revert short reorg limit. Add some comments to dailysnapshot

src/cc/payments.cpp

@@ -292,7 +292,7 @@ int32_t payments_gettokenallocations(int32_t top, int32_t bottom, const std::vec
     /*
     - check tokenid exists.
     - iterate tokenid address and extract all pubkeys, add to map. 
-    - rewind to last notarized height for balances? 
+    - rewind to last notarized height for balances? see main.cpp: line# 660.
     - convert balances to mpz_t and add up totalallocations
     - sort the map into a vector, then convert to the correct output.
     */
@@ -885,6 +885,7 @@ UniValue PaymentsRelease(struct CCcontract_info *cp,char *jsonstr)
                         } 
                         else 
                         {
+                            // NOTE: should make this default behaviour.
                             // truncate off any vouts that are less than minimum. 
                             mtx.vout.resize(i+1);
                             break;

src/main.cpp

@@ -656,8 +656,12 @@ std::vector <std::pair<CAmount, CTxDestination>> vAddressSnapshot;
 
 bool komodo_dailysnapshot(int32_t height)
 {
-    int reorglimit = 10; // CHANGE BACK TO 100 AFTER TESTING!
+    int reorglimit = 100; 
     uint256 notarized_hash,notarized_desttxid; int32_t prevMoMheight,notarized_height,undo_height,extraoffset;
+    // NOTE: To make this 100% safe under all sync conditions, it should be using a notarized notarization, from the DB. 
+    // Under heavy reorg attack, its possible `komodo_notarized_height` can return a height that can't be found on chain sync.
+    // However, the DB can reorg the last notarization. By using 2 deep, we know 100% that the previous notarization cannot be reorged by online nodes,
+    // and as such will always be notarizing the same height. May need to check heights on scan back to make sure they are confirmed in correct order.
     if ( (extraoffset= height % KOMODO_SNAPSHOT_INTERVAL) != 0 )
     {
         // we are on chain init, and need to scan all the way back to the correct height, other wise our node will have a diffrent snapshot to online nodes.