For the past few days, I’ve been delving really deep into Geneva, though it can be really frustrating at times because it takes awhile to learn the definitions. Although I’m somewhat familiar with terms like Relying Party (RP), Identity Provider (IP), and Security Token Service (STS) due to my little research on Cardspace a couple years ago, it’s still a daunting task when more terms are unravelled before you. So far the best article I’ve read is from the beautiful Michele Leroux Bustamante in her article named “Building a Custom Security Token Service”. I tried out her sample app and it shows how to build a custom active STS. However my aim is to build a custom passive STS because the client is accessing through a web browser and all my RPs are web apps. This article was very informative in explaining the gotcha when authenticating using a passive STS. I’m far from completing my little POC but I feel compelled to write what I’ve learnt before I totally forgot about it.
In order for Michele’s active STS to work you have to import a number of certificates into your LocalComputer cert store. Her Certificates Setup document was pretty good in explaining how to create and install the certificates. However I got lost in her section on Certificates Security. The section talks about private keys installed into the LocalMachine cert store must be accessible at runtime by the host running the RP or STS. Since these are web apps, the ASP.NET account must have access. It then talks about running the Certificate Key File Tool (certkeyfiletool.exe) which is located in the \bin directory of the Windows SDK. After searching high and low, I found a webpage on MSDN that says that this tool no longer exists. To replace it somewhat is a sample app called FindPrivateKey.exe. This sample app has to be extracted from the WCF samples archive in C:\Program Files\Microsoft SDKs\Windows\version\Samples. I trust you do not want to extract all the sample apps in this archive. Since the files in the archive are already sorted by its respective path, look for TechnologySamples\Setup\CS\FindPrivateKey.
Build the FindPrivateKey project in Visual Studio 2008. In order to find the private key, you need to get the thumbprint of your certificate. Assuming you have a certificate called “localhost” and it is already imported into your LocalComputer’s Personal cert store. Run mmc, add the Certificates snap-in, choose Local Computer Account, and under Personal, open the localhost certificate.
Copy the thumbprint.
Run the Windows SDK CMD Command Shell as Administrator, and type this at the command prompt:
FindPrivateKey.exe My LocalMachine -t "c1 37 df 00 92 07 40 6b 6f 26 20 10 a8 74 57 bb 35 92 e1 3e" –a
(of course replace the string in quotes with the thumbprint of your certificate).
Running this returns the physical location of your certificate. Copy the entire path from the CMD shell and paste it onto Notepad. You will need to remove the leading spaces on the 2nd line of the path.
Next run the cacls.exe utility from the same CMD shell.
cacls "C:\ProgramData\Microsoft\Crypto\RSA\MachineKeys\51fd35739bae502ad67f4
d4086d32400_3271abe4-895a-4463-82fe-267f8027cb2a" /E /G "Users":R
For starters, I just grant permission to User group on my machine. However the permission should be granted to the user account under which the application code may be running. For example, if the ASP.NET application always runs under a dedicated user account, you will use that account. You can determine the name of the account by calling WindowsIdentity.GetCurrent().Name to find what account you need to grant access. You may put this in your web application to find out the account of the IIS default application pool. If all is default, you should see the computer account as IIS APPPOOL\\DefaultAppPool.
Run the following command again:
cacls "C:\ProgramData\Microsoft\Crypto\RSA\MachineKeys\51fd35739bae502ad67f4
d4086d32400_3271abe4-895a-4463-82fe-267f8027cb2a" /E /G "IIS APPPOOL\DefaultAppPool":R
You may also try the steps above if you get this error:
System.Security.Cryptography.CryptographicException: The system cannot find the file specified.
Source Error:
An unhandled exception was generated during the execution of the current web request. Information regarding the origin and location of the exception can be identified using the exception stack trace below.
My Live Session information was just published on the Microsoft TechDays 2009 website. Here’s a screen clipping of it:
Catch me live tomorrow at 10:30 am (UTC+08:00) Kuala Lumpur, Singapore.
Just released - the first preview of the Visual Studio 2010 and .NET Framework 4.0 Training Kit!
This training kit preview includes presentations, hands-on labs, and demos for Visual Studio 2010 and .NET Framework 4.0. The content is designed to help developers learn how to utilize the Visual Studio 2010 features and a variety of framework technologies including: C# 4.0, Visual Basic 10, F#, Parallel Computing Platform, ASP.NET AJAX 4.0, and ASP.NET MVC Dynamic Data.
Download the training kit - November 2008 Preview today!
For the past 3 days, I’ve been engrossed in finishing an IE8 accelerator for my pet project codenamed as Arc Reactor. The IE8 Accelerator is simply known as Arc Accelerator. None of these names are final yet, as my team mates and I are still deliberating what will be the actual name of this project.
Anyway the Arc Accelerator is meant to allow users to select a chunk of text in IE8, and save it as a tag. The saved tag can be populated into a “tag cloud” (again no actual name decided yet) maintained by the Arc Reactor client, which is developed as a WPF application.
In order to save the tag, a “session” of some sort needs to be created. It needs to be persistently stored (at least until after the saved tag had been retrieved by the Arc Reactor client). In order to have something like this, I need to identify the user, and to identify a user, that requires some sort of authentication. I’ve pondered over implementing my own authentication function by means of using the ASP.NET membership provider. However, that’s too inflexible as it means yet another set of credentials to be maintained by the user. I thought about using Windows Live ID, but I had no clue how to develop such an authentication function. The fear of the unknown can be rather counter-productive sometimes. I went to check out how to create identity-aware applications, downloaded the WebAuth sample which was made available in the Windows Live ID Web SDK (C# code available, woo hoo!), and it was such a simple thing to do.
Before I could use Windows Live ID authentication, I need to create a new project on the Azure Services Developer Portal. To gain access to this, you just have add the Azure Services Invitation in your Connections Dashboard. Yes it is an invitation-only program, hence you might have to wait a bit before access is granted to you. While you’re at it, also add other Connection programs such as Live Framework CTP Developer Connection, Live Mesh Beta, SQL Server Data Services, and Windows Live Dev Connect. Here’s a sneak peak of all the Connection programs I have under my dashboard:
On the Azure Serviced Developer Portal, click to create a new project. The service component to choose is “Live Services: Existing APIs” as shown below. As you could see, I don’t have access to the other bits, which is a bummer. But anyway, let’s take one step at a time. In due time, I will be able to develop and provision my Azure and Mesh application. I can’t wait.
Enter your project details in the following web page:
The 2 most important fields are Domain and Return URL. Return URL is especially important because this is the URL of your web page handler that will process all the goodies that are returned after Windows Live had authenticated your user. The one and only goodie that I’m keen to get is the user ID so that I can associate the saved tag to an Windows Live ID. After creating your project, you will get 2 very important fields which are the AppID and the SecretKey. You will use these properties in your Windows Live ID login control at a later stage.
Pretty soon I’ll either post my web application here on my blog, or include it in the Arc Reactor project on Codeplex. Stay tuned. In my following posts, I will expound on other things which I have developed such as:
- REST and POX service: To return all the saved tags by Windows Live ID
- WCF service that is based on the List-based Publish-Subscribe pattern
- Using LINQ to SQL class to write the SavedTag into DB
- Clients to test the REST service, WCF service, and ultimately the Arc Reactor client implementation.
If you can’t wait to download my slides from the Adopting ALM event, it’s right here:
The other thing I demo’ed at my session was how to set up an interactive TFS Build agent. There’s a good reason why you would want to run an interactive TFS Build Agent. This is the scenario. You have associated a test list that includes a UI functional test case to your build definition, you want TFS Build to bring out the application window so that your tester could perform UI functional testing, and then decide whether the test passed or failed.
It’s pretty simple. Go to the following folder (the installation folder of Visual Studio 2008).
C:\Program Files\Microsoft Visual Studio 9.0\Common7\IDE\PrivateAssemblies
Edit tfsbuildservice.exe.config file. Locate the following element:
<add key="InteractivePort" value="9192" />
Open a Visual Studio 2008 Command Prompt and make sure you run it as administrator if you are using Windows Vista or Windows Server 2008.
Reserve the port using the wcfhttpconfig utility. The commands you could use below:
C:\Program Files\Microsoft Visual Studio 9.0\Common7\IDE\PrivateAssemblies>wcfhttpconfig
wcfhttpconfig.exe <command> <options>
Commands:
reserve <user> <port> Used to reserve a port for a particular user
free <port> Used to remove a port reservation
Examples:
wcfhttpconfig.exe reserve DOMAIN\User 9191
wcfhttpconfig.exe free 9191
C:\Program Files\Microsoft Visual Studio 9.0\Common7\IDE\PrivateAssemblies>wcfhttpconfig reserve tfsrtm08\darren 9192
Then run the TFSBuildService like below. Essentially you are running another instance of TFSBuildService that listens at the port 9192.
C:\Program Files\Microsoft Visual Studio 9.0\Common7\IDE\PrivateAssemblies>TFSBuildService.exe
Press the Escape key (Esc) to exit...
When you queue your build, just select the interactive build agent. The properties of an interactive build agent:
A couple days ago, I presented a session called “Continuous Integration throughout the Entire Development Lifecycle” at an Application Lifecycle Management (ALM) event held at Auditorium East, Level 29, Microsoft Malaysia’s office. As promised I will blog about my findings and what I’d demo at my session. I’ll be explaining how to customize your TFS Build project so that you could override an existing target with a custom task.
TFS Build is essentially built on top of the foundation provided by MSBuild. Having said that, you could customize TFS Build just like how you would customize your MSBuild script. To find out what are the customizable Team Foundation Build targets, there is a good article on MSDN that explains it well. I’m not exactly adept at creating my own custom tasks, so to my delight I found the MSBuild Extension pack which you could download from Codeplex.
After installing the MSBuild Extension pack, the next thing to do is to edit the TFSBuild.proj file which can be found under the default source control folder named TeamBuildTypes. You may have a number of Team Build types already defined. Each Team Build type corresponds to a build definition. Take note that this is the file that you had set along with its version control folder in the Build definition wizard as shown in the following screen clipping:
In the following task, I am editing the TFSBuild.proj that corresponds to my build definition called “ALM111908Build”.
Check out to edit the TFSBuild.proj file. Open up the file in the Visual Studio editor. I’ll be customizing the Team Foundation Build target, BeforeDropBuild, with a task I am borrowing from the MSBuild extension pack called MSBuild.ExtensionPack.Communication.Email. The definition of the BeforeDropBuild target is:
BeforeDropBuild
Called before saving the built binaries, build log files, and test results to the build-drop directory on the release server. Insert your task into this target to run custom tasks before the built files are saved to the drop-directory.
This task I’m borrowing may be similar to the “A build completes” in the Project Alerts in TFS as illustrated in the following screen clipping.
However I have more control over how the notification email is composed such as including attachments, recipients, subject and body of the email. And I could also decide when to send an email notification at any stage along the TFS Build process, not just when a build completes.
At the end of the TFSBuild.proj file, right before the </Project> closing element, I add the following:
<UsingTask AssemblyFile="$(MSBuildExtensionsPath)\\ExtensionPack\\MSBuild.ExtensionPack.dll" TaskName="MSBuild.ExtensionPack.Communication.Email"/>
<Target Name="BeforeDropBuild">
<ItemGroup>
<!-- Specify some attachments -->
<Attachment Include="C:\demo.txt"/>
<!-- Specify some recipients -->
<Recipient Include="darren@tfs.local"/>
</ItemGroup>
<MSBuild.ExtensionPack.Communication.Email TaskAction="Send" Subject="Email After Build 5pm" SmtpServer="tfsrtm08" MailFrom="tfsbuild@tfs.local" MailTo="@(Recipient)" Body="This email is sent after ALM Build" Attachments="@(Attachment)"/>
</Target>
Be sure to check in the TFSBuild.proj into the source control. Otherwise, your current build definition will not be able to invoke the overriden TFS Build target. If you have set up continuous integration, which in other words each check-in will trigger a TFS Build, you will realize that checking in TFSBuild.proj will not trigger any build. This is an expected behavior because only files within your solution will trigger a build. So if you want to see that the customized email task for the BeforeDropBuild target is successfully performed, you can just queue a new build in the Team Explorer Build menu. You can then check out the BuildLog.txt file, and search for BeforeDropBuild. The first thing you see if the following:
Overriding target "BeforeDropBuild" in project "C:\Program Files\MSBuild\Microsoft\VisualStudio\TeamBuild\Microsoft.TeamFoundation.Build.targets" with target "BeforeDropBuild" from project "C:\Documents and Settings\TFSBUILD\Local Settings\Temp\Demo\ALM111808Build\BuildType\TFSBuild.proj".
This means that TFS Build understands that you are overriding the the “BeforeDropBuild” with a custom task. Subsequently when you search for another occurrence of the target, you will see the following:
Target "BeforeDropBuild" in file "C:\Documents and Settings\TFSBUILD\Local Settings\Temp\Demo\ALM111808Build\BuildType\TFSBuild.proj" from project "C:\Documents and Settings\TFSBUILD\Local Settings\Temp\Demo\ALM111808Build\BuildType\TFSBuild.proj":
Using "MSBuild.ExtensionPack.Communication.Email" task from assembly "C:\Program Files\MSBuild\\ExtensionPack\\MSBuild.ExtensionPack.dll".
Task "MSBuild.ExtensionPack.Communication.Email"
Sending email: Email After Build 5pm
Adding recipient: darren@tfs.local
Adding attachment: C:\demo.txt
Done executing task "MSBuild.ExtensionPack.Communication.Email".
Done building target "BeforeDropBuild" in project "TFSBuild.proj".
The final check before you shout VOILA is when you see this email in your inbox:
Just remember that there are plenty of other tasks you could borrow from the MSBuild Extension Pack. For instance I can compress the assemblies after the build, deploy my web app to IIS 7, create an event log in Windows, and basically anything you have in mind that you want to associate to your TFS Build.
FAQ for Innovate On
Q. I have already profiled my solution on Innovate On portal but I didn’t print out any web pages during the solution profiling process. How do I go back to my solution profile details?
A. Login to www.innovateon.com using your Windows Live ID. You will see the following Innovate On homepage. You solutions are listed in the “My Solutions” section somewhere in the upper right-hand corner. Please note that you may have multiple solutions profiled.
Click on the solution link. This brings you to the solution profile web pages. Example of the first page is illustrated below.
You might think which idiot would want to run a business that is not profitable. True. So wouldn’t you want to invest some time to estimate the P&L for new business opportunities using a tool.
This is perhaps the best-kept secret on the Microsoft Partner Program. Check out this tool here.
I’ve often mentioned this to my ISV partners that whether they move into a software-plus-services world, or while they are debating whether to offer an on-premise solution or hosted solution, it all boils down to their own business model. While Microsoft is more ready than ever with the recent announcements about the Azure Services Platform at PDC, the parallel thought process should be about whether they are prepared for this new business model. I found this very good whitepaper that talks about how Microsoft delivers measurable business opportunities for partners in software-plus-services today.
Microsoft Innovation Day MSC Technology Commercialization Center @Cyberjaya (21st October 2008)
This event had attracted 43 attendees. Developers from Shell IT, Jabatan Perkhidmatan Awam (JPA) and Sapura IT were among the attendees.
The technology focus for this month is Windows Presentation Foundation and the next generation browser, Internet Explorer 8. Mr Peter Tam, Senior ISV Manager from Microsoft gave a welcome remark which emphasized on Metro Early adopter program and InnovateOn. Chris Leong, the Technology Advisor from Microsoft then demonstrated to audience on how to differentiate their product from competitors by employing Microsoft's presentation frameworks such as Windows Presentation Foundation (WPF) and Silverlight.
Another hot topic were addressed by Mr Rohan Thomas, a Senior Web Platform and User Experience Technology Advisor who had demonstrated what makes IE8 the best browser to render Web 2.0 web solutions. The session was ended with the entrepreneurial talk by Aila Zafira, the Assessment Manager from MDeC on Technopreneur fund, which is a conditional grant given to Malaysian to develop their business idea into commercially viable ICT projects.
Microsoft Innovation Day MSC CyberCenter, Plaza Sentral @Kuala Lumpur Multimedia Development Corporation (MDeC) @Cyberjaya (5st August 2008) – Repeated event
This event had attracted 87 breadth audiences ranging from project manager, architects, developers, IT professional to students.
With the theme of Software plus Services(S+S), Mr. Aaron Tan Dani, Director of Financial Platform Strategy at Microsoft completely fascinated the audience with his simple explanation on how S+ S evolves to deliver promising IT solutions by leveraging both of Software and Services worlds. His personal touch on the real world scenario used to demonstrate the importance of S+S effectively addressed the true meaning of S+S.
The duo presentation by Patrick Yong, Application Platform Advisor from Microsoft as well as Ervin Loh gave the developers a good start to create S+S enabled application using Visual Studio 2008. The audiences were also given a short entrepreneurial talk by Jaime Haniff, a senior executive from MDeC on Technopreneur fund to develop their business idea into commercially viable ICT projects.
This event report came from my colleague, Rahimah, who is the Microsoft Innovation Center (MIC) Manager at Cyberjaya.
Microsoft® Innovation Day events offer an inspiring and productive day of previews of emerging technologies and trends, resources, and business benefits that will keep you competitive and help you to energize and maximize your technology. See firsthand, real-world demonstrations that will motivate you to create exciting new products and amazing new solutions for your customers!
This monthly event is brought to you by Microsoft Innovation Center Cyberjaya, a physical facility designed to foster strong, self-sustaining local software industries. This center empowers local startup to focus on innovative development, technology and software solutions through technical consultancy, training, industry partnership, references and best practices.
We have successfully conducted 3 events under Microsoft Innovation Day. More information about innovation day can be found at http://www.microsoft.com/malaysia/events/innovationday/
Thanks to a reader of my blog, SharePoint To Go sample application now uses the Sync Framework 1.0 RTM. The kind contributor of the updated version is one Daniel Studer from Switzerland. Thank you very much, I really appreciate it.
Daniel has updated sample application and in his own words:
I have updated the version.
Errors (not fixed):
- IntegerCalculator not available
(C:\Fieldcontent\SCSF-Apr2008\QuickStarts\Offline\DisconnectedAgent\IntegerC
alculator)
Errors (fixed):
- Added the new References to Sync Services
- Changed SqlCeMetadataStore to SqlMetadataStore
I can't test the project, first I must change the credentials to forms. I hoped it's easier to connect to sharepoint for data synchronization. Both way SQL Server SSIS and Sync Framework needs a lot of code. Just the access developer can work easy and fast with sharepoint lists.
You could easily get the IntegerCalculator from the SCSF-Apr2008 Quick Start application. Anyway you don't really need the Integer Calculator because this demo is about consuming the List.asmx web service from SharePoint.
Yes in order to consume the List.asmx, you first need to add the permissions in SharePoint administration. Please refer to some articles on MSDN on how to do that. My bad, I forgot to document the steps required and I have kinda forgotten what I'd done to enable it. But I've also done a quick hack. Search for NetworkCredentials in my solution. I modified the proxy class (generated by VS when I add a Web reference) to properly use the NetworkCredentials of the standard user in my VPC, who is tfsrtm08\darren.
There seems to be a lot of code in order to do synchronization even when using the Sync Framework. However this is not exactly true. When I developed this sample application, I was reusing most of the code in the sample apps that came with the Sync Framework tutorial. Essentially, all I provided was the implementation classes that is derived from the generic sync provider class in the tutorial. Having said that, it is still relatively simple and it does not involve that many lines of code.
Disclaimer: I have not tested the content of fix provided by Daniel. Everything is provided as-is by Daniel.
There's a good article about how to create versioned assemblies for precompiled ASP.NET web sites. You can read it here.
In order to create versioned assemblies, one of the easiest way is to add a compilerOptions attribute to the compiler Element in your web.config file. You will also have to create an assembly information file in your web application so that it can refer to the version information. In the assembly information file, you would have an attribute like the following:
[assembly:System.Reflection.AssemblyVersionAttribute("versionNumber")]
Please note that the each decimal point in the version number has to be an integer, and it has to be in the format of "major version"."minor version"."build number"."revision". This page has all the information about the format of the versionNumber parameter.
By following the conventional developer wisdom means you will hit F5, or Build the Website. Kaboom! An error appears, and it says:
Source file 'path\AssemblyInfo.cs' could not be found
You'd spend the next tens of minutes checking if your path is correct, or if the AssemblyInfo.cs file was spelt correctly. Do not fret. I'm telling you that you need not worry. The version information is not taken into account by the compilation procedure in Visual Studio. Instead, it is taken into account by the ASP.NET precompiler. You could precompiler your ASP.NET Web site using the following command (launch the Visual Studio 2008 Command Prompt):
aspnet_compiler -p physicalOrRelativePath -v / targetPath
This page describes everything you need to know about precompiling ASP.NET Web sites.
To ensure that your assembly was indeed "versioned":
- At the Visual Studio 2008 Command Prompt, navigate to the targetPath that you have specified in the aspnet_compiler command-line options.
- Go to the bin folder of your precompiled and deployed ASP.NET Web site.
- Run ildasm App_Code.dll.
- You will see the version information such as the following:
.assembly App_Code
{
.ver 1:1:2002:2
}
Hence if you want to have versioned assemblies, this should be a step after building (compiling) your ASP.NET Web site in Visual Studio. The precompilation will take care of the versioning of your assembly. If anyone know the proper way of doing this, please do leave a comment in this blog post.
In the past couple of days, I've been delving into prescriptive architecture guidance for Team Foundation Server (TFS). Along the way I learnt something new, something I never knew existed in TFS. It's called cloaking. Allow me to quote the following text from the Team Development with Visual Studio Team Foundation Server prescriptive architecture guidance (search under Chapter 4):
Cloaking
You can use cloaking as a performance optimization when you want to prevent a part of the source control tree from being retrieved. The following are typical scenarios for using cloaking:
• You want to build the project locally and a folder is not needed for the build, for example a documentation folder.
• You are part of a large team project and you only want to retrieve part of the project.
For either of the above scenarios you can cloak folders to stop the client retrieving those folders. You cloak folders on the client by editing the workspace and changing the status of the working folder from active to cloak.
Keep the following recommendations in mind when you cloak:
• Do not cloak individual files. This is likely lead to maintenance problems later in the project.
• For a large project, map out the root folder and cloak sub folders rather than creating multiple workspaces for the project.
Having said that, I went ahead to put this to practice on Visual Studio TFS 2008. Before I got started, I paid close attention to the statement that says, "You cloak folders on the client by editing the workspace and changing the status of the working folder from active to cloak.".
The question that begs is "What is a Workspace?".
Workspaces Explained
A TFS workspace is a client-side copy of the files and folders in TFS source control. A workspace maps source control folders to local file system directories. When you make changes to files within the workspace on your local computer, the local changes, referred to as pending changes, are isolated in your workspace until you check them into the server as an atomic unit. The collective set of changes, checked in as a batch is referred to as a changeset.
As a good measure for practice, I didn't want to edit my existing default workspace, so I created a new workspace. This new workspace will map the source control folders to a new folder in my local file system. These were the steps I took to create a new workspace:
- Create a new workspace.
- In your Source Control Explorer, go to the Workspaces combo box. Click Workspaces... (as shown in the following diagram)
- This brings out Manage Workspaces window.
- Click Add.
- Then, I mapped the solution folder in my source control to a local folder as shown below.
- Since the Doc folder is not required in the building of my solution, I want this folder cloaked. Another reason why I would want the documents folder cloaked is because there may potentially be very large files that I do not want to update from the source control every time I get latest changes.
- After I've cloaked the Doc folder, when I get latest version from the source control, it will not get the files under the Doc folder, and the folder appears as grey in the Source Control Explorer.
- Done!
Please bear in mind that you can only cloak folders that have a mapped parent. Otherwise you get an error message like the following.
In a week's time, I'll be in Guangzhou to assist a managed ISV. Guangzhou is located in the province of Guangdong, and the people there primarily speak Mandarin and Cantonese. I speak Cantonese somewhat fluently, but I know very little of Mandarin. This is because I can't read or write Chinese. To my excitement, there is a new translator service called Microsoft Translator. What's even more interesting is that a new "T-bot" that translates as people chat in Windows Live Messenger and it is part of the Microsoft Translator team's growing success story. Fear no more as when one is lost in translation.
Microsoft translation technology is now available for Live Search, Internet Explorer 8, the Windows Live Toolbar, and the Windows Live Messenger T-bot. The translator, often available as a button in those programs, can help translate Office documents, Web pages, and chat sessions. It also offers the more traditional Web translation, where users type a sentence or paragraph to be translated.
To use T-bot, add mtbot@hotmail.com to your Windows Live Messenger buddy list. When you initiate a chat, T-bot will talk back and walk you through which languages you want to use at both ends of the conversation. Once you select your languages, invite another person to the chat. T-bot will assume you have invited someone who speaks the target language and then translate everything you say. The chat translation time delay is usually less than a second.
With my T-bot, I feel very confident about communicating with the R&D team in Guangzhou. So far they have been copying me in emails. That's great except that the email was written in Chinese! :p
Fear no more!
