I use $ude alot, in loops too, without any problems.... v10 

My next step would be to compare using a hex editor the .prc of the 2 versions(working one and broken one)

Hi

I could loops through hundreds of "objects" like global procedures and UnifAce will compile them.
But the next call to a procedure after the loop crash UnifAce.

FORENTRY "objects"
 $1 = $ude("compile", ...)
ENDFOR
call GP_HELLO_WORLD("")  ; <----------- Crash

Ingo

I did the same test as above(forentity) for 2 global procs... and it worked

*I did the test via the ide and as a standalone application

------------------------------
Osman Shariff
------------------------------

...
Compile Global Proc module: 'GP_DTFELDER'
Global Proc module: GP_DTFELDER
Compile done:  No compilation messages

 (); DETAIL; 11; [0]; V_PE = $procerror
 (); DETAIL; 12; [0]; reset $procerror
 (); DETAIL; 13; [0]; if (V_PE < 0)
 (); DETAIL; 19; [0]; else 
 (); DETAIL; 20; [0]; fieldvideo "ULABEL.UPRC", "DEF"
 (); DETAIL; 21; [0]; C_CHECK.UPRC/INIT = 0
 (); DETAIL; 22; [0]; endif 
 (); DETAIL; 23; [0]; endif 
 (); DETAIL; 24; [0]; endfor 
 (); DETAIL; 6; [0]; forentity "UPRC"
 (); DETAIL; 25; [0]; debug 
 (); DETAIL; 26; [0]; call GP_ABOUT("") ; <- first call to a procedure
-- nothing more to come here --

neither the call itself is made (show me lines in GP) nor an error is returned and the next step is executed
If I change the number of parameters instead, UnifAce reports this correctly when calling GP_ABOUT :
But then UnifAce crashes at the next CALL GP_xyz

Hi Ingo,

IMHO you should compare your Uniface config (INI + ASN) to a working environment like the one from Osman.

Regards,
Gianni

I presume it works if there are no occurrences in "objects" ?

Did you try running it from the ide... compile/test frm ?

did you try :

call GP_HELLO_WORLD("") 

FORENTITY "objects"
 $1 = $ude("compile", ...)
ENDFOR
call GP_HELLO_WORLD("")  

------------------------------
Osman Shariff
------------------------------

proccode:

ENTRY LP_TEST1
  v_CUROCC = $curocc("UPRC")
  v_STATUS = 0
  setocc "UPRC",-1
  setocc "UPRC",1
  putmess "A1"
  CALL GP_ABOUT("")
  putmess "A2"
  FORENTITY "UPRC"
    IF(C_CHECK.UPRC)
      ;message/hint "%%UVAR.UPRC%%%:%%ULABEL.UPRC%%%"
      reset $procerror
      $1 = $ude("compile", "proc","%%ULABEL.UPRC%%%","library=%%UVAR.UPRC%%%")
break
    ENDIF
  ENDFOR
  setocc "UPRC",v_CUROCC
  putmess "B1"
  CALL GP_ABOUT("")
  putmess "B2"
  RETURN(0)
END

Compile/test in IDE which starts a new session with  "ide /tst fromname"
result:

A1
A2
Compile Global Proc module: 'GF_DTF_DTF_INST'
Global Proc module: GF_DTF_DTF_INST
Compile done:  No compilation messages
B1

Crash and no more information

But this "bug"(or feature?) is not new