Post Reply 
WP 31S available at SourceForge
04-15-2014, 07:49 AM
Post: #1
WP 31S available at SourceForge
Hi all,

Seems not everybody interested is aware of the availability of WP 31S. Thus let me repeat:
  • WP 31S is a number cruncher like the WP 34S, but featuring a cleaner, simpler UI and a reduced function set. In particular, it is nonprogrammable.
  • You find calc.bin at http://sourceforge.net/p/wp34s/code/HEAD...realbuild/
  • You find the good old Windows emulator at http://sourceforge.net/p/wp34s/code/HEAD...ndows/bin/
  • For the emulator, download emulator.dll and the .skin, .bmp, and .exe files from said directory; then start .exe
  • For the real thing, flashing works as for the WP 34S (and explained in its manual in detail).
  • There are no adhesive overlays yet. You can get around that by printing a paper overlay using layout (a) as published in the current poll. The overall width of the print shall be 77mm - then you can cut out what you need and it will be to scale.
  • There is no manual published yet. The WP 31S function set, however, is a subset of the one of the WP 34S - except UNDO.
  • WP 31S is the first RPN calculator worldwide featuring a full-fledge UNDO: pressing it will restore the entire calculator status as it was before the last command was executed.
Just download and try. Enjoy! Smile

d:-)
Find all posts by this user
Quote this message in a reply
04-15-2014, 08:03 AM
Post: #2
RE: WP 31S available at SourceForge
(04-15-2014 07:49 AM)walter b Wrote:  You can get around that by printing a paper overlay using layout (a) as published in the current poll. The overall width of the print shall be 77mm - then you can cut out what you need and it will be to scale.

I am and stay impressed...!

Could anyone point to an SVG or a PNG of JUST the overlay please or yet better put it into the source forge directory?

Thank you so much!!!

a.n.
Find all posts by this user
Quote this message in a reply
04-15-2014, 08:14 AM
Post: #3
RE: WP 31S available at SourceForge
Please look here: http://www.hpmuseum.org/forum/attachment.php?aid=406 , take that .png file and get it copied e.g. in a MS Word file, scale it therein, and print it. Take a pair of scissors and do what is explained on p. 170 (of 244) of the WP 34S Owner's Manual. HTH

d:-)
Find all posts by this user
Quote this message in a reply
04-16-2014, 01:15 PM
Post: #4
RE: WP 31S available at SourceForge
(04-15-2014 07:49 AM)walter b Wrote:  
  • There are no adhesive overlays yet. You can get around that by printing a paper overlay using layout (a) as published in the current poll. The overall width of the print shall be 77mm - then you can cut out what you need and it will be to scale.

I have one in progress but cannot finish it until the final layout is decided upon.
Visit this user's website Find all posts by this user
Quote this message in a reply
04-16-2014, 07:36 PM
Post: #5
RE: WP 31S available at SourceForge
(04-16-2014 01:15 PM)Eric Rechlin Wrote:  
(04-15-2014 07:49 AM)walter b Wrote:  
  • There are no adhesive overlays yet.

I have one in progress but cannot finish it until the final layout is decided upon.

Assumed that and I guess I know which. For sake of fairness, however, we should wait until the layout poll is finished at the 19th. Enough time for whoever may have further layout suggestions. And sufficient time for tests using the emulator.

d:-)
Find all posts by this user
Quote this message in a reply
04-16-2014, 08:10 PM
Post: #6
RE: WP 31S available at SourceForge
(04-15-2014 07:49 AM)walter b Wrote:  WP 31S is a number cruncher like the WP 34S, but featuring a cleaner, simpler UI and a reduced function set.

I just installed the emulator and indeed I see a lot of reduction. For instance in the statistical distributions: no matter what the content of J and K is (and the stack, for that matter), the emulator always returns "Invalid PArAMEter". The quantile functions behave differently: they all throw a "RAM is FuLL" error. #-)

What's that? Am I doing somthing wrong?

Quote:There is no manual published yet. The WP 31S function set, however, is a subset of the one of the WP 34S - except UNDO.

So the available functions should work the way they do on the 34s, right?
Since there is no manual: what is the idea behind the greek letters?

(04-15-2014 07:49 AM)walter b Wrote:  Just download and try. Enjoy! Smile

I downloaded and tried. However, the enjoyment currently is a bit limited. ;-)

Dieter
Find all posts by this user
Quote this message in a reply
04-16-2014, 08:21 PM (This post was last modified: 04-16-2014 08:22 PM by walter b.)
Post: #7
RE: WP 31S available at SourceForge
(04-16-2014 08:10 PM)Dieter Wrote:  I just installed the emulator and indeed I see a lot of reduction. For instance in the statistical distributions: no matter what the content of J and K is (and the stack, for that matter), the emulator always returns "Invalid PArAMEter". The quantile functions behave differently: they all throw a "RAM is FuLL" error.

What's that? Am I doing somthing wrong?

(1) An error known already. (2) No, we do. Sad

Quote:So the available functions should work the way they do on the 34s, right?
Since there is no manual: what is the idea behind the greek letters?

(1) Yes. (2) To get faster to the corresponding CONSTants (see p. 147 of 244 of the WP 34S Owner's Manual). Smile

We're working to raise your joy Wink

d:-)
Find all posts by this user
Quote this message in a reply
04-16-2014, 09:10 PM (This post was last modified: 04-16-2014 09:14 PM by Dieter.)
Post: #8
RE: WP 31S available at SourceForge
(04-16-2014 08:21 PM)walter b Wrote:  (1) An error known already.

OK. Do you also know about PRIME? not returning an answer (neither YES nor NO)?

Quote:So the available functions should work the way they do on the 34s, right?
...
(1) Yes.

Compared with the 34s, some command have different names (TSON, TSOFF, CATREG, DMAX, \(\Sigma\)CL, ...), so some additional information would be helpful. BTW why have these names been changed?

Quote:[ greek letters ]
(2) To get faster to the corresponding CONSTants (see p. 147 of 244 of the WP 34S Owner's Manual). Smile

OK. I first thought it was an easy way of entering distribution parameters like \(\mu\), \(\sigma\) or \(\lambda\). Not the worst idea. ;-)

Quote:We're working to raise your joy Wink

Thank you very much.

Dieter
Find all posts by this user
Quote this message in a reply
04-16-2014, 09:13 PM
Post: #9
RE: WP 31S available at SourceForge
(04-16-2014 08:10 PM)Dieter Wrote:  What's that? Am I doing somthing wrong?

Nothing, I introduced some problems when I shrank RAM usage to include UNDO. I've fixed the RAM is full issues I think but there seems to have been another problem that crept in somewhere.


- Pauli
Find all posts by this user
Quote this message in a reply
04-17-2014, 09:34 AM
Post: #10
RE: WP 31S available at SourceForge
(04-16-2014 09:10 PM)Dieter Wrote:  
(04-16-2014 08:21 PM)walter b Wrote:  (1) An error known already.

OK. Do you also know about PRIME? not returning an answer (neither YES nor NO)?

Now I know. NEXTP seems to work correctly though.

(04-16-2014 09:10 PM)Dieter Wrote:  Compared with the 34s, some command have different names (TSON, TSOFF, CATREG, DMAX, \(\Sigma\)CL, ...), so some additional information would be helpful. BTW why have these names been changed?
  • E3ON and E3OFF became TSON and TSOFF; reason: faster access via f DISP T.
  • SHOW became CATREG; reasons: (1) traditional SHOW appears on the 31S, (2) CATREG can be accessed rapidly via f MORE C.
  • DENMAX became DMAX; reason: faster access via f MODE D M.
  • CLΣ became ΣCL; reason: faster access via f CLEAR Σ.

d:-/
Find all posts by this user
Quote this message in a reply
04-17-2014, 12:05 PM (This post was last modified: 04-17-2014 12:24 PM by Dieter.)
Post: #11
RE: WP 31S available at SourceForge
(04-17-2014 09:34 AM)walter b Wrote:  d:-/

That's what I think as well. I think between the 31s and 34s there should be as much consistency as possible. IMHO choosing different names for the same commands is not a particulary good idea.

And now for something completely different: the calendar issue, i.e. the switch between the Julian and the current Gregorian calendar. Years ago this topic has been discussed with regard to the 34s. Finally a flag was implemented that switched between the official date (15 Oct 1582) and the one for the British Empire (14 Sept 1752). Not quite a perfect solution, but at least it works for a large part of the world.

However, the 31s does not have that option. It has a fixed setting that's only valid for the British Empire and its colonies. In particular, the "official" date as stated in Inter Gravissimas is ignored.

Since different regions of the world adopted the new calendar in different years, I think there are only two solution that really make sense:
  1. If no option is given, the fixed switch should be between 4 and 15 Oct 1582, as stated in the official document. This is the solution all calendar programs I know of work with. Even TI did so in their ML-20 program. ;-)
  2. Make the switch date user-selectable, so that everyone can use the correct date for his/her region. For instance with 15,101582 JGSET. This solution has two major advantages:
    • It will work for absolutely every region in the world and not just for two large blocks as it does now on the 34s.
    • It extends the functionality of the date commands because now even calculations based on a so called proleptic Julian or Gregorian calendar become possible. Which means that the Julian resp. Gregorian calendar is assumed for any date.

The 31s now gives us the opportunity to realize an improvement that has not made it into the 34s. On the other hand, if there is no user-selectable switch date, the only valid solution in my humble opinion is a fixed 15 Oct 1582, which is consistent with most other calendar software as well as ISO 8601. However, I think we can do better. Walter, Pauli and all others: let's have the user make his choice. My vote clearly goes to a JGSET command. On the 31s as well as on the 34s. Consider it a bug-fix. ;-)

Dieter
Find all posts by this user
Quote this message in a reply
04-17-2014, 12:24 PM
Post: #12
RE: WP 31S available at SourceForge
JGSET is a good idea. Smile Input shall be a decimal date according to the date mode set. If we're talking about e.g. the 11th of October in 1582, it would be entered 11.101582 in default D.MY, 1582.1011 in ISO standard Y.MD, or 10.111582 in exotic M.DY.

It's a good idea for the WP 34S. In the WP 31S, however, no Julian day numbers are implemented AFAICS Wink No implementation, no problem. Smile

d:-)
Find all posts by this user
Quote this message in a reply
04-17-2014, 12:38 PM (This post was last modified: 04-17-2014 01:14 PM by Dieter.)
Post: #13
RE: WP 31S available at SourceForge
(04-17-2014 12:24 PM)walter b Wrote:  JGSET is a good idea. Smile

Great. :-)

Quote:It's a good idea for the WP 34S. In the WP 31S, however, no Julian day numbers are implemented AFAICS Wink No implementation, no problem. Smile

I suggested this JGSET idea because I noticed a problem: Both ΔDAYS and WDAY use Julian day numbers. Try Johann Sebastian Bach's birthday: the 31s requires the Julian date (21 Mar 1685) for WDAY to correctly return "Saturday", the 34s with SETEUR gives the same answer with the Gregorian date (31 Mar 1685). Or try ΔDAYS with 1 Oct 1582 and 31 Oct 1582 on both calculators (20 resp. 30 days).

Dieter

Edit: Of course DAYS+ is also affected.
Find all posts by this user
Quote this message in a reply
04-17-2014, 12:48 PM
Post: #14
RE: WP 31S available at SourceForge
(04-17-2014 12:38 PM)Dieter Wrote:  Both ΔDAYS and WDAY use Julian day numbers. Try Johann Sebastian Bach's birthday: the 31s requires the Julian date (21 Mar 1685) for WDAY to correctly return "Saturday", the 34s with SETEUR gives the same answer with the Gregorian date (31 Mar 1685). Or try ΔDAYS with 1 Oct 1582 and 31 Oct 1582 on both calculators (20 resp. 30 days).

Oooh, that's an obvious case of imperialism (better: Empirialism) ;-/ which shall be corrected asap. Thanks for reporting Smile

d:-)

(If we don't take care of them some folks tend to regard 'God's Own Country' as the center of the world. Bad education definitly.)
Find all posts by this user
Quote this message in a reply
04-17-2014, 03:18 PM
Post: #15
RE: WP 31S available at SourceForge
(04-17-2014 09:34 AM)walter b Wrote:  
  • E3ON and E3OFF became TSON and TSOFF; reason: faster access via f DISP T.

E3ON and E3OFF are just as fast: Type an E and these the first two commands starting with that letter. ;-)

Yes, there is one more command starting with an E (ENG). It will always stay this way unless each and every single command in a menu starts with a different letter.

(04-17-2014 09:34 AM)walter b Wrote:  
  • DENMAX became DMAX; reason: faster access via f MODE D M.

Following that logic, the other fraction-related settings should be renamed DANY, DFAC and DCONST. #-) Walter, forget it. Pressing D in the MODE menu shows the first of four possible denominator settings and pick the one he prefers. The user can scroll easily through these four options. DMAX simply is inconsistent. Reminds me of photography/sensitometry resp. a German TV station. It should be renamed back to DENMAX.

Accessing commands in menus can be simplified in another way: try grouping similiar commands by having them start with the same initals. For instance like DENANY, DENFAC, DENFIX and DENMAX. This could be done with FitBest, FitLin, FitExp, FitLog and FitPow, with FrcImp and FrcPro, with ClAll, ClRegs and ClΣ... you get the idea. Would have been a nice on the 34s as well. Yes, I know there is a six-character limit.

Dieter
Find all posts by this user
Quote this message in a reply
04-17-2014, 03:39 PM (This post was last modified: 04-17-2014 03:46 PM by Dieter.)
Post: #16
RE: WP 31S available at SourceForge
(04-17-2014 12:48 PM)walter b Wrote:  Oooh, that's an obvious case of imperialism (better: Empirialism) ;-/ which shall be corrected asap. Thanks for reporting Smile

Let's not make this a US vs. Euro thing. I remember earlier discussions on the Julian/Gregorian transition where some remarks made me think the usual switch in 1582 had to be avoided because it was originally decreed by the Pope. I always thought the world had overcome this kind of sentiments. Today we have international standards like the mentioned ISO 8601. If I understand it correclty, it declares the Gregorian calendar as mandatory (and optional for dates before 15 Oct 1582).

ISO 8601 also states the way calendar dates are written: yyyy-mm-dd. Maybe this could become the one and only standard for the 31s and 34s as well. This way a number of commands could be removed. As well as the m.dy indicator that's so annoying for the users beyond the big pond, or the y.md for those in the Far East. A calculator solely based on internationally agreed standards - wouldn't this be a great idea?

Dieter
Find all posts by this user
Quote this message in a reply
04-17-2014, 03:46 PM
Post: #17
RE: WP 31S available at SourceForge
(04-17-2014 03:18 PM)Dieter Wrote:  
(04-17-2014 09:34 AM)walter b Wrote:  
  • E3ON and E3OFF became TSON and TSOFF; reason: faster access via f DISP T.

E3ON and E3OFF are just as fast: Type an E and these the first two commands starting with that letter. ;-)

Yes, there is one more command starting with an E (ENG). It will always stay this way unless each and every single command in a menu starts with a different letter.

You didn't get the point: f DISP A accesses ALL, f DISP E accesses ENG, f DISP F accesses FIX, f DISP S accesses SCI. So E3... had to make way.

(04-17-2014 03:18 PM)Dieter Wrote:  
(04-17-2014 09:34 AM)walter b Wrote:  
  • DENMAX became DMAX; reason: faster access via f MODE D M.

Following that logic, the other fraction-related settings should be renamed DANY, DFAC and DCONST. ... Pressing D in the MODE menu shows the first of four possible denominator settings and pick the one he prefers. The user can scroll easily through these four options.

I tend to disagree. Personally, I hate to scroll through three unwanted items to reach the fourth. Three extra keystrokes ... Sad

(04-17-2014 03:18 PM)Dieter Wrote:  Accessing commands in menus can be simplified in another way: try grouping similiar commands by having them start with the same initals. For instance like DENANY, DENFAC, DENFIX and DENMAX. This could be done with FitBest, FitLin, FitExp, FitLog and FitPow, with FrcImp and FrcPro, with ClAll, ClRegs and ClΣ... you get the idea.

Wrong direction IMHO. Grouping this way slows accessing down. On the WP 31S, catalogs are small enough to allow for each command therein starting with a different letter (though not everywhere).

Real menu trees will become possible with the 43S.

d:-)
Find all posts by this user
Quote this message in a reply
04-17-2014, 03:53 PM
Post: #18
RE: WP 31S available at SourceForge
(04-17-2014 03:39 PM)Dieter Wrote:  A calculator solely based on internationally agreed standards - wouldn't this be a great idea?

For sure. Smile Just our friends in the USA won't buy it. Sad They won't find their units in CONV anymore. Wink

Lesson to be learned: The USA has agreed on many international standards with next to no effect on what they really do. It's not fair but such is real life.

d:-)
Find all posts by this user
Quote this message in a reply
04-17-2014, 05:27 PM
Post: #19
RE: WP 31S available at SourceForge
(04-17-2014 03:39 PM)Dieter Wrote:  ISO 8601 also states the way calendar dates are written: yyyy-mm-dd. Maybe this could become the one and only standard for the 31s and 34s as well. This way a number of commands could be removed. As well as the m.dy indicator that's so annoying for the users beyond the big pond, or the y.md for those in the Far East. A calculator solely based on internationally agreed standards - wouldn't this be a great idea?

The date indicator is not shown if Y register display is enabled (it wouldn't fit on the screen). There's also a compile time option to disable it statically and it wouldn't be difficult to make it a run-time option if needed (although it would increase the size of the firmware).

I use yyyy-mm-dd everywhere because in my opinion that's the only format that really makes sense, and I agree it'd be a good default, but by removing the other options we'd inconvenience very many (if not most) users in order to save a tiny bit of firmware or catalog space. I don't think it's a good idea. International standards should be promoted through leading by example and making them easy to use, not by restricting choice.
Find all posts by this user
Quote this message in a reply
04-18-2014, 12:46 AM
Post: #20
RE: WP 31S available at SourceForge
I quite like the JGSET command suggestion. Julian day numbers are used internally for the various date commands even if they aren't exposed to the user.

The reason we couldn't do this on the 34S was memory. Having the two commonest change over dates cost one bit of non-volatile RAM. Making it fully configurable will cost some bytes. We have the space on the 31S.

We'd also need to update the various locale setting commands and perhaps include them too.


- Pauli
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 2 Guest(s)