Aleph One seems to randomly choose a subset of polygons from the map to render for #checkpoint sections in terminals. Sometimes this is just an annoying glitch, but in other instances it can really make what would be a useful snapshot of the map worthless. To see this problem do this:
- Download EMR 3 RC1 from
http://sourceforge.net/project/showfiles.php?group_id=167399
- Command-option-new game or control-shift-new game to level 16, Over the River and Through the Woods
- Turn around and enter the metal 'future pod' and read the terminal, to the second page that says "You are currently located here," and the next 2 pages.
Notice a subset of the polygons are not drawn, that those polygons not
being drawn don't seem to have any special properties that should preclude
them from being drawn, and that this map's geometry is largely unchanged
from the days when it worked in Marathon Infinity.
This happens in SDL and in NIBs. I also noticed that among 2 checkpoints
that show the same polygon, one checkpoint can display the polygon while
the next doesn't.
See bug report 1657074 at EMR's project page for more information.
Logged In: YES
user_id=373617
Originator: NO
The polygons that don't display aren't special in themselves, but one thing I noticed is that they seem to be blocked from the goal by one or more visible monster trigger polygons.
Using this hypothesis, is it possible to build a small Marathon 2 map that demonstrates the problem? It's ridiculous to ask would-be bug fixers to download 200+ MB to troubleshoot a problem. Plus the demonstration map could be tested against Marathon 2 and Infinity to see what the behavior was in those engines.
Logged In: YES
user_id=810548
Originator: NO
This bug has always been there. The "official" workaround has been to not use #checkpoint pages.
-:sigma.SB
Fixed in r4818. Checkpoint maps might still be incomplete on high-polygon-count levels, because of a fixed-size buffer in the map-building routine. If this impacts you, please start a new bug.