goodboy
goodboy
**Issue by [moises-silva](https://github.com/moises-silva)** _Friday May 15, 2015 at 18:59 GMT_ _Originally opened as https://github.com/sangoma/switchy/issues/6_ ---- Post Dial Delay Average Call Duration
**Issue by [tgoodlet](https://github.com/tgoodlet)** _Tuesday Mar 10, 2015 at 21:06 GMT_ _Originally opened as https://github.com/sangoma/switchy/issues/4_ ---- I'm thinking moving to standard sphinx markup is most desirable? See here: https://github.com/rtfd/readthedocs.org/issues/995 Currently it's...
As per discussion with @vodik we agreed that it would be useful for a variety of reasons to publish [protocol.py](https://github.com/friends-of-freeswitch/switchio/blob/master/switchy/protocol.py) as a separate package most likely dubbed `aioesl`.
As per discussion with @vodik in friends-of-freeswitch/switchy#4, there seems to be a weird requirement when calling `BaseEventLoop.call_soon()` (called from `Future.add_done_callback()`) which has a thread id check only in debug mode....
It's defined [here](https://github.com/friends-of-freeswitch/switchio/blob/449301cd160a074cd538ef352f06c111b674e480/switchy/connection/aioesl.py#L226) but @vodik asks *why?* in friends-of-freeswitch/switchy#4.
As per friends-of-freeswitch/switchy#4 @vodik suggested that once we drop py27 support the [`loop` here](https://github.com/friends-of-freeswitch/switchio/blob/master/switchy/async.py#L27) should be made a [keyword-only arg](https://www.python.org/dev/peps/pep-3102/) to stick with common style in other `aio-` libs.
As per friends-of-freeswitch/switchy#4, @moises-silva made a good point that users should be able to override handlers in the [reactor core](https://github.com/friends-of-freeswitch/switchio/blob/master/switchy/reactor.py#L166) as it fits in better with the UNIX signal handler...
I'm pretty sure the kernel team has now limited write access of MSR registers as per: - https://www.phoronix.com/scan.php?page=news_item&px=Linux-Filter-Tightening-MSRs - https://github.com/torvalds/linux/search?q=MSR&type=commits I haven't been able to get this (very very lovely...
Resolves #104 and is something I wrote a few years ago after a discussion with @RonnyPfannschmidt. I haven't integrated the `_c_multicall()` into the benchmark tests yet (starting to think I...
Example/untested draft implementation for discussion in #50. Ping @fschulze too.