Very glitchy set of results using LA 1005 on the OFB Hillcreek church lot. Tears and leaky water. No flattening to speak of. Oh well.
From what I can see, it looks like you tried the LA Flatten feature on the church, before packaging it for the first time... is that correct?
Can I assume that you didn't restore from an LA-untouched backup before each test? In general, once a lot has been corrupted, there's no use in continuing to test with the corrupted version of the lot. The one exception is if you think that a separate option might fix the corruption; for example, using the Smooth option to try to correct corruption created by using the Flatten option.
The most likely reason why the LA Flatten option isn't working as expected for this lot: the ground level is not at layer 0 in 3D Array Instance 1. This will have implications for the LA Smooth feature as well.
It looks like the water level is actually being flattened correctly. Since the ground level is not being flattened, this results in the water level sometimes being higher than the ground level.
So, the fundamental question is:
Which layer of 3D Arrray Instance 1 actually contains the ground level?
Answer determined through testing: layer 1.
Since I managed to determine the ground level, I flattened the church lot at layer 1 of the 3D Array. Looks good to me, although it still needs to have the excess buildings removed and the interior lot terrain flattened.
Flattened version of the lot for bbear0287 attached to this post.
What I've learned:
- When flattening the edges of a very bumpy lot, it is best to remove stuff from the edge of the lot before flattening. This may also hold for smoothing.
Here are my current priorities:
- Fix the Pool Offset Bug. Mainly, I want to make sure that the new POOL record type is being parsed correctly for all lots.
- Fix the Smooth feature and the new Flatten feature so that they work for Brighten11's lot. The main question now is: Since there is no level=-1 in the WGRA, how does the LA determine the correct 3D Array Instance 1 layer to change? There must be another record type which has a level=-1...