Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ImplicitTiling looping request #728

Closed
AnthonyGlt opened this issue Aug 29, 2024 · 5 comments · Fixed by iTowns/itowns#2419
Closed

ImplicitTiling looping request #728

AnthonyGlt opened this issue Aug 29, 2024 · 5 comments · Fixed by iTowns/itowns#2419
Labels
bug Something isn't working
Milestone

Comments

@AnthonyGlt
Copy link
Contributor

When I load this link https://storage.googleapis.com/ahp-research/projects/sogelink/hackathon/bdnb/sampleset/tileset.json which has implicitTiling
The data is loading as expected but once everything is loaded, the 0_0_0.subtree is fetched again and provoking errors
Maybe it's caused by one of the implementations left todo but I don't see which one it could be

To Reproduce

Steps to reproduce the behavior:

  • I've used the file example/index.js to load the link
  • I've replaced
    geospatialRotationParent.scale.set( 1, 1, 1 ); by geospatialRotationParent.scale.set( 0.01, 0.01, 0.01 ); because the dataset is big
  • I've set errorTarget & errorThreshold at 0 + checked displayRegionBounds
    Screenshot from 2024-08-29 09-59-06
  • Zoom out

Screenshots
Screenshot from 2024-08-29 09-58-51

Screenshot from 2024-08-29 10-00-03

@AnthonyGlt AnthonyGlt added the bug Something isn't working label Aug 29, 2024
@gkjohnson gkjohnson added this to the v0.3.38 milestone Aug 30, 2024
@gkjohnson
Copy link
Contributor

The positioning of the tile set in the index example is a bit odd 😅 Maybe the automatic positioning logic could be improved. But I was able to get it in frame once and wasn't seeing any errors so this may need a more succinct repro.

@AnthonyGlt
Copy link
Contributor Author

index.js.txt
Here is my index.js (to be renamed). Is it because I've put errorTarget & errorThreshold at 0 ? Did you do it too ? Are you sure you got everything in frame ? Video of what happens to me.
I hope it will help to understand what happens :D

Screencast.from.30-08-2024.10.12.31.webm

@gkjohnson
Copy link
Contributor

I see it now - I think this is related to #726 which I noticed last week. The LRUCache unload behavior isn't working correctly so I need to spend some time this next week figuring out what's going on. Setting the error target to 0 causes the buggy lru cache behavior to kick in.

Also I see that this tile set has two extremely far clusters of tiles making it difficult to see the extent / find the geometry:

image

Adding a new debug settings to display parent tile bounds (Both empty and not empty) would make tile sets like this easier to understand.

@gkjohnson
Copy link
Contributor

gkjohnson commented Sep 13, 2024

@AnthonyGlt do you mind retrying your test? The issue should have been fixed in #726, #727. I still want to fix #741 before a new release but everything should be stable again.

@AnthonyGlt
Copy link
Contributor Author

Yes it got fixed, well done ! :D

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants