Ron Frederick

This is a bit of a long shot, but do you see a difference if you set `term_type='ansi'`? This will cause the remote system to allocate a pseudo-tty, which changes...

Happy to try and help. The PTY thing was a long shot, but I figured it was worth a try just to see if the behavior changed. Thinking about it...

Yeah - that looks reasonable. Was that the issue in all of your tests, though? I thought one of your examples which failed was a plain "exit 1" not in...

Closing due to inactivity - feel free to open a new issue if there's anything further needed here.

I must admit I haven't really looked much at how asyncio deals with KeyboardInterrupt. I definitely have seen the behavior in test programs where sometimes you have hit Ctrl-C multiple...

I sort of expected this to be the case. When you hit Ctrl-C for the second time, do you get any kind of traceback? There isn't any handling of CancelledError...

Do you actually need the "done" variable in the latest workaround? It looks like that version has removed the `self.close()` call when exiting the try block, so I would expect...

Sorry for the slow response - I missed your last edit here. Even in the `abort()` case, there could be some activities related to cleaning up the channel that aren't...

Ok - this should now be available as commit 676534b in the "develop" branch.

The `_cleanup()` function is scheduled to run by `_discard_recv()`, which is called from `abort()` when the receive state is not yet 'closed'. Since `_close_send()` is called before that, the sending...