Tamás Gálffy
Tamás Gálffy
Thanks for the PR @warsang! Have you tested this with TickInterpolator? How did it perform?
I'd prefer to pick this up after #321, as it might bring some refactors on the internals of `RollbackSynchronizer`. Can't say exactly how much time it will take for a...
> Let's confirm it after https://github.com/foxssake/netfox/issues/347 Nothing to confirm. This is a design choice. > Since a property - independent of its instance - is to have the same index/hash...
@olivatooo getting there :slightly_smiling_face:
Cleaned up the PR a bit so we can focus on the actual changes. In the future, **please keep in mind the following** for PRs: - One PR should implement...
Ok so, I have a few big concerns regarding this implementation. The way I understand it: 1. Host starts sending full states 2. Client receives full state, ack's 3. Host...
> But the bug was that it should be -1 before any state is set. With this PR, _latest_state_tick starts with -1 and doesn't change until an actual state is...
Added some metrics so we can track how much data we save by diff states. Only reports the number of properties, for actual data usage I don't see any other...
Hi @BtheDestroyer, Thanks for raising this! I see that merging per property instead of replacing the whole input reference does bite us in `_submit_input`. As for having multiple input nodes...
For some reason none of the links show up for me :slightly_frowning_face: Maybe an unlisted YT video? Nonetheless, the displacement code is admittedly, rather finicky, so this is valid and...