@Germano Cavalcante (mano-wii) GP strokes have the same problem. It is less likely to run into it, but once this happens it's actually worse because there is seemingly no easy way to change the radius anymore.

@Edward (edward88), I don't feel safe making changes to Grease Pencil's Shrink/Fatten.
Maybe @Antonio Vazquez (antoniov) or someone else on the Grease Pencil team can take a look and come up with a similar solution.

@Germano Cavalcante (mano-wii) If I have understood properly the bug report, the problem is when you Shrink to 0.0 a point. In GPencil there is already a check to avoid this.... I cannot reproduce the problem and I can expand the point again...just need apply a bigger value.

These are the actual lines of the code:

if (td->val) {
       *td->val = td->ival * ratio;
       /* apply PET */
       *td->val = interpf(*td->val, td->ival, td->factor);
       if (*td->val <= 0.0f) {
         *td->val = 0.001f;
       }
     }

As you can see there is a check to avoid 0 value, adding a small bias of 0.001.

Am I missing something?

@Germano Cavalcante (mano-wii) @Antonio Vazquez (antoniov) Ah my bad, I'm sorry! It's just so small that i thought it would'nt expand anymore.

@Antonio Vazquez (antoniov), for GP's shrink/fatten operator this would basically be the same changes seen in rB2d1fe736fabd and rBb70bbfadfece but now for transform_mode_gpshrinkfatten.c.

• rB2d1fe736fabd Remove the radius limitation
• rBb70bbfadfece Allows zero radius scaling

While I can see why limiting radius scaling on curves can be an issue, I'm not sure if it's a problem for Grease Pencil.

I wish the comment deletion wouldn't leave traces of my stupid ideas x')