The primary advantage to RPN over a conventional mode calculator is that it uses a STACK.
Why so? You can enter all the numbers once and review them as you go. Then you can run a total. In that way, you don't find yourself repeatedly entering and summing until you get answers that you are sure you are right.
Consider me part of the RPN choir.
You'll swear RPN is the dumbest thing ever the first time you try it. But once you start working on some problems and get used to using the stack, using normal calculators will drive you crazy with how cumbersome it is to solve a problem.
As I mentioned in my first post, you can use Newton's method on some crazy polynomial and never have to stop to store parts of your answer in memory. You just move your pieces of the larger problem onto the stack and take them off as needed.
It sure was frustrating when I first purchased my 48S not to be able to add two numbers with it. It kept on saying "Error, not enough arguments". I thought it was trying to pick an argument with me. When the salesman said it used "Reverse Polish Notation" I thought he was making some sort of ethnic joke (I really did). After reading the RPN instructions in the manual I was just about ready to return the thing. Now it is intensely frustrating to have use the cumbersome entry method of other calculators and try to plan how to solve a problem so I don't use up all the limited number of memory positions to hold answers to smaller portions of the larger problem.
RPN and a stack takes a calculator to a whole different (as in better) level.
...
You'll swear RPN is the dumbest thing ever the first time you try it. But once you start working on some problems and get used to using the stack, using normal calculators will drive you crazy...
I'm also in the HP/RPN crowd, with an HP 49g+ sitting on my desk at work and various HP calculators at home. I don't have anything against TI, except maybe their prices, although HP does still sell their 12C for $70. If you learn RPN, it will make the more common infix notation feel like a waste of time. More importantly, I think that HP's graphing calculators are easier to program and to store and recall variables than any of their competitors. They also have a ridiculous number of graph types, and the newer ones have useful GUIs to make graphing easier. They also have a solver that is easy to use and works well, which I use all the time on even the simplest equations, because I can see the results right away. (e.g. I'll enter a simple Ohm's Law equation in to figure out the output of a voltage regulator based on the ratio of the resistors in a divider at the feedback pin. I can change either resistor's value and instantly see the output voltage.
Just thinking about my calculator history. These are the ones I actually used on a day to day basis at some point. There are quite a few others that never made it to "day to day" use.
TI-30 (the original one with blue "denim" belt case) - junior high and part of high school
Sharp EL-5100 - more of high school (very cool calculator)
TI-58 (I think, I'll have to check the model number later) - college (no fancy calculators in calculus class)
HP-41CX - at some point in grad school - my favorite of all of them - complete with card reader [1]
HP-28 - used for a while, but dropped to go back to the HP-41CX
Various and sundry Sharp and Casio scientific calculators (including graphing) because I didn't want to risk my HP-41CX carrying it around anymore.
HP-35S - current
[1] The story is that they carried HP-41Cs on the early space shuttle flights to use as a backup to compute reentry parameters.
...but I was solving problems faster than anyone else in college
When I was in college our mathematics lecturer could calculate faster in his head than anyone in the room with a calculator. Not that there was much call for doing calculations with actual numbers in those classes. That was for the realm of getting results out of physics and electronics experiments where the accuracy of a slide rule was quite sufficient most of the time.
When I was in college our mathematics lecturer could calculate faster in his head than anyone in the room with a calculator. Not that there was much call for doing calculations with actual numbers in those classes. That was for the realm of getting results out of physics and electronics experiments where the accuracy of a slide rule was quite sufficient most of the time.
Yes, indeed - I miss my slide rule and you have to dig to buy one these days. I worked in an engineering office where everyone else was senior to me, so they gave me the worst calculator and the worst chair. Since it would take forever to multiple (yeah, it was mechanical), I brought in my slide rule and just used the calculator for sums.
One certainly doesn't need numbers to 10 places for most tasks.
"RPN and a stack takes ... to a whole different (as in better) level."
"If you learn RPN, it will make the more common infix notation feel like a waste of time."
A word of caution:
If you get used to RPN, you may wish to use stack techniques while programming your prop, and you end up using
I don't know if the forum is composed of strange beings, or the preference is wider, anyway many of us prefer a not more in production calculator, and the HP looks like a preferred option.
For sure the old calculators had a keyboard feeling that no one is offering at the moment. For sure the same applies to PC keyboards, anyway. I would say this prefernce is related to a real quality we are now missing, and not a sentimental feeling.
RPN is probably more complex, and the learning curve is steeper. The new calculators offer a more direct approach, and it is easier to get started. I think if you "plan ahead" your calculation a RPN calculator is a terrific number crunching machine.
So we are probably expressing a preference to a calculator built so well that many of us still use it after 20+ years, and using an approach that pays as long as you climb the first slope.
Prof_braino, I understand you point. I used a lot of Fortran programs for neutron tranport, with Monte Carlo methods, and I have been intrigued by the language, but never dared to try it. To tell the truth I didn't know it used a RPN style approach until you discussed about it a couple of weeks ago....
Now propforth is on top of my learning list.... :-)
I spent a year doing take off estimating for major building foundation excavation and it never would have been possible without a HP41CX. RPN isn't that difficult to grasp and once you begin to see how it saves effort, you love it. We also had a survey crew that would use nothing else.
Yes indeed, quality calculators seem to be a thing of the past. My HP50g in not nearly as nice as the HP41CX was. But then the price is much more reasonable.
I suspect these days, I might just use a spreadsheet for all my calculation if I were still doing the same job. But having successfully spent a year getting 3.5 million USD of work with that calculator, I am sold on HP. There are die-hard TI and Casio users out there and I can't even buy an HP calculator in Taiwan.
HP is aware that people may never use RPN, so most of their calculators now provide for you to choose between RPN or conventional.
The whole point with the RPN is that you don't have to plan ahead, you have usually 10 scratch pad memory registers to save things to - constants, interim results, repetitive dimensions that require reentry.
I think the planning ahead came from not wanting to have to "Roll" a number off the stack. I'd try to plan ahead so I just needed to press Enter" to duplicate a value or "swap" to access the stack one value up.
If you left a number you needed high on the stack, it slowed down the calculation.
I still have my 28S someplace. I think I'll rig up some sort of adapter so I don't have to buy those expensive "N" batteries (that they sell in pairs but the calculator uses three at a time).
RPN is probably more complex, and the learning curve is steeper. The new calculators offer a more direct approach, and it is easier to get started. I think if you "plan ahead" your calculation a RPN calculator is a terrific number crunching machine.
It's the other way around really.. if you think about a sequence of e.g. additions and multiplications, with a non-RPN calculator you have to plan ahead if you wish your parentheses to start at the right place. With RPN you just stack the result when you come to a point where you realize that you have to take priorities of operations into account. Withouth RPN it's too late, you were supposed to add a starting parenthesis somewhere earlier.
That's the real beauty of RPN - you just start right away and sort it out as you go along. For non-RPN calculations I only use something where I can edit the operation before I execute it. In practice that'll be with the *nix 'bc' command line tool: I write out the operations in an 'echo', edit it until it's rigth, then pipe it into bc. If I can't edit then I will only use an RPN calculator.
Okay, I decided to dust off my HP50g and see why I can't seem to enjoy it. It isn't like the good old HP41cx or the HP12c or many of the archtipical RPN calculatiors -- it has a whole new OS for graphic display.
I hate to say it, but HP really messed up and lost its way on this one. I had previously mentioned how I liked the 10 scratch pad registers that were standard in so many models. But in this one, we have an SDcard and a file system that stores variables. You have to go to a lot more effort to key in an item - somewhere between silly and stupid.
Also, none of the documentation for it is easy to read. HP really needs to go back to its roots and produce a calculator for fast fingers rather than lots of overlays of programming features.
Okay, I decided to dust off my HP50g and see why I can't seem to enjoy it. It isn't like the good old HP41cx or the HP12c or many of the archtipical RPN calculatiors -- it has a whole new OS for graphic display.
I hate to say it, but HP really messed up and lost its way on this one. I had previously mentioned how I liked the 10 scratch pad registers that were standard in so many models. But in this one, we have an SDcard and a file system that stores variables. You have to go to a lot more effort to key in an item - somewhere between silly and stupid.
Also, none of the documentation for it is easy to read. HP really needs to go back to its roots and produce a calculator for fast fingers rather than lots of overlays of programming features.
Take a look at it. It's the closest HP has come to their old calculators in a long time. While they aren't quite the same, the buttons almost feel right. It doesn't have all the bells and whistles of the new graphing calculators but it's a good calculator. I have plenty of computers I can use when I want the bells and whistles.
I would like to see them reissue the 11 and 15, but I don't think it will ever happen (unlike the 12 which seems like it will be around forever). I never had an 11 or 15, I went straight from TI to the 41cx.
The best "Calculator" that I ever had wasn't really a Calculator at all. It's the TRS-80 PC4 pocket computer, fully programmable in BASIC! Needless to say, I was always the first one done in my statistics classes lol!
The best "Calculator" that I ever had wasn't really a Calculator at all. It's the TRS-80 PC4 pocket computer, fully programmable in BASIC! Needless to say, I was always the first one done in my statistics classes lol!
I have the PC2 - complete with 4-pen plotter and rs-232 interface. These were made by Sharp. You can see their lineage in the older Sharp calculators like the EL-5100 (which I still have).
Interesting comments. I just downloaded the 887page PDF for my HP50g and indeed they claim it isn't really a calculator at all, but a hand-held graphic computer for mathematics -- something that I really didn't want at all.
So I have to buy a HP35s to get the calculator I want. And, I have to go back to the USA to buy it as Taiwan doesn't sell them.
I guess the 'great divide' between calculator and hand-held computer is that the calculator should allow one to easily do entry without looking at the keyboard.
And yes, I'd love it if HP would revive the HP11 and HP15 - one was for everyday stuff without the financial forumlas and one was for computer programmers with bit, byte, and such calculations. But I guess if something doesn't generate cash, it is done for. I can use all of these one-handed, even with the wide body.
I wore the keyboards out on two HP-41s - an HP-41c and an HP-41cx. There is a niche market for extensive calculation - surveyors in the field are certainly one. I image Libya freedom fighters would like something for artillery as well.
I still have, use, and love my HP11C that I won in a programming contest in 1981 at SF State (one of the benefits of having Hewlett-Packard just down Hwy 280).
On the iPhone, the best calculator by far is PCalc.
Has every function I would ever want to run on a calculator and is $35. If it can not be done on this calculator then you probably need MatLab anyways.
Comments
Consider me part of the RPN choir.
You'll swear RPN is the dumbest thing ever the first time you try it. But once you start working on some problems and get used to using the stack, using normal calculators will drive you crazy with how cumbersome it is to solve a problem.
As I mentioned in my first post, you can use Newton's method on some crazy polynomial and never have to stop to store parts of your answer in memory. You just move your pieces of the larger problem onto the stack and take them off as needed.
It sure was frustrating when I first purchased my 48S not to be able to add two numbers with it. It kept on saying "Error, not enough arguments". I thought it was trying to pick an argument with me. When the salesman said it used "Reverse Polish Notation" I thought he was making some sort of ethnic joke (I really did). After reading the RPN instructions in the manual I was just about ready to return the thing. Now it is intensely frustrating to have use the cumbersome entry method of other calculators and try to plan how to solve a problem so I don't use up all the limited number of memory positions to hold answers to smaller portions of the larger problem.
RPN and a stack takes a calculator to a whole different (as in better) level.
Duane
I agree.
David Carrier
Parallax Inc.
TI-30 (the original one with blue "denim" belt case) - junior high and part of high school
Sharp EL-5100 - more of high school (very cool calculator)
TI-58 (I think, I'll have to check the model number later) - college (no fancy calculators in calculus class)
HP-41CX - at some point in grad school - my favorite of all of them - complete with card reader [1]
HP-28 - used for a while, but dropped to go back to the HP-41CX
Various and sundry Sharp and Casio scientific calculators (including graphing) because I didn't want to risk my HP-41CX carrying it around anymore.
HP-35S - current
[1] The story is that they carried HP-41Cs on the early space shuttle flights to use as a backup to compute reentry parameters.
When I was in college our mathematics lecturer could calculate faster in his head than anyone in the room with a calculator. Not that there was much call for doing calculations with actual numbers in those classes. That was for the realm of getting results out of physics and electronics experiments where the accuracy of a slide rule was quite sufficient most of the time.
But I don't recommend you to buy a TI 89-Titanium now... there are two newer models now.
Yes, indeed - I miss my slide rule and you have to dig to buy one these days. I worked in an engineering office where everyone else was senior to me, so they gave me the worst calculator and the worst chair. Since it would take forever to multiple (yeah, it was mechanical), I brought in my slide rule and just used the calculator for sums.
One certainly doesn't need numbers to 10 places for most tasks.
"If you learn RPN, it will make the more common infix notation feel like a waste of time."
A word of caution:
If you get used to RPN, you may wish to use stack techniques while programming your prop, and you end up using
For sure the old calculators had a keyboard feeling that no one is offering at the moment. For sure the same applies to PC keyboards, anyway. I would say this prefernce is related to a real quality we are now missing, and not a sentimental feeling.
RPN is probably more complex, and the learning curve is steeper. The new calculators offer a more direct approach, and it is easier to get started. I think if you "plan ahead" your calculation a RPN calculator is a terrific number crunching machine.
So we are probably expressing a preference to a calculator built so well that many of us still use it after 20+ years, and using an approach that pays as long as you climb the first slope.
Prof_braino, I understand you point. I used a lot of Fortran programs for neutron tranport, with Monte Carlo methods, and I have been intrigued by the language, but never dared to try it. To tell the truth I didn't know it used a RPN style approach until you discussed about it a couple of weeks ago....
Now propforth is on top of my learning list.... :-)
Massimo
Yes indeed, quality calculators seem to be a thing of the past. My HP50g in not nearly as nice as the HP41CX was. But then the price is much more reasonable.
I suspect these days, I might just use a spreadsheet for all my calculation if I were still doing the same job. But having successfully spent a year getting 3.5 million USD of work with that calculator, I am sold on HP. There are die-hard TI and Casio users out there and I can't even buy an HP calculator in Taiwan.
HP is aware that people may never use RPN, so most of their calculators now provide for you to choose between RPN or conventional.
The whole point with the RPN is that you don't have to plan ahead, you have usually 10 scratch pad memory registers to save things to - constants, interim results, repetitive dimensions that require reentry.
If you left a number you needed high on the stack, it slowed down the calculation.
I still have my 28S someplace. I think I'll rig up some sort of adapter so I don't have to buy those expensive "N" batteries (that they sell in pairs but the calculator uses three at a time).
Duane
That's the real beauty of RPN - you just start right away and sort it out as you go along. For non-RPN calculations I only use something where I can edit the operation before I execute it. In practice that'll be with the *nix 'bc' command line tool: I write out the operations in an 'echo', edit it until it's rigth, then pipe it into bc. If I can't edit then I will only use an RPN calculator.
-Tor
I hate to say it, but HP really messed up and lost its way on this one. I had previously mentioned how I liked the 10 scratch pad registers that were standard in so many models. But in this one, we have an SDcard and a file system that stores variables. You have to go to a lot more effort to key in an item - somewhere between silly and stupid.
Also, none of the documentation for it is easy to read. HP really needs to go back to its roots and produce a calculator for fast fingers rather than lots of overlays of programming features.
That's why I like the HP 35s: http://www.shopping.hp.com/product/calculator/Scientific/1/storefronts/F2215AA%2523ABA
Take a look at it. It's the closest HP has come to their old calculators in a long time. While they aren't quite the same, the buttons almost feel right. It doesn't have all the bells and whistles of the new graphing calculators but it's a good calculator. I have plenty of computers I can use when I want the bells and whistles.
I would like to see them reissue the 11 and 15, but I don't think it will ever happen (unlike the 12 which seems like it will be around forever). I never had an 11 or 15, I went straight from TI to the 41cx.
http://www.old-computers.com/museum/computer.asp?c=1003
I have the PC2 - complete with 4-pen plotter and rs-232 interface. These were made by Sharp. You can see their lineage in the older Sharp calculators like the EL-5100 (which I still have).
So I have to buy a HP35s to get the calculator I want. And, I have to go back to the USA to buy it as Taiwan doesn't sell them.
I guess the 'great divide' between calculator and hand-held computer is that the calculator should allow one to easily do entry without looking at the keyboard.
And yes, I'd love it if HP would revive the HP11 and HP15 - one was for everyday stuff without the financial forumlas and one was for computer programmers with bit, byte, and such calculations. But I guess if something doesn't generate cash, it is done for. I can use all of these one-handed, even with the wide body.
I wore the keyboards out on two HP-41s - an HP-41c and an HP-41cx. There is a niche market for extensive calculation - surveyors in the field are certainly one. I image Libya freedom fighters would like something for artillery as well.
On the iPhone, the best calculator by far is PCalc.
Has every function I would ever want to run on a calculator and is $35. If it can not be done on this calculator then you probably need MatLab anyways.