|
TI & RPN
|
|
|
|
« Next Oldest | Next Newest »
|
▼ 
10-20-2012, 01:12 PM
The TMS1000 was TI's first microcomputer and it powered many TI algebraic calculators. Ironically, the 1975 TMS1000 Programming Manual (10 MB PDF from BitSavers) includes an RPN calculator as the sample application! ▼
10-20-2012, 01:40 PM
I could be mistaken but one of reference books for the TI-58/59 has an RPN program. Eddie ▼
10-20-2012, 03:23 PM
Hi. IIRC, there is also a synthetic programmable keycode that makes the TI58/59 to operate somehow in RPN mode. Joerg, please, is that so? Cheers. Luiz (Brazil) ▼
10-20-2012, 03:29 PM
At least, there was a solid state software module for RPN operation ... ▼
10-20-2012, 03:38 PM
Yes, right, I remember now that I read about it when I was at the University (80's).
10-20-2012, 03:37 PM
Not that I'm aware. TI did sell a TI58/9 ROM module that made it easier to convert HP calculator programs but it did not replace the built-in Algebraic Operating System with RPN. My original post points out an irony that they published assembly code for an RPN calculator design (though much simplified - 6 digit integers only) whereas they made algebraic calculators with that chip. Still it's an interesting look at how calculators were designed in 1975.
10-21-2012, 06:17 PM
Yes, for the TI 58/59, there is/was an 'RPN Simulator' module. How it worked is that you would input the HP-67 keycodes and the module's translator would provide the keycodes for the similar program on the 58/59. The interpreted program would incorporated a series of module/subroutine calls. For example, the keystroke [PGM] 51 [A] would call the ENTER^ subroutine. ▼
10-24-2012, 07:29 AM
Hi Allow me to intervene in this topic because I started very recently to use RPN since I implemented it in my emulator TI58C. Educated in the direct algebraic method (from the feeding bottle!) I had a lot of trouble to make me to RPN. Follower and fan of TI58C I made an emulator which was, naturally, in AOS. But not stopping developing my emulator, a user, whom I would not name, tempted me in the way of the RPN. I had some difficulty to adapt the language TI to the RPN but I believe I arrived to do it after good sweats (thank you Namir!). I did not retain the RPN module of the TI because very decried by the pure RPN followers. To conclude, I believe that there is one right method or one good tool: the one that is suited to our culture, our habits and our personal use. The hammer is a poor tool for the pianist, the treble clef inappropriate for the garage.... unless... Regards Pierre
10-20-2012, 03:36 PM
Hi. If we consider a classic processor core and its arithmetic unit, it actually accumulates the operands prior to perform the operations. This is the essence of RPN. In very earlier discussions at this forum (some years ago), some reasoning about why RPN was chosen led to a simpler firmware, not necessarily because it was some new way to use calculators. Algebraic evaluation demands the existence of an argument stack storing data and operation tokens, and extra OS handling. And this also demands more memory - both ROM and RAM - and extra processor cycles for most user operations. Although I prefer reasoning in RPN structures while programming, I am aware of the fact that it exists for reasons beyond usage. Cheers. Luiz (Brazil) |
