Turbine blade textures garbled after texture reload
octarine-noise opened this issue ยท 5 comments
Reloading texture packs with assembled turbines inside render range will make the turbine blade textures garbled. Breaking and reassembling the turbines will fix it, as will quitting and loading up the world again. I didn't try going beyond chunkloading range and coming back again, but I'm guessing that will fix it too.
Kind of irrelevant in itself, but it may point to some deeper issue that could hit you in the face later.
If a break/reassemble fixes it, then all it needs is a simple re-render. I just need to catch the appropriate event. Shouldn't be hard.
Hmm, I'm not able to reproduce this myself. My steps are:
- Create a turbine (as shown)
- Close MC.
- Install Sphax BR texture pack
- Open MC.
- Verify that only default pack is loaded.
- Open world, stand on top of turbine.
- Open options menu, add Sphax BR texture pack
- Close options menu; system lags briefly while textures are restitched
- Everything looks OK!
Strange. The same steps will reproduce it for me. The exact appearance of the problem varies with the texture pack, and whether it is loaded or unloaded.
For example, if I enable PureBDcraft, the blades will simply disappear. I'm guessing they're still there but rendering with a transparent texture. Disabling PureBDcraft will make them garbled. The result is also garbled textures when both enabling/disabling Frenden's Meringued pack.
However, if I don't reassemble the machine after the pack change, but disable the pack again (and restore the texture atlas to an identical state to when the machine was assembled), the blades will look OK again.
The only mods I have installed:
- Forge 10.13.0.1208
- CoFH Core 3.0.0B5
- BigReactors 0.4.0rc5
- CCC 10.0.3.23
- NEI 10.0.3.51
This is what it looks like when unloading Frenden's (this was the most spectacular)
I suspect I know what this is; if you have a large enough resource pack, the UV coordinates for a given block's texture will not be identical after the reload. That plus the caching I do means that the rotor TESR is using the old UV coordinates instead of remapping to the new ones.
I'll have to do some research; not sure I can just use a texture stitch event handler for this.