Closed ZH-Hristov closed 3 weeks ago
At the very least, we could have the precacheModel function in SF that stops the quota check before it's ran perhaps?
a precacheModel function would be incredibly abusive
Could check if the model exists beforehand, run if doesn't.
Alternatively, maybe the quota pause could be for specific functions like setModel or prop/holo creation, which could cause a precache to happen.
I don't think there's a solution for this. Any addon suffers from it. It's typically models with very large uncompressed textures that lag the most.
I don't think there's a solution for this. Any addon suffers from it. It's typically models with very large uncompressed textures that lag the most.
Is it the textures specifically? I can compress them but I've experienced it happening with a Grigori model with L4D animations and stock HL2 textures ported from the L4D workshop.
Probably a combination. Any ragdoll can be laggy since it has a combination of high res textures and high poly
Doesn't really help find a solution. Only solution I can think of is running the server on an ssd or ramdisk
That's the thing though, I am running it off an SSD, and it can still happen. Awful issue overall, not sure even Rubat could do anything about it. Seems like a deeply rooted engine thing.
Maybe a precache directive could be added --@precachemodel
Is there really nothing that can be done about this? Perhaps pausing the quota checking right before a model is about to be precached and then resuming it afterwards, if that is even possible.
The model doesn't even have to be that high polygon to crash the chip. Simple main character models like Grigori or Eli can crash the chip when being precached.