Yet more component requests/alterations... Please

Hi David

I know you asked others to write there requests in a their own thread so i shall start one too.. as it seems today everyone has requests.

1. Pathmapper: Any chance of making it so it does not clean trees automatically as i have sometimes need this data tree path data but need the paths names shorted also is there any chance of having a graft input as well and a preview option?

2. Tree Branch: Any chance of changing the path input so it can take wild cards like Tree Split?

3. Number of Numbers Count Component: Currently I use a paths and replace branch component workflow to count the number of times a numbers appear in a sorted list though it would be great is we could have a component that does this with out the need to involve paths.

4. Value List : Could this have an input option so we could write the list before streaming it into the Value List Component.

Thanks for all the great work.

Matt

Replies to This Discussion

Permalink Reply by David Rutten on March 30, 2011 at 1:12pm

Hi Mat,

1. Yeah that's a bug, I'll get it fixed.

2. What would be the difference between Tree Branch and Tree Split then?

3. I'm writing a collection of Set Theory components now and I can see this being one of them. They will only work for simple data types (bools, ints, numbers, strings, points, vectors, colours etc, no curves, breps, circles, meshes). This approach will make it easier to analyse the contents of lists of data, including how often a specific value occurs.

4. I don't think this is a good idea. Changing UI based on the solution can result in all sorts of annoying behaviour where you have to go back and reset a value over an over again and having it erased or changed by a specific solution.

David Rutten

david@mcneel.com

Poprad, Slovakia

Permalink Reply by Matt Gaydon on March 30, 2011 at 1:46pm

Hi David

1. Good to know that can be fixed

2. My reason for requesting the tree branch wild card was although it would be very similar to tree split would it not compute faster is it only trying to find certain paths and not dispatch selected paths into different containers though i maybe wrong.

3. The Set Theory components sound like a great additions. I can not see it being too much of a problem is they don't work with non numerical inputs.

4. Ok no worries, that idea was just a thought.

Thanks

Matt

Permalink Reply by David Rutten on March 30, 2011 at 2:07pm

1. I had a good long look at the lexical path matching code and decided it was crap. I rewrote the whole thing and it should now retain paths with zero length. Also, if you don't use index qualifiers on either side, it should run way faster*, so don't do this:

{A;B;C}(i) -> {B}(i)

when you can get away with this:

{A;B;C} -> {B}

* especially if you have few branches with a lot of data in them. The other way around is still slow.

David Rutten

david@mcneel.com

Poprad, Slovakia

RSS