Peter Gerwinski wrote:
According to FPL:
readln(nom_de_fichier); writeln; assign(f, nom_de_fichier);
What about `rewrite'? (See my example.)
Do you mean one should also use rewrite for reading file? Rewrite is supposed to serve only as a prior request to *writing* into file, isn't it? and reset for reading. Well, whatever you mean, I have used rewrite instead of reset, and the end is the same, exit with code one.
What's bothering me is that gdb does not affect correct values to variables when tracing trough the programm. Without changing one's value, gdb will report, now the good value, now an another one, so I have tried inserting writeln(the_variable) to see what happens, and it results in this GPC error. Meanwhile, good new, it is, as you said, only in reporting debugging information that GPC may fails, because, yes, when putting the breakpoint far away, everything works as expected. The actual problem (if one forgett those error of writting files) is that gdb does behave strangely with some part of GPC code. I say some part because I use intensively gdb and for the very greatest part, it works fine.
Cheers.