About DM32

[Vincent Weber]

(05-17-2023 01:04 PM)jonmoore Wrote:  (41 customers who could easily forgive HP for the lack of IO and expansion options in the 42s).

The 42S was originally planned as a 41C successor, but was redefined at latter stage of development as a 15C successor instead, since HP feared it would cannibalise the RPL models. Hence the removal of all I/O and expansion capabilities, save for IR printing.

The 42S was a very adequate 15C successor, doing everything the 15C does in a much more powerful fashion, but 15C users could rightfully feel nostalgic of the legendary HP slanted keys, the two shift keys, the horizontal format, and the very legible display. That's why the 15C is still in demand, and that Maravia is now releasing the 15C CE...

The 32S/32SII line was never meant as a 15C successor, but as a 11C successor, and is also very adequate for that. Same remarks as 15C vs. 42s, though.

The 41C line is the one which never had a proper successor. That would be the 48SX, but this was too bald a departure. Some 41 users embrassed RPL... and some did not, and they did not forgive the lack of I/O in the 42S!

[Marco Polo]

(05-17-2023 08:38 AM)Vincent Weber Wrote:  Plus42 is *not* similar to the 48 model at all.

Then, the extra features such as directories are also optional. You can completely ignore them and keep to the 42 sheer simplicity - that was Thomas' goal, never harm the 42S original simplicity of operation, and he succeeded!

Simply ignoring Plus42 extra features sounds like using Free42 :-)
And using all the Plus42 features imho is not better than using an emulated 48sx

Let me clarify: i strongly appreciate the work (still) carried out by Thomas, i used Plus42 while in beta testing and gave feedback to him.
But when i had to decide whether to adopt Plus42 as my job workhorse i had to consider the lack of MES and the not-so-easy-to-me way to access the new features and decided to stay on RPL calculator.

[jonmoore]

(05-17-2023 01:29 PM)Vincent Weber Wrote:  The 41C line is the one which never had a proper successor. That would be the 48SX, but this was too bald a departure. Some 41 users embrassed RPL... and some did not, and they did not forgive the lack of I/O in the 42S!

I agree with most of your points, but this, not so much.

HP went to huge lengths to compare the 42S to the 41 line of calculators. Sure, the RPL line from the 48 onward, compared in terms of expandability. Specifically that both featured specialist expansions for likes of surveyors that gave the 41 line and the 48GX in particular a super extended shelve life, and robust second had prices long before the rest of the world realized there was gold in them there HP handheld hills.

But RPL being an extension of FOCAL. Mmmm, I've witnessed in the main grumbles from 41 owners attempts to grapple with RPL.

But this thread is about the 32 family tree, so I reckon we should park 42 discussions or take it elsewhere.

[Vincent Weber]

(05-17-2023 04:03 PM)Marco Polo Wrote:  Simply ignoring Plus42 extra features sounds like using Free42 :-)
And using all the Plus42 features imho is not better than using an emulated 48sx

Plus42 has the unique style of equations that came with the 27S and the 17B, an elegant, purely functional language, with LET and GET, conditionals, sigma loops, and enhanced it with FOR loops and user-defined functions with local environment and full recursion. Almost nothing to envy to the powerful BASIC of the 71B. And you can embedd these equations seamlessly in RPN programs. Try this with free42 or with a 48. I'm quite curious
With Plus42 you have the best or simple RPN and powerful algebraic BASIC, without to go into the uselessly complex RPL. It has a great value on its own!

[Vincent Weber]

(05-17-2023 04:10 PM)jonmoore Wrote:  [
HP went to huge lengths to compare the 42S to the 41 line of calculators.

But RPL being an extension of FOCAL. Mmmm, I've witnessed in the main grumbles from 41 owners attempts to grapple with RPL.

There is no way the 42S could be a worthy successor to the 41, the most expandable calculator ever made, with no I/O! In your add HP only claims compatibility with tHP41CV programs, that must be keyed-in, as said, meaning: with your fingers! the 42S was just a much more capable 15C in a less sexy package

And yes, 41 owners grumbled a lot with RPL, that was not an extension of FOCAL (not compatibility). Some grumbled to the point that they sticked with their 41 forever, and to this day are still waiting for a worthy successor...

[Marco Polo]

(05-17-2023 04:56 PM)Vincent Weber Wrote:  

(05-17-2023 04:03 PM)Marco Polo Wrote:  Simply ignoring Plus42 extra features sounds like using Free42 :-)
And using all the Plus42 features imho is not better than using an emulated 48sx

Plus42 has the unique style of equations that came with the 27S and the 17B, an elegant, purely functional language, with LET and GET, conditionals, sigma loops, and enhanced it with FOR loops and user-defined functions with local environment and full recursion. Almost nothing to envy to the powerful BASIC of the 71B. And you can embedd these equations seamlessly in RPN programs. Try this with free42 or with a 48. I'm quite curious

The 48 can do everything you wrote....
And, not wanting to enter in the endless debate RPL vs RPN, I found much harder to use Plus42 than 48

[Vincent Weber]

Of course the 48 can do anything, but not in the style and simplicy of the 17B/27S solver, which is truly unique.
Oh well... Matter of taste

[Don Williams]

(05-17-2023 05:01 PM)Vincent Weber Wrote:  [quote='jonmoore' pid='172863' dateline='1684339808']
[

And yes, 41 owners grumbled a lot with RPL, that was not an extension of FOCAL (not compatibility). Some grumbled to the point that they sticked with their 41 forever, and to this day are still waiting for a worthy successor...

Amen!

RPL was the answer to a question that no one ever asked.

Then to add insult to injury they abandoned keystroke programming.

[Marco Polo]

(05-17-2023 05:26 PM)Vincent Weber Wrote:  Of course the 48 can do anything, but not in the style and simplicy of the 17B/27S solver, which is truly unique.
Oh well... Matter of taste

:-)

[jonmoore]

(05-17-2023 05:35 PM)Don Williams Wrote:  

(05-17-2023 05:01 PM)Vincent Weber Wrote:  [quote='jonmoore' pid='172863' dateline='1684339808']
[

And yes, 41 owners grumbled a lot with RPL, that was not an extension of FOCAL (not compatibility). Some grumbled to the point that they sticked with their 41 forever, and to this day are still waiting for a worthy successor...

Amen!

RPL was the answer to a question that no one ever asked.

Then to add insult to injury they abandoned keystroke programming.

Can I interest you in the exception to the rule and for the fair market derived price of $249. Apparently Surveyors like this one too, as do sneaky EE students. Ignore a few howling bugs and you'll be happy as Larry.

[Marco Polo]

(05-17-2023 07:55 PM)jonmoore Wrote:  Can I interest you in the exception to the rule and for the fair market derived price of $249. Apparently Surveyors like this one too, as do sneaky EE students. Ignore a few howling bugs and you'll be happy as Larry.

I have heard that the 35s price were rising, but this is absurd.
All in all the 35s is a decent calculator, well built, with some flaws.
I have one, but it does not deserve more than 75 Eur, imho

[Valentin Albillo]

.
Hi, Vincent,

(05-17-2023 04:56 PM)Vincent Weber Wrote:  Plus42 has the unique style of equations that came with the 27S and the 17B, an elegant, purely functional language, with LET and GET, conditionals, sigma loops, and enhanced it with FOR loops and user-defined functions with local environment and full recursion. Almost nothing to envy to the powerful BASIC of the 71B.

Seriously, Vincent !? Have you ever used HP-71B's BASIC ?

Does timers, data files, DATA/READ statements, LEX extensions, (BIN) subprograms with parameter passing by value or reference, arrays/matrices, strings, string arrays, IEEE compliance, ON GOTO/GOSUB/RESTORE/TIMER etc. ring a bell ?

Let alone the extensions to BASIC provided by the Math ROM: COMPLEX variables/operations, integrate, solve, FFT, polynomial rootfinding, ... or the ones for extende I/O and device control provided by the HP-IL ROM ..."Almost nothing to envy", you say !! What a joke !

Quote:And yes, 41 owners grumbled a lot with RPL, [...] Some grumbled to the point that they sticked with their 41 forever, and to this day are still waiting for a worthy successor...

Not me. I soon got an HP-71B and never looked back, as it can really run rings around any flavour of 41C.

As for a successor to the 41C, the HP-42S was obviously the one but it was butchered in the craddle by some marketing decisions at HP, who were intent in cramming RPL machines down our throats whether we liked it or not, and the 42S was the innocent victim of the RPL enforced supremacy. Without the criminal 42S's I/O crippling, sales of RPL machines would have plunged into the mud and HP knew it.

Regards.
V.

[Jlouis]

(05-17-2023 01:29 PM)Vincent Weber Wrote:  

(05-17-2023 01:04 PM)jonmoore Wrote:  (41 customers who could easily forgive HP for the lack of IO and expansion options in the 42s).

The 42S was originally planned as a 41C successor, but was redefined at latter stage of development as a 15C successor instead, since HP feared it would cannibalise the RPL models. Hence the removal of all I/O and expansion capabilities, save for IR printing.

The 42S was a very adequate 15C successor, doing everything the 15C does in a much more powerful fashion, but 15C users could rightfully feel nostalgic of the legendary HP slanted keys, the two shift keys, the horizontal format, and the very legible display. That's why the 15C is still in demand, and that Maravia is now releasing the 15C CE...

The 32S/32SII line was never meant as a 15C successor, but as a 11C successor, and is also very adequate for that. Same remarks as 15C vs. 42s, though.

The 41C line is the one which never had a proper successor. That would be the 48SX, but this was too bald a departure. Some 41 users embrassed RPL... and some did not, and they did not forgive the lack of I/O in the 42S!

The curator of this Museum should pin this post in any place that describe HP calculators involved in this discussion, at that period of time.

[carey]

(05-17-2023 05:35 PM)Don Williams Wrote:  RPL was the answer to a question that no one ever asked.

I agree that RPL was overkill for many engineers. However, for engineers whose work required symbolic and advanced numerical answers, and for calculus students, RPL was an answer to a prayer. For the cost of one software package in 1986, one could buy an HP28 with integrated symbolic and numerical capabilities that could often do more and more easily than desktop software at the time.

(05-17-2023 05:35 PM)Don Williams Wrote:  Then to add insult to injury they abandoned keystroke programming.

RPL allows "keystroke programming" by surrounding keystrokes with French quotation marks << >>.

[Vincent Weber]

(05-17-2023 11:02 PM)Valentin Albillo Wrote:  .
Hi, Vincent,

(05-17-2023 04:56 PM)Vincent Weber Wrote:  Plus42 has the unique style of equations that came with the 27S and the 17B, an elegant, purely functional language, with LET and GET, conditionals, sigma loops, and enhanced it with FOR loops and user-defined functions with local environment and full recursion. Almost nothing to envy to the powerful BASIC of the 71B.

Seriously, Vincent !? Have you ever used HP-71B's BASIC ?

Does timers, data files, DATA/READ statements, LEX extensions, (BIN) subprograms with parameter passing by value or reference, arrays/matrices, strings, string arrays, IEEE compliance, ON GOTO/GOSUB/RESTORE/TIMER etc. ring a bell ?

Let alone the extensions to BASIC provided by the Math ROM: COMPLEX variables/operations, integrate, solve, FFT, polynomial rootfinding, ... or the ones for extende I/O and device control provided by the HP-IL ROM ..."Almost nothing to envy", you say !! What a joke !

Hi Valentin,

Yes, I am aware of the powerful characteristics of the HP-71B you described, and I admit I exagerated a bit but I think you exagerate a bit too, my statement is not THAT ridiculous, a case can be made for it...

Here are a few points that you might have overlooked about Plus42:

-Plus42 equation langage does not have GOTO, but has structured programming constructs instead: the C-style FOR loop, with BREAK and CONTINUE statements; and a SEQ statement that allows to easily define a block of instructions (like { } in C). These allow almost anything, including emulating REPEAT... UNTIL and WHILE loops, with no need for the "spaghetti code" implied by GOTO;

-It doesn't have GOSUB, but it allows user-defined functions with full local environment and recursion;

-It doesn't have ON... GOTO nor ON... GOSUB, but it has a list type, those lists can contain any object, including pointers to functions, so you can call a function either by its index in the list, or even from a string with its function name, I have done it !

-These user-defined functions only support passing parameters by value for now, but I have successfully implemented myself passing by reference, using XEQ and indirect STO/RCL(see below for XEQ), and built-in support is being considered for future releases;

-It doesn't have DATA/READ/RESTORE, but supports direct matrix literals, that allow you to define an array of data without the need of writing a loop to read and store this data;

-it doesn't have LEX files, but it supports XEQ of any RPN program within an equation, with parameters pushed onto the stack, so you can easily expand the language with RPN programs (and, as you know, those 42S RPN programs are super fast in the Free42/Plus42 environnement, and they can save and restore the stack and variables environment, so no damage is done). I have done it successfully to implement some PRINT USING and PAUSE statements within equations, and it works seamlessly;

-It does support the full set of the powerful HP-42S math functions from equations: Matrix, complex, statistical functions, as well as INTEGRATE and SOLVE, are already built-in. String functions that were added to Free42/Plus42, with full substring capabilities, are on the way, at my request;

-It does support arrays (matrices) of unlimited-length strings;

-it does support data files, a matrix is a file that can be built, saved, renamed, moved or copied to another directory, imported or exported from/to Excel using intelligent copy and paste.

I think I found alternatives for almost all the points you mentionned, save for timers and IEEE compatibility, for which I admit there is not (yet?) an equivalent, and HP-IL I/O which wouldn't make sense on my phone. Moreover, the Plus42 approach leans more towards structured programming rather than relying on the (in)famous GOTO instruction

[EDIT] IEEE compliance is indeed supported, even if NaNs and inifinites are not fully exposed to the user.

And as far as external devices are concerned, you still get a big LCD screen with customizable size (both numbers of lines and columns), and the 82240 printer simulation, with an infinite "paper" that you can scroll easily with your fingers. Plus42 equations can print to this, and can even be debugged using TRACE mode, so that you can see at once the details results of every single step, with the cursor positionned to the current step, effectively doing single-stepping in one shot. Very convenient!

And equations are pre-parsed (btw an equation can even parse and evaluate another equation contained in a string, which is functionally equivalent to the powerful HP-71B VAL function) , they are fast, and syntax errors can be caught early at parsing time, with the cursor poistionned on the offending statement. With the latest version it is also possible to debug run-time errors this way. The rocks!

I hope that I demonstrated to you that I am not *that* mad, and that at least comparing Plus42 powerful and versatile language to the powerful HP-71B basic is worth a discussion

Best regards,

Vincent

[Paul Dale]

I believe Plus42 is using IEEE quad precision decimal reals.
Or at least can do so.

[carey]

(05-17-2023 08:38 AM)Vincent Weber Wrote:  "...IMHO there is objectively no reason to prefer the 32SII or the 35S over Plus42...", .

Real vs virtual keys?

[Vincent Weber]

(05-18-2023 06:37 AM)Paul Dale Wrote:  I believe Plus42 is using IEEE quad precision decimal reals.
Or at least can do so.

Yes, but I think Valentin was more referring to NaNs and infinites...

[Marco Polo]

(05-18-2023 12:47 AM)carey Wrote:  

(05-17-2023 05:35 PM)Don Williams Wrote:  RPL was the answer to a question that no one ever asked.

I agree that RPL was overkill for many engineers. However, for engineers whose work required symbolic and advanced numerical answers, and for calculus students, RPL was an answer to a prayer. For the cost of one software package in 1986, one could buy an HP28 with integrated symbolic and numerical capabilities that could often do more and more easily than desktop software at the time.

(05-17-2023 05:35 PM)Don Williams Wrote:  Then to add insult to injury they abandoned keystroke programming.

RPL allows "keystroke programming" by surrounding keystrokes with French quotation marks << >>.

It seems to me that many RPL advantages are ignored or minimized in praise of tradition.
As an HP calc user I was born on a 28s in 1989 and evolved passing through 48sx, 48gx, 49g+ and finally 50g.
I appreciate the extreme ductility of RPL and the consistency of operations on different data types.
Furthermore, I find even the UI quite practical and totally customizable to my needs and taste.
On my job I heavily rely upon uom management and MES
I tried to use also less powerful calculator (33s, 35s, Free42, Plus42) but I never got used to the lack of some features and the uom management, in spite of the "elegance" of some solutions.
I have a 35s in my drawer at work, but I almost always end in taking my 50g (physical or emulated)...

[Paul Dale]

(05-18-2023 06:52 AM)Vincent Weber Wrote:  Yes, but I think Valentin was more referring to NaNs and infinites...

They are there too. It's using Intel's decimal library underneath which is full IEEE 754. I don't know if or how well they are exposed.