Carve SOP - split to prims bug (or stupid user)

[Farmfield]

This gotta be a bug, right? After troubleshooting it, I just can't imagine what I can have missed that would mess it up. 

So it's probably something I missed.

procedural.web.v02.issue.hip

 

EDIT:

Seems the dev's added a convert line SOP in H15 that splits a curve/edge into separate prim's - but I would still like to know why the carve SOP doesn't work in this case. 

Edited May 24, 2016 by Farmfield

[Skybar]

Seems to work for me?

 

[Farmfield]

Hehe, how does that work? I see at least three primitives comprised of more than one curve segment.

 

Eller så kan vi växla till Svenska och förvirra skiten ur resten av forumet!

(the text above in no way implies all non Swedes are horrible people, I promise)

Edited May 24, 2016 by Farmfield

[Skybar]

Oh right, you are correct. Really busy at work so just checked quickly . My bad!

Det kan vi gora! For att ge igen nar alla spanjorer och italienare byter sprak och alltid verkar prata om nagot intressant.

[f1480187]

Bug, I guess. PolyCut works, as well as second Carve. They, however, have reversed primitive order to each other.

web_issue.hipnc

[Farmfield]

21 minutes ago, Skybar said:

Oh right, you are correct. Really busy at work so just checked quickly . My bad!

Det kan vi gora! For att ge igen nar alla spanjorer och italienare byter sprak och alltid verkar prata om nagot intressant.

Mjo, fast vi är ju uppväxta utan dubbad TV, det gör - eller iaf gjorde, innan nätet, rätt stor skillnad.

20 minutes ago, f1480187 said:

Bug, I guess. PolyCut works, as well as second Carve. They, however, have reversed primitive order to each other.

web_issue.hipnc

Sound good. Bug is better than me being stupid. 

And how would you do it properly in a PolyCut? I tried my id attribute (same as ptnum) as cut attribute, but messing with the threshold, unpredictable things happen.

[f1480187]

I didn't use any in the scene. It cuts polyline at every specified point then.

[acey195]

this is a very old bug, sometimes it just doesn't split everything the first pass... I hoped they fixed this by now...

 

if you have to use 2 carve nodes its probably faster to use a polysplit or some vexwrangle in primitive mode:

(I typed this from the top of my head so there may be mistakes :P)

////

int closed,i,newPrim,pts[];

closed = primintrinsic(0,"closed",@primnum);

pts = primpoints(0,@primnum);

removeprim(0,@primnum,0);

for(i=0;i<@numvtx-1+closed;i++)

{

newPrim = addprim(0,"polyline");

addvertex(0,newPrim,pts);

addvertex(0,newPrim,pts[(i+1)%@numvtx]);

}

////

also note that you probably need a clean node afterwards, if you want to make sure that you don't have overlapping primitives (this is true whatever method you use)