I found my 48GX a week ago when moving some boxes around and was definitely feeling nostalgic. It pretty much went everywhere with me when I attended college (EE), but that was 25+ years ago. Kinda cool seeing the nameplate on the back with my name on it. Very personal.
That's a fun find. It's funny how attached you get to a little machine that was your companion through a good number of stressful exams and tons of homework. Happy you re-discovered your lost HP48GX! Thanks for leaving a comment.
When I was a student engineer, the previous generation made us learn the slide rule 'out of nostalgia', which is about the same role RPN plays today! HP really make the distinction between 'algebraic' (string representation with nested brackets) and 'textbook' which is most of what you are terming 'algebraic'. You are bang on the money though about 'textbook' making it easier to spot mistakes than RPN and that is maybe worth a few extra keystrokes. The HP28-50 family can do 'textbook' with the equation writer and the Prime does have an RPN mode. Although the latter is a bit brain dead, it is certainly up to doing the examples you show here. Would be interesting to see a shootout between the HP28 and Prime with each using each entry method. Personally I think the 'sweet spot' is the RPN stack with RPL, but for most purposes using the equation writer for data entry (except for parallel resistors of course, or large multiply-add accumulations). HP got so close to a mode-less combination of the best of both systems before sadly throwing in the towel with the Prime.
@@scottcollins7513 Yea, from a marketing perspective they needed someone used to algebraic to be able to pick the calculator up and intuitively do a simple calculation without RTFM. There is no reason I can see why you should not be able to enter an expression on the RPN entry line in Textbook form and on pressing ENTER it evaluates it and pushes the result on the stack. The implied ENTER you get in classic RPN on pressing an operator would be lost and you'd always have to press ENTER after a number entry. But that could be the mode: RPN purists could restore implied ENTER at the cost of having to type a tick mark or some other delimiter when they did want to enter an expression.
RPN is not the best choice only for the number of keystrokes, I like the logic behind, that’s why I prefer using a Rpn calculator. RPL programs can be difficult to read and understand. The fun comes at creation 😉. Nostalgia is a real motivation, true.
The big problem with non-RPN calculators is that they are inconsistent. For example, if you want to work out the square root of two, on some calculators, you type √2 and on others you type 2√. The same goes for sines and cosines etc. and I can never remember which convention any particular calculator uses. Such problems simply don't exist with RPN because operators ALWAYS go after the parameters.
True. I would say that these days, nearly all scientific calculators are using some form of "math/text book" entry where you enter it as you would write it. Thanks for leaving a comment, Chris. I enjoy hearing what others are thinking.
@Scott Collins I grew up loving Ti and Casio calculators (Ti 85 was my computer, it is all I had). I later discovered HP calculators and the RPN/RPL and I love RPN. yes I agree there are times that algebraic or text book entry is easier. what I think is over-looked about RPN is how the person who uses RPN is actually doing the incremental math with the calculator doing the individual computations. that is to say the user has to understand the math, and may actually learn from using the calculator instead of just entering a huge equation and accepting the output. I have heard it described as "A thinking man's calculator" I do agree that RPN is not the best for all scenarios, but it is better if you get numbers and then have to decide what to do with them. if you do have a large equation to use like in your example, the Equation Writer is perfect (on the 48S 48SX 48G 48GX ) (www.hpmuseum.org/hp48s.htm) the HP 48G or 48GX also shines with the built in equation library. the ability to select a formula, and be able to enter the known variables and solve for unknown without having to use algebra to isolate the unknown variable is amazing! (though this is NOT a thinking man's calculator example) I also love my 48SX it has odd usage in my opinon, but it is very powerful. the box for this calculator really under-sold it in my opinion (here is a cool video about it) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-OTPruRVV-e8.html. with the equation library card, I get all of those great equations from that are built into the 48GX, and the periodic table. the most amazing thing about the periodic table is the fact that there are tons of values for each elemet I do think the HP 50G is also a great calculator function wise, though does not feel as nice as my 48SX. I have a couple of shelves of calculator manuals, and I must say that HP really shines in their manuals compared to Ti or casio.
For some reason I never saw your detailed comment until now. Lots of good stuff in it. Thanks! I'm going to check out your video link. My guess is that the build quality on the 48SX is top-tier. I think the color pallet looks better/classier than my 48GX. Agree that RPN requires you to engage your brain which can be good and bad.
I have to agree that the solver is not exactly a thinking man's feature, but an active engenier (HP target) need. With the use of the EqSlvr one can easely found any value of f(x,y,z...), what's is the daily use of calculators. The next step, then, was to write and save all the needed equations for our job. Like TVM on financial calculators. And EqLib card was the collection of most enginering equations for most domains. With that, I've been able to forgot the same collection I use to have on papers, the same way scientific calculators let us forget trigonometric tables and slide rulers. In another hand, when you have to perform long calculation by really thinking on them and knowing what you are doing, it's nice that the same calculator offers the RPN (I found it simpler and more intuitive than algebraic because it's the way I think and work, what you call incremental math.) 48 series just went a step further, giving more possibilities and opening to more worlds including algebraic.
I think RPN is not every time about less keystrokes, but how I'm able to break down formulas while typing and have no need to use parentheses or extra intermediate results.
Totally agree. From my situation, I do a lot of CAS math in controller design, if I use the 48, I need to write down every step and value carefully for some review of the procedure, and bear with its grandma speed. However, Prime revolutionizes this experience, its algebraic style makes the whole procedure smooth and easy to review, and it is so fast, I think Prime represents a modern level of a handheld calculator can achieve with so many years' accumulation, meanwhile 48gx also represents his status at 1990s, they are all kings. I love RPN, it needs you to think more about the math. hope there could be some evolution products on the basis of HP48 with a modern hardware platform.
Appreciated reading your thoughts in the comment and your use case for the Prime. The Prime is blazing fast, but I never warmed up to it, since I do not really have a need for that kind of power. I think RPN is forever relegated to the "enthusiast" market (e.g. SwissMicros), but I still love it.
RPN was the clear choice when I purchased my HP-41CV in 1984. It is amazing how well RPN has held up all these years when compared to modern algebraic calculators.
With text book or formula entry modes, RPN is finished. There is no way RPN is better than entering the formula as written and being able to verify entry and modify just one of the inputs to get an updated result. RPN vs single line algebraic entry, RPN definitely had the upper hand.
Very well said. Obviously RPN lost out, and I think you did a good job succinctly summarizing the "why". I will say that RPN still feels more natural to me when I just need to do a few calculations, but it has to do with having spent so much time in that world. At this point, I keep both types of calculators handy.
@@scottcollins7513 The real irony is that HP was the first ever to offer actual text book style entry with the HP-48SX. It isn't the polished text book entry we see with TI-36x pro and TI-30x Pro MP, as well as most casios, but it was text book entry, slow text book style entry. Today, HP is basically out of the calculators all together. Good example of not adapting fast enough and becoming obsolete.
@@wb.c Good point about hp having that function in the early 1990s. My hp 48GX has the function, and it is SO SLOW that I would never want to use it unless absolutely necessary. They were onto a good idea, but the timing was not right. Happens all-too-often in technology. Thanks for the thoughtful comments/discussion.
Fully agree with your conclusions. I say this still having (using and “playing” with) several calculators, Scientific and Financial, from HP (RPN and RPL) and using AOS (Sharp, Casio).
TI-Nspire CX CAS: 12 + 16 + 20 + 31 = *79* Law of Sines: Lacks dedicated button for sin etc. Sin requires "only" 2 buttons to be pressed, asin 3, atan 5. Eq. Resistance: Lacks dedicated button for 1/x Time value...: 19, if you wrote [5] [EE] [3] Ratio of diag.: Lacks dedicated button for [a/b], instead it's [ctrl] + [ / ] If this calculator is in Automatic or Accurate calculation mode (instead of Approximate), you might sometimes need to pres [ctrl] + [enter] for decimal approximation. Which isn't included to the numbers above.
I have a popup windows app that emulates my HP42S. I just entered the expression at 0:32 in RPN. It took me 54s and I was pushing the buttons with my mouse. I did it twice, each time getting 1.0157 RPN is just that easy. I like to check the intermediate results for sanity. Also easy. Here are the steps. Between every keys, there is one space. 3 x^2 12 x^2 + Sqrt 7 x^2 + Sqrt 10 x^2 33 x^2 + Sqrt 5 x^2 + Sqrt / Display shows 1.0157 Counting each number as 1 keystroke, the above is 21 strokes. Practically automatic for me. It especially satisfying to look at the display and then press the last 3 operator strokes ( + Sqrt and / ). With store and recall to numbered (or named) registers, numeric entry is easy and the program is simplified.
I cannot figure out if RPN feels sort of magical b/c it is what I used in my "formative" college years *or* if there is something inherently superior about it that somehow just feels right, particularly to engineering types *or* (most likely) some combination of the two. Thanks for the 42s info/comment, Mitch.
The advantages of RPN are an illusion. They're not real. Someone didn't sit down looking at all the merits of RPN versus algebraic input and decided that RPN was clearly the best. What actually happened is that HP designed a calculator using the technology of the day. Parsing infix notation, in particular keeping track of order of operations, is a very difficult task with '70s technology. It is far, far simpler to have a stack based environment where you put values on the stack and then execute the operation. That translates 1:1 with the underlying machine code. When a CPU adds two numbers that is exactly what it does. Loads a number into a register, another number into another register, then runs through the addition logic. And after HP took the easy way of designing a calculator, letting the user worry about pesky order of operations, they started coming up with all sorts of reason why RPN is better than algebraic, especially as the years went on and hardware was becoming good enough to enable algebraic calculators to compete with the RPN ones. RPN hasn't been actually beneficial, desirable, or relevant for about 40 years. People stuck with it because HP did make quality devices that were nicer to use. But these days the absolute only reason why one would use RPN over algebraic is nostalgia. There is absolutely nothing to gain from using RPN.
Cristi, thank you for the thorough and well-argued comment. Back in the late 80's, RPN was certainly superior. But, these days, I think I agree with you -- even though I will always have a soft spot for RPN. I'm doing some engineering work again and my go to device is a TI-30X Pro Mathprint (the UK version of the TI-36X Pro) and not one of my HP RPN calculators. I love that on the TI-30X Pro Mathprint you can see what you entered, reuse any expression or result, etc. Just very user-friendly. On the other hand, if I was doing lots of vectors and wanted to do so on a calculator rather than a computer (not sure why I would), I would want to use an HP28, HP48, or even DM42 rather than the TI-30X Pro Mathprint.
hello. thanks for this video. after playing/working with all these calculators , what are your favorites ? dm42 is « simple » and fast with a nice screen, 48G is more complex but RPL give some symbolic hi level logic, and the Prime is definitely the strongest/fastest but it does not have the OS flexibility around a stack ... I do not like the Prime yet...
I like the DM42 for daily driver -- great display, programmable (and relatively straight-forward to do so), and compact size. The HP 48GX would be my go-to if I were still doing engineering b/c of great solver, programmability, unit conversions, and 3D vectors. HP Prime goes generally unused unless I need some quick CAS help. HP 28S has a special place b/c it was the calculator I used in engineering college and b/c I appreciate its simplicity (vs HP 48GX) and excellent keyboard. I also think calculators like the Casio fx-115ES Plus and TI-36X Pro are solid choices. Sort of like asking me to pick the best car -- depends what job I am wanting it to do.
@@scottcollins7513 a pity I sold my 28s for the bucks needed for the 48... Very good memories with this 28S. I agree and understand your comment. It’s nice to meet people here having similar experiences with HP or scientific calculators. DM42 is growing on me...
@@scottcollins7513 currently I don’t really need a scientific calculator, it’s pure for fun and cerebral stimulation. (I was in chemistry engineering at university. So much time on the keyboard 😅)
@@odarge I can relate. It’s just fun to mess around with different calculators. One thing I wish the DM42 (and HP 42S) had was a solver like the HP 28S, HP 48GX, et al.
There's some point making interferance this interesting comparison: the layout… That's where I sometimes get puzzled by manufactures' choices: most often used/needed functions are not the most accessible on most recommanded calculators for many grades, or they are not grouped in a logical way. To speak of RPN calculators, 28 C/S and 48 S/SX/G/GX gave access to some newer features, hence making some of the match examples have more keystrokes than glory 35. Nice videos however.
Yes, it's strange that there is not more agreement on an optimal choice of keyboard layout and which functions should be primary versus shifted functions. Why in the world would Casio dedicate a key to "hyperbolic" when it's hardly ever used -- is there some subset of its users that demand that feature? And why is square root a shifted function on any calculator?
I'm a big fan of analogies. I think yours is a really interesting/apt one. And, you could expand it to say that long ago, when automatics where slow and spongy, sticks offered clearly-superior performance and gas mileage, but modern day automatics w/ dual-clutch transmissions (or just really good traditional automatics) are now at least as good if not better -- even if driving a stick is still somehow more fun.
I never saw your comment -- playing catch-up. The formulas are not saved in a mathprint style. They end up being converted to a single-line when stored.
Like other algebraic calculators, formulas are saved as algebraic objects… Mathprint is displaying (or printing as named), a feature first introduced by HP48 S/SX and copied years later by others.
Not taking advantage of your HP48's deep stack. You can enter the formula in RPN from left to right (numerically). Might cost you more keystrokes, but saves you a bunch of cognitive effort, which is why you're using a calculator in the first place. When I see you start at 3, I know you've already done some of the work in your head. 7, ², 3, ², 1, 2, ², +, √, +, √, 5, ², 1, 0, ², 3, 3, ², +, √, +, √, ÷
I appreciate the different perspective on how to best use RPL (RPN w/ "deep stack"). You are still having to think about the order of operations, but with your approach all the numbers are entered sequentially. I think I generally approach the problem (for better or worse) entering the numbers only as required for the sequential order of operations. I think doing that way feels more natural to me and requires less (not more) cognitive effort. Interesting to get your take -- thanks for taking the time to detail out the specific stack entries. Take care, Brian.
I do agree with Brian. There are simple tips. Start at the deepest level, start at numerator rather than denominator. I used to argue philosophically that RPN is not only more intuitive, but better fits what is happening. You get to leave intermediate results on the stack and they automatically appear when you need them. The philosophy part is that one first takes operands and then uses an operator. For example, to find a right triangle hypotenuse, "Take one side and square it. Take the other side and square it. Add those two squares together. Take that sum and do a Square root. a, ², b, ², + Sqrt No parentheses ever needed!
Another point, your table of keystrokes somehow surprised me. I thought the difference would be favorable to RPN. The only dissent I would have is about the last, larger equation: Although when we are “rusted” the solving sequence of most “4-stack RPN calculator” could be more “delicate” without storing intermediate results as STO 1 and STO / 1, when a RPL calculator (as 28S, 48G, 50G) we count with their infinite stack (or 8-stack from WP34S), making any stack-based calculation much easier.
In the end, I gravitate to RPN (and RPL) just because there is something appealing about it that I cannot quite put my finger on. There is a text editor called VIM that I prefer over the more-modern Notepad++ in a similar way. I'm happy to use a non-RPN calculator or to use Notepad++, but when I get to use RPN or VIM, I am just more energized. Thanks for leaving the comments, EJE.
That is a very good point that I'm not sure I ever talked about. Thanks for mentioning it here, Javier. It definitely takes a little time to get used to the new approach, but worth it.
With the advent of Mathprint, RPN will be relegated to a very minor niche. There are even some on HP forums that are looking to find a way to use algebraic equations in the Free 42, rather than having to create a program to implement an algebraic equation.
Agree. Clearly RPN is going away. Re: the HP42, I'm surprised it is as popular as it is. I feel like the HP28/48 was a lot more powerful, but still easy (enough) to use. But, maybe I am missing something.
@@scottcollins7513 I had the 42S at the same time I had the 48GX and preferred the larger graphical display and user RPL of the latter. If Swissmicros could legally make a 48 then I would expect it would also be one of their bestsellers.
@@dryster123 Good information. I did not realize there were legal impediments to SwissMicros making a "DM48". I think some of the attraction to the 42S is because of its small form-factor and relative simplicity (just due to the fact that it is more limited).
@@scottcollins7513 Yes, the HP48 ROM(s) are allowed for personal use but not commercial. The Smissmicros 42 is based on the reverse engineered free42 which does not use any HP code, and therefore does not need a ROM. I think there are legal issues with system and user RPL also.
@ss free42 is not based on any HP code and does not use a ROM dump from the HP42S - I suppose that is not classed the same way as HP48 and RPL proprietary stuff although I don't think some of the HP48 emulators need a ROM any more, but also don't know if they're compatible with system and user RPL.
Not quite sure what "no reuse or double checking of previous expressions" means for the HP48 series, mainly as every expression can be stored as a reusable, retrievable, variable as many times as wanted. In the same theme every stored expression can be retrieved to the stack to double check or displayed in the equation editor. The amount of stored expressions is only limited by the HP48 memory - unless of course that is the point, that The HP48 series are not true RPN and cheat by forcing algebraic input and editing.
You make a fair point. You could save all your work as you go, though I do not think most users will want to create duplicates of each previous expression 'just in case' they might need it later. Also, at least on my 48GX, the equation editor, though a great idea, has terribly slow performance. I think the reality is that most people using RPN/RPL are not going to use it in the manner you are describing. If someone really likes the equation editor approach on the HP48, they might be happier with a more-modern "text book entry" style calculator like the HP Prime. Appreciate your thoughtful critique.
I love the RPL calculators, but they are not the real native RPN deal. To be fair to RPN you should use a real classic RPN calculator like the HP 42S instead (or some of the other many RPN calculators HP made). If the 42S is hard to get you can buy a new clone from Swiss Micros. Or a cheap way is the Free42 app for mobile or computer.
Actually, I forgot that I do have the Free42 on my iPhone. I need to mess around with that and see if I can figure out what you mean. Appreciate the comment and food for thought.
@@scottcollins7513 Thanks! Would love for you to include a "real" RPN calculator. I love the idea about your comparison. I have nothing against algebraic calculations. I also have both Casio (fx-82EX) and Texas (TI-30X Plus MathPrint) algebraic non graphing calculators (as I have no need for graphs in my work, but sometime appreciates a physical calculator). But I am sorry to say that I feel that your current comparison is not fair, and as long as you do not include a classic HP RPN calculator (very different key strokes compared to RPL calculators like 28/48, even if they share RPN-concept) I think it is really incomplete. Love your basic idea, even if not the execution of it. Please update the series when you have tried a "real" HP PRN calculator! :)
@@martin-ot I think the missing dimension here is that the 48 series are RPN/RPL and you are right that a true comparison would need to be algebraic versus 'real' RPN, however a true comparison would also require calculators of the same era, calculators with the same processing speed etc. I am of the opinion that the usefulness of any machine is often in the hands of the user, not in the manner of input, and I would pick an RPN machine as it suits me.
@@dryster123 Yes, the HP 42 S RPN calculator was made and sold at the same time when I bought my 28C, S and 48SX RPL calculators. So I see no difference in age, and really fail to see your point here (but very open to it if I can grasp it)?
Of course everyone should use whatever tools they prefer. I have and use; RPN calculators from HP (42S, 32S, 11C, 15C, 16C, 35S), RPL calculators from HP (50G), several algebraic calculators from Casio (FX-82EX ClassWiz, FX-82 SOLAR II) and TI (TI-30X Plus MathPrint), and my favorite; the new modern RPN calculator (fast and with highest internal precision of any calculator in the world) DM42 by SwissMicros. They are all good in different ways, and I enjoy them all.
Sorry if the music was distracting. I think if I did the video again I would either have no music or much lower volume. I was trying something out that was probably not great looking back.
I tried something different. I can see where it is distracting for some; in fact, I often wish people would not have background music -- especially when there is something to listen to like a car engine. In this case, I think it is more of the wrong music choice. Thanks for the feedback, Yujia.
