zah's request: add missing comments

nimbus/db/state_db.nim

@@ -80,6 +80,11 @@ template createTrieKeyFromSlot(slot: UInt256): ByteRange =
   # morally equivalent to toByteRange_Unnecessary but with different types
 
 template getAccountTrie(stateDb: AccountStateDB, account: Account): auto =
+  # TODO: implement `prefix-db` to solve issue #228 permanently.
+  # the `prefix-db` will automatically insert account address to the
+  # underlying-db key without disturb how the trie works.
+  # it will create virtual container for each account.
+  # see nim-eth#9
   initSecureHexaryTrie(HexaryTrie(stateDb.trie).db, account.storageRoot, false)
 
 # XXX: https://github.com/status-im/nimbus/issues/142#issuecomment-420583181

nimbus/vm/computation.nim

@@ -231,8 +231,11 @@ proc generateChildComputation*(fork: Fork, computation: BaseComputation, childMs
       applyMessage(childComp, opCode)
   except VMError:
     # weird Nim bug
-    type T = vm_types.Error
+    # cannot use `Error` directly when constructing error object
-    childComp.error = T(info: getCurrentExceptionMsg())
+    # it seems the compiler confused in the presence of static[Op]
+    # param
+    type ErrorT = vm_types.Error
+    childComp.error = ErrorT(info: getCurrentExceptionMsg())
     debug "applyMesage() failed", error = getCurrentExceptionMsg()
 
   return childComp

nimbus/vm_state.nim

@@ -68,6 +68,8 @@ method getAncestorHash*(vmState: BaseVMState, blockNumber: BlockNumber): Hash256
 
   result = vmState.chainDB.getBlockHash(blockNumber)
   #TODO: should we use deque here?
+  # someday we may revive this code when
+  # we already have working miner
   when false:
     let idx = ancestorDepth.toInt
     if idx >= vmState.prevHeaders.len: