HP Spice series simulators for iPhone

[Willy R. Kunz]

(02-21-2016 12:27 AM)Matt Agajanian Wrote:  

(02-20-2016 11:36 PM)Willy R. Kunz Wrote:  Hi Matt,

I entered the Mach formula in RPN-32 on my iPhone 6 exactly as shown here: Mach
(scroll down half the page to Example XXVI)

I found two differing results:
4th to last: 0.579614684 vs. 0.579614685 on RPN-32 (vs. .5796146848 on RPN-35)
Final result: 0.835724536 vs. 0.835724535 on RPN-32 (vs. .8357245352 on RPN-35)

The corresponding 15-digit results are:
0.579614684798178
0.835724535175251

So it seems RPN-32 simply gives more precise results than the original calculators. As does RPN-35.
Sorry about that.

Okay. Here's what I notice:

After going through the HP-45 and onward steps for the Mach calculation, I tried the HP-35 variation--ENTER^ x for x^2. So, after I press 350 ENTER^ 661.5 / ENTER^ x

somehow when I enter 0.2 x, the result is 0. It seems after the HP-35 method of ENTER^ x, the stack lift is disabled and 0.2799 is cleared from the Y register. Just a hunch but, it seems +, -, x, and / disable stack lift. But, like I said, when I tried the HP-35 method of ENTER^ x to square, somehow stack lift doesn't go.

Ah, now I see what you mean. It's not about a calculation error, but a stack lift problem. And yes, it's a bug, plain and simple, caused by not uncommenting a line in the source code. Already fixed for the next update.

What happens is that 5 ENTER +, for example, doesn't reenable automatic stack-lifting (disabled by the ENTER statement), so the next entry will overwrite the result. However, 5 ENTER y^x and most other operations are not affected by the bug.

As a workaround you can insert an x<>y operation. For example, 7 ENTER x<>y + would cause the stack to behave as expected.

RPN-38, while sharing a lot of code with RPN-32, works correctly.

Thanks for reporting the issue. If possible, please use the contact link on my website cuveesoft.ch for comments and bug reports, as I'm not visiting this forum on a daily basis.