Mathias Agopian
Mathias Agopian
@poweifeng the code itself seems pretty inefficient: - `removeReference` does a hash-map lookup while we're iterating the hash-map! That's bad. - there are 2 iterations of the hash-map, maybe we...
ah... but I think for (1) we shouldn't *wait* for the fence, we should just *check* if it signaled and proceed only if it did. If it didn't, we keep...
Thanks for the report. This is a bad one, basically the API is broken.
@romainguy I don't think it will work, because the java (kotlin actually) layer doesn't store the long, it just returns it to the caller: ``` fun createIndirectLight(engine: Engine, buffer: Buffer,...
Okay but we now made all the Java objects bigger just to support that broken API -- And we're breaking the model of java objects being "dumb", now these have...
@yurivict would you be able to sign the CLA so we could submit these patches? I think the branch needs to be rebased as well. Thanks!
@yurivict sorry, but another round of rebase/conflict is needed.
@PsycoTodd would you be able to update the branch? I would like to try and get this PR in again. thanks!
@PsycoTodd would you be able to rebase this PR one more time (it has conflicts again), I'll try to merge it this week.
PCSS is still work in progress.