WP 34S --> WP 31S

[Sanjeev Visvanatha]

I may be that volunteer. Having gone through the entire process, I feel like I could tackle this challenge. I am not a developer, and like many others, I have a busy day job. But, I do have some momentum now, so I would be willing to contribute to this project.

However, Fair warning to all: I may ask questions that seem straightforward to others. I also may require some direction with respect to working with online code management systems (e.g.
Sourceforge), and I probably will not use the correct jargon when discussing the code, in general.

If you feel like this set of provisos will fit the bill, then I will thrown in my name as a candidate.

[walter b]

(02-06-2014 01:15 PM)Sanjeev Visvanatha Wrote:  However, Fair warning to all: I may ask questions that seem straightforward to others. I also may require some direction with respect to working with online code management systems (e.g. Sourceforge), and I probably will not use the correct jargon when discussing the code, in general.

I can live with many questions - hope you can live with my responses.

Our SW department, however, is headed by other folks.

d:-)

[Jonathan Cameron]

(02-06-2014 01:15 PM)Sanjeev Visvanatha Wrote:  I also may require some direction with respect to working with online code management systems (e.g. Sourceforge), and I probably will not use the correct jargon when discussing the code, in general.

I'm willing to help answer such questions too. In fact, if I am given svn access, I can create the branch and help with the coding. That might make your job a little easier.

-Jonathan

[walter b]

(02-06-2014 01:13 PM)fhub Wrote:  1) in the top row I would prefer the 'basic' functions to be grouped together, so I would suggest a small rearrangement of the keys:

Code:


sqrt(x)   x^2   1/x   y^x   e^x   LN
 >POL    >REC    %    d%   10^x   LG

2) as already suggested by someone else I would prefer x<> on the shifted x<>y key, and at the same time CONV under STO (so CONV and CONST are put together)

3) and finally I don't think that PI is sooooo important (and often used) that it deserves a primary key position. So I would swap PI and HYP, because that would also reduce the keypresses for hyperbolic functions (2 or 3 keypresses instead of 3 or even 4).

Everything else looks quite nice ... :-)

You make me wondering what's wrong with our design. Seriously, some answers:

1. The order as suggested by me takes into account that SQRT and 1/x are the functions used most frequently by an average user. Thus they are put at either edge. 1/x will often be used followed by y^x (for roots); thus y^x is placed next to 1/x. x² is put next to SQRT for obvious reasons. e^x is put next to y^x. One space is left for LN. Here we go.
2. Think of STO as x→, RCL as →x. This matches the symbol x↔ (for swapping x and the content of the target register) very well. Now what's wrong with placing x→ (unshifted) and x↔ (shifted) on the same key?
3. And finally I don't think that HYP is sooooo important (and often used) that it deserves a primary key position. I feel a lot better with π. YMMV.
   

d:-)

[walter b]

(02-06-2014 03:52 PM)fhub Wrote:  Well, I should have known better ..... the usual BS.

(02-06-2014 03:03 PM)walter b Wrote:  SQRT and 1/x are the functions used most frequently by an average user. Thus they are put at either edge ((of that row, of course)).

With this idiotic argument you should as well have put +,-,*,/ at the 4 (top/bottom left/right) corners, because they are certainly the 4 most used keys at all.

Franz,

I had hoped you had recovered since our last meeting - alas, obviously that didn't happen. I sign off wishing you all the best.

([:-/

[walter b]

(02-06-2014 05:39 PM)fhub Wrote:  How could I recover when I have to read your nonsense here again and again?

If it's your inner voice forcing you to do that you should really tell somebody else.

([:-/

[Jonathan Cameron]

A new 'wp31s' branch for the WP 31S project has been created on sourceforge.

-Jonathan

[walter b]

(02-07-2014 03:30 PM)Jonathan Cameron Wrote:  A new 'wp31s' branch for the WP 31S project has been created on sourceforge.

In fact, it's a branch of the WP 34S project on sourceforge - just in case you'll be looking for it.

d:-)

[walter b]

For sake of a clean start, this is the target function set (state of today):

  ENTER_4Bc.pdf (Size: 433.18 KB / Downloads: 81)

A total of 350 operations.

d:-)

[Jonathan Cameron]

(02-07-2014 03:52 PM)walter b Wrote:  For sake of a clean start, this is the target function set (state of today):
A total of 350 operations.

Thanks Walter,

I added this file to the wp31s branch in: docs/WP-31s-plan-v1.pdf

-Jonathan

[jwm]

(01-30-2014 09:11 PM)walter b Wrote:  Let's try to reduce the function set of the WP 34S to lower the entry step for possible RPN juniors. This would lead to a WP 31S which is still "scientific" (should read: technical), but drops programmability and the ... hmmh ... more arcane mathematical functions. So it's boiled down to a "basic scientific" calculator of today, containing a pure subset of 128 functions of the WP 34S function set. It's still way more than a 4-banger, but featuring only one shift key it's way less frightening than a WP 34S. I put the keyboard layout and the index of operations (just showing the full access paths to each and every operation) into a little pdf attached. Hope it suits the target group.

As usual, comments, remarks, etc. are most welcome.

d:-)

I had a similar idea, was actually thinking of making it as close to a 15C clone as possible, perhaps omitting the matrix routines. Then I could use the firmware to reflash my 15C LE, since HP doesn't seem interested in fixing its bugs. I also much prefer the 15C's seamless complex number support as opposed to the multi-stack-location version of the 34s or 33s.

one nit about your layout, make the 'f' key right above the 'on' button to be consistent with most other HP calculators.

John

[walter b]

(02-08-2014 12:55 AM)jwm Wrote:  I had a similar idea, was actually thinking of making it as close to a 15C clone as possible, perhaps omitting the matrix routines. Then I could use the firmware to reflash my 15C LE, since HP doesn't seem interested in fixing its bugs. I also much prefer the 15C's as opposed to the multi-stack-location version of the 34s or 33s.

one nit about your layout, make the 'f' key right above the 'on' button to be consistent with most other HP calculators.

About the HP-15C LE as target platform: Sorry, but its LCD is even less flexible than the one of the HP-30b - its a show-stopper IMHO. And I find 'seamless complex number support' in the HP-42S, not in the HP-15C.

About [f]: Please take some time and walk through this very museum. This will make you change your mind.

d8-)

[BruceH]

Thanks to Jonathan and Walter for your replies.

1) Just as a discussion point (I'm not expecting a change but just saying) - the 31S has a strong statistics capability. HP produced just such a calc at one time (the 21S) and it didn't sell very well. So I suggest that strong stats is not something that users need. Can we use the allocated keyboard space for something more popular?

2) I'm slightly confused by the RMDR function. It's missing in #59 & #60; appears in #66 & #71; gone again in #73 and still gone at #96 which is the 'final' version, but no discussion as to why. I was quite happy with it on the [/] key.

3) Finally, one last late suggestion and apologies for not thinking of it sooner. Given the lack of programming, is there any way to have a key that re-executes the last MORE function executed? For example, suppose you want to do a lot of RMDR calculations and RMDR is on the MORE menu. Pressing [SHIFT][MORE][R][ENTER] each time is a lot of keystrokes. On a programmable machine you could simply create a small program to do these keystrokes in one go. Since the 31S is not programmable this is not an option, so a 'redo last command' type thing would be useful.

[Paul Dale]

(02-08-2014 10:20 PM)BruceH Wrote:  3) Finally, one last late suggestion and apologies for not thinking of it sooner. Given the lack of programming, is there any way to have a key that re-executes the last MORE function executed? For example, suppose you want to do a lot of RMDR calculations and RMDR is on the MORE menu. Pressing [SHIFT][MORE][R][ENTER] each time is a lot of keystrokes. On a programmable machine you could simply create a small program to do these keystrokes in one go. Since the 31S is not programmable this is not an option, so a 'redo last command' type thing would be useful.

The 34S firmware remembers the last menu command used and opens that menu there. This is great for using the same command multiple times in a row.

The reason we didn't implement this for each menu separately, was simply that we ran out of RAM. The design proposed here won't have any sych RAM limit and would be able to store the last position in each menu if that is considered desirable.


- Pauli

[walter b]

(02-08-2014 10:20 PM)BruceH Wrote:  1) ... the 31S has a strong statistics capability. ... Can we use the allocated keyboard space for something more popular?

For sure we can't print on the respective keys [more popular], can we? What is it what you want??

Quote:2) I'm slightly confused by the RMDR function. It's missing in #59 & #60; appears in #66 & #71; gone again in #73 and still gone at #96 which is the 'final' version, but no discussion as to why. I was quite happy with it on the [/] key.

IIRC I gave a reason. I've to find it again since I hate retyping.

d:-)

[i]Edit: Here you are: http://www.hpmuseum.org/forum/thread-576...46.html[/i]

[Sanjeev Visvanatha]

Are any talented hands willing to make an emulator skin for the WP-31S? Using the existing skin for 34S as a guide would ensure proper key registering. Would the original creator of the 34S skin have the editable master for this purpose?

The s/w is progressing well at this point

[Jonathan Cameron]

(02-09-2014 01:33 PM)Sanjeev Visvanatha Wrote:  Are any talented hands willing to make an emulator skin for the WP-31S? Using the existing skin for 34S as a guide would ensure proper key registering. Would the original creator of the 34S skin have the editable master for this purpose?

Sanjeev I can do this.

However, if someone can provide the original gimp .xcf file that would be helpful. I tried one of the ones in the QtGui/images directory and gimp could not open it.

Also, it world be very helpful to find out exactly what font(s) are used in the emulator skin.

-Jonathan

[walter b]

(02-09-2014 03:46 PM)Jonathan Cameron Wrote:  Also, it world be very helpful to find out exactly what font(s) are used in the emulator skin.

The emulator skin was handmade pixel by pixel. I can do that again.

Any particular skin we should start with? I like the 'medium' size best.

d:-)

[Marcus von Cube]

(02-09-2014 04:46 PM)walter b Wrote:  The emulator skin was handmade pixel by pixel. I can do that again.

Not from the beginning . Walter preferred Paintbrush over Gimp and thus dismissed my original Gimp masters.

[walter b]

(02-09-2014 05:05 PM)Marcus von Cube Wrote:  

(02-09-2014 04:46 PM)walter b Wrote:  The emulator skin was handmade pixel by pixel. I can do that again.

Not from the beginning . Walter preferred Paintbrush over Gimp and thus dismissed my original Gimp masters.

It isn't Paintbrush but I prefer individual pixel control anyway.

d:-)