On Thu, 2006-01-26 at 15:27 +0000, [EMAIL PROTECTED] wrote:
> Author: xan
> Date: Thu Jan 26 16:27:20 2006
> New Revision: 9836

Hmm, I'm pretty sure that the next change which modifies this is a bug.

Originally, we didn't set hitpoints_, and apply_modifications() looks at
the hitpoints_ value cap it to max is modified.  This makes valgrind
complain that hitpoints_ is uninitialized (true).

Initializing hitpoints_ before apply_modifications() is cool, *except*
that apply_modifications can change (esp. increase) maxHitpoints_; in
which case we end up capping hitpoints to the old top, and end up with
all strong or resiliant units starting under max hp.  Hence the need to
cap it *after* apply_modifications().

Hope that helps!
Rusty.
-- 
 ccontrol: http://ozlabs.org/~rusty/ccontrol


Reply via email to