programs in physical hardware fail

[Angus]

Right now I started to work with the physical prime when I disovered that my user keys did not work.
I checked the program and got a syntax error. I did not see anything so I deleted the program and copied it over from the emulator. Same thing.

Now if I only have the physical prime attached and drag a *.hpprgm onto it I see an empty program in the prime?

What is that? Shouldn't I be able to drag programs to the prime? (right click and sync seems to work, but doesn't update the prime content in the conn kit AND still gives an error with that program) Why would a program that runs on the emulator give an error on the hardware itself?

btw. I need my programs urgently back on the prime.

*confused* :-(

[Angus]

I can give you an example of some short code that does give a syntax error in lnie 5 e.g. when transferred to the physical calculator.

Code:


//SUBROUTINES
EXPORT FIRCORE(N)
BEGIN
  LOCAL CORE:=seq(0,N);
  FOR I FROM 0 TO N-1 DO
    CORE(I+1):=(12*I-6*(N-1))/(N*(N²-1));
  END;
  RETURN CORE; 
END;

EXPORT FIR(V,C)
BEGIN
  LOCAL N;
  LOCAL RES:=0;
  LOCAL LEN:=DIM(C);
  FOR N FROM 1 TO LEN(1) DO
   RES := RES+V(N)*C(N);
  END;
  RETURN RES;
END;

If someone can point out what is wrong I would be glad. But I have NO idea why on the emulator there are no problems at all.

I would kindly ask to look at these things! Essential for any usage I think.

btw there seem to be issues with the con-kit. I restarted everything and was able to drag the program into the prime w/o having a blank program as a result.

[eried]

The line 5 has a ^2, does the error still happens not using that special char?

[Angus]

Yes, the same.
I think it is interessting that I used several programs for some time on the emulator.

[Jsather]

Angus,

I entered your code in my physical calculator and there is no syntax error:
N:=5;FIRCORE returns {-0.1 -0.1 0 0.1 0.2}.
FIR({1,2,3},{4,5,6}) returns 4.

Just today I transferred some code I had developed on the emulator over to the physical calculator and it failed during run time with a "bad argument" error. It turned out that the following straightforward sequence would generate a bad argument on the physical calculator:
RANDMAT(M1,5,5); M2:=subMat(M1,1,1,4,4);

This was persistent, but as I was tinkering with it, the calculator froze up in a Help screen. It was frozen so hard that holding On and Symb would not restart it. I used a paper clip to reset it and the calculator restarted. After this restart, the "bad argument" problem with subMat had disappeared.

I think this is a fairly common occurrence, that the physical calculator or the emulator will get into a malfunctioning or "buggy" state where it needs to be returned to a default condition. I'm not certain this is the problem in your case, but I wouldn't be surprised. You might cross your fingers and try a paper clip reset.

Good luck. I feel the same pain.

[Angus]

Is there a chance the files are somehow corrupt? Any hidden chars - no matter how they got into the conn-kit code...

Is it possible to load the programs, i.e. the textfiles to to calculator w/o using the software? Or is all that blocked due to misabue in class?

A simple paperclip reset does the job or do I need specific keys in addition?

Seriously, you will understand that such flaws are really annoying and hard to tolerate for me. You don't want to fight with your tools; they are supposed to help you.

thank you

[Angus]

So in order to have at least a working calculator I decided to re-type at least the user keys on the physical prime.
Without modified integer base handling on softkeys work is ugly in my eyes (please see petition on rpn changes, some way to autoexecute a returned string in rpn) plus some common simplifications and assignments I often use.

Now I am sitting here typing a medium sized program on the prime itself. What happens right now is that the comma key changes to a semi-colon key once I am in program editing! Is that working as intended or am I missing something??? How are you supposed to type in code - the comma is essential.
User Keyboard is switched off completely to avoid anything from that. I would really(!) like to avoid picking the colon from the char menu. Btw the dedicated semi-colon key is working as usual so I have two ways to type that char.

[cyrille de brébisson]

Hello,

Quote:What happens right now is that the comma key changes to a semi-colon key once I am in program editing!

Are you in ',' mode for decimal separator? if yes, this would be why your PC program generates a syntax error when transfered on the calcualtor.
try inserting this line:
#pragma mode( separator(., integer(h32) )
at the top of your program and transfer it again...

Cyrille

[Angus]

I'll try that, thank you.

edit: yes, good idea that solved it! Thanks a lot that is a bad trap.
I abandoned playing with the new digit grouping instead of adding pragmas to the code. Only thing I don't quite understand is why the text editor is affected by the way numbers are presented and why that way does harm to the transferred programs. Well don't really care maybe a bug with the new feature, maybe I don't understand it. Not so important.