= Nacreon

So here it is, bits will be gotten to work, don't depend on anything in here,
as not a lot of it has been figured out yet.  It is beta software.  There's a
client (Ruby and command-line) that also serves as a reference implementation;
see github.com/pete/terpene for that.

Nacreon is intended as a very simple, lightweight, Platform-as-a-Service
application.  It is written in Ruby, and currently supports only Ruby 1.9
applications, but see doc/TODO for information on future support.

== Prerequisites:

* Ruby 1.9.x (1.9.1 *ought* to work, although I've been working under 1.9.2)
* Some gems:  watts, json, sequel, (and the gems that those
gems depend on).  `rake install_deps` or `bundle install` will take care of
this.

== Getting Ready:

You'll want to run `rake migrate` to update the database.  By default,
it uses an MySQL.  You can change this by making a 'nacreon.json' file,
and pointing the environment variable $NACREON_CONF at it (the one in the
root of the repo is intended as reference and a simple version for devs.)
[TODO:  Document the 'nacreon.json' config file when the other config stuff
get added to it.]

=== OS requirements

The packages required as prerequisites for various distros are listed below.
[TODO: Mostly incomplete.  If you run into a package that Nacreon requires that
isn't mentioned in the command for your OS, please update the command.]  See
doc/deployment.rdoc for information on how to deploy to a server.  Those
instructions are written for Debian/Ubuntu; see below for what is required for
your own operating system.  "MySQL" is assumed; for Postgres or other databases, you'll need a few 

As closely tied to the operating system as Nacreon is, it is difficult to
support operating systems that don't conform on some level to a regular Unix
environment.  Linux, Dragonfly BSD, and OSX work; your mileage may vary
elsewhere.  Will almost certainly not work under Plan 9 or Windows, but one of
the developers loves obscure platforms, so we can't say "never".

==== Debian / Ubuntu:

* Instructions that should work for both are available in doc/deployment.rdoc.

==== CRUX:

* `prt-get depinst ruby mysql-ruby`

==== MacOS X:

* Install (via ports or homebrew, likely) Ruby, MySQL, and the libraries for
accessing MySQL from Ruby.
* rake newb

==== Arch:

* `pacman -S ruby mysql-ruby`

== Running a Dev Server:

There's a rake task for unicorn, but anything that understands a
config.ru ought to work.
`rake unicorn` will start it up.  If you wish to run it as a daemon,
`rake daemon` will work.  You can reload without restarting unicorn by
doing `killall -HUP unicorn`.  See
http://unicorn.bogomips.org/SIGNALS.html for more information.

You'll also need to start up an AppManager daemon if you plan to run any apps.
That can be done with `rake appmanagerd` to run as a daemon with logging to
syslog, or `rake appmanager` to run in the foreground with logging to stdout.

== Poking at the App Interactively:

`rake irb` will start up IRB with everything loaded.  You can poke at
the models in Nacreon::Model, instantiate the application and check
routing, and run requests through it.

== Further Documentation
See doc/overview.rdoc for an overview of how the codebase works, if you want to
hack on Nacreon.

See doc/TODO for where we're going with this.