🚀 go-pugleaf

Brian - following up on our conversation, and this mail, the main barrier to
doing more with HDF from the Khoros viewpoint is time and $$$.  If the
Consortium members make it a priority it will happen; if not, it's a nice
idea that will accidently get done if it's easy or someone in the "community"
does it.  I'd guess that an interface to HDF isn't real hard for a hard-
core Khoros programmer, but otherwise, wading through the code would be
challenging.  

I wouldn't discount a "translator" approach; the data connections between
khoros programs can be implemented either as files (the default) or
through shared memory (mapped memory? it's the choice other than
"file" when you click on a connection in cantata).  So, the
translation program can be viewed as getting the data into memory with
minimal overhead for translating the header.

On the other hand, I exactly agree that Khoros will be more attractive
to NASA projects if HDF is "supported".  

George Huffman