I've been a fan of Eclipse for a long time now, even before FDT and Flash Builder were built around it. The use of workspaces, once you get your head around them, is powerful for focusing your head on the work at hand, keeping all associated projects within the one space.

However there are times when I wish I was able to have multiple workspaces open. And too the rescue comes the command line.

[Path/To/Eclipse]/eclipse -data [Your/Workspace/Dir] &

I've wrapped this up into a little bash script to start eclipse from either the current directory or using the specified directory as the workspace:

#!/bin/sh
if [ -n "$1" ]; then
	echo "Starting eclipse: WORKSPACE: " $1
	/Applications/eclipse/eclipse  -data $1 &
else
	echo "Starting eclipse: WORKSPACE: " `pwd`
	/Applications/eclipse/eclipse  -data `pwd` &
fi

On OSX it's better to use the open command, like:

#!/bin/sh
if [ -n "$1" ]; then
	echo "Starting eclipse: WORKSPACE: " $1
	open -n /Applications/eclipse/Eclipse.app --args  -data $1
else
	echo "Starting eclipse: WORKSPACE: " `pwd`
	open -n /Applications/eclipse/Eclipse.app --args  -data `pwd`
fi

This will work equally as well with the Flash Builder application, just replace the Eclipse.app with the path to Flash Builder.

Happy workspacing!

We've been doing a lot with Native Extensions for AIR lately, both in Android and iOS, and I just wanted to put together a quick "how to" on using other people's Actionscript Native Extensions (ANE) in Flash Builder 4.6.

I'm going to assume you already understand the process of developing and deploying mobile applications using AIR.

The process is quite simple, you'll need the supplied ANE for the extension. In your mobile project you want to add it into, open up the project properties (Right click / Properties) and find the "Actionscript Build Path" (or "Flex Build Path" if it's a Flex project).

You'll see a tab up the top labeled "Native Extensions" in which you can "Add ANE...". Click on this and select the ANE extension file. Once it's added in, you can expand out the details of the extension to see the target platforms supported by the extension. Below I'm showing the details on our Android Camera Extension.

The extension is now included in your project, however before you leave here I suggest you go to the "Build Packaging" for the platforms you are deploying and check the extension is going to be packaged (and available) with them.

You should now be able to return to your code and use the functionality and classes supplied in the extension, hopefully you've been supplied with documentation or example usage of the extension!

Well if you use OSX or any program that produces metadata files on your file-system, it can become a nightmare trying to sort through the files you need to check in to svn... However! Once you discover there is a little hidden global ignore feature in svn, all the horror goes away.

On OSX the file of concern is located :

~/.subversion/config

Linux (sometimes) :

/etc/subversion/config

Windows... No idea sorry, depends on your client, though apparently TortoiseSVN has it as an option...

To add a file or folder to the list just open up that file and find the "global-ignores" option. It's just a space separated list of potential file or folder names that you want to ignore.

Mine now looks like this:

global-ignores = *.o *.lo *.la *.al .libs *.so *.so.[0-9]* *.a *.pyc *.pyo _notes .DS_Store .#* .AppleDouble *:2e_* ._*

Have fun with a cleaner svn...

So we all know and love our Android notification bar. So I just wanted to post a quick example application to show how easy it is to display a local notification.

The process runs something like this:

• retrieve a reference to the NotificationManager
• create a Notification
• notify the manager of the notification you've created

The Notification class, at it's simplest consists of an image icon and 3 pieces of text, one for the ticker, one as the title and the last as the content text.

package com.distriqt.mobile.experiments;
 
import android.app.Activity;
import android.app.Notification;
import android.app.NotificationManager;
import android.app.PendingIntent;
import android.content.Context;
import android.content.Intent;
import android.os.Bundle;
 
public class NotificationExperimentActivity extends Activity
{
    private static final int HELLO_ID = 1;
 
    @Override
    public void onCreate(Bundle savedInstanceState)
    {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.main);
 
        //
	//  Look up the notification manager server
        NotificationManager nm =
          (NotificationManager) getSystemService(NOTIFICATION_SERVICE);
 
        //
        //  Create your notification
        int icon = R.drawable.icon;
        CharSequence tickerText = "Hello";
        long when = System.currentTimeMillis();
 
        Notification notification =
            new Notification( icon, tickerText, when);
 
        Context context = getApplicationContext();
        CharSequence contentTitle = "My notification";
        CharSequence contentText = "Hello World!";
        Intent notificationIntent =
            new Intent(this, NotificationExperimentIntent.class);
        PendingIntent contentIntent =
            PendingIntent.getActivity(this, 0, notificationIntent, 0);
 
        notification.setLatestEventInfo(
            context,
            contentTitle,
            contentText,
            contentIntent);
 
        //
        //  Send the notification
        nm.notify( HELLO_ID, notification );
    }
}
 

The Android developer documentation goes through all of this in detail here.

So I went to update all my CS5.5 last night and ended up spending hours trying to get the application manager to actually allow me to update. I kept getting an error saying "You are already running the Adobe Application Manager. Please close all instances of Adobe Application Manager in order to continue with this installation."

Now I had been getting this error for a while, on two (OSX) machines so I decided that enough was enough and needed to get this running. I knew I only had one instance running, I had checked top multiple times to confirm this. So after much research I figured out that it was simply a permissions problem, so who better to alleviate a permissions problem than my good friend sudo.

sudo /Library/Application\ Support/Adobe/OOBE/PDApp/core/Adobe\ Application\ Manager.app/Contents/MacOS/PDApp


And that was it! The Application Manager ran without that annoying message.

So as many people know, I've been promising an update to the KML Loader for a while now, and even to the point where I've had some local revisions and haven't had the time to check them in (I know, I know, bad practice!)

The major update has been to enable the ability to load linked assets. Initially this just extends to ground overlays and custom icons, but now that I have the structure setup to do this I'll be able to add in linked files and styles easily.

Other updates include a series of bug fixes, features with multiple geometries now load in correctly and I've updated the utility library to correctly parse the newer OGC KML namespace.

Some things to note:

• I've added a trunk directory to keep the main head revision separate from the tag tree
• I've removed the included version of the google maps utility library. I've committed in some changes to this library so I didn't want to have to maintain it in two places, so you'll now have to go to the google maps utility library on google code here (I'll package it all up at the end of this post as well).
• I've tagged the previous version of the library if you wish to keep using it, you'll find it at: http://svn.distriqt.com/public/trunk/gmaps/tags/R20101004_gmaps_utils_v1.0
• I've used the BulkLoader by Arthur Debert and have included it in the repository. This is used to load all the linked assets.

The functionality should be exactly the same as the previous version, so you can still refer to the post on interactions, about the SimpleKMLLoader and the initial post about the KmlLoader.

The major change is that you now need to grab the gmaps utility library separately and put both libraries in your source tree.

svn co http://svn.distriqt.com/public/trunk/gmaps/trunk/ distriqt-gmaps
svn co http://gmaps-utility-library-flash.googlecode.com/svn/trunk/ gmaps-utility-library-flash-read-only

You'll end up with a src directory in both of these checkouts that you need to add to your source path for your projects.

Here are the current packages (v1.2) distriqt-gmaps-kmlloader-v1.2 and google-utility-library-flash.

Let me know if you find it useful!

This is just a quick post on how to retrieve your iPhone/iPad/iPod UUID (Unique Device Identifier).

We use this UUID to allow your phone to use development versions of our applications.

There are several ways to do this but the simplest is using Apple's flagship tool, iTunes.

• Open up iTunes and plug in your device
• Select your device in the "DEVICES" list
• Select the Summary tab and you should see the Serial Number as in the screen shot below:

• Click on the Serial Number label and it should change to reveal the UDID:

• You can then either just hit Ctrl/Command-C, or go to Edit / Copy to copy the UDID to the clipboard

You should then be able to paste it into an email etc so we can add your device into our allowed development device list.

Just assembling a quick list of some of the shortcuts I use everyday in Eclipse. Let me know if you have any other favourites!

• Code hinting: Ctrl + Space
• Change view: Cmd + 3
• Open any file quickly without browsing for it in the Package Explorer: Cmd + Shift + R
• Open a type (e.g.: a class, an interface) without clicking through interminable list of packages: Cmd + Shift + T
• Go directly to a member (method, variable) of a huge class file, especially when a lot of methods are named similarly: Cmd + O
• Go to line number N in the source file: Cmd + L, enter line number
• Go to other open editors: Cmd + E
• Go to a type declaration: F3 or Cmd + click on declaration

Note: Cmd == Ctrl on Windows/Linux

For an upcoming project we needed to stream content from the user's webcam to be recorded on the server. In theory this sounded like a basic operation, having used FMS to this end before, but actually installing and setting up the Red5 server proved to be a little tricky. So I just wanted to outline the process I went through to get a Red5 Server running on ubuntu and how to setup your first recording stream.

Red5 is a powerful Java based video streaming and multi-user solution for the Flash Player. Red5 includes support for the latest multi-user API’s including NetConnection, NetStream and SharedObject’s while providing a powerful RTMP / Servlet implementation. It is basically an open source Flash Media Server.

We're going to go through two major steps here, firstly installing Red5 on Ubuntu and then building a simple application to record the webcam.

Installing Red5

For this installation we're going to use a stock Ubuntu Server installation (8.04 Hardy Heron). We installed LAMP as well for the project but it's not required for Red5. Firstly we need to install a few dependencies including the Java runtime environment and the Java development kit (and yes these are terminal commands so get it up and running):

sudo apt-get install java-package sun-java6-jdk sun-java6-jre
sudo apt-get install subversion ant

Next we need to grab the code base for Red5. We tried the trunk version first but had some issues with it (as you'd expect from a HEAD revision) so ended up using v0.9.1. You can try other versions of the server by browsing the repository and replacing the reference in the svn checkout below:

svn co http://red5.googlecode.com/svn/java/server/tags/0_9_1/ ~/red5

Now we build using Ant:

export JAVA_HOME=/usr/lib/jvm/java-6-sun
export ANT_HOME=/usr/share/ant/
cd ~/red5
/usr/share/ant/bin/ant

Make sure there were no errors during the build. If there were consult the red5 wiki for help. Install:

sudo mkdir /usr/share/red5
sudo cp -R ~/red5/dist/* /usr/share/red5/.
sudo chmod 755 /usr/share/red5/red5.sh

Start the Red5 server:

export RED5_HOME=/usr/share/red5
/usr/share/red5/red5.sh

Next we need to create the startup script. This script will allow you to get the server starting on reboot etc. Put the following text into the file /etc/init.d/red5 and make the file executable sudo chmod 755 /etc/init.d/red5.

#!/bin/sh
# init script for Red5
# /etc/init.d/red5
 
RED5_USER=root
RED5_HOME=/usr/share/red5
RED5_PROG=red5
 
test -x $RED5_HOME/$RED5_PROG.sh || exit 5
 
case "$1" in
     start)
         echo -n "Starting Red5"
         echo -n " "
         cd $RED5_HOME
         su -s /bin/bash -c "$RED5_HOME/$RED5_PROG.sh &" $RED5_USER
         sleep 2
         ;;
     stop)
         echo -n "Shutting down Red5"
         echo -n " "
         su -s /bin/bash -c "killall -q -u $RED5_USER java" $RED5_USER
         sleep 2
         ;;
     restart)
         $0 stop
         $0 start
         ;;
esac

Now you can start/stop/restart from the command line by using:
sudo /etc/init.d/red5 [start/stop/restart]

Lastly once you've got the server running, you can browse to the test page on your server http://localhost:5080 and you should see:

If you can't see this it and you've confirmed the Red5 server is running it could be a firewall issue. You may need to alter the IPTables on your machine but that's not going to be covered here.

Lastly we need to install the actual streaming application on our Red5 Server. Navigate to http://localhost:5080/installer/, select the oflaDemo and click Install.
To test it installed correctly, navigate to http://localhost:5080/demos/ofla_demo.html (correct the server name in the top right) and hit connect. You should see a list of available videos and a green light in the top right.

Congratulations you're all up and running with a Red5 Server!

The following class shows a very simple example of starting a streaming feed to the server and then streaming the same data back from the server into a Video object on the stage.

The fundamentals are very simple:

• Create a NetConnection object and connect to the server
• Create a NetStream when the connection succeeds and publish the camera feed to the stream as a "record" type

package
{
	import flash.display.Sprite;
 
	import flash.net.NetConnection;
	import flash.net.NetStream;
	import flash.events.NetStatusEvent;
 
	import flash.media.Camera;
	import flash.media.Video;
 
	public class Red5Experiment extends Sprite
	{
		//
		//	Constants
		public static const RED5_URL	: String = "rtmp://SERVERNAME/oflaDemo/test/";
 
		//
		//	Variables
		private var _netConnection	: NetConnection;
		private var _netStream		: NetStream;
		private var _recvStream		: NetStream;
 
		private var _url		: String = RED5_URL;
		private var _filename		: String = "test-red5";
 
		private var _video		: Video;
 
		public function Red5Experiment()
		{
			_netConnection = new NetConnection();
			_netConnection.connect( _url );
			_netConnection.addEventListener( NetStatusEvent.NET_STATUS, netStatusHandler, false, 0, true );
 
			_video = new Video();
			addChild( _video );
		}
 
		////////////////////////////////////////////////////////
		//	FUNCTIONALITY
		//
		private function start():void
		{
			//
			//	Setup the camera
			var cam:Camera = Camera.getCamera();
			cam.setLoopback(true);
			cam.setMode(176,144,15);
			cam.setKeyFrameInterval(10);
			cam.setQuality(0,70);
 
			//
			//	Start publishing the video stream
			_netStream = new NetStream(_netConnection);
			_netStream.addEventListener( NetStatusEvent.NET_STATUS, netStatusHandler, false, 0, true );
			_netStream.attachCamera( cam );
			_netStream.publish(_filename, "record");
 
			//
			//	Grab live stream back from server
			_recvStream = new NetStream(_netConnection);
			_recvStream.play(_filename);
			_video.attachNetStream(_recvStream);
 
			//
			//	Alternatively, Use the local camera feed
//			_video.attachCamera(cam);
		}
 
		////////////////////////////////////////////////////////
		//	EVENT HANDLERS
		//
		private function netStatusHandler( event:NetStatusEvent ):void
		{
			trace( "Red5Experiment::netStatusHandler():"+event.info.code );
			switch ( event.info.code )
			{
				case "NetConnection.Connect.Success":
					start();
					break;
			}
		}
 
	}
}

The files are recorded to a directory within the webapps folder on your server. If you've followed the paths above you'll find the file /usr/share/red5/webapps/oflaDemo/streams/test/test-red5.flv on your server.

You now should have a complete streaming / recording media setup. All running off an open source server!

EDIT: I've created a zip of the sample application files: red5_sampleapplication

I've had a few requests asking about how to interact with the KML once you've loaded it onto the map. It's really quite an easy step once you've got the KML on there but I'll show you how in this simple example.

Firstly if you haven't read the Simple KML Loader post then head over there first. It explains how to get that KML file loaded and displayed on your map.

The objective of this simple example is to show an info window popup when we rollover one of the objects in the kml, showing the name of the kml object in the window.

The code for this example is in our svn:

svn co http://svn.distriqt.com/public/trunk/gmaps/examples/kmlloader/

Make sure you have our SimpleKmlLoader library in your project.

To create the loader, load the kml and have it added to your map:

var _loader:SimpleKmlLoader = new SimpleKmlLoader( _map, true );
_loader.addEventListener( Event.COMPLETE, loader_completeHandler );
_loader.load( _file );
 

Now on the complete handler you can add the interactions:

function loader_completeHandler( _event:Event ):void
{
	// At this point generally this objects array
	//   will simply be the base object of the KML
	addInteraction( _loader.objects );
}
 
/**
 * Add interaction to every object in the array
 */
function addInteraction( _objects:Array ):void
{
	for each (var _object:KmlDisplayObject in _objects)
	{
		_object.interactive = true;
		_object.addEventListener(MapMouseEvent.ROLL_OVER, object_rollOverHandler, false, 0, true );
		_object.addEventListener(MapMouseEvent.ROLL_OUT, object_rollOutHandler, false, 0, true );
 
		// Process any children
		addInteraction( _object.children );
	}
}
 
function object_rollOverHandler( _event:MapMouseEvent ):void
{
	// show an info panel...
	var _object:KmlDisplayObject = KmlDisplayObject(_event.currentTarget);
	_map.openInfoWindow( _object.latLng, new InfoWindowOptions({content:_object.name}) );
}
 
function object_rollOutHandler( _event:MapMouseEvent ):void
{
	_map.closeInfoWindow();
}

Now when you run with this code you should see the name of the kml object appear in the info window when you rollover the object. This is of course a very simple example, and much more can be done with this, but it is a good point to introduce the concept of interactions.