Exception while rendering map
Katrix opened this issue ยท 19 comments
Encountered this while rendering on of my maps.
Exception in thread "Thread-8" Exception in thread "Thread-9" java.lang.IllegalArgumentException: The box is degenerate on z
at com.google.common.base.Preconditions.checkArgument(Preconditions.java:122)
at de.bluecolored.bluemap.core.util.AABB.<init>(AABB.java:92)
at de.bluecolored.bluemap.core.util.AABB.<init>(AABB.java:60)
at de.bluecolored.bluemap.core.render.hires.HiresModelManager.getTileRegion(HiresModelManager.java:124)
at de.bluecolored.bluemap.core.render.TileRenderer.render(TileRenderer.java:49)
at de.bluecolored.bluemap.cli.RenderTask.renderThread(RenderTask.java:162)
at java.lang.Thread.run(Thread.java:748)
java.lang.IllegalArgumentException: The box is degenerate on z
at com.google.common.base.Preconditions.checkArgument(Preconditions.java:122)
at de.bluecolored.bluemap.core.util.AABB.<init>(AABB.java:92)
at de.bluecolored.bluemap.core.util.AABB.<init>(AABB.java:60)
at de.bluecolored.bluemap.core.render.hires.HiresModelManager.getTileRegion(HiresModelManager.java:124)
at de.bluecolored.bluemap.core.render.TileRenderer.render(TileRenderer.java:49)
at de.bluecolored.bluemap.cli.RenderTask.renderThread(RenderTask.java:162)
at java.lang.Thread.run(Thread.java:748)
Not sure how I can get more information to help track down the cause, so opening this mostly just to inform that it happened, and get instructions on how to track down the issue.
Did you configure some render-bounds in your map-config?
That might have caused that, i will look into this.
This is confusing me on a new level rn .. xD
@Katrix
Do you get this issue consistently? Does it render any of the world, or does it fail right away.
You don't have changed anything with the tile-size configs, right?
And .. does the world you are trying to render contains chunks with a really high or really low z-position?
Because the only way i can think of how this could happen is if there is some integer-overflow and infinity value stuff going on..
It just skips the world, and continues with the next one. Think it fails instantly. One of the worlds that are failing is a creative flat world if that makes any difference. Stuff from the creative world has also been copied into the other world that's failing.
So i changed some things, but since i can not reproduce the error its quite difficult to say if that helps or makes it worse ^^
But if you still get an error, it will be a different one.
So if you want, you can download this artifact (for 1.13+) or this one (for 1.12), try it again and tell me if you are still getting an error.
Well, it didn't outright crash at least. I'll report back after it's done rendering if anything else broke instead.
Hmm, i fear the only way to track town this issue right now, is if i had a world where this happens to do some live debugging..
For the cases where it shows up as black now, but didn't before, or the world that just didn't render before?
A few months old, but think this one should work. https://mega.nz/#F!5hU2HCJC!IpwU8ZjpwSALOgLjoZtNlA?xlMBkYaR
Oh, also as a note. I'm rendering with caves enabled. No idea if that changes anything.
EDIT: Copied the wrong link. Fixed
Some more info I found on this. When restarting the render after everything was done it found more chunks to render. Didn't fix the problem, but did improve it
So.. for me all the 1.14 worlds rendered fine with the cli .. no unexpected black spots
Another question: do you have the problem only on the lowres models and when you zoom in the hires models get loaded?
Also, have you double checked that your browser is not caching anything?
The caches can be pretty obstinate, i often have to open the console and disable caching there to get the updated tiles ^^
Only issue i found was this:
But that is caused by the light-values not being calculated correctly on the world, i cant do much there..
So I just realized that many requests returned 403 results with the new jar. Seems like it was caused by some files not being gzipped. Is the default set correctly?
So, that solves the problem of why previously fine areas showed up as black with the new jar, but there is still the original maps that failed to render. Will try to get a world I can hand you that showcases those problems. Those did return 404, and not 403, but don't think there was any cases in the world I sent.
As for the shadow thing, think that has to do with those chunks not being loaded since we updated from 1.12.
Alright, yea the gzip issue was my fault, already fixed it with 7fe60fa.
If you get me an example for the other problem, i'll try that as well :)