[From @joshdata: This is part of @toilal's work in #377 and https://github.com/Toilal/mailinabox/commit/1eb77c332bca0a74a37f792460903a1fb660befc. The changes are:

* Separates out the runit configuration from starting Mail-in-a-Box setup so that Mail-in-a-Box setup does not block the starting of runit services and we can assume that runit is running during setup (i.e. we can restart services).
* Adds a SKIP_INSTALL flag so that the container can be restarted without re-running the whole Mail-in-a-Box setup.
* Made containers/docker/run more flexible.
* I'm also adding some "|| exit 0"s to the run script to stop if there are any docker errors.
* I'm also adding the prereqs installs from questions.sh into Dockerfile so we don't have to reinstall each time.

]

Docker support was initially worked on in 2bbb7a5e, but it never really worked.

This extends f7d7434012800c3572049af82a501743d4aed583 which was an old branch for docker work.

[JT added installing netcat-openbsd in system.sh]

add a new autoconfiguration option PRIMARY_HOSTNAME=auto to simply grab the hostname from reverse DNS

drawn from https://github.com/Toilal/mailinabox/commit/5b23a06a7410e4530a56fd6200a6c46c3c6ea9b6.

move more of start.sh into questions.sh to keep start.sh cleaner and encapsulate all of the variable setting in a single script

Based on https://github.com/Toilal/mailinabox/commit/5b23a06a7410e4530a56fd6200a6c46c3c6ea9b6.

fix wrong redirect for automatic www subdomain redirects

$ curl -I https://www.site.co.il/static/images/1.png?a=b | grep Location
Location: https://site.co.il?a=b
but should be something like 
Location: https://site.co.il/static/images/1.png?a=b

This was broken by the ability to have multiple TXT records in 9f1d633a.

fix loop redirecting

server is redirecting the request for this address in a way that will never complete

* Set ssl_stapling_verify to off per https://sslmate.com/blog/post/ocsp_stapling_in_apache_and_nginx ('on' has no security benefits).
* Set resolver to 127.0.0.1, instead of Google Public DNS, because we might as well use our local nameserver anyway.
* Remove the commented line which per the link above would never be necessary anyway.

OCSP seems to work just fine after these changes.

the control panel auth hmac message should also include the user's password so that resetting a password in the database forces that user to log in to the control panel again; also use a sha256 hmac

when the undocumented custom web settings has a redirect or proxy at the root of a domain, use a minimal nginx config template (same as the new default www redirects)

* Split the nginx templates again so we have just the part needed to make a domain do a redirect separate from the rest.
* Add server blocks to the nginx config for these domains.
* List these domains in the SSL certificate install admin panel.
* Generate default 'www' records just for domains we provide default redirects for.

Fixes #321.

This doesn't change anything. Just preparation for the next commit.

if a user sets a custom A record on PRIMARY_HOSTNAME, which is ignored anyway, don't let that cause PRIMARY_HOSTNAME from being dropped from nginx.conf

Could be related to https://discourse.mailinabox.email/t/nginx-lost-admin-record-after-install-ssl-cert-problem/528.

bring the postgrey patches into this repository rather than maintaining them in a separate postgrey fork repository