Hi all.  How about a survey.

We all must do certain tasks over and over again that require in some
cases extensive workarounds because a component doesn't exist.  What
workaround do you use most often?

A few that I would love to see turned into components:

1) Map values to user-defined domain.  Begone foul functions!
2) Filter/Weave component: ability to accept multiple gate values and
filter by index value. I have two streams of identical length.  I want
to filter this into one stream of the same original length but by
picking values in a binary fashion by index number.  That is give a
list of 0,1,0,1,1,1,etc values that determine the piece of data chosen
for index positions 0,1,2,3,4,5, respectively.  As of now I have to
generate cull patterns to eliminate unwanted data and then use the
Weave component.  It seems to me that the Filter stream component
should accept multiple gate indices and use this "one value per index"
behavior.
3) Shift branch - A component that works like the Shift List component
but works on entire branches.   I find myself using a workaround for
this all the time... which involves a lot of flattening, number
series, subsets, and regrouping and/or path extraction and list
shifting.
4) Split data into paths - You want to divide the flat list of 15
values into 5 sets of 3.  No problem, but it involves creating two
number series, a list of intervals, and subsets.  I would love a
component where you just plug in the list and the group size list as a
repeating mask.  In this case, "3".  The component takes the first 3
values and makes it a path, the next three values and makes it a path,
etc, until the list runs out.  Alternatively, you could provide a
series of numbers that work as a repeating set.  "3,4,5" would create
paths with N=3, N=4, N=5, N=3, N=4, N=5, N=3, etc, until the list runs
out.  Good, no?
5) Lists within lists.  I think you know what I'm talking about.
Breaking down lists into flattening, making subsets, doing an
operation, and then rebuilding the original path structure based on
list length components farther back in the tree.  Lists of lists would
make this so much dreamier.

Other little wishes:

1) Duplicate data: option to generate new paths for each duplicated
data set
2) Override icon preview: really helpful for Parameter, and VBScript
nodes, especially. I love icon view but hate that I have to live with
the icons for parameters and script buttons without differentiation.

There are a few others that I'm forgetting, but perhaps someone can
add some to this thread as well.

Marc




Reply via email to