<<>How about just adding an option to DBUTIL so you could enable on a
database
>level whether DBINFO mode 202 returns the "real" (maximum) capacity or
just
>the "current" capacity (as present). This would seem to avoid the
problem
>of compatibility with applications on space-critical discs that don't
want
>to count on that "maximum" capacity.
 
Normally, this is a topic that I personally don't care much about one
way or the other.  But please _don't_ change the default behavior of a
well-established parameter.  While this particular item's modification
might be fairly benign, this type of change isn't one that we want to
set a precedent for, IMHO.
 
If we're going to lobby for a change (and have any expectation of
getting that or any change implemented), let's lobby instead for a
_new_ DBINFO mode, not a variable meaning for an existing one.>>
 
I have to disagree. I would argue instead that the current implementation
of DBINFO Mode 202 is defective and needs to be fixed. To me, the only
operational purpose of the "capacity" value returned from such a call is
so that an application can obtain an answer to the question "how much can
I put into this data set?" (does anyone have any *real* examples of the
information being required for something else?) The current
implementation of Mode 202 returns an incorrect answer to that question
if DDX is enabled for the set. Certainly, utilities that display
statistics about the database, such as a "% full" value for each set,
would have to be changed, but these utilities are likely to be considered
"broken" anyway if they don't display the full information available from
DBINFO mode 205; once users know that they can have something, they tend
to consider anything that doesn't deliver it to be substandard. I suspect
that the folks behind Adager, DBGeneral, etc. have already made the
switch to Mode 205.
 
Steve Dirickson         WestWin Consulting
(360) 598-6111  [log in to unmask]