Fixed that odd bug where PolyObjects sometimes leave a vertex or two in their control sectors

Turns out the seg-searching code falsely assumed it'd find frontfacing segs first, who knew?
This commit is contained in:
Monster Iestyn authored and RedEnchilada committed 2016-01-31 13:05:30 -06:00
1 parent e49f531ba1
commit 2e58f6c4d9
1 file changed
+5
+5
View File
@@ -427,6 +427,8 @@ newseg:
// seg's ending vertex.
for (i = 0; i < numsegs; ++i)
{
if (segs[i].side != 0) // needs to be frontfacing
continue;
if (segs[i].v1->x == seg->v2->x && segs[i].v1->y == seg->v2->y)
{
// Make sure you didn't already add this seg...
@@ -593,6 +595,9 @@ static void Polyobj_spawnPolyObj(INT32 num, mobj_t *spawnSpot, INT32 id)
seg_t *seg = &segs[i];
INT32 polyID, parentID;
if (seg->side != 0) // needs to be frontfacing
continue;
if (seg->linedef->special != POLYOBJ_START_LINE)
continue;