[comp.databases] Progress OF is more than it's cracked up to be

jr@oglvee.UUCP (Jim Rosenberg) (09/30/88)

File relationships can be expressed quite cogently and compactly in Progress
using the OF construct.  According to the places I've yet found in the
documentation, the relationship in

for each foo of bar:

is established by a commonly named field in the two files foo and bar.  But
there seems to be at least one other way to do it.  Examining the Progress
dictionary code reveals numerous lines of the form:

for each _Field of _File:

Trouble is, there is no commonly named field between _Field and _File.
However, there is a field in _File which is of type RECID.  This clearly
seems to be how this OF relationship is being achieved.

Is there somebody out there willing to stand up as a Progress wizard who can
confirm this?  Did I miss something in the manual, or is this undocumented
behavior?  It would be awfully strange for Progress to distribute the code
for their data dictionary programs but then have them rely on undocumented
features of the language.  RECID variables are an obvious performance savings,
but I'd hate to have our own code depend on undocumented behavior.
-- 
Jim Rosenberg                        pitt
Oglevee Computer Systems                 >--!amanue!oglvee!jr
151 Oglevee Lane                      cgh
Connellsville, PA 15425                                #include <disclaimer.h>