AAARGH!!! well for some strange reason the rest of my message was cut (maybe
some invalid character?). now I look like a jerk, so here's (finally) the rest
(2nd attempt):

uint gc_big_killer[1048576];
for (uint i=0; i<1048576; i++) gc_big_killer[i]=(i<<12)+2048;
// now try to allocate some big blocks a let the gc hunt them ...

(of course I know that vars are initialized, it's ok, helps the gc, indeed!) but when a dynamic array of floats grows (using .length=something), the rest is filled with zeroes instead of NaNs, does the documentation mention this? it could be misleading, one would expect the rest to be initialized using the default initializer .init)

strange: I have to wait for very long until my message gets posted, sometimes it isn't posted at all :((

MaR

On Mon, 13 Jun 2005 11:44:38 +0000 (UTC), <k_mar@seznam.cz> wrote:
> AAARGH!!! well for some strange reason the rest of my message was cut (maybe
> some invalid character?). now I look like a jerk, so here's (finally) the rest
> (2nd attempt):
>
> uint gc_big_killer[1048576];
> for (uint i=0; i<1048576; i++) gc_big_killer[i]=(i<<12)+2048;
> // now try to allocate some big blocks a let the gc hunt them ...
>
> (of course I know that vars are initialized, it's ok, helps the gc, indeed!) but
> when a dynamic array of floats grows (using .length=something), the rest is
> filled with zeroes instead of NaNs, does the documentation mention this?

No, because it's a bug AFAIKS.

> it
> could be misleading, one would expect the rest to be initialized using the
> default initializer .init)

Indeed, which is why I think it's a bug.

> strange: I have to wait for very long until my message gets posted, sometimes it
> isn't posted at all :((

What are you using to post with?

Regan