* Use tree index for DB
* Set uncles_hash, tx_root, receipts_root from verified block
* Use Filth instead of a bool
* Fix empty root check
* Flush block queue properly
* Expunge deref
* small cleanups in trie
* Memory trie skeleton
* decode nodes from RLP
* memorytrie -> memorytriedb
* implement Trie for MemoryTrie
* partially implement insert
* implement trie insertion
* don't load whole trie into memory, lookup across memory and db
* re-implement insertion and lazily load necessary nodes from DB
* node removal w/o fixing
* reduce churn in node storage
* finish remove implementation
* committing the in-memory trie
* reload root node after commit
* generate elastic arrays when converting nodes to rlp
* port triedbmut tests over to memorytriedb, fix a few bugs.
* hash count and dirty flag
* initial attempt for node inspection
* back to skeleton
* inspection framework
* implement insertion
* no panic paths in insert
* implement deletion without fixing
* node fixing
* commit nodes to db
* tracing targets and bugfix
* get tests to pass with a lot of tracing
* set playpen iterations to 10
* remove some tracing statements
* make TrieMut::root take &mut self
* replace triedbmut with memorytriedb
* treat empty insert value as removal
* add test for null insert
* fix some style concerns
* trie: use nodehandle for root_node, minor cleanup
* partially done alternate migration scheme
* finish altering migration framework
* migrate tests to new migration framework
* address comments
* remove superfluous newline
[ci skip]
* TempIdx -> TempIndex
[ci skip]
* modify account_db to work on address hash, not address
* add a database migration for new accountdb
* preserve first 96 bits of keys when combining
* handle metadata keys in migration and preserve first 96 bits
* fix comments and hash address instead of hash
* different migrations based on pruning
* migrations mutably borrow self
* batch abstraction for migration
* added missing licence headers
* overlay recent v7 migration
* better error handling, migrate version key as well
* fix migration tests
* commit final batch and migrate journaled insertions
* two passes on journal to migrate all possible deleted keys
* Display information on journal database on startup.
* Minor restyling.
* Client handles the enode message to avoid repeats.
* Avoid unneeded copies *and* reduce code.
* Fix up typo.
* removed old json-tests
* simplify folds in triehash.rs
* removed unused json_aid
* removed unused squeeze.rs
* json branching tests for trie
* removing todos from util
* separated UsingQueue and Table
* further cleanup, removing unused code
* serde serialization of hash moved to rpc module
* uint wrapper for rpc in progress
* serialization of uint moved to rpc module
* updated eth-secp256k1
* updated igd, serde is no longer dependency of util
* loading trie consensus tests
* renamed aliases in rpc imports
