Yi Xu

> To me it seems like `fixed32` is just a small wrapper around the quant API? This indeed requires us to support quantized scalars. Our current APIs cannot be used...

After yet another discussion with @k-ye @ailzhang @jim19930609, I have formed a mental picture of future plans and would like to share it here. ## Task A: Refine current APIs...

> minor nit: for subtask b.2, I wonder if ti.types.vector(n, dtype) where ti.types.quant.int/float/fixed are added to the whitelist of dtype makes it simpler for users? `ti.types.vector` and `ti.types.quant.vector` are different...

> I think there are different ways to handle this: use `__all__` to control the public symbols, use `quant._type_factory`, etc. Ah yes. I was stuck at the assumption that we...

> I feel like in real use cases shared exponents are typically used only in vectors. Do you have an example where you need that in a struct? :-) Not...

> I feel like the user may want to access the components via `[]` - for example `color = (fixed5, fixed6, fixed5)` and the user writes `luminance = a[0] +...

> I'm thinking about this: for ti.types.quant.vector(n, dtype), can dtype be a list of quantized types? For example, we may want to allow something like rgb565 = ti.types.quant.vector(3, [fixed5, fixed6,...

Let me do a quick summary (which contains some personal ideas, though). 1. Local vectors are always homogeneous, and should be translated to native types if possible, which can naturally...

Hi @xuhao1. Sorry for my late reply. The problem itself is not related to offline cache. The root cause for the re-compilation is that for now a field doesn't have...

Another thing to note here is that Taichi requires type annotations of kernel parameters, so you may need to figure out a way to pass them in.