Maybe someone could help me to understand why for instance…
If inside a ForEach node you say create an Attribute (AttribCreate).
Then why does it not appear on the details View list of the ForEach node looking at it from the outside?
How do you get stuff out of a “for each” node?

It seems really not intuitive and extremely confusing.
After being tortured with my first go of it, I thought it only works on things that were created externally, because creating stuff inside never appeared. Is that also true for Attributes I wondered.
Like would I have create it first outside the loop and then set it in the ForEach node with an attribute transfer or copy or it won't exists?
No it's still defaults coming back out again, even in that case!?

This is seeming really very wrong to me?!! aka poorly done.
What is the logic of a ForEach node if anything you do inside is forgotten afterwards outside? (without dumping to a file node? which is a total bad and a rather nasty work around, that's ultimately equally useless).

Yours truly lost and deeply disturbed
:shock:

I think this is also a great example of how unclear the documentation is and how newbies give up relatively quickly on Houdini, because frankly it makes no sense.

In fairness, ForEach is a very advanced SOP, and those of us who have been using Houdini for many years struggle with it too

So glad to hear you say that; makes me feel semi-human again.

Gotta, this is the video that sorted out a lot of the cobwebs for me:

http://vimeo.com/7346925 [vimeo.com]

Yes thanks, I've seen the video, it's pretty much the only one out there.
It is insightful but does not ultimately help me in this context.

Well I now think the ForEach node, although it has it's place, is basically useless and unusable 95% of the time so should be avoided at all cost (a real time waster). :roll:

:idea: I've taken a VOP SOP with an “Import Attribute”, and a “Bind (Export)” to do an attribute copy (without type issues etc.).

:!: You have to set the “OP Input index” to 1 to get the second Input in the attrib import. So Ironically you can just delete the “global1” and “output1” nodes in the VOP SOP .

Still I'm left thinking there has got to be an “Output” node in the foreach node or something I'm missing there?!? How about an Export changes switch (aka Toggle), that is set ‘on’ as a default.

Well now I suppose the next question for me in this case has become how do I put only Nurbs Curves or Groups into the VOP SOP's main input and not all the vertices? :?
It seems if I try to “delete” the vertices the NURBS Curves and or sweepGroups also get destroyed, so how would I have the VOP SOP iterate over those? coding I guess… I'm totally lost down the rabbit whole for sure.

I've definitely gotten lost down the rabbit hole on this one.
I'm going to post a newb thread about the script node now.