The Museum of HP Calculators

HP Forum Archive 13

[ Return to Index | Top of Index ]

Calculators vs Adding Machines
Message #1 Posted by Katie on 25 Aug 2003, 12:42 a.m.

When talking about my calculator collection, I often get asked "what's the difference between a calculator and an adding machine?". It seems like a naive question and usually comes from people that don't much about calculators, but I've come to think that it's a good question that's not so easy to answer.

According to at least one dictionary an adding machine is:

a machine capable of adding numbers and sometimes capable of performing the other arithmetic functions of subtraction, multiplication, and division: such machines are now obsolescent, having been replaced in most applications by electronic calculators.
Which is not particularly satisfying to my mind as it doesn't distinguish it from a calculator and implies that we would not encounter them often, yet modern business are filled with what people still call "adding machines".

I don't really have a good answer to this question, but I think it should be more along the lines of who the intended user of the machine is. To me, an adding machine is a calculating device intended to be used primarily by accountants. Along with this definition comes the usual adding machine entry logic, sort of half prefix/half postfix operators -- but not always. Early mechanical machines only added, some subtracted and still later ones did multiply and divide and even square root but they all used difference operator orders and they were/are all called adding machines.

So, what's the right definition?

Re: Calculators vs Adding Machines
Message #2 Posted by Vassilis Prevelakis on 25 Aug 2003, 1:34 a.m.,
in response to message #1 by Katie

I would have thought that an adding machine is a mechanical calculator.

So at one stage we had mechanical and electronic calculators, but as the former became obsolete, the term calculator evolved to imply electronic.

Of course people will come up with examples that blur the distinction, but like in the case of calculators vs computers there is no clear cut distinction.

E.g is an HP-34C a calculator or a computer? If you say a calculator, then what if I now replace its ROM so that I can enter machine code via the keypad. Now its a computer for sure.

If you insist on input/output devices, then consider an HP-97. This one has both. Is it a computer?


Re: Calculators vs Adding Machines
Message #3 Posted by Pieter Botha on 25 Aug 2003, 2:17 a.m.,
in response to message #2 by Vassilis Prevelakis

Interesting Issue. I've always wondered when a calculator crosses the "great divide" to become a computer.

Let's try to define what a computer should look like, then measure the HP-97 against those requirements:

(*) Input: Yes - Keyboard, Magnetic Cards

(*) Output: Yes - Display and hardcopy

(*) Processing: Yes - Maths primarily, I'd say too weak on strings :^)

(*) Programmability: Yes

(*) Decision making: Yes - Several conditional branchings

(*) Long-term storage: Yes - Magnetic Cards

Looks like, apart from scale (more is better), the Pentium 4 on my desk offers nothing that the HP-97 doesn't. Oh, sure, it can do fancy-schmancy graphics (windows etc), but that doesn't make it more of a computer. Or does it ?

Something like the TI-58, though, I would hesitate to call a "computer", simple because of the lack of off-line storage facilities, though it can produce hard copy with the optional printer. Guess that one straddles the line...

What do other people think ?

Cheers Pieter

Re: Calculators vs Adding Machines
Message #4 Posted by Katie on 25 Aug 2003, 2:19 a.m.,
in response to message #2 by Vassilis Prevelakis

To keep this discussion relevant to the HP world (and this forum), Dave has called the HP-10 an adding machine on this site:

HP-10 Handheld Printing Adding Machine
So, it's not a moot point by any means nor does "adding machine" just refer to mechanical devices. There is a blur between computers and calculators, I agree and there is no doubt a blur between calculators and adding machines. But why do (we) experienced (very experienced in Dave's case) collectors call some devices calculators and others adding machines when they are similar in many ways (e.g., HP-10 and HP-19C).

Dave, how come you called the HP-10 an "adding machine". Did HP use that term in their advertising literature? In the manual they call it a "handheld printing calculator".

Re: Calculators vs Adding Machines
Message #5 Posted by Vassilis Prevelakis on 25 Aug 2003, 6:38 a.m.,
in response to message #4 by Katie

So should we call the HP-10 an adding machine?

Well if I was trying to sell a printing 4-function calculator to people used to buying adding machines, I'd sure call it an adding machine. It is functionally equivalent to a mechanical adding machine, so you might as well call it an adding machine.

Doesn't this discussion remind you of the horseless carriage?

I'd be tempted to say that 4-function non-RPN machines are adding machines, but Dave calls the HP-01 "The Hewlett-Packard Calculator Watch" so that one is not considered an adding machine.

So, since the 19C is a calculator, your (electronic) adding machine should be both brain-dead and print.


Re: Calculators vs Adding Machines
Message #6 Posted by Masao Kinoshita on 25 Aug 2003, 10:44 a.m.,
in response to message #5 by Vassilis Prevelakis

I believe the term calculator originally referred to somebody who performed calculations. Then there were adding machines and then electronic calculators.

Horseless carriage, automobile, car and ride all can mean the same thing.

Therefore, the first "personal computer" was an aid who performed computations for somebody, e.g., king or astronomer, a long long time ago. :-)

Re: Calculators vs Adding Machines
Message #7 Posted by Dave Hicks on 25 Aug 2003, 1:05 p.m.,
in response to message #4 by Katie

That's my attempt to capture the "essence" of the 10 and distinguish it from others on the front page in a short caption. It's not precise but I don't know what is. In the article I use "adding machine logic" but that's a term I invented. It seems that people who sold/sell these machines don't feel a need to label the logic system. They just evolved and everyone expects them to be as they are.

The HP10 is called an adding machine. (?)
Message #8 Posted by Vieira, Luiz C. (Brazil) on 25 Aug 2003, 2:17 a.m.,
in response to message #1 by Katie

Hi, folks;

this was extracted form the MoHPC page dedicated to the HP10:

The HP-10 was modeled after a standard desktop "10 Key" adding machine. Adding machines use their own style of logic. To add numbers on an adding machine, the user clears the machine (either by an explicit clear command or a previous total command,) enters the first number, presses the + key which adds it to zero, enters the second number and presses + again to add it to the previous result. For adding and subtracting, an adding machine is much like RPN, except that the first number is added to zero rather than Entered.

And about being a calculator or a computer, it has a printing device. Is it a unique adding/calculator/computer machine in one single package?

I agree with Katie that there is a lack of definition for adding machines, and I also have my belief, like Vassilis posted, that it is supposed to be a mechanical device as well, but if we take the word "adding" as it is, then addition is what every computer does when performing any arithmetic operation, exception made to inverting (NOT) a number or comparing it (XOR). Before any complains, I know NOT and XOR are logical operations and should not be considered in this case, but logical operations are the heart of a classical computer and I thought I should mention them.

Babbage developed and created the difference machine; was it an adding machine as well? Based on operands and reasoning over the answer, it may perform additon, too.

Another point: should an adding machine be considered as an analog-based device, while calculators should be considered as digital machines? This way, ENIAC was an electronic version of an adding machine. Or not?

No answers from me, I know. Anyway, "add" my name to the list: I'm curious and I'll wait for a reasonable answer.


Luiz (Brazil)

Edited: 25 Aug 2003, 3:03 a.m.

Addign Machine -- Calculator
Message #9 Posted by Randy Smith on 25 Aug 2003, 4:49 a.m.,
in response to message #8 by Vieira, Luiz C. (Brazil)

To me the difference between the two is in the output. I call it an adding machine if it has only a printed output and no display of the answer on the machine. It is a calculator if there is an output of the answer on the machine itself exclusive of having the answer on a paper tape.

Re: Calculators vs Adding Machines
Message #10 Posted by Andrés C. Rodríguez (Argentina) on 25 Aug 2003, 9:04 a.m.,
in response to message #1 by Katie

Just my $ 0.02:

Adding machine:

Main function is add-subtract, perhaps with some kind of memory for totals. It may have multiplication and division functions, but they look as later "additions" (pun intended), usually not well integrated with the main function. Paper tape output is convenient to check both results and sequence (completeness of input).


User-programmable device, must have some kind of "open" Input/Output capability (i.e.: peripherals, RS232) AND the ability to process (at least) two types of data: numeric and character-based.

The HP41 qualifies (to increase its qualification, just add HP-IL), as the HP48, HP75, HP71 do.

The HP42S is not a computer, due to the lack of enough I/O capabilities and openness.

The HP 97 and TI 59 are "almost computers", the HP 97S just fall a little short of my yardstick to be considered a computer.


Specialized device, oriented to mathemathical, financial, statistics, computer science... calculations. May be programmable or not, the preprogrammed functions go from the basic four operations to much many and very complex ones. The HP42S and the HP15C are examples of calculators.

By the way, the HP01 watch can use the real time as a variable for calculations, which sets it far from an adding machine. MoHPC definition "In a class by itself" is the best I can think about.

Re: Calculators vs Adding Machines
Message #11 Posted by Paul Brogger on 25 Aug 2003, 10:56 a.m.,
in response to message #1 by Katie

Generally speaking . . .

Most of the things that I encounter in the thrift stores and that I call "adding machines" have a combined "plus-equals" key, and seem to be optimized for serial summation ("adding").

The things I call "calculators" have separate "plus" and "equals" keys.

. . . except, of course, for RPN calculators, which have separate "plus" and "enter" keys. (But then, I don't encounter those very often -- neither in thrift stores nor anywhere else!)

I don't suppose this is a definition so much as an observation, but it may point in some useful direction.

Re: Calculators vs Adding Machines
Message #12 Posted by Mark Glusker on 25 Aug 2003, 11:08 a.m.,
in response to message #1 by Katie

I think you have to go back fifty years to answer that question, when calculators were all mechanical.

There were rotary calculators, like the Friden STW in the MoHPC collection, that performed all four functions (+ - x ÷) and adding machines that performed addition and subtraction only. Rotary calculators had mechanical displays and adding machines usually had a paper tape.

After WWII, several manufacturers enhanced their adding machines to enable them to multply and divide, giving them the name "printing calculators" to distinguish them from adding machines. Articles from that era are clear about this distinction, because it was a big deal back then.

My definition is that a calculator can add, subtract, multiply and divide. An adding machine can only add and subtract. However, I have heard the term "adding machine" applied to anything that has a paper tape.

Here are the distinctions that I make
Message #13 Posted by Dave Hicks on 25 Aug 2003, 12:30 p.m.,
in response to message #1 by Katie

Machine vs. Electronic
This is pretty simple. Although many people will refer to computers as machines, I think you know the distinction I'm making here. Some machines are electrically driven though the logic is mechanical.

Adder vs. Calculator
This is a little more fuzzy but there are devices that make division and multiplication "easy" and those that don't. Mechanical calculators usually have a carriage that can be shifted. For example to multiply a number by the number 1024, you would enter the first number, then turn the crank 4 times, shift the carriage, turn twice, shift the carriage twice (skipping the zero) and then turn the crank once more. Electrically driven machines can do all this at the push of a button. Even the basic adding machine sometimes provides at least a little support for multiplication however. For example, you enter your number and press the repeat button down and pull the lever 3 times. Push the repeat button back and pull the lever again (number * 4 is now entered). Now enter the number times 10 and you can figure out the rest...

These distinctions form a 2x2 grid with the boxes being

Mechanical Adder (or Adding machine.)
There are many examples of these for example...

These devices frequently (but don't always) print.

Mechanical Calculator
There are many examples of these too but the general public is almost completely unaware. These devices usually don't print. (I guess only a banker needs an "audit trail".)

Electronic Adders.
While theoretically possible, I can't think of a good example. Because electronics are so cheap and calculator functions are now generally a matter of software, there seems to be little reason to leave multiply and divide out. Having said that, someone's probably got one for sale on ebay. Perhaps built into a ruler, keychain or a mouse ;-)

Electronic Calculators.
I'm sure you know what these are.

The above distinctions are different than the question of what logic system is used (though there's a lot of overlap in practice.) Most machines are essentially postfix with around two to three registers (not really a stack). One of the registers is frequently the keyboard or the key levers.

Electronic calculators come in Algebraic (a term covering many logic systems), RPN/L, and what I call "adding machine logic". The latter refers to electronic calculators for accountant-type functions. These calculators (usually with printers) do + and - in a postfix way (clear x + y +), * and / in an infx way (x * y =) and are usually now electronic though with a mechanical impact print head that produces more or less the same printout that adding machines have produced for centuries.

As for computers vs. calculators, I assert that almost all modern calculators are computers which are programmed to behave like calculators. This trend began when Busicom contracted Intel to build the chips for an early electronic calculator. Intel thought the design was too complex (27 different chips) and counter-proposed a 4 chip design (ROM chip, RAM chip, I/O chip and... the first microprocessor that would later be called the Intel 4004.)

Re: Does anyone remember the link to the "minimum calculator"?
Message #14 Posted by Paul Brogger on 25 Aug 2003, 9:48 p.m.,
in response to message #13 by Dave Hicks

I'm trying to chase down an example of "Electronic Adder" to complete the table implied by Dave's analysis (above).

It seems to me that, not too long ago, there was a discussion thread that devolved to a topic something like: "What is the minimum number of keys ever implemented in an electronic calculator?"

Someone posted a link to a checkbook-balance tracker that had VERY few keys -- maybe as few as fourteen. I remember something like plus, minus, clear and ten digits . . . Anyway, I've been searching the Forums for an hour and haven't found the post -- maybe someone remembers?

Re: Does anyone remember the link to the "minimum calculator"?
Message #15 Posted by Katie on 25 Aug 2003, 10:28 p.m.,
in response to message #14 by Paul Brogger

Here's the device that you're referring to, I think:

Not quite a calculator, not quite an adding machine either.

Edited: 25 Aug 2003, 10:30 p.m.

Thank you, Katie.
Message #16 Posted by Paul Brogger on 26 Aug 2003, 1:06 a.m.,
in response to message #15 by Katie

Yezzz, that's it.

I picked up a TI-2500 "Datamath" a few months ago -- four operations and no memory, but at least it could multiply & divide. Just recently got a Litronix "Checkmate 1700" -- same story.

But that Checkmaster thingy is about as simple as they can come.

Thanks for remembering!

Re: Does anyone remember the link to the "minimum calculator"?
Message #17 Posted by Victor Koechli on 26 Aug 2003, 2:57 p.m.,
in response to message #14 by Paul Brogger

Out of the buttons of my Nokia cell phone, only 13 are used to accomplish the four basic functions. Quite minimal, don't you think? Let's see if someone can come up with even fewer...

Cheers, Victor

Re: Calculators vs Adding Machines
Message #18 Posted by christof (NoVA US) on 25 Aug 2003, 1:44 p.m.,
in response to message #1 by Katie

I've been thinking about this off and on all day.

I think, growing up, that there are a few related differences-

Adding Machine was something my father use for accounting. paper tape, LED display, desktop, generally no carrying system. It stayed on a desk.

From the time my unlce started buying me cheap radio shack pocket calculators in the late 70s, 'calculator' meant a pocketable machine, sometimes with programming capability, almost always with 'special' functions (statistics, 'scientifics', and business were the distinctions I made then. many models had combinations of those)

When I first saw a desktop scientific, I knew it wasn't an adding machine- it was a calculator. I had not codified the differences at the time, but I knew.

Printing alone doesn't make an adding machine, nor does financial functions. Even the desktop form factor doesn't do it. All three together do, though.

regarding computers- I disagree that a computer must have rs-232 and/or a card reader to be a computer.

The HP42S is most defintiely a computer. It has all the requisite parts, and they all work quite well. Lest anyone forget- keyboard entry, videographic display, and printed paper output via IR *are* all forms of input or output.

It may be slower to recover your backups from printed paper than from a magnetic card, sure. That doesn't mean it doesn't exist as output.

I'm not arguing that the 42S wouldn't be better with serial IR bidirectional. If I could get my hands on a properly disassembled 42 ROM image, I'd be tracing code to help find a way to get a good serial IR addition. But that's not this thread :)


Re: Calculators vs Adding Machines
Message #19 Posted by Andrés C. Rodríguez (Argentina) on 25 Aug 2003, 7:15 p.m.,
in response to message #18 by christof (NoVA US)

The HP42S is perhaps the nicest machine I have (I like my 41C and my 25 too). Certainly it has some interface capabilities, but (in my humble opinion) it lacks some kind of OPEN I/O to be considered a computer.

RS232, USB, HP-IL, IrDA, SCSI, IEEE 488, an Ethernet port, a Compact Flash slot, etc., etc., are just some examples of OPEN I/O: a place where SOMEBODY else could connect an external device. The proprietary infrared stream HP uses for its printer (in my humble opinion) does not qualify as open enough.

Sure, there are forms we have discussed here to use these IR pulses, decode TONEs, and simulate keyboard operation to get some I/O, but I think that is not enough.

Again, it is just my opinion.

Re: Calculators vs Adding Machines
Message #20 Posted by christof (NoVA US) on 25 Aug 2003, 7:26 p.m.,
in response to message #19 by Andrés C. Rodríguez (Argentina)

I understand the need and importance of open standards. But I must say I disagree that an OPEN input/output system or lack thereof determines whether or not a device is a computer as opposed to a not-computer.

And i've worked on several other computers that lacked everything you mentioned :)

(how much more open you can get than english text on thermal paper, I don't know- HOw the data gets to the text isn't the important part. the printed output is.... output)

Trivial, perhaps. But I'm occasionally frightened by the One True Way approach to defining *all* computing using only *one* set of standards (OPEN)

Cheers- Christof

Re: Calculators vs Adding Machines
Message #21 Posted by Andrés C. Rodríguez (Argentina) on 26 Aug 2003, 9:26 a.m.,
in response to message #20 by christof (NoVA US)

Perhaps I hadn't found the proper way to express it in English, but I was not supporting any *one* set of standards. In fact I mentioned various interfacing methods, some more standarized (RS232, USB, IEEE 488), some more proprietary (HP-IL)...

I think (and that is just my opinion) that a computer is a general purpose, programmable digital device; and as such the ability to connect to external devices, even from other manufacturers, is an important feature. A proprietary I/O bus, well documented, may be enough. The user hands and eyes are not sufficient (in my humble opinion). Of course, there will always be "border" cases.

Again, I was just trying to suggest one criterium; and my intention was far from suggesting there is only *one* way to look at *all* things.

While I am surprised because of such unintended side effect of my posting, I apologize if my comment produced a "frightening" effect.

Re: Calculators vs Adding Machines
Message #22 Posted by christof (NoVA US) on 26 Aug 2003, 6:22 p.m.,
in response to message #21 by Andrés C. Rodríguez (Argentina)

You have to tak einto account that I work in a field populated with large numbers of religious folk who overcapitalize words like OPEN, SOURCE, STANDARDS, GPL, and etc. Many of these people would like to make illegal any form of computer or software they disapprove of based on their spec ific value of sufficient 'openness'

So that probably explains it part of it :)

The other part is- well, the CS definition of a general purpose computer *sodes* include a requirement for input and output, but all the old CS classes also stress that printouts and text entry are valid forms of input and output.

True- making a computer networkable generally requires more, but networkable isn't a defining characteristic of a general purpose progammable computer. (special purpose and non programmable computer exist) Actually- let me state that it wasn't in 1988. Nowadays, maybe it is.

I'm just being a computer nazi. It's my job, actually. :)

no worries-


[ Return to Index | Top of Index ]

Go back to the main exhibit hall