On 06/20/2011 08:18 PM, Moritz Moeller wrote: > This thing I reported is a completely different issue. It is about > particle data ending up in one GeoInfo but belonging into another. And > again, this works or at least works better when I render the frame 1st > with ScanlineRender. When I render with my renderer I get the munged up > data (position and attributes).
With that I mean that the data I get from Nuke is different, depending on whether the Geo has been accessed by ScanlineRender before or not. I.e. if I connect ScanlineRender and AtomKraft to a Viewer and switch from one to the other, in that order, "id"s are (mostly) fine. If I however do not render with ScanlineRender first, I get the garbled data. I'm more or less certain this is a bug in Nuke. The only problem is to build a case to replicate this. My guess is it has to do with the call sequence I use to get to the GeoInfo. I'd think ScanlineRender uses a different one that gets the right thing to happen. When my node accesses the same data again, I get the correct data from the cache since ScanlineRender triggered its creation, not my node. If I run my node w/o ScanlineRender before, I trigger the bug and get bad data. Just a guess still but everything seems to point to that so far. .mm _______________________________________________ Nuke-dev mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-dev
