View Full Version : DF 20 - error 50 - Expression Internal error / but runs fine in 18.x releases

Samuel Pizarro
30-Jul-2021, 03:20 PM
I had an expression that was running fine in 18.x but when migrated to df 20 is now returning Error 50

BigInt ibResult ibValue ibShift

Move 0 to ibResult
Move 1 to ibValue
Move 0 to ibShift

Move (ibResult ior ((ibValue iand $1F) * (2^ibShift)) ) to ibResult // <<<---- this line raises error 50

If I break the expression in 2 steps.. it goes fine.

//Move (ibResult ior ((ibValue iand $1F) * (2^ibShift)) ) to ibResult

Move ((ibValue iand $1F) * (2^ibShift)) to iTmp
Move (ibResult ior iTmp) to ibResult

Wasn't sure if it's 64bit related or not (maybe). Feel free to move it to the new forum if that is the case.
I certainly missed the right forum. This was supposed to be posted in the regular windows forum, not in the web one, sorry!
Please DAW, feel free to move it


30-Jul-2021, 04:01 PM

Problem in DF20 in both 32 bit as well as 64 bit, but runs without a runtime error in DF19.1

Nice find.

Samuel Pizarro
30-Jul-2021, 04:17 PM
thanks for confirming it!

Vincent Oorsprong
31-Jul-2021, 03:36 AM

Duplicated and logged.

PS: Also tried with |CI$1F and 15 (decimal representation of $F) instead of $1F which is a shortcut.

Samuel Pizarro
31-Jul-2021, 07:41 AM
Thanks for the ack

but regarding the hex token $1F, I got it from the help. If this is not the expected syntax, the help should be updated.


thanks again