Integrating Lync 2010 and Asterisk

Lync 2010 is a great product and can stand on its own to deliver a good PABX system for the education or business sectors. Where Lync falls down is flexibility, tying in 3rd party scripts, connectivity with products like Skype and even voice based menus. Fortunately we can easily route calls between Asterisk and Lync to produce a much more flexible outcome.

I am a massive fan of Asterisk and have used it numerous times in the past for both my home PABX and several businesses I previously supported. Asterisk is not the easiest product for the end user to connect to, a strength that Lync defiantly has.

In my environment I am using an Asterisk box with advanced scripting based dial plans to handle the incoming PSTN to SIP traffic.



The Integration Process

Obviously you need a working Lync and Asterisk installation before you can begin the process. I'm not going into the Asterisk/Lync server setup process in this tutorial but AsteriskNOW is a great place to start if you want an easy to install Asterisk product. I recommend installing the command line variant and not using the limiting AsteriskGUI, it's harder to learn but also gives you many more possibilities.

1. First we need to configure the SIP trunk in Asterisk's sip.conf. Open /etc/asterisk/sip.conf in your favourite Linux editor and add the following text to the end of the file.

[lynctrunk]

type=friend

port=5068

host=192.168.1.10

dtmfmode=rfc2833

context=fromlync

qualify=yes

transport=tcp,udp

The host= variable needs to reflect your Lync front end, for example my front end IP address is 192.168.1.10. Then save and close the file.

2. Next we are going to add an entry to our /etc/asterisk/extensions.conf file to give us an extension we can call from Lync as a test. Open /etc/asterisk/extensions.conf and add the following text to the end of the file.

[fromlync]

;echo test

exten=>9999,1,Answer()

exten=>9999,2,Playback(en/demo-echotest)

exten=>9999,n,Echo

exten=>9999,n,Playback(en/demo-echodone)

exten=>9999,n(eof),Hangup()

Save and close the file. This 9999 extensions will allow us to do a simple echo test, the [fromlync] header is important as this is the context that our sip.conf entry refers to in step 1.

3. Issue the reload command from the Linux command line, this will force Asterisk to load the above changes.

asterisk -rx "reload"

4. Now we can move onto the Lync side, jump on your Lync server and fire up the the topology builder. "Download the latest toplogy from existing deployment" and then jump straight to the "PSTN Gateway" section.

5. As far as Lync is concerned, Asterisk is a PSTN gateway, so right click PSTN Gateways and click "New IP/PSTN Gateway..."

6. On the "Define New IP/PSTN Gateway" screen we need to specify the settings for our Asterisk box. You will need to enter the IP Address, set 5060 as the port and select TCP as the SIP transport protocol as per the image below.

7. To publish the topology changes we have just made, right click on the Lync Server 2010 menu option and select "Publish Toplogy...". After a few seconds these topology changes should be complete, but it can take some time for them to propagate to the Lync server.

8. Now our SIP trunk is in place, Lync needs a dial-plan and associated routes so it knows what numbers it should route to our new Asterisk trunk. We will start by adding a dial plan entry, open the Lync Control Panel and click on the "Voice Routing" option, then select "Dial Plan"

9. Click on the "Global" entry, then select "edit" and "show details".

10. Under the "Associated Normalization Rules" select "New" to add a new rule.

11. You can call the rule anything you want, for the other settings:

"Starting digits" field put "9"

"Length:" field select "Exactly" and "4"

"Digits to remove:" field select "0"

"Digits to add:" field remove the "+" to leave the field blank

You can copy my settings as per the example photo below, then click "OK"

12. You now need to commit the changes. Click "Commit" and then select "Commit all", on the following page click "Commit" again.

13. Now we need to define a "Voice Policy" to forward any 9XXX dialled numbers in Lync to Asterisk. To do this select "Voice Routing" from the left hand side menu and then select "Voice Policy".

14. Again we need to select the "Global" rule, click "edit" and "show details".

15. Under the "Associated PSTN Usages" select "New" to add a new rule.

16. Give your rule a name and then click "New" to add a new route.

17. On the "New Route" screen, give your route a name and then and the in the "Build a Pattern to Match" section add 9 into the box and select "add". This is an extremely board match, but you can tighten it up later it you want.

18. Then move down to the "Associated gateways" box and add the PSTN gateway we earlier added to our topology builder. Click "OK" in the top left hand corner to accept the route. Then "OK" again on the "New PSTN Usage Record" screen to accept. Finally click "OK" on the Edit Voice Policy screen to accept your new route entry.

19. Again we must commit the changes by selecting "Commit" and clicking "Commit all", then click "Commit" on the following screen.

20. The last piece of the configuration puzzle is to go to the "Voice Routing" options tab from the left hand menu, select "Route" from the top menu. and commit these changes. To commit the changes select "Commit" and click "Commit all", then click "Commit" on the following screen.



Now for the fun part

The dreaded configuration is out of the way, now lets test it out. If things don't work immediately please wait 15 minutes and try again, Lync can take some time to propagate changes.

Before we attempt the call it is a good idea to open an asterisk console by SSH'ing into the Asterisk box and typing asterisk -cr on the command line. This will allow you to want the traffic coming in from Lync to Asterisk and how the dial plans are accessed in real time.

 

1. Open your Lync client and dial 9999

2. If you opened the Asterisk console you will see the call passed from Lync to Asterisk and the dial plan take effect. You should hear the echo test begin though your Lync connected head set or phone.

I have blanked out some of the sensitive information but here is an example of what you should see on the Asterisk console (click to enlarge).

Congratulations! You now have a Asterisk/Lync monster that you can manipulate however you want. Of course if you want calls to go in the other direction you can use a dial plan entry such as Dial(SIP/1000@lynctrunk,35,tr) in your Asterisk extensions.conf to call SIP extension 1000 on your Lync box.



Some troubleshooting tips

• First troubleshooting step is to wait 15-30 minutes. It may take some time for the topology changes to propogate.
• Sign out and then back into Lync.
• Do a hard restart of Asterisk with the /etc/init.d/asterisk restart command. This will force a reload of the configuration.
• Re-read step 7 and ensure you did publish the topology changes, then wait 15 minutes.
• If you have any firewall enabled on your Lync or Asterisk box take it down until you get things working. If you require a firewall after all is working, then you need to enable the following port exclusions.

Port 5060 TCP/UDP should be open on your Asterisk box.
Ports 10000-20000 should be open for RTP traffic on the Asterisk box.
Port 5068 TCP/UDP should be open on your Lync box.  

Compiling app_swift for AsteriskNOW or Fedora

For those building Asterisk interactive voice menus the app_swift plugin is a must. Giving the administrator the ability to directly interface with Cepstral voices from within the dial plans is a massive time saver. 


The problem 

Unfortunately many users are having issues with compiling the app_swift module on AsteriskNOW as the appropriate build tools are not included in the operating system.

If you don't have the correct tools installed you will probably see an output like this.

/usr/include/string.h:66: error: declaration for parameter âmemchrâ but no such parameter
/usr/include/string.h:62: error: declaration for parameter âmemcmpâ but no such parameter
/usr/include/string.h:59: error: declaration for parameter âmemsetâ but no such parameter
/usr/include/string.h:51: error: declaration for parameter âmemccpyâ but no such parameter
/usr/include/string.h:43: error: declaration for parameter âmemmoveâ but no such parameter
/usr/include/string.h:38: error: declaration for parameter âmemcpyâ but no such parameter
/usr/lib/gcc/i486-linux-gnu/4.3.2/include/stddef.h:214: error: declaration for parameter âsize_tâ
app_swift.c:569: error: expected â{â at end of input
make: *** [app_swift.o] Error 1

The result is no app_swift.o module and a frustrated administrator. 


Installing app_swift

The following instructions will work with Asterisk 1.6.x on AsteriskNOW or Fedora distributions.

1. Before you install app_swift you need to download and build the Cepstral swift package. Start by downloading the Allison voice package, alternatively you can choose a voice you prefer from Cepstral.com.
wget http://downloads.cepstral.com/cepstral/i386-linux/Cepstral_Allison-8kHz_i386-linux_5.1.0.tar.gz

2. Extract the package .
tar -zxvf Cepstral*

3. Enter the directory.
cd Cepstral_Allison-8kHz_i386-linux_5.1.0

4. Run the installer.
./install.sh

5. Load the library into ld.
echo /opt/swift/lib > /etc/ld.so.conf.d/cepstral.conf

6. Now run ldconfig to finish the ld configuration process.
ldconfig

7. Its time to get app_swift compiled, first start by downloading the source.
wget http://pbxinaflash.net/source/app_swift/app_swift-1.6.2.tar.gz

8. Before we can compile swift, the appropriate build tools and libraries must be installed into the OS. Issue the following commands sequentially to install the tools with yum.
yum install gcc
yum install build-essential
yum install asterisk-devel
yum install kernel-devel
yum groupinstall "Development Tools"

9. Extract app_swift in preparation of compiling.
tar -zxvf app_swift-1.6*

10. Go into the app_swift directory.
cd app_swift-1.6.2

11. Now we can make the package.
make ; make install

12. Copy the configuration file to the asterisk directory.
cp swift.conf.sample /etc/asterisk/swift.conf

13. We also need to make a symbolic link to the swift binary for easier access.
ln -s /opt/swift/bin/swift /usr/bin/swift

14. Finally you need to edit swift.conf to reflect the voice you have installed. I always choose Allison, but the choice is yours.
sed -i 's|David-8kHz|Allison-8kHz|' /etc/asterisk/swift.conf

15. We finish with restarting Asterisk
/etc/init.d/asterisk restart  


Confirming all is well

Once the installation is complete the module should be loaded into Asterisk. You can check this by starting asterisk console and issuing the show application command.

1. asterisk -cr
2. core show application swift

If everything is running smoothly you should see an output similar to below.

asterisk*CLI> core show application swift
  -= Info about application 'Swift' =-

You can now start swift()'ing your way to a flexible, intuitive and interactive dial plan that suites your requirements.

Home automation possibilities using X10 and Asterisk

A couple of years ago I was seriously sick, I knew I was going to be in hospital and then couch bound for several months and decided a home automation project was going to help me get through the tougher times.

After looking at all the automation options, I decided on using the X10 standard. It's affordable, extremely flexible and most importantly I don't need to make any physical modifications to my rental house to implement it. X10 using existing power lines to communicate by using adapters that sit between the power socket and the device to remotely turn devices on and off. Also available among a huge list of optional adapters are light socket adapters that similarly interface between the light socket and the light, providing remote control lighting.

I purchased and quickly implemented a number of X10 power and lighting adapters and a remote to control them. They worked great, but I wanted more. I wanted to control more devices and I wanted to do it remotely, these are all things that X10 does not offer.



X10 is cool, but I need more

Enter Asterisk, the open source, Unix based, PABX solution. I had worked with Asterisk before for several businesses I supported, but only with simple dial plans and voice mail. While Asterisk is designed as a PABX, it has all the required framework to be a great automation conduit. The key features I want to leverage are  the support for complex menus, ability to trigger Unix scripts and most importantly the ability accept remote commands via a VOIP dial-in number.

My plan was to setup Asterisk to handle all the triggering of the automation, then if I was at home I could use my portable VOIP phone or off-site I could dial in with my mobile.



The first steps to home automation

The first step is to setup your Linux based PC with the X10 computer interface module, install and setup Asterisk and register your VOIP phone number with the Asterisk box. I have been using VOIP at home for a number of years, so it was easy for me to connect my handset to Asterisk and then let Asterisk connect to my VOIP provider.

What I love most about Linux is the way things so easily connect together with scripting tools such as grep and awk. I set about making a number of scripts that when requested from the Asterisk dial plan, would go about performing my tasks. I was able to automate the following tasks via Asterisk.

• Getting a temperature reading from inside my house remotely by using a USB based temperature probe
• Turning on/off lightning and my coffee machine
• Inputting scheduled recordings into my Media Center by the use of a small C++ MCE tool I coded (available here)
• Starting and controlling my air conditioner via the use of a USB infrared transmitter
• Activating a web cam in my house and triggering an iptables port forward of the webcam. I could then view the RTSP feed via my mobile phone.
• Accepting all commands via voice input using the Sphinx speech recognition framework
• Voice based SSH port knocking
• Getting the Asterisk box to tell me my IP address for remote access
• Controlling the current state of my house via external scripting. For example, after 10:30PM, my house would set it self to a sleep state, redirect all incoming VOIP calls to voice mail.
• The ability to schedule wake up calls
• All incoming numbers would be announced over a speaker before the phone rings

What you need to build a similar setup

• A Unix based system - I went with Debian Linux and a MSI Titan 700 Via C7 based system
• Asterisk PABX installed with swift installed if you want text 2 voice
• X10 power/light modules
• X10 to rs232 adapter for PC based control of X10 devices
• A TEMPer USB temperature logger for temperature awareness
• External speakers for announcements
• The Sphinx speech recognition toolkit and sphinx.agi for Asterisk if you want speech recognition
• The Linux heyu X10 control software installed

The beauty of Asterisk is that via the "asterisk -rx" command, you can issue commands directly to Asterisk via the Linux command line. For example you could have a X10 movement sensor setup with the heyu daemon listening to trigger a specific script when any movement is detected. That script could then initiate a dial-plan that calls you and reports there has been movement detected. You could then, via Asterisk and your VOIP number, remotely dial in and trigger your webcam stream to be started which you could view on your mobile phone.

To tie everything together I wrote two sets of scripts that worked in separate ways.

One set of scripts are triggered by Asterisk events. Any time I call Asterisk and speak or dial a command, these scripts are triggered to perform tasks such as turning on a light or reading back the current temperature via text to voice.

The second set of scripts are triggered by events such as time of day, a motion sensor being triggered or an iptables logging event. This second set of scripts then initiate Asterisk commands via "asterisk -rx" or performed other tasks such as turning light on or off.

I took things a step further by creating an Asterisk variable called "house mode". I included lists of settings for modes such as phone, night, day, sleep and do no disturb. Based on the current house mode, speaker volume can be automatically adjusted, asterisk can choose whether to ring the phone or send the call directly to voice mail and lights can be dimmed or toggled.

You can potentially store any variables you want in Asterisk via the database put function, but for reliability it is best to store them somewhere more "permanent". For example I keep the current house mode in a flat file and use that file to populate the database as per the example below.

asterisk -rx "database put x10 housemode `cat /etc/heyu/state/housemode.curr`"

Some code examples

This is by no means a small project, it requires extensive bash scripting to link everything together. I have included some code examples below that will hopefully give you some ideas and help get you started.

• Asterisk Extensions.conf examples - this is where all the magic takes place
• A bash script that Asterisk uses to check current house temperature and announce it
• A bash script that Asterisk uses to make an announcement over the speakers
• A bash script that Asterisk uses to open a port no the firewall, VOIP based port knocking
• A bash script to schedule Asterisk to make a phone call
• A bash script that gets the current IP and converts it to a WAV file, ready for Asterisk to play back

This system worked amazingly well for me and was both challenging and fun to build. While I was in hospital I could keep watch of my empty house and when I returned home I hugely benefited from the automation while I was confined to my bed.

Hopefully you can use some of these ideas to get started but please don't hesitate to contact me for more information.

Why virtual desktop admins are going to love VUEM

V.U.E.M. stands for VirtuAll User Environment Manager, a product recently released by VirtuAll Solutions that has the potential to handle delivery of complicated settings to virtual desktops.

It works in a client/server based model. The client is installed onto the virtual desktop and can be hidden from the end user if you require, an application based server is replaced by a simple SQL database that the client connects to directly. This is a perfect solution for most virtualization admin's who will already have access to a MSSQL server and if they are anything like me, they will hate adding new services that require patch management to the network.

Where VUEM really excels is the plethora of different settings that can be deployed with the click of a button, many of which would traditionally require complex and time consuming logon scripts. Via the VUEM  console the administrator can deploy a very diverse range of settings including printers, network drives, registry entries, environment variables, file associations and USER DSN database records.

 

Not only are a large number of settings available for deployment but an equally impressive array of rules and filters can be used to apply the right settings to the right user. I particularly like the XenDesktop and XenApp options that allow me to apply specific settings to specific desktop groups with ease.

Taking things a step further, VUEM has CPU, memory and process management options that enable the administrator to black list applications and protect against CPU/memory spiking. I am particularly looking forward to the "fast logoff" functionality, it is currently only available to XenApp but soon coming to XenDesktop. Fast Logoff does just as its name suggests, instantly disconnects the session when the user clicks log off, allowing another user to immediately log back onto the thin client.

 

It is also worth mentioning there is no reason VUEM needs to be restricted to virtual desktops, it could potentially replace your vbscript or kixtart based logon scripts. VUEM is still in its infancy, only being released in July 2011, but it is extremely developed for its age. This project is certainly worth a look, so head over to http://www.virtualdesktops.info/ and check it out.

Nostalgia Overclocking: MSI Wind U100 meets LN2

A few years ago Josh Collins (former technical writer for Atomic Magazine) and I had a stupid idea. Take the overclockable MSI Wind U100 netbook, add LN2 and a tub full of insulating grease and see how it overclocks.

After a failed first attempt that instantly killed the Wind, our replacement arrived and a 2nd attempt was more successful. The Wind didn't overclock extremely well, but we still were able to push to what was at the time, the highest Atom frequency recorded.

After an hour or so of fighting with CPU-Z we were able to achieve a frequency of 2315 MHz, a full 715 MHz over the stock frequency.

Here is a link to the Engadget article
Also a link to our forum post at the time, dated 25/11/2008

Pointless? Sure. Waste of resources? For sure. Fun? 100%. I tried to convince Apple to let me do the same thing to a Macbook, they weren't so keen.

Building a interactive multi touch surface on the cheap using a Wiimote

Interactive projectors are beginning to come a critical part of both education and presentation, I certainly enjoy using them to play out different networking scenarios, draw over network diagrams and give demonstrations. The key drawback to installing this technology is the expensive, interactive whiteboards such as smart or starboards are very expensive and multi touch interfaces are even costlier.

The next step in the interactive whiteboard revolution is the interactive table. Manufactures like Smart Tech are beginning to get into this field more aggressively but the technology is even more expensive than its wall hanging cousin.

The table brings the projector to a new level, you can write as if you were writing on paper and draw images with much more detailed brush strokes. It also has the potential to be more engaging with people sitting closer to the presentation and being more involved.

Thanks to the pioneering work of Johnny Chung Lee we can all have access to this great technology at a fraction of the cost. While it is not as polished as a commercial product, it is very usable and more than suitable for a wide variety of  uses.



What you need

• A Windows 7 system (preferably 32bit) with bluetooth adapter.
• A projector - I am using an old, end of life projector
• Some form of mount for the projector.
• A Wiimote.
• Software: MultiTouchVista - WiiTUiO - WiiScan
• An infrared pen, you can buy one from a site such as infraredpens.com or build your own.
• Optional: A remote on/off switch and AC power modification to the Wiimote
• Optional: my WiiSync script to automate the launching process

The Challenges ahead

The most challenging part of this configuration is making it easy to use for the end user, if its not easy, they won't use it. As I am installing this unit into an area where multiple users want to use it, it must be extremely easy to use.

The main problems that we need to address are:

Problem
How can we ensure the Wiimote stays paired with the PC's bluetooth adapter? 
Solution
Thanks to a project called WiiScan, we can make the Wiimote remember the bluetooth connection. In most other Wiimote projects a fresh synchronization is established every time the Wiimote connects to the PC. In this scenario the Wiimote never attempts to connect to the PC, the PC always initiates the connection, this adds an extra configuration step for the end user. WiiScan helps us save the synchronization settings in the Wiimote, so every time you turn on the Wiimote and press 1+2 it will attempt to connect to the PC.

Problem
How does the end user turn on and synchronize the Wiimote without misaligning it?
Solution
If the user has to touch the Wiimote no matter how careful they are, eventually it will misalign. I am going to run an extension lead to the Wii's power input via an on/off switch. This will let me connect an AC adapter to power the Wii and allow the user to turn it on and off without ever touching the Wii itself. Furthermore I am going to "jam" on the 1+2 buttons, so as soon as the Wii is turned on, it initiates a synchronization with the PC. 

Problem
How does the end user connect the Wii to the PC with minimum interaction or technical skill?
Solution
The multi-touch interfaces relies on a piece of software called WiiTUiO. WiiTUiO works extremely well but requires the user to first launch it, then tick 2 boxes and press connect in the right order. I have chosen to write a simple AutoIt script that will handle the configuration of WiiTUiO.




Hardware Modifications and Setup

Further to the power/sync modifications I decided it was necessary to "jam" on the 1+2 buttons, this means as soon as someone turns the Wiimote on with the remote on/off switch it will try to connect to the PC.

The Pens
I have built my own infra-red pens, which is really simple to do. You can buy a  infra-red LED from your favourite electronics store (the more powerful, the better), some wire, a AA battery and a click switch and then fit them into a pen or highlighter. The pen builds are out of the scope of this tutorial, but there are lots of great tutorials explaining how to do it.

Further Modifications
Boon Jin has some great tutorials on adding extensions to both the power and red sync button.

Included are some pictures of my further modifications.

Battery lead extension for AC power and switch

Solder points for remote sync button

Remote On/Off and Sync switch

Software Setup

1. Boot your system, copy WiiTUiO, MultiTouchVista (works with Windows 7) and WiiScan onto your computer. Ensure you have a bluetooth adapter installed.

2. Our first task is to installed the MultiTouchVista driver and configure it. MultiTouchVista is the driver that takes our Wiimote from a humble single touch device into a multi touch monster. Under the "multitouchvista" folder, navigate to "Driver" and then "x32" for 32bit or "x64" for 64bit.

3. Run the "Install driver.cmd" to install the base driver.

4. Once the installation is complete. Open "Device Manager", navigate to "Human Interface Devices" and select "Universal Software HID device". Right click "Universal Software HID device" and select "Disable", then immediately "Enable" it again.

Disabling and re-enabling this driver triggers the appropriate services to restart and lets us move onto step 5. If you fail to disable/re-enable it, the following steps will fail.

5. Next we need to configure MultiTouchVista to use the TUIO protocol. To do this navigate to the "multitouchvista" folder and start the "multitouch.service.console.exe", leave it running and also launch "multitouch.configuration.WPF.exe".

6. The default configuration will be "MultipleMice". Select the "Tuio" option from the Available Devices list, then click the blue arrow. At this point "Tuio" should be listed under "Active Device", click "Restart Service". You can now close both windows.

7. Next we need to use WiiScan to create a semi-permanent connection between the Wiimote and the PC. Before we proceed, if you have already connected your Wii to the PC, go into the bluetooth devices list and delete the Wiimote or the next step will fail.

8. Open Wiiscan by launching "wiiscantray.exe". Next promptly press the red synchronization button on your Wiimote and select "Register Wiimote" from WiiScan.

After waiting for about a minute, the confirmation box should appear saying the connection was successful. If it does not, please repeat the process a few times until it pairs successfully.

9. The hard work is done, the Multi Touch driver is configured and the Wiimote has a semi-permanent pair with the PC. Now we need to launch the WiiTUiO software to process the multi touch events.

Launch WinTUiO and tick both "TUIO Multi Touch Evnets" and "Win 7 Multi Touch Events" then press "Connect". If it reports the connection has failed, press the 1+2 buttons on your Wiimote to ensure it is connected to the PC and then try pressing "Connect" again.

10. The last step in the process is to hit "Calibrate" and go through the calibration process. At this point you can also adjust your Wiimote to get the best coverage of your touch area.



Making life easy for the end user

As I mentioned earlier I am planning to put this multi touch table in a heavily used area where different users will be using it all the time. To make things easier I have written a very simple AutoIt script. I don't claim for a minute to have any scripting skills so I am sure this can be cleaned up nicely.

Click here to download the script

Please ensure you put WiiTUiO in a folder called c:\wiimote, also put beyond logic's process.exe in the folder. For this to work perfectly you should also apply the following registry edit, this edit ensures the WiiTUiO icon doesn't get hidden in the systray, which is something my dodgy script relies on!

reg add HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion\Explorer /v EnableAutoTray -t REG_DWORD /d 0 /f

The process for the end user is extremely simple:

1. Click on the compiled WiiSync script.

2. The WiiSync script asks you to flick the "on/off" switch and then press "OK".

3. The WiiSync script briefly disables keyboard/mouse input and sets up WiiTUiO.

4. The user is then informed they can begin using the touch interface.



The Cost

The biggest cost is obviously a projector if you don't already have one, but an old projector that is too weak for other areas will work perfectly.

The other costs for my project are:

Build cost of pens - $10
Wiimote - $50
Bluetooth Adapter - $5
Remote on/off switch - $7
Materials for future mount - $30
Total - $92



Troubleshooting

I found with certain bluetooth adapters I simply couldn't get the connection to be consistent. On the HP laptop I was initially using, the built-in bluetooth adapter would only give me 1 minute of connectivity before I had to repeat the pairing process. I tried a number of different bluetooth stacks, driver versions and configurations, none of which worked.

As soon as I changed to a $5 bluetooth adapter from ebay, things worked perfectly. If you are having issues with pairing you will need to experiment with different bluetooth stacks (bluesoleil seems to be the stack of choice for most people), change the adapter or both.

Some users also reported issues installing VistaMultiTouch if they had previously uninstalled it. This is due to the windows service that VistaMultiTouch leaves behind, your should be able to resolve that problem by opening a command prompt and typing: 

sc delete unisofthid

Then try the driver installation again.



In Action

Here are a few pictures and a youtube video of my set-up in actio. The Windows 7 Multi Touch Pack is a great way to show off this technology, Microsoft Onenote and even Paint also work great.

Youtube video demonstration