Jump to content
IGNORED

Fortran99 Challenge


dhe

Recommended Posts

VORTICON,

 I see a couple of things, I don't know, if they are wrong, but would try differently, just for a comparison.

 1) I try to never use multiple disk i/o on the same drive, so I would use:

      For compile:

          dsk2.{source}

          dsk3.{object}

          dsk5.{listing}

      For linking:

          dsk3.{object}

          dsk4.{executable}

          dsk5.{listing}

 

Also, I watched twice and didn't see you load the file, I've find when loading the file from the fortran99 menu, it sometimes will give you good tips about things that it doesn't thank our good.

 

 

 

Link to comment
Share on other sites

47 minutes ago, Vorticon said:

Create a blank disk image (I use TI99Dir) and set it up as DSK3. Start up XB with the compiler disk in DSK1 and one of the library files in DSK2. It will autoload Fortran 99. Press 1 for Edit from the menu then press 2 to start the editor. Paste the program source I posted earlier using Paste (not Paste XB to preserve formatting) and press Fctn-9 when done. Press 3 from the menu to save the source file to DSK3. Now press 7 to exit the edit menu, then go back into it again by pressing 1. Load the source file from disk by pressing 1 then press 2 to edit again. The file will be truncated and sometimes "rearranged". None of this is an issue when using a FIAD.

Thanks! I gave it a try but it's working okay here? Are you on the latest? (399.018?) I recently fixed a bug that caused the emulator to incorrectly parse files on disk images files with more than two fragments. Shouldn't happen on a new disk, but maybe something else was also fixed by that...?

 

(Also, I assume you are using the "DSK Image" option, and not "TI DSK Controller". The latter has no checks for improperly sized disk images, for instance, and that might be biting you if the image is greater than 180k...)

 

Edited by Tursi
Link to comment
Share on other sites

2 hours ago, Tursi said:

Thanks! I gave it a try but it's working okay here? Are you on the latest? (399.018?) I recently fixed a bug that caused the emulator to incorrectly parse files on disk images files with more than two fragments. Shouldn't happen on a new disk, but maybe something else was also fixed by that...?

 

(Also, I assume you are using the "DSK Image" option, and not "TI DSK Controller". The latter has no checks for improperly sized disk images, for instance, and that might be biting you if the image is greater than 180k...)

 

Yup, I am using version 399.018 and the DSK Image option for all the drives 1-3. Below is a video of the issue. Note that the source file is truncated at the end when I reload it. Weird that you are not able to duplicate the issue...

 

 

Link to comment
Share on other sites

4 hours ago, dhe said:

PS.. A clue to the way the file, is the way it is... I tried to cut and paste the list from textpad, and it said - sorry, no can do, this listing has embedded NULLS.

 

 

 

If you use TI99Dir to view the file, you should be able to cut and paste without issues.

 

Here's the compiler output for the PICALCFT program

 

1 PICALCFT                      Version 4.42                Page   0001                    .
0      0001       PROGRAM PICALCFT
       0002
       0003       INTEGER *4 N, I, TOTAL
       0004       REAL *8 X, Y, D, PI
       0005
       0006 C GET NUMBER OF ITERATIONS
       0007       READ(6, 10, END=9999, ERR=9999) N
       0008
       0009 C MAIN PROGRAM LOOP
       0010       I = 1
       0011       DO WHILE(I .LE. N)
       0012         X = IRAND(10) / 10
       0013         Y = IRAND(10) / 10
       0014         D = X**2 + Y**2
       0015         IF (D .LE. 1) THEN
    **Warning- Mixed Mode Arithmetic
       0016           TOTAL = TOTAL + 1
    **Warning- Mixed Mode Arithmetic
       0017         ENDIF
       0018         WRITE(6, 20) I
       0019         I = I + 1
    **Warning- Mixed Mode Arithmetic
       0020       REPEAT
       0021
       0022 C FINALIZE RESULT AND DISPLAY
       0023       PI = TOTAL / N * 4
    **Warning- Mixed Mode Arithmetic
       0024       WRITE(6, 30) PI
       0025
       0026 10    FORMAT('+', C12, M1.1, 'Number of iterations? ', I8)
       0027 20    FORMAT('+', M12.1, 'Iteration # ', I8)
       0028 30    FORMAT('+', M23.1, 'Approximate PI = ', F1.6)
       0029 9999  STOP
       0030       END
 
1 PICALCFT  Allocation          Version 4.42                Page   0002                    .
0 Statement Labels:
 
    0010 01DC         0020 0208         0030 0228         9999 024C
0 Subprogram References:
 
         IRAND             DI$               DADD$             DCMP$
         J4DIV$            J4MUL$            STOP$             CALL$
         IDCVT$            JDCVT$
0 Local Data Area:
 
  000C j TOTAL      0010 j I          0014 j N          0018 d PI
  0020 d D          0028 d Y          0030 d X
0 Logic Size:   0254
  Data Size:    0038
0 Module Size:  028C
    0004  Warning(s)
    0000  Error(s)

 

I should also mention that 80 column works perfectly fine with Classic99 and on real iron with the F18A. Just set it up from the Fortran 99 Utilities menu.

Link to comment
Share on other sites

OK I figured out what the problem was: The FORMAT statement for the initial READ statement was incorrect. This really takes some getting used to...

I also optimized the program a bit as I got more familiar with the language. Here's the new listing:

 

      PROGRAM PICALCFT
 
C GET NUMBER OF ITERATIONS
      WRITE(6, 10)
      READ(6, 15) N
 
C MAIN PROGRAM LOOP
      WRITE(6, 20)
      I = 1
      DO 1 I = 1, N
        X = IRAND(10)
        Y = IRAND(10)
        D = (X / 10)**2 + (Y / 10)**2
        IF (D .LE. 1.0) TOTAL = TOTAL + 1
1       WRITE(6, 25) I
 
C FINALIZE RESULT AND DISPLAY
      PI = (TOTAL / N) * 4
      WRITE(6, 30) PI
      WRITE(6, 35)
 
10    FORMAT('+', C12, M5.1, 'Number of iterations? ', M5.23)
15    FORMAT(I8)
20    FORMAT('+', M12.1, 'Iteration # ')
25    FORMAT('+', M12.13, I8)
30    FORMAT('+', M20.1, 'Approximate PI = ', F8.6)
35    FORMAT('0 ')
      STOP
      END

And here it is running. It's pretty fast! This also proves that the library files are not corrupted.

 

 

  • Like 2
Link to comment
Share on other sites

Great Job VorTIcon.

  You beat me, and in iteration 2, did all of my thoughts plus.

   1) D is a real and the compiler didn't like D .LE. 1 - it really wanted 1 to be 1.0 - real vs interger.

   2) You declared a number of variables in program iteration 1,  and in 2, you went with the flow, leaving variable types as it's default cast.

               I-N being int, otherwise real.

   I think, like most compilers, this one doesn't like being forced to do things.

 

   You did do something that I didn't think would work, in that I thought the format commands had to appear before their use.

   Your program *PROVES* that isn't true with this compiler! ?

 

   And I like the trick of C12 to clear the screen!

 

d.

  • Like 1
Link to comment
Share on other sites

I got the C12 trick from the sample program in Beard's introduction :)

Incidentally, I did end up needing to add the following types to the program in order to allow for an iteration number > 32767, and the compiler complained but still complied :)

 

INTEGER *4 N, I
REAL *8 TOTAL

I must say, Fortran is a beast to use... But I'm starting to warm up to it he he...

Link to comment
Share on other sites

10 hours ago, Vorticon said:

Yup, I am using version 399.018 and the DSK Image option for all the drives 1-3. Below is a video of the issue. Note that the source file is truncated at the end when I reload it. Weird that you are not able to duplicate the issue...

Yeah, I'd like to fix it if I can...

 

Below is my video of trying - did I miss a step?

 

 

Link to comment
Share on other sites

2 hours ago, Vorticon said:

Not that I could tell. So on a hunch, I deleted the disk image I was using for DSK3 and created a new one, and now everything works as it should. I must have had a bad image for some reason. Sorry about the fuss! Should have done that from the get go. 

Erf. I wish you still had the bad disk... I'd like to know if it's another fragmentation issue. ;)

 

Link to comment
Share on other sites

Incidentally, I wasn't too happy about the value of PI I was getting, so I checked the random number generator function in Fortran 99 and got this:

 

RND_FORTRAN.thumb.jpg.da899343f60fbe1422f2d1d7dab2a533.jpg

 

The expected values are:

MEAN = 4.5

STANDARD DEVIATION = 2.87

 

What still trips me on occasion is the range of variables. For example, if an operation brings a simple integer variable out of range (32767), it goes negative but no overflow error is generated. This is not something we had to worry about in loosely typed languages like Basic or Python...

Link to comment
Share on other sites

6 minutes ago, dhe said:

no fair! post the code! ?

Here you go :) It's based on Moulinaie's XB program.

 

 PROGRAM RANDTEST
 
      INTEGER IRAND,  X
      INTEGER *4 T(10)
      REAL S, S2, SD
 
      CALL RANDOM
      PRINT ,'Number of data points?'
      ACCEPT ,N
      DO 10 I = 1, N
        X = IRAND(10)
10      T(X + 1) = T(X + 1) + 1
      S = 0
      S2 = 0
      DO 20 I = 0, 9
        S = S + (T(I + 1) * I)
20      S2 = S2 + (T(I + 1) * I * I)
      S = S / N
      S2 = S2 / N
      SD = SQRT(S2 - S * S)
      PRINT , 'MEAN = ', S
      PRINT , 'STANDARD DEVIATION = ', SD
      STOP
      END

 

Link to comment
Share on other sites

I modified the example 1 program from navetrad. I also got my first experience and debugging run time errors.

 

C234567                                 
C     MCADAMS  (A SAMPLE FORTRAN PROGRAM
      INTEGER X, SUM, I                 
      DIMENSION X(15)                   
    1 FORMAT (I5)                       
    2 FORMAT ('+','ANSWER ', I8)        
      I=1                               
    3 READ (6,1) X(I)                   
      I=I+1                             
      IF (I.EQ.16) GO TO 10             
      GO TO 3                           
   10 SUM=0                             
      DO 20 I = 1,15,1                  
      SUM=SUM+X(I)                      
   20 CONTINUE                          
      WRITE(6,2) SUM                    
      STOP                              
      END                               
<EOD> (Version 4.4)                     
                                        
                                        
How, do you get the nice printouts, that copy above is just a copy screen from Classic99, and a CTRL-V here at atariage.

It isn't nearly as read able.

 

 

                                     
                                        
                                        

 

Link to comment
Share on other sites

As I read the manual Errata, I was very pleasantly surprised to find that bitmap support had been added to the package, including the ability to draw lines and circles as well as place text anywhere on the bitmap screen and individual pixel plot routines. This made Fortran 99 quite a powerful package!

As an exercise, I coded the chaotic logistic equation and plotted it on the bitmap screen. Requires the GL, ML and FL libraries. Fun:)

 

 

      PROGRAM LOGISTIC
 
      INTEGER X, Y
 
      CALL FILES(1)
      CALL CLEAR
      CALL SETMOD(4)
      CALL SCREEN(2, 16)
 
      WRITE(6, 20)
 
      P = 0.4
      R = 2.5
 
      DO WHILE (X .LE. 250)
        P = P * R * (1 - P)
        Y = IFIX(191 - (P * 191))
        X = IFIX((R - 2.5) * 170)
        CALL SETPIX(X, Y)
        R = R + 0.0001
      REPEAT
 
10    CALL KEY(0, KEYC, ISTATUS)
      IF (ISTATUS .EQ. 0) GOTO 10
 
20    FORMAT('+', M24.1, 'Logistic equation')
      STOP
      END

 

  • Like 3
Link to comment
Share on other sites

Sorry if this was asked before, how does the speed of the compiled program compare to other languages like c99 or xb(compiled)? 

 

(Edit) I’m not asking about a specific program on this thread, just in general, how does FORTRAN compare to the other commonly used languages in speed?  I knew there was FORTRAN for the TI but I didn’t realize how complete it was.

Edited by DuaneAL
Link to comment
Share on other sites

10 hours ago, DuaneAL said:

Sorry if this was asked before, how does the speed of the compiled program compare to other languages like c99 or xb(compiled)? 

 

(Edit) I’m not asking about a specific program on this thread, just in general, how does FORTRAN compare to the other commonly used languages in speed?  I knew there was FORTRAN for the TI but I didn’t realize how complete it was.

It compares very favorably to assembly. Probably vey close to the speed of compiled XB, but with major advantage of direct bitmap access, access to the VDP (memory and registers), and double precision floating point operations when needed. There is even a sound facility! With the exception of speech, it's pretty much complete. An impressive feat and a very under-appreciated package. 

  • Like 3
Link to comment
Share on other sites


fortran99 does not use an intermediate assembly source file like say Clint Pulleys c99.

The procedure to create a program, is compile (produce an object file) and then link.

There are extensive facilities to include things like symbol files for debugging, as well as a whole chapter devoted to debugging the assembled output.

Most people will only need to worry about compiler errors and run time errors.

 

Link to comment
Share on other sites

3 minutes ago, dhe said:


fortran99 does not use an intermediate assembly source file like say Clint Pulleys c99.

The procedure to create a program, is compile (produce an object file) and then link.

There are extensive facilities to include things like symbol files for debugging, as well as a whole chapter devoted to debugging the assembled output.

Most people will only need to worry about compiler errors and run time errors.

 

From that I guess the only way to see the generated code would be with a dis assembler then.

Good to know.  Thanks.  It sounds like an amazing development system.

Link to comment
Share on other sites

3 hours ago, dhe said:

 

Most people will only need to worry about compiler errors and run time errors.

 

Unfortunately, I feel this is an area of weakness with this package. The compiler is not very good at catching all errors, and to date all the run time errors I have encountered only give the address of the error in the compiled code and not a pointer to the source code line causing the error, which makes debugging slow and painful. 

Link to comment
Share on other sites

I agree, fortran99 absolutely expects your code to be 100% correct and only the grossest of syntax errors are flagged - which is unfortunate when your getting acquainted with a language.

 

Having said that, my next personal challenge is a snippet of code to open and read from a file.

 

 

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...