More work cleaning up and testing the use of charge waivers, as well as security deposit utilizations. Much of this was just around the concept of determining balances, which wasn't / isn't very straightforward. It's not hard to calculate a balance for a particular sitation, but it is difficult to generalize. I think it's reasonable as it now stands, but the StatementEntry::stats algorithm is certainly subject to change. Also, I didn't double check every case which called the stats() function, so I highly suspect we'll find cases where the code is not expecting the new return values, either symantically or logically.

git-svn-id: file:///svn-source/pmgr/branches/yafr_20090716@480 97e9348a-65ac-dc4b-aefc-98561f571b83

db/schema.sql

@@ -956,7 +956,6 @@ CREATE TABLE `pmgr_transactions` (
                               'CLOSE',
 --                               'CREDIT',
 --                               'REFUND',
---                               'WAIVER',
 			      'TRANSFER')
                            NOT NULL,
 
@@ -1065,10 +1064,10 @@ CREATE TABLE `pmgr_statement_entries` (
   `type`                 ENUM('CHARGE',
                               'PAYMENT',
                               'SURPLUS',
+                              'WAIVER',
 			      -- REVISIT <AP>: 20090730
-			      --  VOID is just to test out a theory
-			      --  for handling NSF and charge reversals			      
-                              'WAIVE',
+			      --  VOID is used for handling NSF and perhaps charge reversals.
+			      --  It's not clear this is the best way to handle these things.
 			      'VOID')
                            NOT NULL,