MS2 - Joint infos

[HENDRIX-ZT2]

mod_rf_stairwell has 1 jointdata, with 1 joint, with 1 hitcheck, with 1 bbox

[HENDRIX-ZT2]

tree_birch_himalayan_01 has 1 jointdata, 6 joints, with 1 hitcheck and with 1 capsule each

[HENDRIX-ZT2]

making good progress here

[HENDRIX-ZT2]

The global structure of joints has been figured out. Some collision types are still missing. The purpose of many joints structs is still unknown.

[harlequinzeg0]

bounding box colliders are bugged when exported with edit bones checked - temporary fix by deleting such colliders prior to export, issue seems to be in generating the rotation 3x3 matrix

[HENDRIX-ZT2]

Fixed by doing abs() on bbox extent, due to swizzling.

[HENDRIX-ZT2]

Right now, all PZ and JWE2 models load, some JWE1 models do not load.

[ilodev]

Can we update this one with the current joint status?

[HENDRIX-ZT2]

Need to investigate JWE1 joints, run a walker. There are some differences in the joints, like it does not have the strings per hitcheck.

For PZ and JWE2, the data layout is good, but our interpretation of the data has huge gaps still.

[ilodev]

should we split hitchecks and joints? looks like they are related by completely different topics

[HENDRIX-ZT2]

should we split hitchecks and joints? looks like they are related by completely different topics

Hitchecks are part of the big joints struct.

[ilodev]

Ok then, maybe it helps if we define the hitcheck struct for bones that done have any and we settle that down?

…

On Fri, 28 Oct 2022 at 11:06, HENDRIX-ZT2 ***@***.***> wrote: should we split hitchecks and joints? looks like they are related by completely different topics Hitchecks are part of the big joints struct. — Reply to this email directly, view it on GitHub <#77 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AG42R76RAPGWYZM2PGNIY3LWFOJRZANCNFSM4VOEDBBA> . You are receiving this because you commented.Message ID: ***@***.***>

[HENDRIX-ZT2]

Lots of stuff figured out by @harlequinzeg0 again. We now know how ik chains are defined. Need to implement it for the plugin.

[ilodev]

Quick question and I'll run a test, can we just add one bone to the armature that is not connected to any other bone without issues?

[HENDRIX-ZT2]

Should be able to, yes

[HENDRIX-ZT2]

Offset has been added back to convex hull bv - not sure why it had been commented out.
Orientation / transform of mesh collision appears to be wrong, need to look at more samples.

[HENDRIX-ZT2]

Convex hulls and mesh collision seemt to correctly import and export now.
Mesh collision optimizer data is partly understood.