Andrew Clayton

This has been handled in commit 5665838b680c

Heh, I think they can be squashed into a single commit...

You mean squash the two brackets patches into one?

Ah, OK I see now, the 'add more brackets' patch is touching a lot of 'other' code, I did wonder why there was two 'bracket' patches...

[Here](https://github.com/nginx/unit/pull/734) is the start of a more complete cgroup support. It supports per-application (within unit) cgroups.

> ``` > $ git diff > diff --git a/src/nxt_main_process.c b/src/nxt_main_process.c > index 0527d08f..4e1df140 100644 > --- a/src/nxt_main_process.c > +++ b/src/nxt_main_process.c > @@ -1147,10 +1147,21 @@ nxt_main_listening_socket(nxt_sockaddr_t *sa, nxt_listening_socket_t *ls)...

> Unix domain sockets are normally backed by files in the filesystem. This has historically been problematic when closing and opening again such sockets, since SO_REUSEADDR is ignored for Unix...

It's a bug, in xterm! I've seen this before, and it just occurred to me again now. Lets see xterm at 80 chars wide. ``` $ ss -l | grep...

> #if (NXT_LINUX) > > /* > * Linux unix(7): > * > * abstract: an abstract socket address is distinguished by the fact > * that sun_path[0] is a...

I have no strong preference, I guess showing '@' in the config is more 'user friendly'. If we do that, then which ever way we go (I think I'm leaning...