The calculator I always wanted

[vanLudwig]

I’m guessing I’m the only person on this forum that learned RPN on a TI-89. rpn89 by Lars Frederiksen was a short-lived project, but it was the first time I realized that HP graphing calculators were RPN because some people actually liked it. The way rpn89 displayed both questions and answers in a step-by-step manner made the 89’s native console input feel tedious and error-prone. I was devastated when rpn89 was discontinued, and now the internet seems to have been thoroughly scrubbed of it.

Acron RPN Calculator started as my attempt to recreate what I loved about rpn89. I had the advantage of not being trapped by TI’s hardware, so I could have a hi-res color output, faster computation, and a button layout that made sense in RPN. But with that, I wasn’t able to take advantage of TI’s CAS system. Acron RPN’s perfect number engine is better than your average scientific calculator, but can’t compare with what rpn89 using the 89’s CAS could do. It seemed like this was going to be the best I could achieve.

Then in 2018, I discovered Symja. Not only can Symja compete with TI’s CAS, it blows it out of the water. And so began the next phase in my journey to create the calculator I’ve always wanted.

So, without further ado, I’m thrilled to announce Acron Calculator is now available for public beta (notice ‘RPN’ has been dropped from the title – it supports both RPN and WYSIWYG input). Click here to become an Android tester (requires at least Nougat). There are currently no plans to support iOS.

[Nigel (UK)]

Good luck! I still love Acron RPN. Sadly I don't have a device running Nougat so I'll have to miss out on this one. I hope it goes well.

Nigel (UK)

[vanLudwig]

(09-09-2019 10:04 AM)Nigel (UK) Wrote:  Good luck! I still love Acron RPN. Sadly I don't have a device running Nougat so I'll have to miss out on this one. I hope it goes well.

Nigel (UK)

That's a shame; I always valued your input on Acron RPN. Unfortunately that Symja engine requires too much horsepower for it to be worthwhile supporting older platforms.

[lrdheat]

Same boat for me...just me and my Apple iPad...

[johnksellers]

Quote:I’m guessing I’m the only person on this forum that learned RPN on a TI-89. rpn89 by Lars Frederiksen

I cut my teeth on HP's first RPN hand calculator the HP-35. It was an exciting and wonderfully designed device, having only one memory register and a short RPN stack. But it was theoretically impossible to do the function calls, calculations, and data entry with fewer key presses. What an early marvel! It replaced the old desktop cathode-ray tube Friden.

When I was just entering college me and my parents were entertained by a college professor who showed off his multi-thousand dollar desk top Friden Calculator. It had one memory register and displayed 4 registers on a cathode-ray tube screen which was an RPN stack. He sat there with his bulky paper ledger where he recorded his calculations for his high-contrast microscope optics design, completely designed using the Friden.

[vanLudwig]

Hey, I recognize that username - welcome to the forum!

Didn't know you were old enough to have owned an HP-35 though. I bought that first-edition TI-89 when I was in college.

[vanLudwig]

A few features of Acron Calculator you may have missed…

Switch to RPN using 2nd -> MODE and change the Input Mode

Change the number text labels (millions, billions, trillions, etc.) in 2nd -> MODE -> Alternate Answers. You can change the text boxes to the long-scale million, milliard, billion, billiard, etc. (or anything else you want). I like to fill in quadrillion, quintillion, sextillion, septillion, and octillion.

There’s a ton of mathematical functions in the CATALOG that I didn’t have space to add to the buttons. The most-recently used catalog functions are displayed at the top. If you know the function you're looking for, you can use the search.

You can apply algebraic transforms to an answer by selecting it. For instance, if you enter x^2+2x+1 as the question, it will tell you x^2+2x+1 is the answer. But you can select it and simplify to see (x+1)^2. The list of algebraic transforms is automatically pruned to only show you useful commands. You can use the undo button to step backward through any algebraic transforms you’ve applied.

In RPN, some functions use the selection to determine how many parameters to pop off the stack. For instance, roll up/down only roll 2 with nothing selected; if you select something further up, the selected item and everything higher (newer) will roll. The list { } will only take one item if nothing is selected; for larger lists, select the highest item you want in the list. Integrate can usually guess whether you want two or four parameters, but you can use selection to override its guess. In the catalog, anything that uses the selection to determine parameters is marked with an ‘…’

To make a matrix in RPN, create a list of lists. e.g. {{a, b}, {c, d}}.

[vanLudwig]

I wondered for which integers n is the sum of the first n primes itself prime? 2 is prime. 2+3=5 is prime. 2+3+5=10 is not. 2+3+5+7=17 is prime. 2+3+5+7+11=28 is not. Turns out, it happens a lot.

$$\texttt{Select}\left ( \left \{ n \right \}_{n=1}^{2500}, n \mapsto \sum_{i=1}^{n} p_{i}\in \mathbb{P} \right )$$
{1,2,4,6,12,14,60,64,96,100,102,108,114,122,124,130,132,146,152,158,162,178,192,​
198,204,206,208,214,216,296,308,326,328,330,332,334,342,350,356,358,426,446,458,​
460,464,480,484,488,512,530,536,548,568,620,630,676,680,696,708,734,762,768,776,​
780,784,808,814,820,836,844,848,852,926,942,984,992,1024,1026,1030,1036,1070,
1098,1118,1136,1142,1148,1150,1178,1190,1192,1196,1222,1240,1256,1296,1314,1322,​
1338,1356,1386,1394,1400,1408,1418,1420,1446,1460,1466,1474,1476,1478,1494,1500,​
1502,1510,1524,1526,1546,1606,1630,1632,1638,1664,1690,1726,1746,1752,1762,1766,​
1770,1774,1784,1842,1844,1888,1912,1916,1922,1944,1948,1950,1956,2032,2060,2084,​
2094,2106,2128,2130,2132,2184,2186,2200,2210,2222,2262,2268,2288,2310,2324,2342,​
2360,2372,2380,2432,2436,2438,2444,2448,2450,2460,2466,2496}

(This uses catalog functions Select, Table, Function, Sum, Prime, Element, and Primes)

[Valentin Albillo]

(09-16-2019 08:33 PM)vanLudwig Wrote:  I wondered for which integers n is the sum of the first n primes itself prime? 2 is prime. 2+3=5 is prime. 2+3+5=10 is not. 2+3+5+7=17 is prime. 2+3+5+7+11=28 is not. Turns out, it happens a lot.
$$\texttt{Select}\left ( \left \{ n \right \}_{n=1}^{2500}, n \mapsto \sum_{i=1}^{n} p_{i}\in \mathbb{P} \right )$$

Very nice. You can make it run faster by noticing that, apart from the initial "1", n must necessarily be even.

V.

[EdS2]

(09-16-2019 08:33 PM)vanLudwig Wrote:  I wondered for which integers n is the sum of the first n primes itself prime? 2 is prime. 2+3=5 is prime. 2+3+5=10 is not. 2+3+5+7=17 is prime. 2+3+5+7+11=28 is not.

Some resources at the OEIS:
https://oeis.org/A013916

[vanLudwig]

I pushed out BETA 2 of Acron Calculator this morning. I implemented a lot of the framework to support 2nd -> APPS, but there currently aren't any apps available, so this isn't very interesting. I also fixed a draw issue with the catalog search and made some other minor bug fixes.

All the BETA builds expire after 30 days, so you need to update to BETA 2 before BETA 1 expires on October 8th.

[vanLudwig]

I just pushed out BETA 3 of Acron Calculator.

• Dialogs have been machine-translated to 10 languages
• Added answer hints to clarify obscure or ambiguous functions
• Lots of bug fixes and usability improvements
  

If you are multilingual, I would really appreciate you looking over the translations. These were all done by Google Translate, so I'm guessing they are pretty mediocre. Available languages are Spanish, French, German, Italian, Portuguese, Russian, simplified Chinese, traditional Chinese, Korean, and Japanese. If you wanted to make me super duper happy, you could volunteer to touch up any of the resx files.

I think the new answer hints are pretty cool. I've got a pretty wide breadth of math knowledge (though not necessarily a lot of depth), but I wouldn't recognize functions like li and Shi without the hint to tell me they are LogIntegral and SinhIntegral. That at least gives me something to google.

BETA 2 had a horrible error where the antiderivative engine would die whenever the app was put to sleep. This made it appear as if the app couldn't solve even basic integrals. This has been rectified in BETA 3.

I consider BETA 3 to be the full suite of functionality that will be released in version 1.0. I'm pretty much only looking at bug fixes (especially translation bugs) until that release. If you come across anything that needs to be addressed before that public release, please bring it to my attention ASAP.

[cdmackay]

I've just been giving it a try, looks very nice, thanks very much.

One quibble: when running on my Samsung Note 10+, the bottom row of keys are cut-off about halfway. My display is set to 2280x1080.

[jpcuzzourt]

(10-12-2019 08:42 PM)vanLudwig Wrote:  I consider BETA 3 to be the full suite of functionality that will be released in version 1.0. I'm pretty much only looking at bug fixes (especially translation bugs) until that release. If you come across anything that needs to be addressed before that public release, please bring it to my attention ASAP.

I just downloaded it and ran into what seems to be a problem, but I may just be missing something.
By default it comes in WYSIWIG mode. I select yellow-shift->modes and select RPN. Up to this point, fine. But then I cannot seem to get out of the 'modes' menu. Pressing the Android < button pops the wysiwyg/RPN dialog up again.
What's the correct way to exit the menu and preserve my changes?

Thank you. This looks sharp and promising.
Next I need to find and RTFM.

[vanLudwig]

(10-13-2019 01:48 PM)jpcuzzourt Wrote:  But then I cannot seem to get out of the 'modes' menu. Pressing the Android < button pops the wysiwyg/RPN dialog up again.

Shoot, I thought this was only a bug in the Android emulator; I don't see it on my physical devices. Tap the background of the dialog to deselect the picker control. Then the back button will work to exit the Mode dialog. I'll have to find a better solution before final release.

(10-13-2019 10:03 AM)cdmackay Wrote:  One quibble: when running on my Samsung Note 10+, the bottom row of keys are cut-off about halfway. My display is set to 2280x1080.

Weird, this is coded to take up 100% of the screen regardless of resolution. I'll connect to Sumsung's testing cloud and try to figure out why it is getting cut off.

These are exactly the kinds of bugs I'm interested in. Thank you so much, and keep them coming.

[cdmackay]

(10-13-2019 04:57 PM)vanLudwig Wrote:  Weird, this is coded to take up 100% of the screen regardless of resolution. I'll connect to Sumsung's testing cloud and try to figure out why it is getting cut off.

thanks very much; let me know if you need me to test/do anything.

[vanLudwig]

The buttons getting clipped at the bottom of the screen were caused by improper handling of the camera cutout. I have a fix in hand for this issue. As a bonus, the app now consumes the whole screen rather than leaving the space above the camera black.

I'm pretty sure this is the problem with the back button: https://github.com/xamarin/Xamarin.Forms/issues/7311. Looks like they have it fixed in Xamarin Forms, but I don't think they've released a patch yet.

[cdmackay]

(10-14-2019 06:46 PM)vanLudwig Wrote:  The buttons getting clipped at the bottom of the screen were caused by improper handling of the camera cutout. I have a fix in hand for this issue. As a bonus, the app now consumes the whole screen rather than leaving the space above the camera black.

great! thanks very much indeed.

[vanLudwig]

I pushed out BETA 3a with both of these issues resolved

[cdmackay]

(10-15-2019 02:02 AM)vanLudwig Wrote:  I pushed out BETA 3a with both of these issues resolved

Confirmed, key clipping issue resolved for me, thanks very much indeed!