0.1 release
  event sources
  rate limits
  NIP-40 expiration
  NIP-33 param replaceable events
  fix sync
  when disk is full it should log warning but not crash
  ensure DB upgrade flow works
  ? why isn't the LMDB mapping CLOEXEC

features
  finish syncing
    * logging of bytes up/down
    * up/both directions
    * error handling and reporting
    * way to close sync request
    * limit on number of concurrent sync requests
    * full-db scan limited by since/until
    * `strfry sync` command always takes at least 1 second due to batching delay. figure out better way to flush
  less verbose default logging
  make it easier for a thread to setup a quadrable env
  multiple sync connections in one process/config
  relay block-list events
  NIP-42 AUTH
  procmail-like API for event filtering

rate limits
  ! event writes per second per ip
  slow-reader detection and back-pressure
  max connections per ip (nginx?)
  max bandwidth up/down (nginx?)
  ? limit on total number of events from a DBScan, not just per filter

misc
  ? periodic reaping of disconnected sockets (maybe autoping is doing this already)