I myself would add a few suggestions, but based on the gigantic thread that followed the first appearence of the HP35S in this forum, I felt I'd not add significant contribution, though.

Please, read this as my private view of the rare chance we are (supposedly?) having to interfere in HP's choice of features to be offered in a new calculator. If there are so too many suggestions, maybe none of them will be caught, neither considered, and frustration may arise after the sensation that we were not heard after all.

I'd not be surprised if the HP35S is introduced exactly as it was shown in the first place, no changes applied. And we know what happened with the first HP12C Platinum and also with the HP49G+. Both carried many problems and had a negative impact, being replaced by their new versions.

I think we should support fewer suggestions, only one if possible, a consensual proposal. I do not know which one, because I have my own needs for a personal computing device, but I'd surely try to support the one with more people supporting.

I just want a new, good-looking, reliable, resourcefull RPN calculator. It seems to me that the HP35S will match my needs.

Sorry writing too much... (did not spell check, though)

Suggestions? Flames? Blames? Etc?

Cheers.

Luiz (Brazil)

Edited: 7 June 2007, 11:01 a.m.


#95

Quote:
.. I remember that some classmates of mine actually got the main ideas in the many subjects after trying them out with an RPN calculator. Creating a program to solve the problems was a passage to understanding both problem and solution. I actually experienced this.

I did too, in a different sphere. For several years after learning to program a computer, I would turn to the 41 to help me understand abstract problems that came up. These problems would entail algorithms I hadn't seen before, or new constructs in a new language or - especially - algorithms I would have to invent to solve a particular problem. In either case, I would code the problem in Forty One CAlculator Language (FOCAL) and the answers would become clear to me. Since it was the first device I ever programmed, the 41Cs user language was like my native tongue. And moving data around registers, manipulating stacks and so forth are very concrete activities. For one example, when I first encountered pointers in Pascal, and then in C, I thought to myslf "Aha! Indirect addressing."

Quote:
Time has gone and now HP seems to listen to us, users, prior to introduce a new product. Interesting, we all had to wait till HP decided to show us what it was hiding that was astonishing, new and desirable. Now we are trying to get from HP what we believe is mostly important, and there are many close, yet different proposals. ..

I think that "HP" is both listening to us and not listening to us.I believe that the first order of business for HP is to make a profit on the new machine. (It was a surprise to me to hear that the 33S led its market category in sales.) That means they need to listen to all their users. Thus we see features designed to increase acceptability with test administering bodies.

I have always thought - and I still do - that we enthusiasts represent a small segment of the pool of potential customers for HP's calculators. It's true that we tend to have strong opinions, and we like to think that others we come in contact with are positively influenced by those opinions, so it may be that we have a little more leverage in the marketplace than our numbers would indicate. But I don't think that's the reason why we are listened to at HP, if indeed we are.

So are we listened to? Well, we just saw a case and keyboard design that incorporated many elements that are near and dear to our hearts. We also know that Sam and Cyrille are aware of this community. So it's tempting to make the leap into believing that the 35s is the way it is due to our direct influence. But Sam, Cyrille and others at HP are not empty vessels to be filled up with our concepts of what is right and proper in an RPN machine. I think Sam listened to us, and to others, trying to garner good ideas. I think he heard some, but also that many of those ideas resonated with what he believed was right and proper. I think that he then turned around and applied his ideas to the real-world problem of producing a calculator that embodied them.

Quote:

..I'd not be surprised if the HP35S is introduced exactly as it was shown in the first place, no changes applied..


I think that's right, for two reasons. First and foremost, I think that it must be too late to add features to the design. Remember there would be software development and testing, as well as keyboard and case retooling to do, all some three months before the machine is scheduled to ship. Second, I think that what we see in the 35S is HP's best compromise between their idea if the ideal machine, and the realities of bringing such a product to market.

Quote:

I just want a new, good-looking, reliable, resourcefull RPN calculator. It seems to me that the HP35S will match my needs.


Once again, I think that is exactly right.

Regards,
Howard


#96

Howard

I agree. The proposed 35s appears to be a very well thought-out machine and I will certainly buy one. And, (OT), I would like to think Mr. Hurd perhaps, had something to do with this.

tm

#97

Nice work. It must have taken you a lot of time to do that!

However I must comment on it.

1. I do not think that your design matches the features which are likely to be included in the 35s. This is especially the case for the 41x type stuff you have included such as ASSIGN.

2. I feel that you have smothered one of the key features of the 32sii-333-35s family: equations. The controls for them are all shifted (except parentheses) and scattered all over the keyboard.

Considering that the equation feature, especially if improved with full editing, is a forward-looking feature with equal to superior algebraic handling to that of "algebraics" it really aught to take a primary role in the keyboard layout.

But your waork is very nice and I am grateful for the effort and the chance to criticize :-)


Possibly Related Threads...
Thread Author Replies Views Last Post
  Availability of scientific HPs Matt Agajanian 4 769 08-28-2013, 07:40 AM
Last Post: Eddie W. Shore
  JOT scientific calculator Richard Berler 4 857 08-21-2013, 05:25 PM
Last Post: Richard Ottosen (Denver, CO, USA)
  Our 11C Scientific Calculator Android app is on launch sale, 0.99$ only John 1 582 06-22-2013, 01:38 PM
Last Post: Marcus von Cube, Germany
  Vicinno 11C Scientific Calculator app is on Android now John 0 464 06-14-2013, 02:28 PM
Last Post: John
  Great news - Vicinno's HP 15C Scientific Calculator iPhone app is FREE now John 21 2,787 06-07-2013, 05:49 AM
Last Post: Mike (Stgt)
  HP35s scientific calculator GREG W THOMAS 4 832 03-22-2013, 06:49 AM
Last Post: Thomas Radtke
  Yet another new scientific HP calculator ! Mic 24 2,645 11-04-2012, 04:53 PM
Last Post: NateB
  This is not a calculator, It is an instrument Christopher Johnson 71 5,260 05-10-2012, 07:32 PM
Last Post: Nick R
  O.T. Sinclair Scientific Lars Bergström 14 1,828 04-08-2012, 07:51 AM
Last Post: ClausB
  Scientific version of 30b by HP? kc 26 2,742 02-13-2012, 07:24 PM
Last Post: db (martinez, ca.)

Forum Jump: