Alexey Knyazev
Alexey Knyazev
There's another statement in the spec that is not fully compatible with enforcing zero-filled accessor state > When neither `sparse` nor `bufferView` is defined, `min` and `max` properties MAY have...
> It would still have to be an accessor, with the specified number of elements and so on. Yes, a mesh with no-data accessors would still define its vertex formats...
> Should an accessor with missing data be considered valid? Yes, because extensions like Draco may provide its data by other means. Given that the Draco extension is defined on...
> it makes more sense to have them named e.g. `KHR_xmp_json_ld.adoc` Yes and I think we should replace the contents of all `README.md` files with a short standard boilerplate describing...
Excellent question. I think it's a case where _the spirit_ and _the letter_ of the spec have slightly diverged. The underscore attributes are akin to `extras` properties - with no...
@MarkIngramUK From the [Section 3.8.4.1](https://www.khronos.org/registry/glTF/specs/2.0/glTF-2.0.html#_overview): > Client implementations SHOULD follow specified filtering modes. When the latter are undefined, client implementations MAY set their own default texture filtering settings.
> Is that difference intended? Somewhat yes. The MUST is needed because the wrapping mode choice is not up to the implementation.
JPEG images used for textures in glTF assets should not use such EXIF features because exact image decoding is out of application's control in many cases. In this particular issue,...
The short answer is that "unsigned integer" written as text means any unsigned type (`u8`/`u16`/`u32`), while `unsigned int` specifically means `u32`. It's an OpenGL quirk that we should fix in...
On (1): Draco data does contain number of vertices and type/componentType information in some form. Accessors are expected to be used to build VAOs ahead of time, before decompression ends....