[Bf-committers] Porting modifiers > 2.8: hook

[Ronan Ducluzeau]

No, Hook modifier is working without vertex group.
Weight values of a vertex group are used as a way to modulate Hook modifier
effect.
Here, a .blend file as an example.
http://pasteall.org/blend/index.php?id=49473

2018-05-07 19:37 GMT+02:00 Christian Hubert <christian.hubert at dstribe.com>:

> Further investigation about hook modifier.
>
> - Some code in "deformVerts" and " deformVertsEM" seems to be unnecessary
> (it aim to obtain a valid mesh with vertex group from the object if the
> mesh
> is null). So I propose to skip this code, as it seems that the ported
> "Simple Deform" does not need it.
>
> -  But it seems that hook cannot work without a vertex group today (and in
> 2.79). But is that expected? There is a " indexar" field in the modifier
> data which seems to be never set to any value and which is the cause of
> that.
> I propose to remove this field (as it as no usage, probably) and hook all
> vertices if there is no vertex group set.
>
> Can you confirm as there could be other consideration that I don't know
> about it?
>
> Thanks
>
>
> -----Message d'origine-----
> De : Bf-committers <bf-committers-bounces at blender.org> De la part de
> Christian Hubert
> Envoyé : lundi 7 mai 2018 17:17
> À : 'bf-blender developers' <bf-committers at blender.org>
> Objet : Re: [Bf-committers] Porting modifiers > 2.8
>
> OK Sybren.
>
> So concerning hook, boolean and triangulate, few questions as I'm facing
> issues that seems to be external to the direct code of these modifiers:
>
> - Hook
>
> mti->deformVerts(md, ctx, mesh, vertexCos, numVerts); from modifier.c is
> called with a null mesh. OK.
>
> But later, this call fails (exception) because there is no vert data in the
> derived mesh:
>
>         if (deformedVerts) {
>                 CDDM_apply_vert_coords(*r_deform, deformedVerts);
>         }
>
> Additionally, this returns no deform vert: modifier_get_vgroup_mesh(ob,
> mesh, hmd->name, &hd.dvert, &hd.defgrp_index);. Is it the good API to call?
>
> - Triangulate: ok, I can submit a diff
>
> - Boolean
> Is OK when the modifier is active, but get_mesh_eval_for_modifier does not
> return a ob->mesh_evaluated (it is null) when the modifier is applied
> ("apply" button).
>
>
>
> To summarize, the direct job is done (or partially), but I'm afraid I
> cannot
> go further by myself without your help.
>
> Please tell me what I should do (submit diff as it is or wait for your
> answer about the above questions).
>
>
> Christian
>
> -----Message d'origine-----
> De : Bf-committers <bf-committers-bounces at blender.org> De la part de
> Sybren
> A. Stüvel Envoyé : lundi 7 mai 2018 10:47 À : bf-committers at blender.org
> Objet : Re: [Bf-committers] Porting modifiers > 2.8
>
> On 06/05/2018 10:29, Christian Hubert wrote:
> > I don't know if all are already ported. But if I can help, feel free
> > to propose me one or two to work on (I'll give a look if I can be able
> > to do it before accepting).
>
> I'm keeping https://developer.blender.org/T54737 up to date. Currently I'm
> working on the Mesh Deform modifier. Maybe pick Boolean, Triangulate, or
> Hook?
>
> Sybren
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers
>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers
>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers
>