You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since commit 859b1a7 capturing a single frame with spector.js no longer works. I assume it is the same problem as BabylonJS/Spector.js#115. This issue is about different ways to handle that as spector.js is quite useful.
Can we go back to doing the rendering directly in the requestAnimationFrame callback or are there technical issues against that approach? I'll add a pull request with my experiments in this direction.
If not, can we find a way to still capture a single frame with spector.js and document that?
The text was updated successfully, but these errors were encountered:
The main purpose of the linked commit was to remove callback hell from the code.
It might have been an overkill to do that for requestAnimationFrame as it is not used widely, but generally speaking, I would like to keep it as is in terms of APIs.
I would like to receive a better definition of the problem as the linked issue is not specific enough to maplibre, and I would like to get a better reproduction of the issue in terms relevant to this repo and library if possible, i.e. link to jsbin and better description of the issue.
This is a basic map using v4.5.2 and enabling the spector ui: https://output.jsbin.com/tikepev/1
if you click on the red circle to start a frame capture the map remains responsive and after a while you get the error "No frames with gl commands detected. Try moving the camera.".
The same example with v3.6.2: https://output.jsbin.com/tikepev/2
here a click on the red circle blocks the map and after a while you get the capture of a single frame.
Since commit 859b1a7 capturing a single frame with spector.js no longer works. I assume it is the same problem as BabylonJS/Spector.js#115. This issue is about different ways to handle that as spector.js is quite useful.
Can we go back to doing the rendering directly in the requestAnimationFrame callback or are there technical issues against that approach? I'll add a pull request with my experiments in this direction.
If not, can we find a way to still capture a single frame with spector.js and document that?
The text was updated successfully, but these errors were encountered: