* don't broadcast to yourself
* todo
* Also listen for incoming streams
* send to a specific peer
* add godoc comment
* handle EOF as a normal thing
* gazelle
* add test for subscribing via direct comm
* recover from a panic in p2p
* Add safely process function
* print proto message
* add status to the span, if available
* add status to the span, if available
* WIP - with TODOs
* interface w/ test
* basic test for broadcast
* Add computeStateRoot funciton
* remove custody challenge
* resolve TODO lint issues
* more TODOs
* revert new line in types.proto
* broadcaster comment
* one of several failure condition tests
* Add test cases
* handle compute state error test case
* fix config in validator helpers
* fix tests too
* fix conflict
* partial PR feedback
* remove p2p
* gazelle
* package comment
* Better godoc
* add unit test for test broadcast
* execute goimports
* add import peerstore
* refactor and check message equals expected
* remove not needed sleep
* add function to connect hosts to host
* refactor test for better reading
* refactor to only use one channel
* refactor subs to channel
* close channels instead of sending boolean value
* use goimports
* copy value to avoid race condition
* remove TODO comment
* fix issues with gometalinter
* run gazelle
* reduce sleep time
* add comments to goroutines
* /healthz part1
* lint, fix
* lints and todos
* add p2p status test
* Add test for service registry Statuses
* fix my fake errors for the linter
* Test healthz handler
* Run gazelle
* first commit, remote att types
* no more agg attestation proto
* regen mock
* only attestations
* proto
* att process
* fix att references
* more tests passing
* use att protos
* complete
* change visibility
* use gogoprotobu
* Preliminary support for relay nodes
* lint
* Add comment, remove TODO
* work on relay address factory
* dial relay node, if available
* forgot new files
* fix service registry breakage
* added logging
* Added a peer count with prometheus
* always start mDNS
* fix lint
* Added prometheus client and p2p metrics
* Avoid run the adapter if the metrics are disabled
* fix visibility issue
* Fix invalid p2p.Message sent to Adapters
The middlewares (adapters) must receive the complete message to avoid
problems and the main Handler must get the values from middlewares
Also, added tests and comments for metrics package
* Added logrus hook collector
This collector is used to collect counters of log messages.
The main purpose of these metric is to know how many warnings and errors
the system are getting.
* Add hook when register the prometheus service
* update bazel builds
* fix emit tests and remove unused imports
* gazelle --fix
* remove unused logger
* move prometheus package to shared directory
* better metric names and fix metric paths
* improve metric tests and start to use promauto
* added prometheus initial documentation
* fix tests
* fix type differences with go get and bazel
* Fix service test
* start sequence for node to connect to bootstrap nodes
* lint
* maybe warning is more appropiate?
* fix gazelle
* preston's feedback
* add go-ipfs-addr to workspace
* move flag to shared
* removed routed host
* Remove error checking for type assertions on p2p messages. The code should fail earlier in the p2p layer. Also fixed format on some TODOs
* format TODO again
* Add a bit more tests for p2p stuff
* goimports on feed_mock
* remove unused import
* add mutex on top of p2p feeds map
* lock read too
* add concurrent write test and enable race detector
* restrict race detection to concurrent feed map write test
* gazelle fix
* increase ram and I/O in bazel's local_resources
* reverted local_resources changes
* address review comments
* do not disable sandboxing on travis
* address review comments
