praise for the 32sii



#2

OK, I know some folks don't like it, but I do, having recently acquired a couple of them. Why I like it:

  1. It has the perfect function set, for me anyhow.
  2. It has a crystal clear, large display.
  3. The keyboard has that classic HP quality, and the key legends are very easy to read.
  4. Most functions are not hidden in menus (although the 32s uncluttered keyboard is desirable also).
  5. It uses named (A-Z) registers rather than numbered, which improves program usefulness.
  6. The INPUT and VIEW commands are great, rather than relying upon the user to remember what to put on the stack when running programs.
  7. It has decent programming space, especially integers between 1 and 254 requiring only 1.5 bytes.
  8. Navigating program lines shows actual commands, rather than keycodes.
  9. It allows programming indirect access to the 6 stats registers.
  10. It has relatively fast ISG/DSE looping commands (compared to the slow ones on the 33s).
  11. The algebraic equation solver is a pleasure to use.

Nits:

  1. It requires two keys to navigate to each line of code, which requires you to look at the keyboard each time.
  2. It would have been nice if the L.R. menu (linear regression) didn't kick you out of the L.R. menu when you view the slope (m) and y-intercept (b). The backspace key can be used in some menus to not kick you out, but not L.R.
  3. It would have been nice if the designers had not placed the i variable between the regular registers A-Z (1-26) and the 6 stats registers (28-33), so you could have had a continuous block of 33 indirect-access registers.
  4. It would have been nice to support infrared printing, like the 17bii.

Having used the calculator for only a few months, I can see why it is highly valued.

Edited: 13 Dec 2010, 8:20 p.m.


#3

Don --

Nice summary about the HP-32SII; it's my primary calculator.

Some comments about a few items:

Quote:
It has decent programming space, especially integers between 1 and 254 requiring only 1.5 bytes.

384 bytes (and 390 bytes for the HP-32S) total for programming and data is "decent", albeit less than the 469 bytes offered by the HP-15C. However, even though the HP-32S/SII programming paradigm is good enough to use for permanent-residence programs and equations, its user RAM is insufficient to retain them all while still being able to use SOLVE and INTEG. I wish that these two models had the 2 kB RAM of the original HP-28C, even if the 33 registers A-Z, i, and Stat were permanent.

Quote:
The algebraic equation solver is a pleasure to use.

It's the one from the HP-22S. One annoyance is the lack of insert/delete editing within the equation, as opposed to the end of it. That was remedied on the HP-35s.

Quote:
It would have been nice if the L.R. menu (linear regression) didn't kick you out of the L.R. menu when you view the slope (m) and y-intercept (b). The backspace key can be used in some menus to not kick you out, but not L.R.

I don't quite understand. Each menu on the HP-32SII's one-line display is only one level deep, and will exit upon execution so that the full function name and result can be displayed. The regression variables m, b, and r are calculated separately.

Quote:
It would have been nice if the designers had not placed the i variable between the regular registers A-Z (1-26) and the 6 stats registers (28-33), so you could have had a continuous block of 33 indirect-access registers.

Indirect access to the six stat registers and the indirect register (i) was added to the HP-32SII; it was not present on the HP-32S. Any register can be "cleared" by setting its value to zero; the six stat registers can be cleared en masse by a single command (Clear Sigma). I suppose there's no reason why the indirect register could not have been the last one (33rd).

Quote:
It would have been nice to support infrared printing, like the 17bii.

Only the high-end Pioneers (HP-42S, HP-17B/BII, and HP-27S) got that feature, along with a full-matrix display and a faster processor. Also, the printing commands are too numerous and lengthy to be shown clearly on a lesser display.

The HP-32S/SII and HP-14B were mid-grade models.


For a list of most of the other differences between the HP-32S and HP-32SII, please see this archived post:

HP-32S vs. HP-32SII differences

-- Karl


Edited: 14 Dec 2010, 2:08 a.m. after one or more responses were posted


#4

Thanks Karl, great comments and I agree with them all. And thanks for that link to the post with the differences between the 32s and 32sii, I've come to really like both of them recently.

Here's what I meant about the L.R. menu. I've spent a lot of time recently using it to prepare problems for my 8th grade students who are studying linear equations. I enter the ordered pairs for two points in the stats registers, then I want to determine the slope and y-intercept of the line that includes those two points. I press L.R. and then m and it displays the slope. What I'd like to do at that point is press the backspace key and have it bring up the L.R. menu again so I can press b and see the y-intercept, but that won't work. Backspace just clears the display, it doesn't take you back to within the L.R. menu so I could press b. I must press L.R. again and then b to find the y-intercept.

Compare that to the MEM (memory) menu. You press MEM and you see buttons for VAR and PGM. If you press VAR (and you actually have some variables defined), you will see the variables and you can then press backspace and end up back in the MEM menu, so you could then press PGM immediately if you wanted to, without having to bring up the MEM menu again. That's the behavior I'd like to see in the L.R. menu too, it would just make it a tad easier to get slope and y-intercepts. But it's just a nit.

When I see how the 32sii (and 32s) were designed to just work so well, I am amazed at the professionalism of its designers back in the 1980's I guess. In another post I (with Gerson's help) recently tracked down Roy Martin, one of the principal designers of the HP92 I think it was, and I wrote him a letter with some questions and he responded with a letter to me. That was so great. Some of those old guys are still around, and if any of them read this forum I hope they know how much we admire what they did decades ago.

Don


#5

Quote:
I press L.R. and then m and it displays the slope. What I'd like to do at that point is press the backspace key and have it bring up the L.R. menu again so I can press b and see the y-intercept, but that won't work. Backspace just clears the display, it doesn't take you back to within the L.R. menu so I could press b. I must press L.R. again and then b to find the y-intercept.

Compare that to the MEM (memory) menu. You press MEM and you see buttons for VAR and PGM. If you press VAR (and you actually have some variables defined), you will see the variables and you can then press backspace and end up back in the MEM menu, so you could then press PGM immediately if you wanted to, without having to bring up the MEM menu again. That's the behavior I'd like to see in the L.R. menu too, it would just make it a tad easier to get slope and y-intercepts.


I just can't agree with you there, Don.

MEM is a non-programmable command that brings up a menu to allow individual variables or program sections to be interactively deleted, via the VAR or PGM lists.

"L.R.", however, brings up a menu of five programmable functions, each of which calculates and returns a value to the x-register. If the execution of one of these functions did not exit the menu, the function would not work the same way as any other programmable function accessed via menu.

Keystroking for repeated regression analyses can be reduced by using a display program:

Detailed:                 Austere:

LBL L LBL L
m (via L.R. menu) m
STO M R/S
b (via L.R. menu) b
STO B R/S
r (via L.R. menu) r
STO R RTN
VIEW M
VIEW B
VIEW R
RTN

-- Karl

(Added some detail and minor wordsmithing)


Edited: 14 Dec 2010, 9:53 p.m. after one or more responses were posted


#6

Thanks Karl. Yeah, you're right, a simple program can be used to see slope and y-intercept with minimal effort. If I do this regularly I will use a program, as you suggested.

Don


#7

The programming paradigm of the 32sii is IDEAL for ad-hoc programs of just this sort. You MUST use the programming memory for that purpose! Elaborate, long programs have no place on a machine with no I/O, no back-up, and tedious interface, by modern standards.

It is one thing to play with these machines in the mind set of 1978. It is another to think of their utility today, for which I think you will agree, the 32sii, and indeed, even Mike Morrow's sacred cow 42s, is a shabby, less than practical or useful tool for any significant programming.

Put it another way: In 1982 when I got my 11c, there was no reliable portable programming of the sort today. You had Sharps, HP, Timex Sinclair, Commodore Vic 16, Atari. Only the Sharps and HP were portable. The 11c/15c and their progeny, the 32s/42s, were never designed for power-programming: the 41C (and later 28S, heir to the 55/65/67) was intended for that purpose.

Trying to justify one's use of these non-I/O machines for serious programming, rather than ad-hoc, would be akin to showing one's blinders to cost/benefit, or trying to justify the unjustifiable. (Not you, but just a point.)

This was even obvious to me in 1982. The longer programs, so beautifully written in the spiral-bound manual seemed, to me, an awful lot of overhead for something that could not be backed up or loaded later. I knew of the 67 and the 41 then--I had used a 67 once--and so I stayed away from heavy duty programming but relished the ad-hoc simple but extremely useful programs one could construct on the fly.

In the mid-70s when there was NO Timex, or Atari or anything, then trying to squeeze a long program into a non I/O machine had real benefit. But by the early 80s that was no longer the case.

Edited: 14 Dec 2010, 2:16 p.m.


#8

I don't disagree with you, Bill. Nobody's going to write a payroll program for the 12c (I don't think you could quite do that in 99 lines of program space anyhow).

Gene always said that, for the 30b, the programming capability can best be used to bring commonly-used functions within menus to the keyboard, for quick execution. I would agree.

I've used programs I developed on calculators for simple utilities (like base conversion), games (like hi-lo guess), number theory things (like prime factorization), or just the challenge of implementing the solution to a puzzle or riddle on something like a 12c with its limited programming space. To do that successfully is just eminently satisfying, even though it may have no practical value.

Serious programming work requires a PC or a mainframe.

#9

Did you try to enter the L.R. menu twice? At least this works on the somewhat similar HP 42S.

HTH

-- tom


#10

Thomas, I had to enter the L.R. menu twice if I wanted to see both slope and y-intercept, and my point was that it would have been nice to not have to do that. That would have required the backspace key to take you back into the L.R. menu after choosing m or b and viewing them, but it doesn't work that way. Karl pointed out that a simple program will do the trick, and I will probably use that if I find myself needing to find m and b repeatedly.

I don't know if the 42s has similar behavior, I've never owned or used one.

This is certainly not a deal-breaker, however. It's still a great calculator.

Don


#11

Quote:
I don't know if the 42s has similar behavior, I've never owned or used one.

Just checked the behavior of the HP-32SII and it's NOT the same as that of the HP-42S: When you know you will use e.g. the PROB-functions a lot you can select that menu twice. Then you won't exit the menu after choosing a function. Instead you have to use the EXIT-key explicitly. From your description I thought you were looking for something like that. I think that this is quiet clever: when you use a function only once you don't want to be forced to use the EXIT-key as well but when you use a group of functions often you don't have to select that menu every time.

You could download Free42 to get an idea how it works. And I agree: both are great calculators.

Best regards

Thomas


#12

Thomas, that does sound like neat functionality in the 42s.

Free42, maybe I'll play with that, although today in the mail I got one of the new Casio color Prizm graphing calculators that Namir mentioned recently (math teachers get those free to tryout, on the hopes that hundreds more will be ordered!).

Don

#13

Major problems:

(1) It has absolutely unspeakably horrible complex number support, just like the later HP33S and HP35S abominations. It seems to have established HP's paradigm for complex number support on small scientific calculators, very unfortunately. That is a huge disappointment from a scientific calculator that appeared four years after the HP42S and ten years after the HP-15C set the standard for natural complex number handling.

(2) It has an extremely small amount of RAM for a calculator from the mid-1990s...less than the HP-15C of ten years earlier.

One positive: It seems to be about 35 percent faster than the HP42S for problems that it can handle.

I've kept that HP32SII mistake...it's like new today, from lack of use. But I acquired a few HP42S units as soon as possible after the HP32SII disappointment. (Excellent machines in daily use 18 years after they were made.)


#14

The 42s, offering a convenient subset of the huge 41c system, is well suited to RPN users who want to write their own programs on a pocketable platform. It's deeper than a typical calculator.

The 32sii, on the other hand, fills the role of a traditional calculator, a go-to device for quick calculations with its functions mostly all on the keyboard, but with the capability to go a little deeper with either short programs or solver equations.

So, complex number support aside, the 32sii fills its niche nicely, as many have testified. It ain't a 42s, but not meant to be one.

#15

Mike, I never use complex numbers, so that is not important to me.

Yes, more memory would be nice, but we can say that (and I think you have) about the 30b too.

I've never had the urge to get a 42s. I know (or have read) it is similar to the 41, but I've never had the urge to get one of those either.

I'd be glad to take that pesky 32sii off your hands for a reasonable price.

:)

Don

#16

Quote:
(1) It has absolutely unspeakably horrible complex number support, just like the later HP33S and HP35S abominations

I agree about the 33s, but don't think the 35s should be included in the "abominable" group. In my opinion, the 35s implemented a major leap forward regarding complex number support: the inclusion of an "i" key. With a few tweaks it could have been darn near perfect, but even as-is, the 35s does a better job than the 32sii or 33s in handling and calculating with complex numbers.

#17

Quote:
It has absolutely unspeakably horrible complex number support, just like the later HP33S and HP35S abominations. [...]That is a huge disappointment [...] ten years after the HP-15C set the standard for natural complex number handling.

Mike, can you elaborate on this? What do you like about the 15C that isn't in the 32sii and 35s? What would be your ideal complex number support?

I'm not disagreeing with you, I'm just curious to know what you (and others) think is the right way to handle complex numbers.

One thing I like about the 35s that isn't available on the 15C is the ability to store complex numbers in a register.

Thanks,
Dave


#18

Quote:
What would be your ideal complex number support?

Not trying to speak for Mike, but he as already said (many times) - 42s.

This 32sii is great/horrible debate has occurred more times than I can count. It seems to be a perennial topic to liven up the Forum when it gets a little dull. Thanks, Don.

#19

Quote:
What do you like about the 15C that isn't in the 32sii and 35s?

The ability to calculate asin(2). Once you're in complex mode everything just works as expected.

The HP-32SII doesn't claim full support of complex numbers. You may call it clumsy but it may be usefull.

On the 35s a complex data type is supported but it's restricted to certain functions. And that's something I can hardly understamd after having seen what was possible before with the HP-15C or HP-42S.

Quote:
One thing I like about the 35s that isn't available on the 15C is the ability to store complex numbers in a register.

One thing I like about the 15C that isn't available on the 35S is the ability to store complex numbers in a matrix.

Best regards

Thomas


#20

Quote:
On the 35s a complex data type is supported but it's restricted to certain functions. And that's something I can hardly understand after having seen what was possible before with the HP-15C or HP-42S.

Thomas --

The historical explanation is that the complex-number functions offered by the HP-32S, HP-32SII, HP-33s, and HP-35s seem to have been based upon the set of keystroke programs on the HP-41 Math Pac of 1980 (and on the Advantage Pac of 1985). What was present on the HP-41 Math Pac is present in microcoded form on the latter-day models; and what wasn't, isn't.

-- Karl

Edited: 15 Dec 2010, 2:07 a.m.

#21

The 42s is an abysmal failure as a serious programming device, as it has no I/O.

Perhaps it is the Holy Grail of RPN calculators--but it is certainly *not* the holy grail--not even close--of *programmable* calculators (that would have to go to the 41C of course).

For persons in need of the following:

Complex number calculations,
Matrix handling,
nested numerical integration and other calculations,

then the 42s is the obvious choice of non-or-limited programmable *calculator*.

For those who need none of those in a *calculator* (that same person may be doing those tasks on a computer) then the 32sii is in fact Superior to the 42s.

And a small subset of historically RPN users actually find the 27s to be the best machine of all--for the 32s utility level. If only I had bought one back when they were new!


Edited: 14 Dec 2010, 2:26 p.m.


#22

Quote:
The 42s is an abysmal failure as a serious programming device, as it has no I/O.

Even the $1 walmart job has I/O. Or you're ignoring the HP42S keyboard, display, tone, and IR printer I/O devices. :-)

I've used several HP42S units for the past 15 years, including one that goes where I go daily. Even in the early days when I was experimenting with the debugging/system memory display/execution feature that's provided by the HP42S, I have yet to corrupt memory or cause a memory loss on any HP42S.

It would be great if a modern version of the HP42S had a bigger display, much more RAM, a real-time clock/calendar, financial functions, and a micro-SD card. But I don't think that writing to the SD card would be practical in a device powered by button- or coin-sized primary batteries. I suppose that I could live with a thicker design that used N-cells, as long as the battery compartment was better designed than that of the unfortunate HP-28C/S.

Quote:
Perhaps it is the Holy Grail of RPN calculators--but it is certainly *not* the holy grail--not even close--of *programmable* calculators (that would have to go to the 41C of course).

I've owned several HP-41C models since it first appeared. I still have a HP-41CX with mag card reader and IR print modules that I bought new in 1984. It's a nice machine, in a third-of-a-century-old-technology sort of way. But for RPN programming, the HP42S is far better human factored for accessing functions, it is much much faster, it presents results with several orders of magnitude better precision and accuracy, its built-in firmware supports many functions that the HP-41C could only be programmed to do slowly and less accurately (including numerical integration, a numerical solver, a complete spectrum of complex number handling and matrix functions), it has IR printer output and a two-line "real" alphanumeric display, it comes with several times more RAM than the HP-41C could ever support, it's smaller, and it's much much more battery frugal.

OTOH, the HP-41CX has the best keyboard feel of any calculator I've ever used. The industrial design isn't bad (better than the Pioneers), except that HP decided to eliminate the classic black/blue HP logo found on earlier models. I won't be getting rid of my HP-41CX in my lifetime, but for actual RPN programming, years ago I converted my HP-41C stuff to run on the HP42S and if I write anything new, it goes on the HP42S.

With respect to the ultimate programmer's calculator (that exists today), I believe that honor clearly and unambiguously falls to the HP50G. This includes TI, Casio, and all other HP competition. The superiority of the hardware of the HP50G by itself (including the SD card and the display clarity) is enough to cause wonder about how, say, the latest TI-89 Titanium Hardware Version 4 would be more attractive to any knowledgeable person. (Now, if only the HP50G had the speed of the HP30b, which is about ten times faster.)


Edited: 15 Dec 2010, 10:04 a.m.


#23

Although I don't have a 50G, I think I agree with you for the current ultimate. I have 48GX and that is pretty darn close.

I didn't know the 30b was *that* fast. Holy smokes!

I knock on wood for you and your 42s data. I've corrupted data in lots of machines, including voyagers...and a 27S which is of course the 42S predecessor in development (at least by dates on Finseth's hpdatabase).

One interesting development is that if you can accept the "virtualness" of emulators and simulators, the 42s and all the other pioneers can be run on a PC now--or in the 42s case on an iphone/ipodtouche--and in that case, you do actually have back-up!

(But like you, I cannot get past the need to feel the keys under my fingers...)

Edited: 16 Dec 2010, 1:19 a.m.

#24

Agreed, all. And really fast with keyboard calculations!

#25

I like it, but I don't require matrix or complex numbers in my line of work.

I REALLY like it because of the cluttered colourful keyboard ;-) It has a wow factor that the menu calcs do not have. I mean, look at the HP 67 with THREE, count them, THREE function keys.

Yep, the 32Sii is the prettiest, most colorful of the Pioneers and therefore, esthetically my favourite Pioneer. My most useful pioneer is the 42S but that is because it takes up less room and is less prone to ESD then my all time favourite; the 41CX.


#26

Quote:
I REALLY like it because of the cluttered colourful keyboard ;-)

Yep, the 32Sii is the prettiest,


Well, pretty is in the eye of the beholder, but strangely, I agree with you that the cluttered keyboards have an appeal. It bespeaks more capability than the simpler keyboards, although the reverse is often true.

#27

Thats the point,

The average Joe or Jane sees an HP32SII and a 42S together and think the 32 is the more capable machine.

My posting was a bit tongue in cheek but hey, look at the mega-commodores with their 60 key plus the colours. Now they look CAPABLE!

Cheers


#28

I wish that HP had stayed with placing the +,-,/, and x keys on the left side directly underneath the big ENTER key. Does anyone know when and why they changed this?


John


#29

Voyagers, with landscape, moved to the right and followed the standard Ti arrangement. Then Pioneers continued that. All the previous machines had the HP arrangement with arithmetic on the left.

#30

I think that having the +-*/ keys on the right side is actually better from an ergonomic standpoint, if like most one is right handed.

Although putting the ENTER key above the +-*/ keys wasn't possible with the (IMHO terrible) landscape layout of the Voyager experiment, all subsequent models could easily have placed the enter key on the right side too, above the re-located +-*/ keys just like the calculator dieties intended.

It is a horrible ergonomic layout for efficient number entry to have the ENTER key located on the side opposite the +-*/ keys.

At least the HP49G+/HP50G have all these keys on the right (and correct) side, it would be an improvement to place the ENTER key ABOVE the +-*/ keys, instead of below them. And make it wider too!


Possibly Related Threads...
Thread Author Replies Views Last Post
  Praise for the 34S Egan Ford 23 850 09-27-2011, 02:17 PM
Last Post: Steve Simpkin
  In praise of dirty calculators-- again. Michael Meyer 34 1,271 02-15-2010, 07:54 PM
Last Post: Joerg Woerner
  In Praise of Dirty Calculators Michael Meyer 2 146 11-30-2009, 11:45 PM
Last Post: Vieira, Luiz C. (Brazil)
  In praise of hard-copy manuals Luca 40 1,361 09-19-2009, 05:59 AM
Last Post: Bart (UK)
  Slightly OT: Praise the pocket calculator! Frank Boehm (Germany) 1 144 08-29-2008, 09:20 AM
Last Post: John B. Smitherman
  Unusual praise for hp-calculators :-) Maximilian Hohmann 3 202 01-25-2007, 05:15 PM
Last Post: Les Wright
  In Praise of EMU71 Emulator and JPCROM Namir 20 743 01-19-2006, 03:58 PM
Last Post: Namir
  32Sii BROWN vs. 32Sii BLACK NH 8 365 10-10-2003, 04:44 AM
Last Post: NH
  In praise of hp 15c :-) R Lion (Spain) 8 341 03-05-2003, 05:31 AM
Last Post: R Lion
  In praise of the HP-34C Ex-PPC member 10 375 03-04-2003, 12:00 AM
Last Post: Norm Hill

Forum Jump: