[Gtkradiant] [Bug 417] q3map: vis bug or stack corruption, you choose

gtkradiant@zerowing.idsoftware.com gtkradiant@zerowing.idsoftware.com
Tue, 12 Mar 2002 17:31:41 -0600


http://zerowing.idsoftware.com/bugzilla/show_bug.cgi?id=417





------- Additional Comments From ydnar@rasterproductions.com  2002-03-12 17:31 -------
My fix is rather dodgy--increased MAX_POINTS_ON_FIXED_WINDING to accomodate 
empirical winding maximums in test case maps where vis errors were previously 
occurring.

The stack ($M!@^*!@*&^!!!!) winding_t copy in visflow.c was corrected to never 
copy more than MAX_POINTS_ON_FIXED_WINDING, nor raise the read exception when 
trying to copy more than the winding has allocated.

There could be certain cases where a portal with > 24 sides may fail to be 
chopped correctly leading to incorrect vis. But we all know nobody makes maps 
like that, right? :)



------- You are receiving this mail because: -------
Whoops!  I have no idea!