{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiha7eolvfcyzlv74iygu2zeqqrjlpq2ir26ukf3kik666cfcw27ky",
    "uri": "at://did:plc:dxjzgxe7cvirxkwfjr2tjspt/app.bsky.feed.post/3mnoxdghlppk2"
  },
  "path": "/t/jme-3-10-0-beta1/49603?page=2#post_33",
  "publishedAt": "2026-06-07T09:21:14.000Z",
  "site": "https://hub.jmonkeyengine.org",
  "textContent": "To make it part of the material pipeline we would need to standardize a rendering pipeline with one or more compute stages, probably not something we would want to do in jme, since computer shaders are not available, or convenient, in all our target platforms.\n\nI think you’d better implement it as a custom `RenderPipeline` (see com.jme3.renderer.pipeline.ForwardPipeline ) , basically instead of relying on `renderManager.flushQueue()` to do the rendering, you would skip that call and implement your own logic through compute shaders.",
  "title": "jME 3.10.0-beta1"
}