- fixes https://github.com/lbryio/lbry/issues/2244
- reduce the max DHT datagram size to 1400 bytes
- truncate `contacts` field of find_value response datagrams to the k closest (8)
- truncate peers in find_node response datagrams to the 2k closest (16)
- remove `contacts` field from find_value responses beyond `page` 0 (the first/default)
- deterministically shuffle the peers for a blob in a find_value response
- add optional `page` argument to `find_value` and `p` field to find_value responses containing the number of pages of k peers for the blob
- test one blob being announced by 150 different peers to one peer
- speed up pylint and remove some disabled checks
instead of recursive bytes2unicode use a proper JSONEncoder to conver bytes->unicode for json.dumps()
removing excessive isinstance(data, bytes) checks
py3: / -> // and list() around .items() that gets modified in loop
moved lbrynet.undecorated to where its actually used and hopefully we can delete it eventually
removed build/upload_assets.py, travis can do all this now
Move parameter handling into its own function and
better use the `.get()` function for dictionaries.
Early return on the failed checks is more readable.
The lambda function in the callback was long and
hard to read so moved it out.
