"not defined" variable IS defined...

[Brian Alvarez]

 Hello Charles, what could be causing this? (FREEFILE is also defined, a function)

[Charles Pegge]

#recordof ff

will display the details of ff at o2 compile-time. Very useful for tracing corrupted definitions.

[Brian Alvarez]

I placed it right before:

Code: [Select]

#recordof ff
ff = (FREEFILE())

[Brian Alvarez]

Dont worry. I am probably doing something wrong. I will keep looking.

[Charles Pegge]

#recordof long

have you accidently redefined long ?

[Brian Alvarez]

No, not this time. But the error dissapeared when i modified another part of the program. I Suspect it was a stray parenteses i was incorrectly generating.

Edit:

Charles, i tracked it to the following statement. I was generating an extra closing parenteses (4 instead of 3 just before the last &) and that had repercutions few statements ahead.

Code: [Select]

    LONG ff
'        [...]
        filename = (TRIM$(PATHNAME$(2, TRIMMED$(EXE.chr(@comp, 37333, 255))), 1, ("/\ " & CHR$(byte {-1, 0, -1, 32}, countof))) & "\" & PATHNAME$(3, TRIMMED$(EXE.chr(@comp, 37333, 255)))) & ".RC")
'        [...]
            ff = (FREEFILE())
 Oxygen is finally compiling PluriBASIC's for 64bit... I mean, it finishes with no errors, but i am getting no output. No executable or DLL is generated. No error reports or anything.

 It takes 55 Seconds...

 I feel i am so close... It finally compiles ALL of the statements, pointers, ASM, TYPES... very complex macros, arrrays...  all the complex stuff!! All 82778 very complex lines of code...

 But no output... no clues. It is correctly compiling complex stuff 1/10 the size of this, but not this big one.

What can i try?

[Charles Pegge]

It could be a silent GPF (Windows 10).

My approach would be to split the code down to isolate the problem.