Correct layout of e^/LN/10^/LOG?

[C.Ret]

(10-23-2024 11:52 AM)naddy Wrote:  What is the one true sacrosanct layout of the e^, LN, 10^, LOG functions relative to each other on a calculator keypad? And which other placements are heretical abominations that call for burning at the stake?

It is precisely all those which have so many prefix keys that must be put on the stake, INV 2nd Shift DEF Sml Alpha ALPHA f f-1 g h , etc... all witches.

Nothing like the excellence of simplicity, to each god his unique temple:

This, obviously, expresses without any seriousness all my own beliefs.
Yes, yes, of course, trolls exist!

[Maximilian Hohmann]

Hello!

(10-23-2024 10:57 PM)Thomas Klemm Wrote:  It is an external http link from within an https page.
Browsers block these for security reasons.

This is the first time I have encountered that browser behavior! On my devices (MacBook, iPad, mobile phone with different browsers: Firefox, Safari and Chrome) I always see the pictures.

Just to confirm that this is the reason, I have uploaded the same photo to another of my domains which has SSL encryption:



Regards
Max

[Thomas Klemm]

(10-24-2024 09:03 AM)Maximilian Hohmann Wrote:  This is the first time I have encountered that browser behavior! On my devices (MacBook, iPad, mobile phone with different browsers: Firefox, Safari and Chrome) I always see the pictures.

It was already discussed here: Test to img on my site

In Safari Version 18.0.1 (19619.1.26.111.11, 19619) on MacBook M2 Pro I get:

Quote:[Warning] The page at https://www.hpmuseum.org/forum/thread-22557.html requested insecure content from http://www.bombie.de/tmp/Privileg_583D-E_800px.jpg. This content was automatically upgraded and should be served over HTTPS. (thread-22557.html, line 656)

[Warning] The page at https://www.hpmuseum.org/forum/thread-22557.html requested insecure content from http://www.bombie.de/tmp/Privileg_583D-E_800px.jpg. This content was automatically upgraded and should be served over HTTPS. (thread-22557.html, line 1247)

[Error] Failed to load resource: An SSL error has occurred and a secure connection to the server cannot be made. (Privileg_583D-E_800px.jpg, line 0)

[Steve Simpkin]

(10-24-2024 08:07 AM)C.Ret Wrote:  

(10-23-2024 11:52 AM)naddy Wrote:  What is the one true sacrosanct layout of the e^, LN, 10^, LOG functions relative to each other on a calculator keypad? And which other placements are heretical abominations that call for burning at the stake?

It is precisely all those which have so many prefix keys that must be put on the stake, INV 2nd Shift DEF Sml Alpha ALPHA f f-1 g h , etc... all witches.

Nothing like the excellence of simplicity, to each god his unique temple:

This, obviously, expresses without any seriousness all my own beliefs.
Yes, yes, of course, trolls exist!

C.Ret, That brings back memories! My first scientific calculator was a Commodore SR-4148R like the photo you posted. I bought it from the Sears catalog in early 1976 for around $65. I had it for about a year and a half before buying my HP-25.

[Maximilian Hohmann]

Hello!

(10-24-2024 09:45 AM)Thomas Klemm Wrote:  It was already discussed here: Test to img on my site

I missed that one, probably because my browser doesn't behave that way. The two solutions presented: 1) insert a link instead of the image or 2) use "Let's Encrypt" to add https to one's site don't really appeal to me. Option 1 is not nice for the viewers of the posting and option 2 is not allowed by my provider. He wants to sell his own SSL stuff at 7€/month/domain instead...
So maybe someone can come up with yet another solution!

Regards
Max

[AnnoyedOne]

(10-23-2024 10:57 PM)Thomas Klemm Wrote:  It is an external http link from within an https page.

Yeah. Similar issue here I suspect.

https://www.hpmuseum.org/forum/thread-22...#pid193984

Don't post http links! Webmasters--update your site(s)!

A1

[Eddie W. Shore]

(10-23-2024 11:52 AM)naddy Wrote:  Since you all—okay, we—are an opiniated bunch here, I need to ask:

What is the one true sacrosanct layout of the e^, LN, 10^, LOG functions relative to each other on a calculator keypad? And which other placements are heretical abominations that call for burning at the stake?

Because, you see, even Hewlett-Packard couldn't figure it out. Different models use different combinations of shifted and unshifted keys, and there's even the groupings e^/10^ and LN/LOG. (I smell smoke.)

My preference would probably be LOG/10^x and LN/e^x as two keys and shifts, but as long as the functions are there and I can find them, I'm cool.

[AnnoyedOne]

(10-24-2024 08:07 AM)C.Ret Wrote:  It is precisely all those which have so many prefix keys that must be put on the stake...

Early computers didn't have any shift keys either. UPPERCASE only.

Those designers are likely already dead so no burning at the stake for them.

A1

[C.Ret]

(10-24-2024 01:35 PM)AnnoyedOne Wrote:  Early computers didn't have any shift keys either.

Oh! Yes! I know it too well, some very pioneering personal equipment never even had their own keyboard!

And with that, to calculate a transcendental function even elementary, it was not enough to simply press the right key!

At the same period, some were already very privileged:

And there too, with a few exceptions, each mathematical goddess has her dedicated altar. But where is \( 10^x \) ?

[vaklaff]

(10-23-2024 10:51 PM)dm319 Wrote:  I'm going to go with putting Log and Ln on buttons and the Antilogarithms shifted. I justify it because anti- should be shifted.

For consistency '-' and '÷' should also be shifted.

Hear, hear! :-)

[AnnoyedOne]

(10-24-2024 06:15 PM)C.Ret Wrote:  ...some very pioneering personal equipment never even had their own keyboard!

Yeah, I owned one.

I was thinking even further back though.

https://en.wikipedia.org/wiki/UNIVAC_I

Zoom in on the keyboard.

https://en.wikipedia.org/wiki/File:Univa...HM.agr.jpg

A1

[naddy]

(10-23-2024 11:52 AM)naddy Wrote:  there's even the groupings e^/10^ and LN/LOG

Nobody bit on that... It's the 32S that has e^ and LN keys, shifted 10^ and LOG.

[naddy]

(10-23-2024 12:06 PM)rprosperi Wrote:  IMHO, which is 'best' depends on what the user actually uses often - some users will never, ever use e^x and LN, while others can see no reason to even have base-10 logs at all, so how do reconcile the 'best' layout to accommodate both users?

While its pretty much guaranteed no single answer will gain unanimous support, these disucssions remain interesting in they reveal sometimes very different use cases and calculating habits of the various folks that reply.

A tangentially related issue is the inclusion, or lack thereof, of the hyperbolic functions:

• It looks like they first appeared on the 32E, with separate labels for SINH, COSH, TANH, and the respective inverse functions. They were dropped again from the programmable models 33E/C and 34C, presumably because there weren't enough keys.
• They are completely absent from the 41C.
• They reappeared on the 11C and 15C by way of HYP and HYP⁻¹ prefixes.
• The 32S had them, now with a single HYP prefix, so for arsinh you would type [SHIFT][HYP][SHIFT][ASIN].
• The 42S had the functions in its catalog, but they were not assigned to any keys nor in any menu.
  

[dm319]

(10-24-2024 10:28 PM)naddy Wrote:  A tangentially related issue is the inclusion, or lack thereof, of the hyperbolic functions:

Ok now is as good a time as any to ask how did we decide on sin, cos and tan? Why not sec, sin and tan (and co-sec, cosine and cot on shifted functions?). Is it simply that cos/sin/tan are more useful? Actually I don't even know what the 'co' means come to think of it.

[rprosperi]

(10-24-2024 10:58 PM)dm319 Wrote:  

(10-24-2024 10:28 PM)naddy Wrote:  A tangentially related issue is the inclusion, or lack thereof, of the hyperbolic functions:

Ok now is as good a time as any to ask how did we decide on sin, cos and tan? Why not sec, sin and tan (and co-sec, cosine and cot on shifted functions?). Is it simply that cos/sin/tan are more useful? Actually I don't even know what the 'co' means come to think of it.

"The prefix ''co'' links sine with cosine, tangent with cotangent, and secant with cosecant. These pairs are called cofunctions. Cofunctions are linked through complementary angles, meaning they add to 90o. Two angles are complementary if they add to 90o" (blatantly stolen from google search results)

As for hyperbolics, the only engineering application I recall is that for a chain between posts (aka the 'catenary problem') the depth of the drop in the center due to gravity is calculated using hyperbolic cosine.

Are there practical applications which can be solved using hyperbolics in Electrical and Electronic engineering?

[naddy]

(10-24-2024 11:59 PM)rprosperi Wrote:  Are there practical applications which can be solved using hyperbolics in Electrical and Electronic engineering?

The fact that \(\frac{d^2}{dx^2}\sinh x = \sinh x\) and \(\frac{d^2}{dx^2}\cosh x = \cosh x\) means they will pop up in solutions to certain differential equations. In particular they come up when modeling transmission lines, I believe.

[AnnoyedOne]

(10-24-2024 11:59 PM)rprosperi Wrote:  Are there practical applications which can be solved using hyperbolics in Electrical and Electronic engineering?

I vaguely remember using them when studying EE but not for what. Perhaps with complex numbers [as an aside 'j' is used in EE since 'i' is for current]. Or as naddy said transmission line theory. Not my favourite subject. I'm pretty sure that I haven't used them since. Useful in civil/mechanical engineering?

Sin/cos are frequently used in EE. Tan not as much. Cos is just "phased shifted" sin to a EE. As for Sec/Cosec/Cotan I recall them being mentioned but forget the context. Probably not very "important". Perhaps to mathematicians. Then again some of them like to calculate pi to a million decimal places. Why I have no idea. 3.142 works just fine for me. As does 2.718 for 'e'.

A1

[Johnh]

I can't see any point I having dedicated functions for sec cosec and cotan. It's better to just use the 1/x key after cos, sin or tan. This helps to keep in mind the basic relationships between the sides of a right triangle and I think we dont need the extra names.

In civil and structural engineering im not aware of any common use for hyperbolic functions either, though catenary geometry was mentioned above. So I've learnt that here!

[dm319]

(10-25-2024 09:39 AM)Johnh Wrote:  This helps to keep in mind the basic relationships between the sides of a right triangle and I think we dont need the extra names.

True, though I was wondering how we ended on cos/sin/tan as the ones placed directly on the buttons. Bob's reply stimulated to me to do a bit of digging. It looks like we could just as well have sin, sec and tan and be relatively happy given sec is the inverse of cos, you can still calculate your angle-adjacent-hypotenuse relationships without problem.

But! there are reasons which I found in this maths stack exchange question. The most convincing ones that argue against it are:

1. sin x and cos x represent x,y coordinates, which is convenient
2. cos is continuous, sec has undefined spots.
   

[Geoff Quickfall]

Being a Fortran user I would use:

LN
LOG
E^X
** not 10^X


:-)