I can not comment on when, but it will become available, someday.
Vision Builder has been doing this for years.
The LV State Diagram Editor (which is wonderful by the way) can
obviously edit diagrams and is apparently implemented as a VI!
THe SD Editor (which is great by the way, I love this thing, NO MORE
PowerPoint to design) is an incredibly small add-on. This leads me to
suspect much of this functionality (i.e. VI's to edit VI's) is already
implmented in LV, we just do not have mechanisms to get at it (other
than using the very restricted tools mentioned above).
I stongly agree with Wiebe in that this functionality will be rather
challenging to master.
It will also be dificuly to support!
Imagine the type of code that would result. here is one mind bogling
possiblity.
Write three VI's
"A.VI" alternately calls "B" and "C".
"B.VI" Executes a "random VI" evaluates the reuslts and clones a
modified version of itself as "C.VI" based its evaluation of the
random code results and any prior experiments it has performed.
C.VI then duplicates the actions of B.
This could lead to a type of artificial intellegent code that is able
to adapt it own algorithm.
What happens if the code after learning and growing decides it does
not want to do math anymore and would rather just listen to the same
albumn over and over again?
How does a support engineer address that?
The reason I present this extreme case was that this functionaly, when
it comes, it will be limited at first (as we see in dynamic event
registration).
I seem to remember Rolf Kalbermatter (on Info-LabVIEW) saying
something like "Self modifying code is a bad idea and went out style
years ago".
Just my thoughts,
Ben
