Skip to content

Desuq Cafe
Documentation menu

Components at a Glance

Weft’s package documentation includes a full component map, every single piece you can add through Unity’s Add Component menu, grouped by the job it does. This page is a condensed preview of that map: the components a first-time reader is most likely to reach for, not the whole list. The full map ships with the package.

Simulation front-ends

These own a garment or strand’s simulated shape. Exactly one of them drives any given piece of cloth or hair.

  • Hair strand. Walks one bone chain, like a ponytail or a bang, and hands it to the solver. This is the default single-strand starting point.
  • Chain group. Hosts several strands together in one shared world instead of one world each, useful for a skirt’s ring of hem panels or a hairstyle with many locks, and can stitch neighboring strands together so they move as one piece of fabric.
  • Mesh cloth. Simulates a full triangle mesh garment, a skirt or top with no bone chain, by binding directly to its renderer.

Touch and grab

Everything involved in letting a hand, real or scripted, reach in and grab, gated by consent.

  • Touch zones. Holds the painted regions on a strand or garment, each carrying its own consent rule. This is what a grabber points at.
  • Grabber. Drives the actual grab lifecycle for a hair strand or chain group member, and raises events on contact, release, and tension so you can wire a reaction.
  • Mesh grabber. The same grab lifecycle, for a mesh cloth garment.

Scene infrastructure

Colliders, wind, and the scheduler every front-end reads from each frame. None of these simulate on their own.

  • Scheduler. Coordinates per-frame solver stepping across every front-end in a scene.
  • Body colliders. A set of capsules, auto-fitted from a humanoid rig or hand-placed, that keep strands and garments from clipping through the avatar’s own body.
  • Wind sources. Bounded, local wind volumes (a fan, a heat vent, a draft) layered on top of ambient wind.

Authoring assets

Reusable assets you create once and share across garments.

  • Feel profile. A shareable asset capturing a tuned feel, so a headpat feel or a bouncy tail feel becomes something you can save, apply, and share across garments instead of re-tuning every time.

Which components do I actually need

For a single avatar with hair and a garment, the baseline is: a hair strand per hair chain, a mesh cloth for a full mesh skirt or top, body colliders so neither clips through the avatar, and touch zones with a grabber (or mesh grabber) per zone you want grabbable. That covers the whole quickstart.

To add touch response on top of that baseline, wire a reaction hook to the grabber’s events, no extra component needed. To make a zone feel more intimate, change that zone’s consent rule on the touch zones asset itself.

For a crowd or background scene, background characters that do not need full cloth physics can use a much cheaper forward-kinematic sway component instead of a full hair strand.

Where to go next