Many months ago we had threads about btrees.  There seems to be a problem
with btree wildcard find & chained reads on a master -- if you try to
(regular) find & read detail under each of the masters found in step one, it
loses its place on the master, and the next read on the master says
end-of-chain.

HP was aware of this too, and didn't seem to plan to fix it.  I guess it was
going to be ugly because of only having one set of pointers somewhere (?)

I would dearly love to have this fixed, to avoid the following scenario:
1:    Code new app with btree find on a master
2:    Temporary amnesia
3:    Wait till Jan, 2007
4:    'Enhance' the app to look at detail under the master
5:    Wonder why it stops working.
6:    Re-code to use 2 dbopens.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *