| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This way the issue fixed by the previous commit won't happen again.
|
|
|
|
|
|
|
|
|
|
|
|
It can cleanup all messages from a pid, or just the ones from a
specific stream. The latter is useful in combination with cancel/2.
|
|
|
|
Thanks David Cartt for reporting it.
|
|
|
|
For some reason this only happened with SSL. I suppose the behavior
is a little different from TCP.
|
|
We branch out and don't try to use NPN when the function
ssl:negotiated_next_protocol/1 isn't exported.
|
|
|
|
|
|
'gone' is more explicit than 'too_many_retries'.
|
|
The basic idea is that we do an async call and then optionally wait.
If we already have a monitor for this connection, then we can reuse
it across all await* calls, otherwise one is created automatically.
|
|
|
|
|
|
Support for HTTP/1.1 and compatibility with HTTP/1.0.
|
|
|
|
Was due to incorrect documentation before R16B03. Thanks Ingela
for finding it!
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
SPDY unfortunately is still completely client-server
and we can't do that according to the specs.
|
