HP Calculator Highlights since the 1980

[grsbanks]

(04-01-2018 02:39 PM)Maximilian Hohmann Wrote:  [...] and teachers wouldn't have liked to teach 1/2 of the students RPN and the other half algebraic depending on what the parents could afford to buy their children. […]

When I was at school and of an age when we were starting to use calculators (basically throughout the '80s), teachers did not teach students how to use them. They allowed students to use them but it was up to the students to learn how to.

It was probably OK for that to be the case at the time because calculators were a means of verification of what we'd done either manually or using trig/log tables. We never used calculators as our primary tool for getting answers.

[Helix]

(04-01-2018 03:43 AM)emersone12 Wrote:  Hypothetically. would you be against the production of a DM35, DM45, DM55, DM65 etc.....or a universal single DM_Multi-calculator accommodating all the available range of calculators, involving a collaboration/arrangement using your boards/kits. The existing DM10,11,12,15,16 are pretty much close replicas of the original HP's in function, layout, physical appearance, size & case shape, & equivalent identical-shaped replicas of the old original HP LED calculators as modern LED reproductions would be a welcome addition if upgraded using modern technology, LED's & battery systems. The old HP case shapes are attractive well designed works of art & very comfortable to hold compared to the ridiculous sharp-edged design of the recent modern Chinese designed HP-30B.

While I have the same wish as you, my understanding is that those tiny bubble LEDs are not available any more. Teenix had to use a big LED display in his multicalculator kit (http://www.teenix.org/), which seems not as beautiful as the Classic LED display.

[Harald]

(04-01-2018 03:43 AM)emersone12 Wrote:  

(03-31-2018 08:13 PM)Harald Wrote:  The Classic Upgrade Kit is designed in such a way that the modification is completely reversible. The hole in the battery compartment is optional, and if you don't cut it, all you have to do is open the calculator and swap the main PCB.
Leving the keyboard and display PCB alone sacrifices the power savin shut down. This is something I thought about long and hard. In the end I thought sacrificing functionality to have a completly reversible modification seemes like the right compromise.

Harald

Hypothetically. would you be against the production of a DM35, DM45, DM55, DM65 etc.....or a universal single DM_Multi-calculator accommodating all the available range of calculators, involving a collaboration/arrangement using your boards/kits. The existing DM10,11,12,15,16 are pretty much close replicas of the original HP's in function, layout, physical appearance, size & case shape, & equivalent identical-shaped replicas of the old original HP LED calculators as modern LED reproductions would be a welcome addition if upgraded using modern technology, LED's & battery systems. The old HP case shapes are attractive well designed works of art & very comfortable to hold compared to the ridiculous sharp-edged design of the recent modern Chinese designed HP-30B.

Well, that's a question you'd have to ask Tony, not me. My part in the classic upgrade kit is the PCB design and I programmed the controller that interfaces between the HP display drivers and Tony's PIC. None of that would be needed for a reproduction calculator. It would essentially be Tony's hardware and software in a nice shell.

Cheers,
Harald

PS: I would definately buy one if anyone cares to make it.

[teenix]

I looked into OLED display types early on.

Something like the attached display is almost ideal and it may even fit inside a stripped Classic case, being about 70mm wide. The lower custom 5 x 5 digit size is about 0.5mm higher than the original LED displays. On average it would draw about 20mA at 50% duty cycle which is not too bad.

The downside, is the cost per display is about 5 times that of a 15 digit 5mm high LED display and a whole line would be wasted. I haven't been able to find a single line display similar to this.

cheers

Tony

[pier4r]

(04-01-2018 01:55 PM)Maximilian Hohmann Wrote:  But I can see a development there. In the 80ies it was decisions from the mathematics geniuses (not engineers and certainly not marketing people for sure) who put HP calculators onto that RPL track which moved them away from the all-important education market and from a big portion of their "fan base" for good

For the little understanding that I collected, I think that the RPL was designed as natural evolution of RPN, that was the core of most if not all HP calculators. I guess the idea was exactly to retain the user base while moving on something a bit more expressive.
The problem is that while RPN and RPL are easier for the calculator to execute, as the parsing of a statement is mostly done by the user, they are harder to read and use (more info: here ).

What surprises me though is the recent realization that the HP did a quite great math environment equipped with basic, the 71B (and the 75 if I am not mistaken). The 71B was also based on the saturn chip that then was improved and used in the 48 series.
So while I appreciate the work on RPL, that is quite a nice language but frustrating to learn at first, I don't understand why they didn't continue with the 71B basic.
I do not own a 71B, but as far as I read the 71B basic is even more readable than ti basic. Could be that with it HP could have dominated the calculator market in general.

Anyway, now they made the Prime also as smartphone app, they couldn't do a better move in my opinion.

[John Keith]

The Prime is arguably more of a successor to the HP71B, its programming language is similar to modern BASIC dialects. It's nice that the Prime is also about 500x as fast!

[rprosperi]

(04-02-2018 10:39 AM)pier4r Wrote:  I do not own a 71B, but as far as I read the 71B basic is even more readable than ti basic. Could be that with it HP could have dominated the calculator market in general.

The BASIC dialect in the 71B is easily the most friendly and flexible of any I've used; the only thing it really lacked was allowing long variable names, but this was not done to keep program memory usage down (remember those days?).

But, note that the BASIC in the 75 was quite different, based on the BASIC in Series-80, which shared the same CPU design. While there are many similarities, there is no question that 71B BASIC is more flexible, powerful and easier to use.

Sadly, AFAIK, there was no follow-on to the 71 BASIC in any other machine family. Some folks have commented that some of it's better features are common to Rocky Mountain BASIC which pre-dated the 71B (and also survived well after) though I've never used it so can't comment specifically. RMB came from HP in CO (hence the name) and 71 BASIC from Corvallis, so if there is any commonality, it's likely by inspiration rather than evolution.

[pier4r]

For my little searches the RMB (Rocky mountain basic) evolved in htbasic, available here: http://transera.com/downloads/

But as wrote in another topic http://www.hpmuseum.org/forum/thread-10318.html , it is unsure whether RMB has a math library like the 71B + roms or not.

Surely the 71B + rom impressed me, sadly there was no follow up. Well there is some sort of follow up, the 71B and Basic by hrastprogrammer for the 50g.

[rprosperi]

(04-02-2018 09:28 PM)pier4r Wrote:  Surely the 71B + rom impressed me, sadly there was no follow up.

You may not realize it but the 48/49/50 series, plus all the Pioneers, all used the core math routines written initially for the 71B when the Saturn CPU was new. There have certainly been many tweaks, enhancements, bug-fixes, etc. but the core IEEE Math routines created in Saturn assembler form the basis of the actual math routines used for the next 30 years. One of the original developers has commented that it's likely many of those routines survive untouched all the way through the 50g.

So, yes I agree it's too bad there was no follow-on BASIC product, but the core proved to be a useful framework to bolt RPL onto, and build Pioneer machines around (both Algebraic and RPN families.)

[pier4r]

(04-02-2018 09:55 PM)rprosperi Wrote:  You may not realize it but the 48/49/50 series, plus all the Pioneers, all used the core math routines written initially for the 71B when the Saturn CPU was new. There have certainly been many tweaks, enhancements, bug-fixes, etc. but the core IEEE Math routines created in Saturn assembler form the basis of the actual math routines used for the next 30 years. One of the original developers has commented that it's likely many of those routines survive untouched all the way through the 50g.

So, yes I agree it's too bad there was no follow-on BASIC product, but the core proved to be a useful framework to bolt RPL onto, and build Pioneer machines around (both Algebraic and RPN families.)

I was pretty sure that the work done on the 71B and the 15C was ported in newer versions (even in the Prime? I read somewhere that they port the code slowly).
As you correctly identified, though, my point was about a basic system with such libraries. It would have been superior to the current ti/casio basic versions in my opinion.

[John Keith]

(04-02-2018 10:14 PM)pier4r Wrote:  I was pretty sure that the work done on the 71B and the 15C was ported in newer versions (even in the Prime? I read somewhere that they port the code slowly).
As you correctly identified, though, my point was about a basic system with such libraries. It would have been superior to the current ti/casio basic versions in my opinion.

The Prime's CAS is an updated version of the CAS on the HP49/50, and the Prime's Home screen duplicates much of the functionality of the RPL machines. Since the Prime does not use the emulated Saturn architecture, though, I would think that the internal code is quite a bit different.

[pier4r]

(04-03-2018 12:52 PM)John Keith Wrote:  The Prime's CAS is an updated version of the CAS on the HP49/50, and the Prime's Home screen duplicates much of the functionality of the RPL machines. Since the Prime does not use the emulated Saturn architecture, though, I would think that the internal code is quite a bit different.

As I wrote, I read somewhere from an HP dev member that they go through the quite time consuming process of checking the sources (even from machines like the 15c) and rewriting them.

If what I remember is true, they have a wealth of solutions right there (15c, RPL up to 50g, 42S, 41, etc..), but it takes time to port them (and test them).
And sincerely I love the fact that the prime:

(a) Is optimized for the new hardware, instead of simulating hw (that is necessarily slower). Indeed see the hopes for newRPL on the 50g. While hpgcc3 is already "finished" but there are less people using it.
(b) it takes time to port the functionality of previous models but slowly it will be more and more mature, since the hw has a lot of power.
(c) can be compiled on multiple platforms, having even more power (android I am looking at you).
(d) has an easier to handle language (pascal like rather than basic like), although I love RPL. There is no doubt that basic/pascal proved more user friendly to large groups of users.

And my point was. If they would have kept a sort of 71B branch in their models (basic), likely HP would have gotten way more market share.
I mean, in my opinion they could have kept producing the 71B itself, I don't see why it was discontinued in the first place.

[rprosperi]

(04-04-2018 09:02 AM)pier4r Wrote:  ...If they would have kept a sort of 71B branch in their models (basic), likely HP would have gotten way more market share.
I mean, in my opinion they could have kept producing the 71B itself, I don't see why it was discontinued in the first place.

The 71 was a not a big seller, and it was discontinued relatively early like any other model that doesn't sell well. It was not popular with a lot of HP's customers because it was expensive (nearly twice as much) and also due to its large, landscape form factor. HP intended the 71B to replace the 41C family as a platform for VARs and custom application developers, which is why they spent a huge amount of time and money creating the IDS volumes to fully document everything, but most VARS stayed with the 41, as can be easily seen by the number of 41 custom ROMs, compared to the number of 71B custom solutions (many hundreds vs. tens).

Its hard to imagine that HP could have seen how far the Saturn CPU would go (early 80s to the 2000s) though RPL, based on '71 FORTH was created specifically as a platform for 'all future calculators'. So while the 71 itself did not last very long, its DNA served HP well for a long time.

[pier4r]

I still miss a lot of history, indeed I recently appreciated the 71b. Anyway I'm surprise there even with basic and a lot of ram available, people kept using the 41c .

I mean the 41c is great and all, anyway basic proved being more user friendly. I would have imagined the 71b selling like bread .
It is really a surprise.

[rprosperi]

(04-04-2018 01:31 PM)pier4r Wrote:  I still miss a lot of history, indeed I recently appreciated the 71b. Anyway I'm surprise there even with basic and a lot of ram available, people kept using the 41c .

I mean the 41c is great and all, anyway basic proved being more user friendly. I would have imagined the 71b selling like bread .
It is really a surprise.

We view the 71b from today, with appreciation for documentation, flexible I/O capabilities, easy expansion to beyond what people can afford, etc. most of which were new innovations that were not as appreciated at the time.

Also, many (most?) VARs did not trust selling their valuable and expensive programs written in BASIC, as the security was easily broken and programs (and even full ROM modules) were easily copied, so it was very hard to prevent theft and copying by unscrupulous customers, many of whom had no idea they were breaking the law.

[Maximilian Hohmann]

(04-04-2018 01:31 PM)pier4r Wrote:  Anyway I'm surprise there even with basic and a lot of ram available, people kept using the 41c .

As someone already wrote above, the 71B was very expensive. Breathtakingly expensive. I remember from my university days in the early/mid 1980ies (when I would not have been able to buy a 41C which cost around 600-800 German Mark depending on which accessories one took) that I used to stop at a surveying supply shop which was on my way and look at their display of HP calculators. When the 71B came out it had a price tag short of 2000 German Mark, three times as much as a 41. No wonder they sold so few. The competition (Casio, Sharp) had BASIC programmable calculators similar in size and basic features of the 71B which sold for 1/4 or less.

[pier4r]

(04-04-2018 02:43 PM)rprosperi Wrote:  Also, many (most?) VARs did not trust selling their valuable and expensive programs written in BASIC, as the security was easily broken and programs (and even full ROM modules) were easily copied, so it was very hard to prevent theft and copying by unscrupulous customers, many of whom had no idea they were breaking the law.

Thanks is a big point indeed

(04-04-2018 04:21 PM)Maximilian Hohmann Wrote:  As someone already wrote above, the 71B was very expensive. Breathtakingly expensive. I remember from my university days in the early/mid 1980ies (when I would not have been able to buy a 41C which cost around 600-800 German Mark depending on which accessories one took) that I used to stop at a surveying supply shop which was on my way and look at their display of HP calculators. When the 71B came out it had a price tag short of 2000 German Mark, three times as much as a 41. No wonder they sold so few. The competition (Casio, Sharp) had BASIC programmable calculators similar in size and basic features of the 71B which sold for 1/4 or less.

Woah! That puts things in perspective. I thought that was as expensive as the 41C, being the 41C already expensive

[Waon Shinyoe]

(04-04-2018 04:21 PM)Maximilian Hohmann Wrote:  The competition (Casio, Sharp) had BASIC programmable calculators similar in size and basic features of the 71B which sold for 1/4 or less.

I think the competition from Japanese pocket computers (ポケコン) is one reason for 71B’s market failure.

At lease this is the situation in my country (China) during the 80s... it was all about SHARP PC-1500 and you couldn’t find one HP... PC-1500 was so widely used in surveying/civil engineering/military that old PC-1500s are still flooding the market today for 100 ~ 200 yuan (20 ~ 30 US dollar)

And 71B (Saturn) is way slower than 75C (Capricorn)... And the screen is smaller... It just “seems” like a step backwards in technology...

[rprosperi]

(04-04-2018 09:31 PM)Waon Shinyoe Wrote:  And 71B (Saturn) is way slower than 75C (Capricorn)... And the screen is smaller... It just “seems” like a step backwards in technology...

It was slower (for most, but not all, number-crunching than the 75C, but for I/O it was MUCH faster) to keep battery life long; more than twice that of the 75C. But it didn't take long for PPC-ers to figure out how to jack up the clock speed.

The LCD screen was kept small to keep it portable, but I really, really wish they had included a 2-line display.

Both of these changes came partly due to customer 'feedback' (aka complaints) about the 75C introduced a year earlier.

But the introduction of the PC and its explosive growth in the same time period as both these machines, radically changed people's view of what 'computer' means.

[pier4r]

Hmm. Are there sharp 1500 clones?

my English is already broken but touchscreens and autocompletion don't help.