GitLab

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@whyrusleeping

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@whyrusleeping @jbenet

since the logger is created with package scope, don't need to specify
the package name in event messages

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@whyrusleeping

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@jbenet
@whyrusleeping

Let me know if you want to direct the eventlog output to _both_ the file
and stderr. Right now it goes to file. Perhaps this is just a minor bip
in the larger discussion around log levels.

https://github.com/jbenet/go-ipfs/issues/292

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@whyrusleeping This appears to be a timing issue. The asynchronous
nature of the new structure provides has the bitswap waiting on the
context a bit more. This isn't a problem at all, but in this test, it
makes the functions return in an inconveniently timely manner.

TODO don't let the test depend on time.

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

vanilla:
    21.57 real        45.14 user         8.51 sys

short:
    14.40 real        31.13 user         5.56 sys

License: MIT
Signed-off-by: Brian Tiger Chow <brian@perfmode.com>

@jbenet

remove 'adapter' concept
instead, describe the component as the bitswap network

it's still an adapter, but it's just not necessary to describe it as
such

https://github.com/jbenet/go-ipfs/issues/270#issuecomment-61826022

to emphasize idempotence

comes at the cost of O(n) Blocks() method.

so there's one less name to think about

A nice invariant for bitswap sessions:

        Senders and receivers can trust that messages do not contain
        duplicate blocks or duplicate keys. Backing the message with a
        map enforces this invariant.

        This comes at the cost of O(n) getters.