|
From: | GFotios |
Subject: | Re: different results at subsequent runs |
Date: | Tue, 2 Nov 2010 21:06:06 +0100 |
On Nov 2, 2010, at 7:24 PM, Jordi Gutiérrez Hermoso wrote:
On 2 November 2010 11:23, GFotios <address@hidden> wrote:A common source of puzzling errors is the reuse of previously defined variables.Or the script could be reading a file that is changing. Or it could be using the output of toc in a computation. Or it could be calling an improperly written .oct file that has undefined behaviour. Or it could be running into a race condition during an asynchronous operation (e.g. network communication). Or there could be a call to a function that does invoke a RNG and wasn't expected to do so. Or the operating system itself may have memory/stack corruption. Frankly, I still prefer my software gnomes theory. ;-) - Jordi G. H.
I could almost bet (not really just emphasizing :) that the problem is the one i mention or the first from your list (the real one not the magic land). Btw is operating system memory corruption included in <common>....? Having the code will indeed help, i just mention what i ve learned through educational experience (90% of the problems with "random" behaving scripts i ve seen (written by students) are related to forgoten variables definitions).
/Fotis
[Prev in Thread] | Current Thread | [Next in Thread] |