char []funk(char [][]s){
int M[256];
char []ARRAY[256];
return "";
}

Tends to chrash some times. It crashed about in 1 time of 8.  It just crashes, I
cant catch any exceptions.
Actually, it was:

char []funk(char [][]s){
int M[256];
char []ARRAY[256];
my code follows........
}
and it crashes.
Then I put

char []funk(char [][]s){
int M[256];
char []ARRAY[256];
return "";
my code follows........
}
It also crushes.

Thend I put
char []funk(char [][]s){
int M[256];
return "";
char []ARRAY[256];
my code follows........
}
It didn't crash

Then I put

char []ARRAY[256];
char []funk(char [][]s){
int M[256];
return "";
my code follows........
}
it worked, and then I put

char []ARRAY[256];
char []funk(char [][]s){
int M[256];
my code follows........
}
And it worked fine.
I think, that if "char []ARRAY[256];" is inside a function, maybe the garbage
collector misanderstands it. Isnt "char []ARRAY[256];"  filled with crap at the
beginning, and if the gbc looks at it, maybe it tries to free random arrays?

M.

M wrote:
> char []funk(char [][]s){
> int M[256];
> char []ARRAY[256];
> return "";
> }
> 
<snip>

Not sure if it is an error, but this is the only part of syntax that didn't make sense to me.
The D declaration should be:

char[][256] ARRAY;

(if not plz correct me).

I don't know Walter allows the old C style to work also?
Since D arrays are very un-like C arrays, and C libraries are already linkable, why allow it?

I realize it is not syntactically ambigious(ie. allowing it doesn't require any overhead/lookahead) but it just doesn't 'fit' with the rest of the clean syntax.

Just my .02.
Thanks for the great job as always Walter.