request for comments - SDE

[Stephen Lime]

Hi All: The comments so far have been helpful. One 
calrification I'd like to add is that even as just
another data source I still plan to allow a single 
SDE based attribute query that accompanies the 
necessary spatial query. The results will always be
processed through MapServer classes. If you take a
second to think about it this is *really* flexible. 
For example you can use an Oracle "decode" statement
to pre-classify entirely in Oracle/SDE and then use
simple MapServer classes to asign colors. So this is 
not quite as all-or-nothing as I had mentioned earlier.

MapScript may eventually open even more possibilities 
such as opening up OSVecDB/SDE spatial operations like
buffering to shapes retrieved from anywhere.

Steve

>>> "Stephen Lime" <steve.lime at dnr.state.mn.us> 12/29/00 22:33 PM >>>
Greetings all: I'm screwing around with new code to
allow broader support for queries etc. At the heart of 
this discussion is whether consider SDE as simply a
data source or data processor. As it sits the current 
code utilizes much of the SDE processing engine (eg. 
SDE based clipping, RDBMS expressions). This leads to 
very non-general code and potentially confusing
configuration. SDE is sort of special in this case. I'm
leaning to using SDE simply as another datasource,
relying on mapserver functions to project, clip, classify and query. I believe consistency is important
and I really don't want to change the rules for
different data sources.

Understanding that MapServer expressions are ripe for
optimization and expansion (eg. functions), how many
users would this decision impact? Given that a full 
SDE implementation hasn't been out there anyway I would
hope not many.

Have a good new year...

Steve