Richard Hodges

This is a common occurrence when user programs do not respect the preconditions. Remember that streams are not thread safe and you may only have one async write operation in...

If you can send me a minimal, complete example (i.e. a 1-file program that demonstrates the problem) then I'm happy to debug it. What we normally find is that when...

Hi there, I've just had a quick look through the documentation to be sure we had not forgotten anything. Please see for example: https://www.boost.org/doc/libs/1_76_0/libs/beast/doc/html/beast/ref/boost__beast__websocket__stream/async_write.html > The program must ensure that...

If it doesn't happen in an MCVE, or in any of our test programs, then it's likely to be a bug somewhere else. Common errors are: - more than one...

Are we in a position to close this issue now?

I'd be interested to see the minimal code that implements this. I do a similar thing (including a timeout to close the underlying socket if the async_close does not complete...

We can work with this issue. I'll post a fix as soon as I get a moment.

If the tests pass I'll merge it now

``` cd git checkout master -u origin/master git pull git submodule update --init --recursive (cd libs/beast && git checkout master && git pull) ./bootstrap ./b2 ... etc ```

``` $ ./b2 --prefix= variant=release [any other options that matter to you] install ```