I have had a problem with my visualization code which uses Appgraphics since switching some systems to Windows 11. I lived with this on other systems for a while be since switching my main workstation to Win11 it has become really annoying. Aside from placement of text in the window, when selecting a check box to force the program to plot circles for particle sizes, the execution of the plotting loop grinds to a very slow pace. In addition, my mouse action on the desktop becomes erratic when this happens. I then have to basically shutdown the program for the mouse to behave properly. Now, I have not touched this code in over a year and it is the poster child for spaghetti code so i am not looking forward to a debugging session because I not sure what is going but the mouse action away from window being erratic is some sort of clue.

I have updated my SF to the latest this afternoon and the problem still exists. Is there anything new in Appgraphics calls that might be causing this?

Oh, one more thing, the program works fine in Windows 10, no change in behavior!

Thanks in advance.

Rod

Thanks,
Frank

I noticed that selecting Project File --> Add File(s) --> Cancel appears to update, correctly, the code Syntax and Project Issues in the program. Would it be possible to add a menu option that does this in a single selection to easily update the code syntax and project issues?

Frank

This issue occurs only in the Simply Fortran compiler on macOS. Whether using the trackpad or mouse to scroll the page, the scrolling speed is excessively fast, not smooth enough, which affects the user experience.

I can confirm this issue is unrelated to my computer settings (M3 MacBook Air), as page scrolling works perfectly fine and smoothly in all other software.

Could you please check and fix this problem?

I have m_io.mod which I downloaded using sfpm.

   use stdlib_kinds
   use m_io, only: basename, dirname

IDIR=-IC:/Users/truep/AppData/Local/sfpm/32/include -IC:/FORTRAN/PROJECTS/TPPOST/modules

yet I get error:

C:\FORTRAN\PROJECTS\TPPOST/./src/tppost.f90:221: undefined reference to `__m_io_MOD_dirname'

Any idea why it's not linking?  I'm also using stdlib and no issues there.

I've had to add the m_io.f90 source to my project and compile it to my ./modules folder to compile/build.

Trying to use libraries as this project is getting huge...

Thanks,
Bryan

I don't know what the compiler does with that. By calling the function, I'm adding overhead. But making the functional "tool" makes the program more tidy.  If we are talking about and execution difference in seconds or a minute for the whole loop to process the function call, I can live with that and a "cleaner" program. But if it adds 10 minutes, I'd leave the code in the main area.

Simply Fortran has been working great cross platform WINDOWS 11 and LINUX (UBUNTU x86).

I am using the version from the Snap Store on LINUX the majority of the time, but when I use my WINDOWS computer, it is nice to have the debugger.

Is there a version of LINUX SF with all the windows features operational which I could purchase or register?

I'm interested in the debugger and package manager subscription.

Thanks,
Bryan

I've been having trouble with the Build level of past programs, or at their Run Time.
I opened the HelloWorld Sample, Clicked on hello.f90 and Saved, using the second icon in the tool bar (Save).
Then, under the Project menu I selected Save the Project and it's default path - it creates a Fortran folder in my Users/MyName/folder and put the Hello folder there.
Then I clicked the Build (gear) icon and got this:

==============================================================================
Compiling hello.f90
f951: Error: unknown value 'apple-m1' for '-march'
f951: note: valid arguments are: armv8-a armv8.1-a armv8.2-a armv8.3-a armv8.4-a armv8.5-a armv8.6-a armv8.7-a armv8.8-a armv8.9-a armv8-r armv9-a armv9.1-a armv9.2-a armv9.3-a armv9.4-a native
f951: note: did you mean '-mcpu=apple-m1'?
Error: Last command making (build/hello.o) returned a bad status
Error: Make execution terminated

* Failed *

In my "real life" program, I get a Complete on the Build, but when I run it, it exits apparently before reading any records.
It looks like this:
OPEN (UNIT=20, FILE = "permutationsBase10.txt", STATUS="OLD", ACTION="READ", IOSTAT=ierror)

Print *, "let the fun begin"

ierror = 0
icount = 0
resultw = ""
reorder = ""

!*******************************************
WRITE(*,*)
WRITE(*,'(A)') 'Before Read Loop'
WRITE(*,*)

DO
    icount = icount + 1
    Read (20,10,IOSTAT=ierror) O, T, S, H, N, I, A, C, R, Y
    10 format (10I1)

WRITE(*,*)
WRITE(*,'(A)', ADVANCE = 'NO') 'READ RECORD AT '
WRITE(*,'(1I8)') icount
WRITE(*,*)

    IF (ierror > 0) THEN
        WRITE (*,'(A)', ADVANCE = 'NO') 'I/O Error at '
        WRITE (*,'(2I8)') ierror, icount
        GOTO 100
    END IF

And ends with:
END DO
100 Write (*,'(A)', ADVANCE = 'NO')  ' read all '
    WRITE (*,'(I9)')  icount
    Write (*,*)

This is what I see:
let the fun begin
read all         1

So it does the Print "let the fun begin"
Then it appears to jump immediately to "read all" at the end of the program.

Those "Write" statements are not permenant code. They are just something I put in to see if the program got that far. If I did my homework, there is probably a diagnostic option to execute by stepping through the code one line at a time.

But if that simple HelloWorld example had a problem, then there might be other demons at work.

I wondered if I had downloaded the wrong version, one NOT For Silicon Mac, but looking at the website, all versions (Mac, Window, SiliconMac, have the same version numbers).

As an aside, I see that two dimensional Plotting (Aplot) is an option for Mac. It would be nice if an program that showed its use would be added to the New Projects examples.

It appears from my simple program (below) that SF Do Concurrent statement does not accept LOCAL(variables).
It would be better if the variable "R" was locally specified (Local(R)) within the Do Concurrent loop.
It's possible that the code is in error, but if I'm correct it would be a nice to add Local()  to the SF compiler.
Btw, the variable "I" stays local by default even though it's not defined as local.

Frank

Routine and Results:
!--- Example DO_CONCURRENT
!--- The following shows a DO CONCURRENT construct with a mask-expr
!--- and no locality specified for variables:
PROGRAM DO_CONCURRENT
    IMPLICIT NONE
    INTEGER, PARAMETER :: N = 3
    REAL,DIMENSION(N) :: A, B
    REAL              :: C, R
    INTEGER :: I

    A = [4.0,5.0,6.0]
    B = [1.0,2.0,3.0]
    C = 15.0
    R = -1.0
    I = -1

    PRINT *, 'ORIGINAL VALUES:'
    PRINT *, 'A=', A
    PRINT *, 'B=', B
    PRINT *, 'C=', C
    PRINT *, 'R=', R
    PRINT *, 'I=', I
    PRINT *

    WRITE(*,'(1X,A)') 'RUN DO CONCURRENT LOOP:'
    WRITE(*,'(1X,5(A,5X))') 'I', 'A', 'B', 'MOD(A,B)', 'R'

    DO CONCURRENT (I = 1:N, A(I) > 0.0) ! LOCAL(R)
        R =  MODULO (A(I), B(I))

        A(I) = A(I) - R

        C = -15.0

        WRITE(*,'(1X,I1,1X,4(F6.2,2X))')  I, A(I), B(I), A(I)-FLOOR(A(I)/B(I))*B(I), R
    END DO

    PRINT *
    PRINT *, 'DO CONCURRENT FINAL VALUES:'
    PRINT *, 'I=', I, ' (DOES NOT GET REDEFINED!)'
    PRINT *, ' R=',  R, ' (GETS REDEFINED!)'
    PRINT *, 'A=', A
    PRINT *, 'B=', B
    PRINT *, 'C=', C, ' (GETS REDEFINED!)'
END PROGRAM DO_CONCURRENT

ORIGINAL VALUES:
A=   4.00000000       5.00000000       6.00000000
B=   1.00000000       2.00000000       3.00000000
C=   15.0000000
R=  -1.00000000
I=          -1

RUN DO CONCURRENT LOOP:
I     A     B     MOD(A,B)     R
1   4.00    1.00    0.00    0.00
2   4.00    2.00    0.00    1.00
3   6.00    3.00    0.00    0.00

DO CONCURRENT FINAL VALUES:
I=          -1  (DOES NOT GET REDEFINED!)
  R=   0.00000000      (GETS REDEFINED!)
A=   4.00000000       4.00000000       6.00000000
B=   1.00000000       2.00000000       3.00000000
C=  -15.0000000      (GETS REDEFINED!)

Would it be possible to add a APLOT_STYLE_DASH plot series as continuous dashed line to APLOT?

Frank

MEAN.FOR as follows:

Module MEAN_MODULE
    USE ISO_C_BINDING
    IMPLICIT NONE
CONTAINS
   FUNCTION MEAN(A,B) BIND(C, name=“mean”) RESULT (M)
      REAL(C_DOUBLE), VALUE :: A, B
      REAL(C_DOUBLE) M
      M = (A + B)/2.
   END FUNCTION
END MODULE

MEAN.DEF as follows:

LIBRARY MEAN
EXPORTS
   MEAN=“MEAN@16”

This gets … Generating Makefile Okay
But then ….  ld returned 1 exit status
               Last command making (target.exe) returned in bad status.

Any clue as to what I am missing, please.