design "your" HP calculator



#48

I guess this topic has come up in the forum before, but anyway.
I would like people to suggest a design of their ideal HP calculator: sneak under the skin of an HP-designer/developer and work out the calc that would be ideal for you from the point of capabilities, looks, user interface etc.
My personal calculator would be a remake of the HP-41 series, primarely for their form factor, color scheme and it's exclusive yet still modern looks. Of course it would be an RPN only calculator. A disadvantage that I encounter from time to time when I work with the 41 is the fact that some of the functions of its total function set (which itself is somewhat limited according to modern standards) are accessible only by typing their names in full. I would ommit this problem partially by introducing a 2nd shift key (of course in blue) and write additional functions on the slopes of the keys instead of the alpha characters. The alpha characters would simply be printed to the right of their corresponding keys in the same (creamy) white colour as the main key-functions. I don't think this will result in an "overcrowded" keyboard, after all the 34C and the 67 also have 4 function for one key. W.r.t. the function set I would copy some if the functionality of the 15C such as solve and integration tools. Furthermore a complex mode, also like the 15C, but handling complex numbers rather in a 42S/48xx manner (displaying something like (0.0000;1.0000) rather than using the "hidden" imaginary stack of the 15C). Of course this requires a matrix dot display.
For the mathematical purist there would also be hyperbolics in a 15C fashion (just a simple HYP prefix). Furthermore ABS, INT, FRAC, N! (combined with gamma-function) and maybe calculation of combinations and permutations, all directly accessible via the blue shift key. Expandibility would be maintained, allowing for connection to a PC and modular extension of the function set. Some graphing functions can be added via a separate plug-in (matrix-dot) display approximately the size of the 41's card reader. You plug it in when you need it or leave it at home when you just need a calc for simpler calculations. Graphing functions accessible from the keyboard through a menu (invoked via blue shift key).
Well, this is my 1st suggestion for an ideal HP calculator. Everybody feel free to come with ideas of their own.


#49

Hi, Jan,

you start with a really nice description. How about a layout? So you can show where you place all the valuable functions you mentioned.

#50

I was thinking of something like the 28C, in terms of having a clamshell case, but having a landscape rather than portrait layout.

The "top" half would have the normal scientific function buttons (sin, cos, exp, ln, etc.), but would have an extra large screen, because this calculator would need to have a CAS.

The bottom half of the calculator could be a qwerty, 71B style keyboard.


I figure the clamshell design would take care of one problem of graphing calculators: The tend to be bulky. In this case, when not in use, its footprint is only half that of what is needed for the keyboard and screen.

I'd also hope that having some more scientific functions instantly available on the top half would overcome another weakness of graphing calculators; namely, needing to go through layers of menus to find needed functions. It's much easier to do certain things (linear regression being a good example) on a scientific than a graphing calculator; I'd like a calculator with the power of a grapher, and the convenience of a scientific.

I also think a "built-in" case has its advantages.

One thing I notice when I use an RPL calculator is that I frequently need to switch between two two different sets of soft keys. Say I'm writing a program. Well, then of course I might want the program soft keys. But I'm also writing a program to DO something -- I might want soft keys related to the functionality of the program (whether probability or calculus or whatever).

So, maybe two rows of soft keys under the screen would be good (assuming this doesn't make things too cluttered).


Some form of expandabiity (eg., adding extra RAM) would be good, since you don't want an expensive calculator to become quickly obsolete.

Beyond that, in my opinion, the more power, the better.


I think it would be reasonable to have (over the 50g) online help for every single function in the calculator's catalogue. In a dreamworld, I'd like a much more comprehensive equation library (something to rival math- and scienceworld).

#51

Hello!

Quote:
I guess this topic has come up in the forum before, but anyway.

Just search this forum for "Open RPN" :-)

Speaking for myself, if I had to design an ideal calculator today, it would be something like the iPhone ( http://www.iphone.org/ ) that Apple has presented earlier this week. Or better: "My" calculator would simply be an application that runs on the iPhone. A simple one with big, easy to hit, keys and basic functionality for everyday use (maybe like a HP-21) and a more complicated, programmable one for those who need it for work (whoever that may be, with notebook and/or desktop computers on everybodys desk...).

If it needed to be a hardware solution, "my" computer would be a modern version of the hp-25C. Same size (but only half the thickness of course), same layout, same keyboard, but with an additional row of assignable keys with appropriate labels on the display. For the display itself, it would be a backlit inverse LCD (or better still, an OLED display) that would mimick the red LEDs of the original hp-25 in normal mode, but with additional capabilities like alpha-prompts and function-key labels. I would leave the writing of more complex programs to an application that runs on a PC/Macintosh so as to keep the keyboard of the calculator uncluttered. This PC based application could contain a lot more functions than are accessible from the calculator keyboard so that everybodys needs would be taken care of. A flash-memory card could be used for storing the programs and data and for transfer between PC and calculator.

And finally, the whole thing must not cost more than 100 (or even 50?) Euros/Dollars, otherwise nobody would buy it ...

Greetings, Max

#52

Hello, Jan --

As Maximilian stated, these discussions have taken place in the past, usually under "OpenRPN".

Quote:
My personal calculator would be a remake of the HP-41 series, primarely for their form factor, color scheme and it's exclusive yet still modern looks.

Of course it would be an RPN only calculator.

A disadvantage that I encounter from time to time when I work with the 41 is the fact that some of the functions of its total function set (which itself is somewhat limited according to modern standards) are accessible only by typing their names in full. I would [address] this problem partially by introducing a 2nd shift key (of course in blue) and write additional functions on the slopes of the keys instead of the alpha characters.


I, too, find the HP41 aesthetically pleasing, but it's too chunky for modern tastes. Of course, it had to be, in order to accommodate the four ROM ports and four "N" cells or rechargable pack.

For non-graphing calc's without Equation Writer, I agree that RPN is the way to go. But I'm biased, and in the societal minority...

I agree with the concept of using two shift keys and beveled keys, but there would still be too many functions -- menus are needed, as well.

Quote:
The alpha characters would simply be printed to the right of their corresponding keys in the same (creamy) white colour as the main key-functions. I don't think this will result in an "overcrowded" keyboard, after all the 34C and the 67 also have 4 function for one key.

W.r.t. the function set I would copy some if the functionality of the 15C such as solve and integration tools.

Furthermore a complex mode, also like the 15C, but handling complex numbers rather in a 42S/48xx manner (displaying something like (0.0000;1.0000) rather than using the "hidden" imaginary stack of the 15C).

Of course this requires a matrix dot display.


We wouldn't want to use the HP-67 and HP-34C as an example. Those are non-alphanumeric, non-menued models. The alphanumerics of the HP-41 were developed in part to eliminate the clutteredness of the HP-67.

The successor HP-32S, HP-32SII, and HP-33S have SOLVE and INTEG, which of course should be available. What was missing from those models that the HP-15C had was matrix functionality, as well as complete and convenient complex-number math.

Speaking of complex numbers, I think you mean the ability to handle complex numbers as a single object instead of two separate parts. The 42S' display of complex numbers is ideal -- different and better than the parenthesized-pair of the HP-48 you cited. E.g.,

2.71818 -i3.14159
instead of
(2.71818, -3.14159)
For more, please read my post from yesterday ("P->R and complex numbers"). It includes a link to an archived post of mine on the topic of complex numbers.

Absolutely -- the dot-matrix display makes possible the alphanumerics, meaningful prompts and error messages, and menus.

Quote:
For the mathematical purist there would also be hyperbolics in a 15C fashion (just a simple HYP prefix).

Furthermore ABS, INT, FRAC, N! (combined with gamma-function) and maybe calculation of combinations and permutations, all directly accessible via the blue shift key.


That's how it was on the HP-15C, by necessity of its limited 7-segment display. All of these functions (except perhaps x!, which combines n! & gamma) are seldom used, or are more useful in programming.

The HP-32S and HP-32SII put most of these into menus. Annoyingly, the HP-42S lacked positions or menus for the hyperbolics, even though the HP-27S had a hyperbolic-function menu and the same display.

Quote:
Expandibility would be maintained, allowing for connection to a PC and modular extension of the function set. Some graphing functions can be added via a separate plug-in (matrix-dot) display approximately the size of the 41's card reader. You plug it in when you need it or leave it at home when you just need a calc for simpler calculations. Graphing functions accessible from the keyboard through a menu (invoked via blue shift key)

Well, this is my 1st suggestion for an ideal HP calculator. Everybody feel free to come with ideas of their own.


It's fine to develop a concept that includes everything one finds "cool", but a target should be chosen and boundaries must be imposed.

I'd find the development of a brand-new, high-end calculator of limited value. The HP-50 and TI-89 are already out there and well-established. As PDA's continue to improve, I'd say the best platform for a new super-capable calculator is downloadable software for a PDA. Take advantage of the speed, RAM, and multi-color, high-resolution LCD.


Still, I'd still like to see the perfect non-graphing calculator that is easy to use and which combines the best functionality and attributes of the HP-15C and the HP-32SII, but adding more speed, RAM, and excellent complex-number math -- in short, what the HP-33S ought to have been.

My concept would start with the HP-32SII paradigm, with each letter of the alphabet serving as a variable name or label. The HP-32SII also includes equations and fraction math (not included on the HP-32S from which it was derived). The Pioneer-series models -- even inside their slipcover -- fit inside a man's shirt pocket, although they protrude slightly.

  • A longer (and higher-resolution) LCD unit would accomodate at least 18 large characters -- enough for both components of a complex number, longer equations, and more menu items.

  • Full matrix capability with complex-valued entries and easy editing (a la HP-42S) would be included.

  • Complex numbers would be easy to enter in either rectangular or polar coordinates: e.g., 3.14159[i][+/-]2.71828 or 6.3[shift][angle]30. A menu of complex-number operations would be provided (real, imag, conj, R->C, C->R, P->R, R->P).

  • The equation editor would include insert/delete editing, implicit multiplication, and syntax checking.

  • The keyboard and menus would be well-organized, unlike those of the HP-33S and HP-32SII (to a lesser extent), which evolved from the HP-32S' menu-driven concept.

All of this would probably require a landscape layout, like the Voyagers. Size and shape would be like the Pioneers, to accommodate the wider keys for numbers and arithmetic. The landscape layout is better-suited for desktop use than for handheld use. This is good for students in class -- and for others whose computer keyboard tray isn't pulled out...

Some unresolved issues: How would a matrix be distinguished from a letter variable? How would "SHOW" work for a complex number using a one-line display? It's certainly not easy to develop a complete functional specification for a commercial-grade product!

-- KS


Edited: 14 Jan 2007, 11:32 p.m.

#53

Hmmm, I was thinking more on the lines of a chip implanted in my brain that would enhance my maths capabilities. Perhaps it could
interface with an optic nerve for display. The keypad could be
implanted into my fingertips or receive input through the aural canal or thoughts. Software upgrades could be accomplished visually or by a low RF transmitter. If RF is used additional memory could be accessed via Bluetooth.

#54

When calculators first appeared, we were happy to have a hand held programmable device with 8 to 10 memory registers that performed simple common tasks. As new models were introduced, we had the privilege of more memory, more programming space, and more functions. Modules added more memory, functions, and code. I think there is a natural capacity of memory registers, program steps, and functions that hand held calculators can handle efficiently. I think we are currently well beyond that limit. To me, Excel is the new kind of calculator that suites my needs. I can see several columns, rows, and tables of data. I can use VBA to create program that further automate my tasks. Whenever I want to test some calculations or algorithm, I use Excel and not an HP or TI calculator. I think the spreadsheet has sentenced the calculator to the inevitable oblivion.

Namir


#55

Don't tell that tale to too many folks, Namir. It will destroy the value of your collection! 8)

Seriously, I think the handheld format has life in it, and some of that is in calculation. I agree Excel is a powerful tool, and can replace a good deal of what we used to use calculators for. However Pocket Excel sucks wind for those kinds of tasks. Excel also lacks a CAS, though its graphing capability is highly developed.

I don't think anyone has figured out the "ultimate" handheld device yet. For this calculator geek, Qonos was pretty close to the mark. They had the goals of extended battery life, calculating power and keyboard excellence firmly in view. The hybrid nature of the device was nothing short of brilliant. The failure to get commercial backing for that project was a cold shower to many of us.

On the other hand, Eric's hobbyist device shows that such an animal is possible. Devices like the MLDL2000 and Clonix also show that there is room for successful calculator hardware projects, and that it doesn't take a corporation's budget to gain modest success among fellow enthusiasts.

I think these projects could serve as a bridge to the time when desktop hardware design and manufacturing become possible, a time that may be closer than you think. Then hobbyists may start addressing niche markets the nature of which is hard to predict. But here's hoping cool calculators are among them!

Regards,
Howard


#56

Howard,

Perhaps positioning the calculator as an extended tool for the PC may work. In HHC2006 I told the HP team that it would be nice to be able to transfer data easily from (say Excel) to the calculator and then perform data processing (regression, basic statistics, and so on). Wireless transfer would be very nice, but using SD card is still acceptable. The main point is to have a piece of software that performs a two-way conversion between Excel tables and HP calculator matrices.

Namir


#57

Or rather in the Open Document Format ;-)


#58

Quote:
Or rather in the Open Document Format ;-)

Hah! See below. I posted that before I saw your contribution. 8)

Regards
Howard

#59

I think your idea regarding the calculator as an extension of the spreadsheet (for a moment, let's pretend the Redmondium Empire won't own all our mathematical data in 10 years) is sound. But I don't think that's going to be the whole story. Why don't we all use PCs for phone calls? The PC, while a powerful general purpose computing platform is not convenient as a voice conversation endpoint, at least for most of us. Given that people want mobile communications, the cellular (or wireless) telephone will be a long-lasting, if not permanent feature of most peoples lives in both the developed and developing worlds.

Since that's true, a whole lot of other applications find an opportunity to "piggy-back" on the mobile phone's ubiquity. Games are popular, and a lucrative market for software developers. In the last couple of years, mobile email and text messaging have become popular in the US, though these applications haven't yet achieved the success they have in Asia and Europe. More to the point, capable software calculators on these devices are also popular, despite basic calculator functionality being bundled in to the various platforms.

So there is demand for stand-alone calculator functionality in a hand-held device. Interoperability with desktop apps can only add to the appeal of such systems. XML data standards should make this more or less automatic going forward, except that Microsoft's XML standard may be too complicated for a mobile device. This is an area where ODF probably has an advantage.

But alas, none of these developments particularly drives vendors to deliver mobility with a nice calculator keyboard. Touch screens are a poor compromise for this, I believe, even though thumb-optimized systems work pretty well for other types of input.

So here's my idea for the ideal (calculator) mobile device. It consists of a CPU, RAM, nonvolatile memory and various buss connections in a package no more than 2 inches square and a half inch deep. This processing module is built on the latest Silicon fab technology, and packs the optimum compute cycles and memory capacity with the best power management profile possible. This module can then be dropped into a variety of shells with various application-specific forms. These shells can contain additional memory, radios, math coprocessing and what have you. I can now shop for the shell(s) that best meet(s) my mobile computing needs. And I get enhanced control over my mobile life. This last important point is why this prospect won't appeal to folks like the cellular network providers. So there are significant barriers to getting something like this off the ground.

However, what if you took eight of these modules and dropped them into a server chassis, or a home entertainment center? What if they were completely general? The economies of scale in their production would be enormous. I can dream, can't I? 8)

Regards,
Howard

#60

My perfect calculator is pretty simple: an HP-12C with trig, log10, R<->P, H<->H.MS, backspace, lots more memory, and 11C programming. "User" keys would be nice, too.

A souped-up and programmable HP-27 packaged as a Voyager, basically, and they could do it with a little more silkscreening and a firmware update. Some of us don't need that much calculator!


#61

Ooo-la-la!

That's exactly my need! I only add some solving capabilities, that could easily be added, I guess!

-- Antonio

Edited: 15 Jan 2007, 5:24 a.m.


#62

IMHO, one does not need 2300 built-in functions; a large number of functions is a clutter and a distraction. A calculator should have a small number of functions built-in, and the possibility to seamlessly add new functions by downloading them from the manufacturers' web site. Once added, the new functions should be undistinguishable from the built-ins. Keep it small and simple.

#63

Quote:
My perfect calculator is pretty simple: an HP-12C with trig, log10, R<->P, H<->H.MS, backspace, lots more memory, and 11C programming. "User" keys would be nice, too.

Perhaps something like this?

(from a layout started by Antonio:http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv016.cgi?read=95249 )

In run mode DRG would toggle among angle modes. In Program Mode a parameter would be required (DRG 0 for DEG, DRG 1 for RAD and DRG 3 for GRAD). This idea could be extended to BEG/END and D.MY/M.DY (Toggling between the settings in Run Mode and working with a parameter in Program Mode, thus freeing two more positions). DSP would work the same way (DSP 0 n for ALL, DSP 1 n for SCI, DSP 2 n for ENG and DSP 3 n for FIX, n=0..9).

HYP would work like in the HP-32S (except that SIN is not a shifted function on the 32S): f HYP f SIN for SINH and f HYP g SIN-1 for ASINH. DSE and ISG would work like in the HP-15C or HP-41 (no dedicated indirect register).

TEST would work like in the HP-15C, except that two of the logical tests would to be left out to give room to the basic x<=y and x=0.

I'd buy two or three of them!


Edited: 15 Jan 2007, 4:47 p.m.


#64

OK, maybe we need a luckier number than 13, but that's PERFECT.

I'd buy more than one for sure.


#65

There was never an HP-14C (though there was an HP-14B). But it could be named just HP-12CE (Enhanced), as it's been suggested.

Seen Apollo XXIII? :-)

As they say in Spain "Yo no creo en brujas, pero que las hay, las hay!". ("I don't believe in witches but exist they do", in a translation supplied by Valentin).


Edited: 15 Jan 2007, 5:25 p.m.


#66

Hello!

Quote:
As they say in Spain "Yo no creo en brujas, pero que las hay, las hay!". ("I don't believe in witches but exist they do", in a translation supplied by Valentin).

Interestingly, for many years the Gotthard Road Tunnel in Switzerland (with nearly 17km the second longest road tunnel worldwide) initially had no "km 13" sign, but two "km 12" signs in sequence. I know for sure because I drove through it over 100 times. Later, someone decided to put an end to this nonsense and a "km 13" sign was installed.

Regarding the HP-13/14 calculator itself: I would also take one, but not with the focus on financial functions as it is suggested here (I need none of them, what the bank prints on my account statements is enough for me :-( ).
Maybe an interchangeable keyboard (like the one on my Apple PowerBook) would be a solution...

Greetings, Max

Edited: 16 Jan 2007, 10:01 a.m.


#67

Hello Max,

Quote:
Later, someone decided to put an end to this nonsense and a "km 13" sign was installed.

I've heard there are buildings with no 13th floor, the 12th beeing followed by the 14th. I am not sure this is true though.

Quote:
Regarding the HP-13/14 calculator itself: I would also take one, but not with the focus on financial functions as it is suggested here (I need none of them, what the bank prints on my account statements is enough for me :-( ).

I don't need the financial functions either. Actually my knowlegde of them doesn't go beyond the TVM functions. The key idea here is the following: if the only remnant Voyager is the 12C it would be nice if it at least the trigonometric functions were available. The lack of these functions made sense when the HP-12C was introduced, as this would have made the calculator needlessly expensive. I don't know the innards of the present 12C and 12C Platinum, but I guess there might be some unused ROM space that could be filled up. Of course these functions have no application in Business activities but they might be useful to Business students during their first college year. And it could be also useful to some of us who'd like to have a cheap brand-new usable Voyager calculator today :-)

Regards,

Gerson.


#68

There is more witchcraft in this modern world:

In most airplane cabins, you will look for the seats of line 13 in vain. And many of the ICE express trains in Germany do not carry a coach 13, but have coach 12 followed by 14.

As you said, Gerson, one never knows... d;-)


#69

Quote:
In most airplane cabins, you will look for the seats of line 13 in vain. And many of the ICE express trains in Germany do not carry a coach 13, but have coach 12 followed by 14.

Sounds like you need to check out http://www.seatguru.com for more information on this. As far as I can see each airline has a row 13.

Cheers,

Steve

#70

Yes that is true about 13th floors. The hotels I was in in Hawaii this year had no 13th.

The shipyard in Germany that I worked with for 2 months back in 1998 had a launching coming up, and the work flow put it on Friday the 13th of November. Well, that wasn't a good idea so they moved it up to Wed the 11th. The customer was P&O and I'm sure that they would have been quite alarmed, had their new feeder ship been launched on such a day!




Edited: 17 Jan 2007, 8:32 a.m.

#71

The building I currently work in has no 13th floor; neither did the apartment building I once lived in. I think it's quite common - maybe it's just a Canadian thing? I hope not...


#72

Hello!

Quote:
The building I currently work in has no 13th floor; neither did the apartment building I once lived in. I think it's quite common - maybe it's just a Canadian thing? I hope not...

No no, you Canadians are not alone with this :-) But its not as bad in Germany, I think. I lived in a house with number "13" for many years and I have never notices that 13 is left out in hotel rooms or appartment numbers.

Interestingly, there a slight differences among European countries: In Italy, where I lived for 20 years, the real bad day is not Friday 13th, but Friday 17th. I once made the mistake to ask a lady "What day is it?" and she looked at me with horror and said: "Now you have ruined my whole day!". When I asked her why, she replied: "Because you reminded me!" That was about the worst outbreak of superstition I have ever encountered...

Greetings, Max

NB: I would be happy enough with an HP-13 :-)


#73

Some time ago (like 400 years) "13" was a holy number, because there were 13 persons at Jesus' last supper. Maybe that is the reason why Italy doesn't fear these black Fridays.


#74

Quote:
Some time ago (like 400 years) "13" was a holy number, because there were 13 persons at Jesus' last supper. Maybe that is the reason why Italy doesn't fear these black Fridays.

That's probably why we (italians) do not like to have 13 people sit around a table for a dinner...

I personally don't care about but you'll find many who really do!

Massimo

#75

Quote:
OK, maybe we need a luckier number than 13, but that's PERFECT.

I don't think I'd go that far, it has PI on the keyboard twice :)

Still it does look pretty neat.

- Pauli


#76

Oops! And I had removed Integrate because I couldn't find a free position. I'll fix that. Unless someone prefer the ALG/RPN key :-) (But then we'd need a pair of parentheses)

Thanks,

Gerson.

Edited: 15 Jan 2007, 5:32 p.m.


#77

Of course, it HAS to have algebraic mode and parentheses.

Really, today's market is such that an RPN only calculator just isn't going to be produced and sold.

We may not like it, but if we can get RPN at the expense of an algebraic mode and parentheses keys, we should be glad.

So, find room for ALG and RPN functions and an open and close parenthesis. Makes you give up 4 key positions, but that's the price to play the game.

Gene


#78

Since ALG and RPN are not programmable, the mode selection could be done by means of an ON key operation, like ON/ENTER. So we actually need two positions. And the canditates to be left aside are SOLVE and INTEGRATE. A real pity, but we still have Valentin's Gaussian Quadrature for integrate (or Les Wright's version). Ideally, the parentheses keys should be unshifted. This is starting to get complicated...

You are absolutely right. There's no market for an RPN-only calculator nowadays.

Gerson.


#79

Quote:
The mode selection could be done by means of an ON key operation, like ON/ENTER. So we actually need two positions. And the canditates to be left aside are SOLVE and INTEGRATE. A real pity, but we still have Valentin's Gaussian Quadrature for integrate (or Les Wright's version). Ideally, the parentheses keys should be unshifted.

Not that complex, indeed: as in ALG mode the X<>Y and RDN keys are useless, you can overload them as ( and )... Only the keyboard will look quite cluttered...


#80

Quote:
in ALG mode the X<>Y and RDN keys are useless

I'd say the parentheses are useless in RPN mode. I have noticed the x<>y and Rv keys are shifted on the HP-17BII but in RPN mode they are accessible withouth pressing the shift key. However these keys appear to have a function when in ALG mode (history of past results). This is an option though.

#81

Quote:
Since ALG and RPN are not programmable...

This would only be true if the 13C were to slavishly duplicate one of the dumbest design decisions in HP calculator history.

By adding algebraic mode to the 12CP, HP effectively created a machine that had two different programming languages which happened to be representable by the same 'alphabet' of commands. In doing so they introduced the potential to run an RPN program in algebraic mode, or vice versa, which would then produce the wrong results. Unfortunately, they omitted a way to designate a program as 'RPN mode' or 'algebraic mode', and also omitted to allow the mode-switching commands to be included in a program.

Consequently, it becomes the user's job to deal with this, which is just the wrong way to handle the problem, since there is no benefit to the user in being able to run a program in the wrong mode and get the wrong answer.

So, taking this thread at face value as a design exercise, I would strongly encourage that the machine's programmability be tarted up to allow every program to be designated as RPN or algebraic, that the default be whatever mode the machine was in at the time the program was entered, that every program automatically runs in its associated mode, and that the user's choice of operating mode is not altered by running a program.

That way, there would be no penalty for users to enter RPN programs from books or other sources while continuing to use the machine in algebraic mode for their everyday calculations. Further, if the programmability of the machine allowed it, RPN-mode programs could call algebraic ones, and vice versa, without everything going horribly wrong.


#82

Interestingly, I have an early 12cp that allows ALG and RPN commands in a program.

:-)


#83

ALG and RPN are programmable on my 12CP 25th Ann. Edition. Somehow I had forgotten this...

Gerson.

#84

ALG and RPN are programmable on the HP-12C Platinum indeed. I just had forgotten about that. But one thing that stopped me from using this feature is exactly the inability to switch to the default mode in case of an abnormal program exit, as due to an error condition. This might confuse the user despite the change in the mode annunciator.

The way you have suggested is nice (the inclusion of only one mode command and automatic switch to the user's default mode, if I have understood correctly). But to programs would have to be located in different areas, as in the HP-41/42S so that ALG and RPN programs could coexist.

Gerson.

#85

Hi, Frnk --

Quote:
By adding algebraic mode to the 12CP, HP effectively created a machine that had two different programming languages which happened to be representable by the same 'alphabet' of commands. In doing so they introduced the potential to run an RPN program in algebraic mode, or vice versa, which would then produce the wrong results. Unfortunately, they omitted a way to designate a program as 'RPN mode' or 'algebraic mode', and also omitted to allow the mode-switching commands to be included in a program.

Consequently, it becomes the user's job to deal with this, which is just the wrong way to handle the problem, since there is no benefit to the user in being able to run a program in the wrong mode and get the wrong answer.


I agree with the principle you stated. It's one way of giving the user a fine means of screwing up, and then -- rightly or wrongly -- blaming the calculator. It's not easy to distinguish at a glance an RPN keystroke program from an AOS keystroke program.

Mode-setting via "RPN" or "ALG" in the HP-33S is programmable. I wrote a "Two-bus AC Power Transfer" program for the HP-33S in RPN mode, then developed an ALG-mode version that does calculations in RPN, but user I/O in ALG -- for those who might prefer ALG mode. These programs are available in the MoHPC Software Library, along with versions tailored for the HP-32S and HP-32SII.

-- KS

Edited: 16 Jan 2007, 4:51 p.m.


#86

They are programmable on the 25th ann edition of the 12cp.

Also, keep in mind that very very few people program their 12c/12cp. I'd guess less than 1 in 500, probably much much smaller than that.

#87

Of course, it HAS to have algebraic mode and parentheses.

Really, today's market is such that an RPN only calculator just isn't going to be produced and sold.

Are you sure about that? They seem to be selling enough 12Cs to justify making both the platinum and the 25th anniversary editions. Yes, I know the 12CP has ALG, but it's pretty obviously not the preferred way to use the calculator.

Also, the 33S has an ALG mode, of course, but the manual sure does seem to favor working in RPN.


#88

IMO, The 12c is bought by most users in SPITE of the RPN-only modes.

A 12c with RPN allows RPN nuts (like me) and users used to the RPN mode to continue using it.

Adding ALG mode allows HP to appeal to people who passed it up because they never understood RPN or just didn't want to change their way of thinking.

I'd be shocked if we ever see a new HP with RPN only.

No, that's no insider information, just reading the cards.

Gene


#89

I think it's time for HP to start up a new line of RPN-only models and get back to the golden age.

The market is full of algebraic calculators, very good and sometimes low-cost (TI, Casio, Sharp, Citizen), and HP has its hands into many affairs (printers, computers, UNIX, photo and video cameras, software, etc.), so the only way to get up with little effort and small risks (the calculators market is surely minoritary) is to differentiate their calculators again: if one doesn't absolutely want to deal with RPN, well, he just look at another manufacturer...

HP would have as customers the TOTALITY of the RPN users in the world!

-- Antonio

Edited: 17 Jan 2007, 2:36 a.m. after one or more responses were posted


#90

Hallo!

Quote:
HP would have as customers the TOTALITY of the RPN users in the world!

You mean, all 99 of us ;-) !

Saluti, Max


#91

Counting me, it would be 100.

#92

Two PI's on the keyboard?

That reminds me of the Simpsons episode where Homer "helps" his
half-brother design an automobile.

Homer said something like this...

"Horn buttons! More Horn buttons! I can never find a Horn button when I'm looking for one!"

#93

Quote:
maybe we need a luckier number than 13

I hope you realize that it's bad luck to be superstitious!

#94

Quote:
TEST would work like in the HP-15C, except that two of the logical tests would to be left out to give room to the basic x<=y and x=0.

Why not use TEST 0 .. TEST 5 for comparisons against 0 and TEST . 0 .. TEST . 5 for comparisons against y? Or vice versa.


- Pauli


#95

That would be fine. Something that could be done was grouping the financial and the scientific funcions in different areas, but I didn't want to make many changes. A problem in changing the programming model would be backward compatibility. Almost all old 12C programs would have to be rewritten.

Gerson.

---------

Only one pi key now. Two is too much. And if there was no room for pi, 180 -> RAD would do (Is there a simpler solution?)

Edited: 15 Jan 2007, 5:57 p.m.


#96

Quote:
Something that could be done was grouping the financial and the scientific funcions in different areas, but I didn't want to make many changes. A problem in changing the programming model would be backward compatibility. Almost all old 12C programs would have to be rewritten.

I'd have thought that the change from line number addressing to label addressing would cause significantly more problems than regrouping key functionality.

- Pauli


#97

Indeed, the calculator could operate in two modes, for instance:

In program mode:

GTO 0034 could go to the number line (adapting the existing HP12C programs is easy: add two 00s in from of the number line), while GTO LBL 0 would jump to label 0.

This way, the compatibility with HP11C and HP12C program could easily be maintained (adapting HP11C program is easy: for any GTO 0 change it to GTO LBL 0).

In run mode:

GSB/GTO 0034 could jump to the proper number line (and thus maintaining the HP12C behavior), while f LBL 0 could jump to the indicated label and execute it. Of course this is different, regarding the HP11/15 usage, since they used GSB, but I think this is a little difference (the program would return on the first RTN). With this is mind, it would be nice having the LBL command directly on the key (Gerson, what about it?)

-- Antonio


#98

Ciao Antonio,

I fear the programming compatibility issues are the most difficult problem to solve. Anyway, most 12C old programs would be easy to convert to the new calculator. If really needed, a compatibility mode could be included. This may not be an elegant solution but a MODE key should be include to save lots of positions. It could have the format MODE n n. For instance,

MODE 0 0   RPN
MODE 0 1 ALG
MODE 1 0 END
MODE 1 1 BEGIN
MODE 2 0 M.DY
MODE 2 1 D.MY
MODE 3 0 DEG
MODE 3 1 RAD
MODE 3 2 GRAD
MODE 4 0 Label addressing
MODE 4 1 Line number addressing (for program compatibility with older 12C)
.
.
.

The second digit is 0 in the default modes. DRG could remain in the keyboard as a shorcut, toggling among modes 3 0, 3 1 and 3 2. All of these modes could be programmed. The modes tables would be printed in the back of the calculator. That's what the blackplates are for :-)

Regards,

Gerson.


#99

Hi, Gerson --

Hmm I dunno... the two-input MODE scheme seems rather unintuitive, even with information on the backplate. It is reminiscent of Casio's multi-level MODE menus.

The HP-41, HP-42S, and RPL-based models address this issue with multiple system flags. For those, one must consult the manual.

-- KS


Yes, I agree with Karl. Probably the MODE feature is too complicated for old HP-12C users and even for ex HP-11C users which could adopt the "HP-12CE".

-- Antonio

"BRAVO" I will order two of them. "Todo esta bien chevere" All is fabulous.

Andy
Aerospace Design Engineer
currentluy using National Semiconductor 4520 Scientist (RPN calc)

I'll take an HP-42S with a little more speed and a flash card memory port.

I like alphabetics in my programs.

"ELEV 1 = ?" or "R = 123.76"

as a prompt or result can't be beat. No question about the program you are running or what you need to input or the answer. I can hand the calc to anyone and show them how to execute a program and they can use it.

One of the main shortcoming of the typical RPN programmable calculators is the use of numbered (and more recently with the HP32S, 32SII, and 33s, single-letter) memory registers. When you write re-usable subroutines that store data in memory registers you very easily run into memory conflicts.

What I would like to see in the next generation RPN programmable calculators (HP I HOPE YOU ARE LISTENING) is the ability of dynamically defining a new scope of memory registers. This scheme then allows a subroutine to take input from the stack AND/OR GLOBAL memory registers, create a new block of memory register for its own use, and not worry about over-writing data in registers that belong to other subroutines. Each subroutine can push results into the stack or save them to the GLOBAL memory registers.

I have been thinking about writing an RPN programmable engine with such a memory scheme.

Namir


Quote:
I have been thinking about writing an RPN programmable engine with such a memory scheme.

Hmmh, and what came out of it?


As I contemplated the scheme I saw different variants, such as named variables, named arrays, and so on.

That's why I like RPL because it allows you to declare local variables.

Namir

Please let me collect:

  • Karl wants to see proper handling of complex numbers,
  • Gerson does not need too prominent financial functions,
  • the world requires algebraic,
  • all of us agree RPN is best (but we need these ALG customers, too, so we shall have both modes),
  • many members of this High Forum want the HP15C to return,
  • after the Pioneers, menus and soft keys are the (only reasonable?) way to cope with the amount of functions people want (or calc designers say they want),
  • nevertheless, most of us do not like to dig into deep menus but want to oversee as much as possible easily on the keyboard,
  • the HP15C has a pretty nice function grouping (perfect according to Karl),
  • Forrest wants to have alphabetics, so named variables are requested,
  • Paul requires one PI only,
  • Voyagers have just the right size,
  • nowadays, dot matrix is a must,
  • and the world is full of PC's for the tough math.

Thanks to all of you for your inputs. So how about something like this (finally!)?

  • It's the size of a Voyager.
  • It has the <BIG> horizontal ENTER we have missed so much.
  • It has an "i" key working as "+i*" in rectangular mode. It takes care of an unambiguous display of complex numbers.
  • It features 2 prefixes (f, g) to bring over 90 operations on the 39 keys a Voyager provides.
  • It has a 3rd prefix (h) for 30 lean menus and some more functions. The menus include one FINancial (containing TVM etc.) and one for PARTS (containing also RE, IM, CONJ). In the picture, the menu TRIG is displayed. The menu HYP looks almost identical.
  • It features 8 soft keys. With a menu displayed, you can still access the top line functions via "g". For example, right now unshifted key "7" would execute "ASIN", while "g 7" would insert the number.
  • It works in RPN (of course) and ALG (see "=" and the parentheses).
  • You may set 13 modes via keys directly (see for example line 2 from "Cx" to "16"), and more via the menus SET and DISP. And "->" allows to use the same keys for conversions. Modes include CPRES, fraction (see "h ."), 4 integer bases and alpha.
  • It features an Equation Editor (see EQN, "=", "(", ")", and ";"). And there is a library of predefined formulas (F.LIB) and constants (C.LIB).
  • It provides a UNITS menu for unit conversions. This menu may be overwritten by something more useful, but I heard some North Americans need it desperately. So let them have it.
  • It features a full set of alpha functions (also in X.FCN) and characters (in C.LIB, F.LIB, MAT, TEST, FIN), also beyond English (yes, there are other languages still).
  • You may customize menus and key plate operations via ASN. There are 2 separate assignment layers for calculation and alpha mode. Accessible menu contents will depend on the calc's mode.
  • It may display up to 4 lines of results or 4 stack levels (fixed stack depth of up to 16 levels is settable via SET).
  • It contains the full sets of operations of HP42S, HP32Sii, HP21S (see DISTR for statistical distributions), HP16C (e.g. LOGIC, X.FCN and Boole's operators in column 10), HP12C (see FIN), and more.

OK, it works on paper only, so far, but at least it shows the opportunities. And it proves it's possible to bring it all on a Voyager in a reasonably grouped way (IMHO). So there may be small scale calculator life beyond the realms of finance.

Feel free to ask and criticize at your will. A special invitation goes to Karl for checking the layout. Further explanations are available on request. d:-)

(Edited to embed the image. Thanks a lot, Dave!!)

(Edited to add remarks about menus and modes.)


Edited: 21 Jan 2007, 6:44 p.m. after one or more responses were posted


Quote:
Gerson does not need too prominent financial functions,

Actually I need no financial functions at all, but if that was the price to pay to have a (new) usable Voyager again... :-)

Quote:
In the picture, the menu TRIG is displayed.

I haven't seen the picture yet, but I wouldn't like the trigometric functions hidden in a menu. They ought to be unshifted and positioned in the central upper row of the keyboard, as in the HP-11/15C. That was my greatest disappointment with the HP-28S, by the way.

Quote:
Feel free to ask and criticize at your will.

About criticizing :-)

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

From Amadeus (1984):

Emperor Joseph II to Mozart:

"My dear young man, don't take it too hard, your work is ingenious. It's quality work. And there are simply too many notes, that's all. Just cut a few and it will be perfect."

Mozart:

"Which few did you have in mind, Majesty?"

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

Best regards,

Gerson.

Edited: 16 Jan 2007, 9:09 p.m.


Bom dia Gerson,

Thanks for your kind response and the brilliant quotation of Amadeus! I've seen it then, but didn't remember anymore. Thanks for recalling!

Quote:
Actually I need no financial functions at all

Nor do I. There was space left. And I didn't want to discriminate financial people d;-)
Quote:
I haven't seen the picture yet, ...
Thanks for the feedback. You should see it now.
Quote:
... but I wouldn't like the trigometric functions hidden in a menu. They ought to be unshifted and positioned in the central upper row of the keyboard, as in the HP-11/15C.

So you have to kick out 3 unshifted functions:
Quote:
"Which few did you have in mind, Majesty?"

Comprimentos de Alemanha,

Walter


Hello Walter,

Sorry for the late response. It's just that I've been away for a few days visiting my parents (no computer!)

Now that I am able to see it your layout looks really great. I dare not suggest any function to be left out to give room for trigs :-)

Nice Portuguese too: muito bom! (sehr gut!)

Greetings from Brazil,

Gerson.

Quote:
  • nowadays, dot matrix is a must,

I don't think I agree with this. One of my favorite things about my 12C is the incredible legibility of the numbers compared to my 33S and 49G+.

Quote:
  • and the world is full of PC's for the tough math.

Indeed, which is yet another reason for not larding a 12CX/13C/whatever with alphanumerics, crazy sovers, etc. Just the basics plus enough flexibility to take out some drudgery by writing short programs.

My 49G+ is gathering dust because, IMO, it is neither fish nor fowl: it's too big to drag along anywhere (I'm not in school and don't carry a backpack!) to be useful as an everyday calculator, and it's too small and too screen-challenged to compete with my various computers for anything even half-serious.


The reasonable way to go seems to be to have a "virtual" calculator with built in phone, email, excel, word, calendar, contacts, maps, sync'ing etc. I need a reasonably good screen, full QWERTY and capability to download want I want. I do not want to have too many machines in my pockets.
I have a Qtek 9000 (aka I-Mate Jasar) which is as close to what I find ideal as can be had at the moment (a little too thick). For calculators I have downloaded HP 12 C and HP 15 C (download purchase) and HP 41 C (ev41) and HP 42 S (Free 42) from the net, in addition to various other calculators and converters available on the net.
BUT: I still cannot find as convenient a virtual calculator with SOLVER function as I have on my HP 200 lx (the 42 S is quite cumbersome).
Any suggestions where I can find that ?


Quote:
I still cannot find as convenient a virtual calculator with SOLVER function as I have on my HP 200 lx (the 42 S is quite cumbersome). Any suggestions where I can find that ?


I would suggest Christoph Giesselink's Emu42 loaded with the HP-19BII ROM (quite complete and similar SOLVER), but then you'd have to get an HP-19BII. An option would be the HP-17BII ROM, but it lacks trigs. A problem might be extracting the ROM image, even if you have the real calculators at hand (I still need to extract the ROM image from my HP-17BII).

As of "virtual" calculators, I prefer Thomas Okken's Free 42 with Ehrling42sm skin on my Palm TX. I still have to try Power 48 (last updated on Nov/2004. Hope it works on the TX (it used to work on the Zire 72). Having the 42S (RPN) and the 48G/49G (RPL) in the same device is pretty good to me.

Gerson.


Quote:
I still have to try Power 48 (last updated on Nov/2004. Hope it works on the TX (it used to work on the Zire 72). Having the 42S (RPN) and the 48G/49G (RPL) in the same device is pretty good to me.

Power48 (the standard currently available-from-the-author version, not a modded version) works marvellously on the TX, provided you run a program I found here:

http://www.PalmPowerups.com

Specifically this page here.

The program is called "MemUnfragment". Before I ran this, I could not get the 49G emulation to load, but afterwards, I have no problem.

I use the combination of Free42 and Power48 for almost all the calculating I do away from my desk. Since I'm a nut, I also have around 20 other calculator programs for the Palm... the lovely P41CX, Parens, Palm16C, powerOne Graph, MathU Pro, Easy Calc, CplxCalPro, LyME, SC-123PU, and several others...


Thanks for the links! I will bookmark them in case I decide to go back to the original author's version.

I managed to set up Power48 on the Palm TX yesterday. I replaced version 1.5.1 with the modified program in the link provided by Gene Wright one year ago (Thanks, Gene!):

http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv015.cgi?read=85318#85318

The 49G emulation is running just fine. It's about seven times faster than the real calculator. The Power48/Free42 combination quite fits my needs but I'll take a look at some of what you have listed.

Best regards,

Gerson.

Knut wrote:

The reasonable way to go seems to be to have a "virtual" calculator with built in phone, email, excel, word, calendar, contacts, maps, sync'ing etc. I need a reasonably good screen, full QWERTY and capability to download want I want. I do not want to have too many machines in my pockets.
[end quote]

You forgot to include MP3 player, GPS, multi-band radio, TV and digiCam.

B^)

Ren

dona nobis pacem


Quote:

You forgot to include MP3 player, GPS, multi-band radio, TV and digiCam.


It's called an Apple iPhone (unless Cisco beats them in court).


Hello!

Quote:
It's called an Apple iPhone (unless Cisco beats them in court).

Rumour has it, that the iPhone (*) will be able to run "Dashboard Widgets" for MacOS 10.5 (out in a few weeks). A programming environment for these widgets called "Dashcode" can be downloaded for free from the Apple developer site. It is a very intuitve drag&drop thing (a bit like MS Visual Basic) where one can "click together" the his/her personal dream calculator in record time. Programming the functionality should not be a problem either, because JavaScript is used.

Greetings, Max

(*) as of now, Apple does not intend to support writing direct application programs for this device!

I love 7-segments numbers!

-- Antonio

Hi John,

Quote:
... another reason for not larding a 12CX/13C/whatever with alphanumerics, crazy sovers, etc. Just the basics plus enough flexibility to take out some drudgery by writing short programs.

You may delete as many menus and other orange commands as you want to by reassigning the keyboard using ASN. Just CAT and CLEAR are permanently resident.

Regarding displays: dot matrix allows for readable messages, clear display of programs etc. I agree 7 segment LCD's have very good contrast, but they are not flexible at all.

Regarding size: Fully agree. Therefore, size shall not excess Voyager's dimensions.

Quote:
  • Voyagers have just the right size,

    and

  • It's the size of a Voyager.

  • The contributors to this forum may believe that the Voyagers have the right size and shape and layout. I suggest that the younger generations do not find that to be true. That is because the younger generations operate their hand-held devices with their thumbs. That makes a vertical layout with a chevron configuration for the keys the way of the future and probably the way of future hand-held calculators, whether from H-P, TI, Casio or whatever.

    Those of you who read the comics may remember that the ZITS strip of January 15 addresses the thumbs issue. The first frame shows the father looking in at a group of youths who are mostly operating hand-held devices with their thumbs. The second frame shows the father aa a bewildered grumpy old man (Luddite? troglodyte? Whatever!) telling the mother "I predict that our grandchildren will be giant-thumbed, multi-tasking beings."


    Quote:
    The contributors to this forum may believe that the Voyagers have the right size and shape and layout. I suggest that the younger generations do not find that to be true. That is because the younger generations operate their hand-held devices with their thumbs. That makes a vertical layout with a chevron configuration for the keys the way of the future and probably the way of future hand-held calculators, whether from H-P, TI, Casio or whatever.

    I am just barely young enough to be Generation X, so I could be lumped in with the whippersnappers you're talking about.

    Now, a big reason I like the Voyager layout is because it makes it easier for thumb-twiddling. Think about it: thumbs don't have as much "vertical" range of motion as "horizontal". I just tried it with my 33S and 12C. The 12C is much more thumbadelic.


    Possibly Related Threads...
    Thread Author Replies Views Last Post
      HP-80 History, Design and Interesting Facts BShoring 1 556 11-30-2013, 08:50 AM
    Last Post: Xavier A. (Brazil)
      AME Design Port Extender for the HP41 Dane Spillers 2 600 03-14-2013, 07:58 PM
    Last Post: dane spillers
      12C in 14 Masterpieces of Gadget Design Egan Ford 10 1,391 07-31-2012, 10:18 PM
    Last Post: Palmer O. Hanson, Jr.
      HP-28--better design? Matt Agajanian 7 991 06-27-2012, 09:27 AM
    Last Post: David Tellet
      hp design influence db (martinez, ca.) 5 783 03-01-2012, 03:23 PM
    Last Post: Bunuel66
      HP30b design choices David Griffith 1 502 02-13-2012, 06:41 PM
    Last Post: Tim Wessman
      Design for a simple, entry-level RPN calculator Eddie W. Shore 84 5,999 11-04-2011, 12:48 PM
    Last Post: Bart (UK)
      A design rule for WP34s library software? Dieter 10 1,252 08-15-2011, 05:40 AM
    Last Post: Paul Dale
      Difference in design for 25 and 25c exschr 2 535 07-04-2011, 04:36 PM
    Last Post: exschr
      Calculator key mechanics - Which design is more durable? Timo 3 614 03-30-2011, 09:03 AM
    Last Post: Walter B

    Forum Jump: