Reinventing the wheel?

[Rob Seaman]

Brian replies to Roy who replied to Arnold who replied to ... who  
replied to Doug:

> I hate to see this coming up in this context. This is work that has  
> already been broached by the work by the Quantity group. I suggest  
> taking look at the various online documentation.

By this point I suspect I'm not the only one who is clueless about  
what is actually being discussed.  Doug's message described several  
pragmatic observations about handling multidimensional arrays in  
astronomy.  (The folks from Arecibo were talking to us, too.)  Each  
successive message has become more theoretical in nature.  Is Brian's  
XML schema (modestly versioned at v0.05) proposed as some solution to  
the problems outlined by Doug?  Certainly image processing (whether 2  
or 3 or higher dimensions) is not a new topic in astronomical software.

All design involves compromises, simplifications and (over)  
generalizations.  What is the problem being discussed?  What is its  
proximate solution?  What are the implications of that immediately  
practical solution for larger issues of VO architecture?

Rob
seaman at noao.edu