HP Forum Archive 16

[ Return to Index | Top of Index ]

Is RPN a pattern of HP? Message #1 Posted by Timespace on 6 Aug 2006, 6:02 p.m. Is RPN a pattern of HP? If another company starts making and selling calculators with RPN mode of operation without algebraic, does it have to pay HP fees?

Re: Is RPN a pattern of HP? Message #2 Posted by Chan Tran on 6 Aug 2006, 6:20 p.m., in response to message #1 by Timespace I don't know about the word RPN, but the RPN logic is not HP Pattent. There were several calculator back in the 70's that had RPN besides HP. So if you make and sell an RPN calculator, it won't be an infringement. I am not sure about calling it RPN though.

Re: Is RPN a pattern of HP? Message #3 Posted by Timespace on 6 Aug 2006, 7:11 p.m., in response to message #2 by Chan Tran I think if RPN is a HP patent or trade mark, then it is tied to that mode of operation. If I called my calculator's mode of opeartion say RSO for "Rolling Stacks Operation" which is identical to RPN, then it is still an infringement. Isn't it?

Re: Is RPN a pattern of HP? Message #4 Posted by Timespace on 6 Aug 2006, 7:03 p.m., in response to message #1 by Timespace I meant "Patent"

Re: Is RPN a pattern of HP? Message #5 Posted by Bill Wiese on 6 Aug 2006, 8:17 p.m., in response to message #4 by Timespace There were other RPN calcs besides HP in the 1970s: APF, Omron, National Semiconductor/Novus. However, these calcs didn't necessarily have the same setup with 4 level stack and T-register copy-down behavior we are accustomed to with traditional HP calc RPN. Rolldown (and, later, rollup) keys were present on HP, not sure about other calcs (???). The LastX feature is also kinda associated only with HP's version of RPN. In any case if there are patents on any features related to standard HP RPN they've beeb long expired now (it's been a lot longer than 17 years since the 1970s). calcinfo.org has lotsa calculator patent information... I believe you could legally make an HP-style RPN calc with 4-level stack, T-copy-down, roll up/roll down and Last-X features without infringing. The 'RPN' name itself may possibly have been trademarked, which might still offer some protection - but it could be argued that that is not a trade name but a technical description of operation: in the worst case you could always call it 'operand prefix notation' or 'operation postfix notation' or other similar nonsense... Bill Wiese San Jose, CA USA Edited: 6 Aug 2006, 8:21 p.m.

Re: Is RPN a pattern of HP? Message #6 Posted by Howard Owen on 6 Aug 2006, 8:34 p.m., in response to message #5 by Bill Wiese The proof of that (if any is needed) is that there are software calculators such as Neocal that implement exactly the XYZTL stack. They don't include any HP trademark notices, for example. And patents do apply to software, as we in the USA know, sometimes to our sorrow and consternation. It's unlikely that HP would allow those software implementations (without royalties, at least) if they had enforceable patents. Neocal is a very nice RPN calculator, by the way. I think that MathU Pro is more powerful, but correspondingly more complicated. It's not quite an RPN vs RPL sort of thing, since MathU Pro extends RPN in a different, less complicated direction than RPL does. Regards, Howard

Re: Is RPN a pattern of HP? Message #7 Posted by Vassilis Prevelakis on 6 Aug 2006, 8:51 p.m., in response to message #6 by Howard Owen Howard Owen wrote: > And patents do apply to software not in the 70s. Software patents is a modern concept. Look at the patent for the 9825 where everything (incl. the ROM source) is included in the patent. Also copyright was generally thought not to apply to software (and in any case till the US signed the Berne convention, you had to assert copyright explicitly to have protection). > It's unlikely that HP would allow those software implementations > (without royalties, at least) if they had enforceable patents. even if HP had patented the stuff, the patents would have expired by now.

Re: Is RPN a pattern of HP? Message #8 Posted by Howard Owen on 7 Aug 2006, 12:37 a.m., in response to message #7 by Vassilis Prevelakis Quote: > And patents do apply to software not in the 70s. Software patents is a modern concept. Yes. And as this community knows better than most, "modern" doesn't necessarily mean "good." Quote: > It's unlikely that HP would allow those software implementations > (without royalties, at least) if they had enforceable patents. even if HP had patented the stuff, the patents would have expired by now. A point Eric made in the parent. That's why I said enforceable. This was offered as another angle on the "HP doesn't hold patents on RPN" theme. (And also as a springboard for me to plug a couple of nice software calculators. 8)

Re: Is RPN a pattern of HP? Message #9 Posted by db (martinez, ca.) on 6 Aug 2006, 11:13 p.m., in response to message #1 by Timespace Timespace; There have been lots of rpn units made by companies other than hp and two (Friden and Anita) pre-date hp's entry into the calculator market. Some were really poor attempts but some just plain out hp-d hp. For instance Corvus, Privileg, and Qualitron had better battery & charging systems than the entire Spice and Woodstock lines - the 5 or so calcs using the Mostek chipset had better trig accuracy than anything hp made for another ten years (plus last x and replicating t registers) - and the Garrett 2000 series may be the prettiest calculators ever built. Some were programmable and one, the Heathkit 1401 with National Semiconductor chips, has a five level stack (with t register). As of a couple of months ago; the Aurora fn1000 is still being sold at Fry's. With any luck; The Eric Smith / Richard Ottosen "DYI" RPN programmable scientific will be on the market soon too. If you are interested you can see some at non hp rpn Page #1 will show up but you have to hunt a bit to find the following pages. There are over 50 examples. As to your question on payment of fees to hp for use of rpn; rpn seems to have been invented by the Austrailian computer pioneer Charles Hamblin in the 1950's. Lawyers can, and for a fee will, claim anything but i don't think that hp has tried or even wanted to try this. Edited: 6 Aug 2006, 11:16 p.m.

Re: Is RPN a pattern of HP? Message #10 Posted by GE on 7 Aug 2006, 4:15 a.m., in response to message #9 by db (martinez, ca.) I think I remember that Friden or some other pioneer had something that would be later called RPN before anyone else, and that HP had to actually *pay* them some non-negligible license to be able to use 'RPN'. Maybe this information was on Rick Bensene's site (always worth another visit...). Agree / disagree anyone ?

Re: Is RPN a pattern of HP? Message #11 Posted by Namir on 7 Aug 2006, 4:21 a.m., in response to message #10 by GE Should'nt YOU be paying GE some chas to use their names in this forum??? After all, some genius visiting this web site might thing that you are representing GE corp in the discussions??? :-) Namir

Re: Is RPN a pattern of HP? Message #12 Posted by Palmer O. Hanson, Jr. on 7 Aug 2006, 11:36 a.m., in response to message #11 by Namir National Semicoductor and NOVUS offered a fairly wide range of Reverse Polish calculators in the 1970's. If you go to Katie Wasserman's site (www.wass.net/manuals/) you can find that page 4 of the manual for the NSC 4615 contained a section titled "Reverse Polish Logic and the Stack System" which started with the sentence "Your calculator uses Reverse Polish logic with three registers called X, Y and Z. .." The manual for the more primitive NS 600 contains no mention of RPN or "Reverse Polish Logic". The manual for the equally primitive NOVUS 650 doesn't mention RPN or "Reverse Polish logic" but does include the following text: "... If you've used other calculators you'll probably find this one a little different. Your Novus calculator uses the same mathmatical "language", or logic system, as many computers and advanced scientific calculators. We chose this powerful language for your calculator because people who who use math in their professional lives believe it's the most logical, consistent and efficient way to solve problems. Your Novus calculator lets you approach each problem the same direct way, displays each intermediate step you take, lets you easily correct wrong entries and often takes fewer keystrokes to get an answer. After you do a few simple calculations, you'll quickly see how easy it is to use. Once you're used to it and discover the accuracy and oomputational power it gives you, we're confident you'll come to prefer the way your Novus calculator tackles math." Does that sound like the standard RPN line or what? There was another issue involved in decisions related to patents and copyrighting back in those days. The early players such as RemRand and IBM in computers (and I suspect Sharp and others in calculators) had patents, copyrghts and reduction to practice in many areas of technology fundamental to any application. Late comers like Honeywell in computers and H-P in calculators had to walk a careful line with respect to intellectual property rights lest they would enable the early players to exercise their earlier rights to the exclusion fo the later entrants.

Re: Is RPN a pattern of HP? Message #13 Posted by unspellable on 15 Aug 2006, 9:34 a.m., in response to message #9 by db (martinez, ca.) At the risk of giving away my age I will state that I was familiar with the term "RPN" for "Reverse Polish Notation" back when calculators had hand cranks only four functions. The term "Polish Notation" as I recall predates WWII. I can still recall the way it was written on a black board. I would be quite certain that no one has either a patent or copyright on the term "RPN" or "Reverse Polish Notation". "Polish Notation" originated as a term used by mathematicians. I suspect he same may be true of "Reverse Polish Notation". BTW: ALL scientific calculators use RPN at least in part. Try finding the sine of 25 degrees without using RPN. Bet you can't find a calculator that will do it. (I'll take that back, my slide rule will do it.)

Re: Is RPN a pattern of HP? Message #14 Posted by John Gustaf Stebbins on 15 Aug 2006, 10:39 p.m., in response to message #13 by unspellable The Casio fx-115ES I have sitting on my desk will give me the sine of 25 degrees without RPN. Key strokes: sin, 2, 5, = PN, no? -jgs

Re: Is RPN a pattern of HP? Message #15 Posted by Karl Schneider on 15 Aug 2006, 11:31 p.m., in response to message #13 by unspellable Quote: BTW: ALL scientific calculators use RPN at least in part. Try finding the sine of 25 degrees without using RPN. Bet you can't find a calculator that will do it. Indeed, as John Gustaf's previous post identified, today's EOS calculators (which include even lower-end models from Sharp and Casio) operate in that manner: [sin] [2] [5] [=] gives the answer to sin(25) after putting the symbolic expression "sin(25)" in a buffer. AOS calculators, once common but now reserved for the cheapest of models, used RPN-style postfix for unary operations, because there was no buffer. -- KS Edited: 16 Aug 2006, 5:57 p.m. after one or more responses were posted

Re: Is RPN a pattern of HP? Message #16 Posted by unspellable on 16 Aug 2006, 12:15 p.m., in response to message #15 by Karl Schneider Well learn something new every day. The last cheap calculator I've acquired is an HP 6S. Sine of 25 degrees is [2][5][SIN]= on the 6S. Maybe I should have said cheap calculators instead of scientific calculators. But we are excluding 5 function calculators. BTW: I have an old how to book for finding trig functions, etc. on a four function calculator.

Re: Is RPN a pattern of HP? Message #17 Posted by Egan Ford on 7 Aug 2006, 6:38 p.m., in response to message #1 by Timespace Quote: Is RPN a pattern of HP? If another company starts making and selling calculators with RPN mode of operation without algebraic, does it have to pay HP fees? I have OCR'd a few manuals and searched for RPN and RPL. No trademark or patent statements exist. That does not mean that there are none, but usually you would expect them to be in the manuals. My search did yield the following from the a 42S manual: "HP's operating logic is based on a mathematical logic known as 'Polish Notation,' developed by the Polish logician Jan Lukasiewicz (1878-1956). ..." That said RPN would be prior art, and it should not be allowed to have a US patent. However, lawyers can be crafty and get anything patented (Google for the Australian that tried to patent "the wheel"). However, *how* something is implemented is patentable. Even if RPN was patented, US patents expire after 20 year unless you can find a new use for your invention. Trademarks are a different matter. Disney has succeeded in extending the trademark for Mickey Mouse. I am no lawyer, but I believe that trademarks can be implied. My $0.02US, you can make an RPN that works like an HP, but you may be unable to label it "RPN SCIENTIFIC". And if you are in the right and HP in the wrong, it can cost you. (Ever seen the 12L ('L'awyer), most of the keys end in '$' and are exponential functions). Edited: 7 Aug 2006, 7:31 p.m. after one or more responses were posted

Re: Is RPN a patent of HP? Message #18 Posted by Katie Wasserman on 7 Aug 2006, 7:15 p.m., in response to message #17 by Egan Ford The Heathkit OC-1401 uses a 5-level stack (X,Y,Z,T and S) that they call RPN and provide the SWAP, ROLL-DOWN, and LAST X functions. The Heathkit OC-1401 Manual devotes a good amount of space to describing RPN and how to use it. It reads a little like HP calculator promotional literature: Quote: Calculating is faster and the operational stack and the RPN used in this unit provide the most efficient way known to computer science for evaluating mathematical expressions. The Heathkit OC-1401 was made starting in 1978 and if HP had trademarks or patents covering RPN, LAST X, etc. they must have still been in effect so Heathkit may have paid some licensing fee (or just got lucky). In any event they managed to not mention HP in their manual. Edited: 7 Aug 2006, 9:48 p.m.

Re: Is RPN a pattern of HP? Message #19 Posted by Trent Moseley on 7 Aug 2006, 9:29 p.m., in response to message #17 by Egan Ford My two cents is that $0.02 should look like this: 2¢. tm

Re: Is RPN a pattern of HP? Message #20 Posted by Trent Moseley on 7 Aug 2006, 9:44 p.m., in response to message #19 by Trent Moseley Or better yet. Some years ago in a US Post Office after a one cent increase in the postage rate, they posted a hand written sign: Temporarily out of $.01¢ stamps! tm

Re: Is RPN a pattern of HP? Message #21 Posted by Vacationer on 16 Aug 2006, 2:35 p.m., in response to message #20 by Trent Moseley Good idea :-) Is that a reason for a UNITS key?

[ Return to Index | Top of Index ]

Go back to the main exhibit hall