Raimo Niskanen

For every receive size 0, yes. :-(

All `active` modes use receive size 0 so they exhibit the same speed optimization / memory waste. Whenever the libc call recv(2) returns an active data message is delivered. For...

> What I meant here is instead of allocating 1 large buffer (512MB), we allocate 16kb or 160k or something, and keep > ... Alright! The `{otp,buffer}` option allows setting...

Sry. This option does not do precisely what I thought. For `socket:recv/1,2,3,4` with `Length = 0` and `Timeout = integer()>0 | infinity`, that is a read-me-everything with a timeout >...

With `inet_backend = socket`, `gen_tcp:recv/2,3` translates into `socket:recv/2`, via a buffering process (that implements `packet` modes, if needed). If you start nicking data from under the feet of this `gen_tcp_socket`...

Is there up to date documentation somewhere? E.g `usrsctp_init(uint16_t)` in Manual.md does not match well what is in `usrsctp.h` which matches even less the its call in `inet_drv.c`. Anything on...

Have you tried to get any of your `usrsctp` chages accepted upstream at sctplab?

Pull requests are accepted, also for documentation, and are, for documentation, fairly easily approved. In the documentation for `gen_statem` I think there is an explicit mention somewhere about using `throw/1`...

@d1jang: I am afraid @garazdawi is on vacation until mid August. Maybe someone else in the team can find the time to have a look, but do not hold your...

@max-au: `gen_server`, (the obsoleted `gen_fsm`), `gen_event` and `gen_statem` all implement this "feature" of equating a thrown value with a regular return. I suspect, as @mikpe, that this stems from using...