Software and configuration for the Ardent Mobile Cloud Platform's Raspberry Pi
Python C Shell
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
effects
layout
media Ignore converted wav files Aug 16, 2013
osc-server
system
.gitignore Untested but feature-complete cloud effect core Aug 7, 2013
README.md Minor bugs, some more debugging logs, step by step from raw Raspberry Pi Aug 13, 2013
avahi_announce.py
requirements.txt
server.py add saveload to server Aug 24, 2013
setup.py
test_amcp.py Better structure for working with systems. Aug 7, 2013

README.md

amcp-rpi

Software and configuration for the Ardent Mobile Cloud Platform's Raspberry Pi

System Setup

  • Use the latest version of Raspbian
  • Use rpi-update to install the latest firmware
  • Install extras with system/install.sh

Building the Server

The server is written in Python, but it includes C-language extension modules. For convenience, you can build it in-place:

$ ./setup.py build --build-platlib=.
running build
running build_py
running build_ext

Running it:

$ ./server.py
2013-08-07 14:40:56,536 - amcpserver - DEBUG - action="init_server", port="8000"
2013-08-07 14:40:56,540 - amcpserver - DEBUG - action="init_soundout", system="Darwin"
press enter to quit...

Shairport

The shairport project is an open source AirPlay-compatible server. It runs really nicely on the Raspberry Pi. The original version was a mishmash of Perl and command line tools, with a C-language daemon to handle audio decompression.

There's a new 1.0 release of shairport under development which is much friendlier and more efficient. It's been rewritten fully in C, and it has a much nicer architecture. This code still hasn't been merged back to the main branch, but I've been testing it and it seems nice and solid. The repository is on Github:

https://github.com/abrasive/shairport/tree/1.0-dev

Important: By default, this will compile without any optimization flags. Compile-time optimization is really important for the audio decompression code here! In my testing, CPU usage went from 40% down to about 15% by compiling with optimization:

./configure
CFLAGS="-O3 -ffast-math" make

Raspberry Pi Steps

install SD card
boot up and change the default password and expand the filesystem, reboot (raspi-config is the tool)
sudo rpi-update
sudo apt-get update

sudo apt-get install libssl-dev

sudo apt-get install mplayer
sudo apt-get install screen vim
git clone https://github.com/abrasive/shairport.git -b 1.0-dev (clone the shairport branch that's better)
cd ~/shairport
./configure CFLAGS="-O3 -ffast-math"
make
make install
cd ~
git clone https://github.com/ArdentHeavyIndustries/amcp-rpi.git
python setup.py build --build-platlib=.
cd ~/amcp-rpi/system
./install.sh