by default it blocks queries to private zones, can be extended to match anything in the query content (i.e. water torture style attacks)

before root hints were hardcoded to the resolver,
now they are present in form of a cut in the resolution
context, and the modules can add/remove/replace them
on the fly

the Redis backend supports UNIX, TCP and custom ports
no real transactions nor iteration, but it’s not needed at the moment
transactions keep replies in a freelist, which is purged on transaction
commit / abort

also some cleanup and version bump to Q2

this prepares cache/txn structures to hold API as well, so we can get
rid of the global api

this fixes a problem when module is unloaded at runtime
and active

memcached is a distributed caching system, it is a good fit
for building resolvers with shared and replicated cache

this hooks into query resolution and collects counters on answers, queries and iterator

this allows for continuation of a function in the background,
a good use case for this is loading of chunks of data or I/O