ENTER↑ is a misnomer, long live ENTER:

[hansklav]

First there was the anonymous large key with the upwards pointing arrow ↑ in the HP 9100A and 9100B, then when the HP-35 was created and this key's function changed a bit it had to be given a name, and it was christened "ENTER" and this name appeared next to the upwards pointing arrow on the largest key of the HP-35: ENTER↑

At the time (1972) this was a perfectly logical name and certainly not a misnomer for this key, because it paralled the use of the other key with the word "Enter" in its name, the "Enter Exponent" key "E EX", which was called "ENTER EXP" in the 9100 models. Because "E EX" should be pressed before keying in the exponent of ten it is always used in a prefix way. And ENTER in a classical RPN calculator should also always be used in a prefix way, before keying in the next number (clever tricks like using it for doubling or squaring a number left aside).

Suppose you want to add 2 and 3 using a classical RPN calculator:
2 ENTER 3 + (Answer: 5)
If you use ENTER in a postfix way (not too far fetched in a calculator that uses RPN or postfix logic) you get the wrong results for your calculation, without warning:
2 ENTER 3 ENTER + (Wrong answer: 6)

So in 1972 ENTER↑ was not a misnomer for the largest key of the HP-35.

But even then misunderstanding was lurking. In the (in many respects admirable and beautifull) HP-35 Operating Manual one can read sentences such as: But at the start we've got to get the first number into the machine. To do this you key in the first number and press ENTER. Now key in the next number and then press +, -, x or ÷. This already could give the impression of the wrong postfix meaning of ENTER. And the concept of an "automatic ENTER" was introduced to explain what was later called the 'stack lift': an "automatic ENTER" was supposed to be performed after every operation when a new number was keyed in. This also gave a postfix feeling to ENTER.

People at HP felt that ENTER was not the right name for the large key after all, and in the second handheld calculator, the HP-80, the key was rechristened "SAVE". This was the first real misnomer, because saving keyed-in numbers and results from calculations in the RPN-stack is always done automatically. And SAVE has an even more postfix connotation and could more easily lead to wrong postfix use of the key by newbies. So SAVE was dropped as name for the key in the next model and ENTER has remained for all later models (until INPUT was used in the 20b and 30b).

At the time when the first RPL calculator (the HP-18C) was developed in 1986 the use by many people of personal computers, with their Enter keys, gave most new users an intuitive postfix feeling for an ENTER key on a calculator, and HP made the logical switch by changing the function of ENTER such that 2 ENTER 3 ENTER + gives the same result als 2 ENTER 3 +. This use of the ENTER keys is also known as 'Entry RPN'.

Now ENTER has become the proper name for the key in RPL and other 'Entry RPN' calculators, such as the 20b and 30b. So to me it is an incomprehensable decision by HP that they have renamed the key in the 20b and 30b to INPUT.

But ENTER now has become a misnomer in classical RPN models. A better name would possibly be 'E NXT #' (for 'ENTER NEXT NUMBER'), but of course it is too late for such a drastic change. A more subtle change would be ENTER: , which to new users conveys the prefix meaning very well thanks to the colon after the word ENTER, and which would mean only a subtle change from the upwards pointing arrow ↑ after ENTER.

So this is a plea for ENTER: in stead of ENTER↑ on the WP 43S and all future classical RPN calculators to make clear the distinction between the ENTER used in these classical RPN models as opposed to the other version of ENTER in RPL and other 'Entry RPN' models.

In this way we provide more clarity and so might convert more people to the church of RPN!

Hans

[Katie Wasserman]

Interesting take on ENTER. I mostly agree, but just assume have a big, unlabeled key in a different color from all the rest.

I think you have a typo in your post:

Quote: first RPL calculator (the HP-18C)

This should be: HP-28C.

[hansklav]

(07-06-2014 04:38 PM)Katie Wasserman Wrote:  (…)
I think you have a typo in your post:

Quote: first RPL calculator (the HP-18C)

This should be: HP-28C.

It was not a typo. My source is the MoHPC.
According to the MoHPC the HP-18C was launched in 1986, and the HP-28C in 1987.

But I read that the HP-18C only used 'System RPL', and I don't know anything about the difference in the behaviour of the ENTER key between 'System RPL' and 'User RPL', so you might be right that the HP-28C was the first with 'Entry RPN'.

Hans

[Katie Wasserman]

(07-06-2014 05:18 PM)hansklav Wrote:  But I read that the HP-18C only used 'System RPL', and I don't know anything about the difference in the behaviour of the ENTER key between 'System RPL' and 'User RPL', so you might be right that the HP-28C was the first with 'Entry RPN'.

That's correct, user RPL was introduced on the 28C. But it's the user ENTER, not system ENTER, that's relevant to your post.

[rprosperi]

(07-06-2014 05:41 PM)Katie Wasserman Wrote:  

(07-06-2014 05:18 PM)hansklav Wrote:  But I read that the HP-18C only used 'System RPL', and I don't know anything about the difference in the behaviour of the ENTER key between 'System RPL' and 'User RPL', so you might be right that the HP-28C was the first with 'Entry RPN'.

That's correct, user RPL was introduced on the 28C. But it's the user ENTER, not system ENTER, that's relevant to your post.

You're both correct, sorta. The 18C came first and it's OS is SysRPL, although this was not accessible to a user for programming, while the 28C was the first machine with user access to RPL for programming.

[Thomas Klemm]

From the Hewlett Packard Journal (August 1987, p. 9):

Quote:Of all the decisions made by the design team regarding
the user interface, the one that was by far the most difficult
(as well as the most controversial) was the one that made
the HP-18C operation algebraic, rather than RPN.

The HP-18C has an [INPUT] key.
There's no ENTER command in UserRPL nor in SysRPL. But there's DUP.

Cheers
Thomas

[walter b]

I respectfully plea for the right to disagree. ENTER↑ is a name fitting the classic RPN function well (compare e.g. R↓ and R↑, and look at all the pocket calculators from the HP-35 to the HP-41CX). Then the arrow was dropped since it didn't fit the Voyager keyboard; and it didn't return with the Pioneers for whatever reason. The real problem, however, arose with RPL (ab)using the RPN label ENTER for a slightly different function. IMHO it's the duty of the abuser to refrain from abuse - not of classic RPN to make way for it.

Luckily noone ever used ENTER↑ on an RPL or so-called 'Entry RPN' calculator. Thus I think ENTER↑ is perfectly correct on the WP 34S, while ENTER should be replaced by something else on the HP-30b. I doubt, however, that we'll get that - so returning to ENTER↑ on the WP 34S, 31S, and 43S is the minimum differentiation we can do to make people aware of what we're talking about.

d:-/

[hansklav]

(07-06-2014 06:21 PM)Thomas Klemm Wrote:  From the Hewlett Packard Journal (August 1987, p. 9):

Quote:Of all the decisions made by the design team regarding
the user interface, the one that was by far the most difficult
(as well as the most controversial) was the one that made
the HP-18C operation algebraic, rather than RPN.

The HP-18C has an [INPUT] key.
There's no ENTER command in UserRPL nor in SysRPL. But there's DUP.

OK. My fault, I should have looked better. Also the MoHPC says it all (although not in the overview on the home page): the HP-18C has an algebraic logic system and is no RPN-calculator at all.

So the HP-28C is the first one with Entry RPN.

Thanks for the hint,

Hans

[hansklav]

(07-06-2014 07:28 PM)walter b Wrote:  I respectfully plea for the right to disagree. ENTER↑ is a name fitting the classic RPN function well (compare e.g. R↓ and R↑, and look at all the pocket calculators from the HP-35 to the HP-41CX). Then the arrow was dropped since it didn't fit the Voyager keyboard; and it didn't return with the Pioneers for whatever reason. The real problem, however, arose with RPL (ab)using the RPN label ENTER for a slightly different function. IMHO it's the duty of the abuser to refrain from abuse - not of classic RPN to make way for it.

Luckily noone ever used ENTER↑ on an RPL or so-called 'Entry RPN' calculator. Thus I think ENTER↑ is perfectly correct on the WP 34S, while ENTER should be replaced by something else on the HP-30b. I doubt, however, that we'll get that - so returning to ENTER↑ on the WP 34S, 31S, and 43S is the minimum differentiation we can do to make people aware of what we're talking about.

I agree with you that ENTER↑ has the historical right for use on Classical RPN calculators, and ENTER↑ conveys the difference between Classical RPN and Entry RPN quite well.

But this is only when we look back into history. And looking back into history we can also see that HP and RPN might have done better when the key would have had a more future-proof name.

Looking into the future it is important that the key has a name that speaks for itself. Imho ENTER: gives future users of Classical RPN a much better intuitive feeling for the semantic of this key in Classical RPN calculators than ENTER↑.

Only by providing maximum clarity to future user can we hope to attract more users to Classical RPN calculators.

Hans


P.S.

(07-06-2014 07:28 PM)walter b Wrote:  (…) , while ENTER should be replaced by something else on the HP-30b.

The HP 30b already has no ENTER but an INPUT key (with the meaning of Entry RPN ENTER), so ENTER (without ↑ or : ) would have been a suitable name.

[Didier Lachieze]

(07-06-2014 04:30 PM)hansklav Wrote:  At the time when the first RPL calculator (the HP-18C) was developed in 1986 the use by many people of personal computers, with their Enter keys, gave most new users an intuitive postfix feeling for an ENTER key on a calculator, and HP made the logical switch by changing the function of ENTER such that 2 ENTER 3 ENTER + gives the same result als 2 ENTER 3 +. This use of the ENTER keys is also known as 'Entry RPN'.

If you're interested in HP calculator history, we can trace back the origin of the "Entry RPN" before the HP-28C to the HP-41 translator pac for the HP-71b. Bill Wickes who developed the HP-41 translator pac as an "after-hours" project was also the software project manager for the HP-28C/S [1].

In an article published in January 1985 in CHHU Chronicle ( THE HP-41 TRANSLATOR PAC FOR THE HP-71) Bill Wickes introduced the command line and the removal of the stack lift disable:

Quote:The HP-41 is strictly a key-per-function calculator. HP-71 FORTH and BASIC both use a command line approach instead […] The HP-41 translator allows you to use either method. The default mode is command lines. You can type in up to 96 characters of functions or numbers together, each entry separated by a space, then press [ENDLINE] to execute the whole sequence. […] But if you prefer key-per-function operation, you can make the built-in file KEYS41 the active keys file, so that in user mode, any HP-41 function can be immediate execute. [...]

The HP-41 Translator makes a bold break with tradition (here’s how you make a possible drawback into a feature) by not implementing any stack lift disable. Yes, that’s right, the ENTER^ key becomes a vestigial organ not worth including in the built-in keys file. [...]

Consider how a traditional HP RPN calculator handles number entry. [...] To terminate the number entry, you just press any non-numeric function key. The problem arises when you want to enter two consecutive numbers, with no operation in between. Hence the ENTER^ key. But here the HP-35 designers overshot the mark—instead of just having ENTER^ terminate digit entry, they (who knows why?) made it also carry out the unrelated task of duplicating the number into the Y-register, then disabling stack lift. [..]

The HP-41 translator does what the HP-35 and its descendants should have done: it has a special key (in this case the [SPC] key) that has no role in lift except to terminate digit entry (or to separate commands).

This article which is sub_tittled "Bridging the gap" is also interesting because it was published while the HP-28C was being finalized and it illustrates some of the intellectual steps on the road from the 41C to the 28C:

Quote: FORTH is a logical “next language” for HP-41 user language aficionados.

[1] From "HP28 Source, SYSEVAL's and Wickes"
Newsgroups: comp.sys.handhelds
Subject: HP28 Source, SYSEVAL's and Wickes
Message-ID: <17980006@hpcvra.CV.HP.COM>
Date: 4 Dec 89 18:13:54 GMT

[Katie Wasserman]

Quote: But here the HP-35 designers overshot the mark—instead of just having ENTER^ terminate digit entry, they (who knows why?) made it also carry out the unrelated task of duplicating the number into the Y-register, then disabling stack lift.

I don't really know the answer to this, but being an original HP-35 user I thought (and still think) it was to make squaring easier. Since there is no x^2 function on the HP-35 using ENTER * does the trick. In the HP-35 manual they suggest doing exactly this (see page 12). To this day, I still square like this even on RPN calculators that have an X^2 function (well, unless I'm programming and need the extra stack level or memory space).

[walter b]

(07-06-2014 10:40 PM)hansklav Wrote:  The HP 30b already has no ENTER but an INPUT key (with the meaning of Entry RPN ENTER) ...

Sorry, I look at too many (have been) 30bs showing ENTER↑. My fault - or call it my contriobution to get that problem solved.

d:-)

[walter b]

(07-07-2014 02:54 AM)Katie Wasserman Wrote:  

Quote: But here the HP-35 designers overshot the mark—instead of just having ENTER^ terminate digit entry, they (who knows why?) made it also carry out the unrelated task of duplicating the number into the Y-register, then disabling stack lift.

I don't really know the answer to this, but being an original HP-35 user I thought (and still think) it was to make squaring easier. Since there is no x^2 function on the HP-35 using ENTER * does the trick. In the HP-35 manual they suggest doing exactly this (see page 12). To this day, I still square like this even on RPN calculators that have an X^2 function (well, unless I'm programming and need the extra stack level or memory space).

So do I.

d:-)

[John R]

(07-07-2014 02:54 AM)Katie Wasserman Wrote:  

Quote: But here the HP-35 designers overshot the mark—instead of just having ENTER^ terminate digit entry, they (who knows why?) made it also carry out the unrelated task of duplicating the number into the Y-register, then disabling stack lift.

I don't really know the answer to this, but being an original HP-35 user I thought (and still think) it was to make squaring easier.

Another (related) intention may have been to simplify stack preparation for constant arithmetic -- that is, operations subjecting the X register to repeated addition to, or multiplication by, the same constant value. The traditional four-level RPN stack is prepped for this by keying in the constant value, followed by ENTER ENTER ENTER, followed by the X value. If ENTER served only to terminate digit entry without duplicating stack entries, then the user would be faced with the tedious and error-prone process of keying in the same constant value three times in succession.

In a parallel universe, HP might have used ENTER solely for digit termination, but also provided a separate key (DUP?) for stack entry duplication -- in which case we would presumably now be discussing the HP-36 instead of the HP-35.

[walter b]

(07-06-2014 10:40 PM)hansklav Wrote:  

(07-06-2014 07:28 PM)walter b Wrote:  ENTER↑ is a name fitting the classic RPN function well (compare e.g. R↓ and R↑, and look at all the pocket calculators from the HP-35 to the HP-41CX). ...

... Looking into the future it is important that the key has a name that speaks for itself. Imho ENTER: gives future users of Classical RPN a much better intuitive feeling for the semantic of this key in Classical RPN calculators than ENTER↑.

Sorry, I continue disagreeing. While ENTER: looks like a label to me (no operation), ENTER↑ describes what the operation does - it enters a number in the stack. Even the arrow points in the right direction (although HP talks about pushing values on the stack, it pushes them below the stack in reality - but that's another story). I'm pretty sure you'll also find ENTER↑ on the 43S.

d:-)

[hansklav]

(07-06-2014 11:35 PM)Didier Lachieze Wrote:  If you're interested in HP calculator history, we can trace back the origin of the "Entry RPN" before the HP-28C to the HP-41 translator pac for the HP-71b. Bill Wickes who developed the HP-41 translator pac as an "after-hours" project was also the software project manager for the HP-28C/S [1].

In an article published in January 1985 in CHHU Chronicle ( THE HP-41 TRANSLATOR PAC FOR THE HP-71) Bill Wickes introduced the command line and the removal of the stack lift disable:

Thanks for this very interesting article!

It gives insight into both the primordial history of what would become the RPL lineage, and into the fact that two functions were merged into the ENTER↑ key of the HP-35: digit entry termination and duplication on the stack.

To the novice user of RPN especially the first function, digit entry termination, is important to be able to learn when to push the ENTER↑ key (and moreover when NOT to push it).
The label ENTER: on this key would stress this function, "ENTER NEXT:", so only press ENTER: when you have to key in another number before pressing a function key.

To the seasoned user and programmer the aspect of stack duplication may be more interesting, but then this user hardly needs the labelling of the key as a mnemonic.

Hans

[hansklav]

(07-07-2014 10:50 AM)walter b Wrote:  

(07-06-2014 10:40 PM)hansklav Wrote:  ... Looking into the future it is important that the key has a name that speaks for itself. Imho ENTER: gives future users of Classical RPN a much better intuitive feeling for the semantic of this key in Classical RPN calculators than ENTER↑.

Sorry, I continue disagreeing. While ENTER: looks like a label to me (no operation), ENTER↑ describes what the operation does - it enters a number in the stack. Even the arrow points in the right direction (although HP talks about pushing values on the stack, it pushes them below the stack in reality - but that's another story). I'm pretty sure you'll also find ENTER↑ on the 43S.

The problem is that you, as a seasoned user of RPN, only see one function of ENTER: duplicating a number on (sorry: under) the stack. But the ENTER↑ key has another function that is far more important to someone who tries to learn how to succesfully use an RPN calculator to get the right answer for his calculations, and that is the function of digit entry termination.

First the novice must learn when to push then ENTER↑ key, and especially when it should NOT be pushed.

The label ENTER: on this key would stress this premier function, "ENTER NEXT:", so only press ENTER: when you have to key in another number before pressing a function key.

To the seasoned user and programmer the aspect of stack duplication may be more interesting, but then this user hardly needs the labelling of the key as a mnemonic.

So if you're interested in clarity for new users, then you will follow my argument; if you're only interested in tradition, if the fact leaves you cold that a lot of new users find it hard to understand why in one RPN calculator you get 6 as result and in another 5 when you key in 2 ENTER 3 ENTER + then don't change anything.

Hans

[walter b]

(07-07-2014 05:37 PM)hansklav Wrote:  The problem is that you, as a seasoned user of RPN, only see one function of ENTER: duplicating a number on (sorry: under) the stack.

How do you know? If you'd take a look to the WP 34S manual instead of telling me what you guess what I see then you'd see how I explain ENTER in the start section ("elementary stack mechanics"). And also HP was aware of that 'novice' problem 40 (!) years ago already: please find below their explanation of 1974.

I won't go into your other insinuations. Please do better research next time.

d:-/

[hansklav]

(07-07-2014 07:20 PM)walter b Wrote:  

(07-07-2014 05:37 PM)hansklav Wrote:  The problem is that you, as a seasoned user of RPN, only see one function of ENTER: duplicating a number on (sorry: under) the stack.

How do you know?

Because this is what you wrote in your reply:

(07-07-2014 07:20 PM)walter b Wrote:  (…) ENTER↑ describes what the operation does - it enters a number in the stack. Even the arrow points in the right direction (…)

And you did not mention the other function (digit entry termination).

(07-07-2014 07:20 PM)walter b Wrote:  If you'd take a look to the WP 34S manual instead of telling me what you guess what I see then you'd see how I explain ENTER in the start section ("elementary stack mechanics").

My apologies. I did not realize that you are the author of that decent piece of work! And I must admit that the beautiful lay-out of the WP 34S Owner's Manual was an inspiration for me when writing an RPN Tutorial (to be published soon).

(07-07-2014 07:20 PM)walter b Wrote:  And also HP was aware of that 'novice' problem 40 (!) years ago already: please find below their explanation of 1974.

Ah, the famous RPN Algorithm from HP’s ‘ENTER↑ vs. =’ brochure! In that brochure HP called it ‘a diagram of the RPN method’, but it is a modification of the original RPN Algorithm first published in the HP-35 Operating Manual and later also in the HP-45 Owner's Manual.

That diagram certainly showed clearly the number separation function of the ENTER↑ key. But it has nothing to do with the point I would like to make in this thread about the naming the ENTER key in classical RPN calculators of the future.

However the algorithm depicted in this diagram has another interesting problem: it doesn't work for some quite simple calculations with nested parentheses… Because it was told by HP to start at the left side of the calculation. And then you will soon get stuck (that is: get wrong answers without any warning), unless you have ‘an operational stack of unlimited length’. So the ‘ENTER↑ vs. =’ brochure was the last time HP used this diagram (as far a I know), although it is perfectly valid in the later RPL calculators with (virtually) unlimited stack length.

(07-07-2014 07:20 PM)walter b Wrote:  I won't go into your other insinuations.

I didn’t mean to insinuate anything, just wanted to make my point most clearly. If you felt that as insinuations, again: my apologies.

(07-07-2014 07:20 PM)walter b Wrote:  Please do better research next time.

I’ll do my best!

Hans

[d b]

Hans;
Good thread you started.
RE: Your upcoming RPN tutorial - Show it to us here if you will. And remember:
Except for memory size and lines of display; the way the 41 does it is best and right.