Development Blog

There are a few blog posts out there already about setting up the SharePoint Information Worker Demo machines in a "dual boot from VHD" mode, but I found that I needed to pull a number of them together to get it all up and running to my satisfaction.

Firstly, download the latest 2010 IW Demo VM from Microsoft (at the time of writing, this was 2010-10, released 26 March 2012).

For basic SharePoint demos, you can get away with just Server A, which gives you everything except Exchange (Server B) and Lync (Server C).

This walkthrough assumes you don't have access to Hyper-V - which will have implications later on unfortunately.

Ensure you've also got yourself:

• Optional: VirtualBox - a great little app for creating and running Virtual Machines - we're going to use this to get the basic configuration up and running before we set up the boot-to-VHD.
• Drivers - You'll want to ensure that you have suitable drivers for a few bits of hardware - especially Graphics Cards and Wireless adaptors if you're running this on a laptop.

Configure your Virtual Machine

If you want to save yourself a little bit of pain later on, once you've downloaded and unpacked the Virtual Machine, fire up VirtualBox and create a new Windows 2008 Server. When prompted to for the hard disk, attach the one you've just unpacked.

Before firing up the VM, you need to change the hard disk from SATA to IDE otherwise you'll be bluescreening during boot:

1. Select the VM, and click on the Settings button.
2. Switch to the Storage page, and remove the VHD from under the SATA controller - make sure you KEEP the actual files when prompted.
3. Then, add a new Hard Disk under the IDE Controller (the stacked disk icon), and point it at the VHD.

You're now good to go. Fire up the VM, and log in, (using the admin password "pass@word1" - remember that it's currently using the US keyboard layout, so the @ symbol is on the 2 (or quote mark).

Let the warm-up scripts run, but don't try and view SharePoint yet, we haven't configured the networking correctly, but you should be able to open the Central Admin site to confirm things are correct.

You can either configure standalone networking now, or move on to the Dual Boot options.

Setup Networking

Open up the device manager, select the root computer node (demo2010a), and then from the Actions menu select "Add Legacy Device".

In the wizard, you want to add a "Network adapter", then select "Microsoft" from the list of manfacturers, and "Microsoft Loopback Adapter" from the list of network adapters.

Then open up the Network and Sharing Center, and go through to Configure Network Connections. Find the new adaptor (it'll have a description of "Microsoft Loopback Adapter"), right click on it, and select "Properties". Then select the "Internet Protocol Version 4 (TCP/IPv4)" item, and press the Properties button.

Set up the connection as follows:

• Use the following IP address:
  • IP address: 192.168.150.1
  • Subnet Mask: 255.255.255.0
  • Default Gateway: Leave Blank.
• Use the following DNS server addresses:
  • Preferred DNS sever: 192.168.150.1

And click OK. You'll then get a warning that the IP address you've entered is already assigned to another adapter (Microsoft Virtual Machine Bus Network Adaptor) which is no longer present in the computer. Select "Yes" you do want to remove the static IP configuration for the absent adapter.

You should now be able to view the empty Intranet Team site site on http://intranet/ (the default homepage of Intranet.contoso.com may not work at this point, however everything else should now be inplace, including the profiles, and search - try searching for "Tad" or "Erika".

Dual Boot

Next it's time to set up our Dual Boot setup - I've been happily following Scott Hanselman's advice on this for some time, and you should too: Less Virtual More Machine - The magic of Boot to VHD.

If you've left the VHD in the default unpacked location, there's a space in the folder path - when using bcdedit you need to ensure that the path is quoted, for example:

bcdedit /set {guid} device vhd="[C:]\VHDs\2010-10a\Virtual Hard Disks\2010-10a.vhd"

Note also that these VHDs are expandable to 125GB, and when you boot to VHD they need to be able to fully expand, even if they're not using it all - so make sure you've cleared off the .rar files, etc before continuing.

Enabling the Wireless Network

Once you're up and running, if you're on a Laptop, you'll probably want to enable the Wireless Networking Service:

The key thing to do here is to install the Feature from the Server Management view, called "Wireless LAN Service" - once that's configured, you can install your WiFi card drivers and connect to a wireless network (assuming you've configured some AV software, etc.

Fixing the Locale

If you're like me and not based in the US, you may find typing things like the @ symbol tediously complex, in which case you will probably want to change the system locale from English (United States) to something more in line with your keyboard layout.

This is fairly easily accomplished through the "Region and Language" control panel, but don't forget to switch to the Administrative tab and copy the keyboard settings over to the "Welcome screen and new user accounts" - you'll need to restart for these changes to take effect.

Once you've done this however, you will probably notice that nickname and phonetic searching no longer work, so take a look at my short summary of helpful links for enabling SharePoint People Search.

Limitations

The core limitation that I've found with doing it this way is that the content packs aren't installed by default, and unless the VM is attached to a HyperV system you can't install them using the downloadable content packs as these work by attaching a (very) virtual DVD to the VM - and it won't work with Virtual Box or dual booted systems.

I've currently managed to get around this by ripping the content database out of an old instance of the VM (hunt around for 2010-07a) which had the content pre-configured, however I'm still missing some of the walkthrough content - most notably the KPI and Business Connectivity configuration.

Filed under: SharePoint

There are a number of useful blog posts out there that talk about applying Custom Master Pages to SharePoint Team Sites, Blogs, etc., and some, like Bugra Postaci's post, even use stapling to do this automatically.

However, at the end of the day, they all have one thing in common: They claim you need to activate the Publishing Features to do this, even when you're doing it programmatically.

Assigning custom Master Pages to Out-of-the-Box SharePoint sites without activating Publishing Features

Obviously this all needs to happen within a SharePoint project, and assumes that you've created a customised Master Page in your Master Page Gallery.

Starting from a Blank SharePoint project, that is set to "Deploy as a Farm Solution" I added a new feature, and scoped it to the "Web - Activates a Feature for a specific web site.":

Right-click on the feature in the solution explorer, and add an event receiver:

Implement the FeatureActivated method (note that we're also setting the Site Master as well as the System Master, that way if the Publishing Features are enabled later they will be set correctly):

public override void FeatureActivated(SPFeatureReceiverProperties properties)
{
  // Get the SPWeb we're being activated in.
  var web = (SPWeb)properties.Feature.Parent;

  if (null != web) {
    // Get the Site Collection root path to get the master page gallery.
    string siteCollectionRoot = web.Site.RootWeb.Url;

    // Set the Site Master to Custom.master
    var siteMaster = new Uri(siteCollectionRoot +
                             "/_catalogs/masterpage/" +
                             "Custom.Master");
    web.CustomMasterUrl = siteMaster.AbsolutePath;

    // Set the System Master to Custom.master
    var systemMaster = new Uri(siteCollectionRoot +
                               "/_catalogs/masterpage/" +
                               "Custom.Master");
    web.MasterUrl = systemMaster.AbsolutePath;

    // Clear the Alternate CSS
    web.AlternateCssUrl = string.Empty;

    // Save the changes back to the web
    web.Update();
  }
}

And FeatureDeactivating to set things back to the default master page:

public override void FeatureDeactivating(SPFeatureReceiverProperties properties)
{
  // Get the SPWeb we're being activated in.
  var web = (SPWeb)properties.Feature.Parent;

  if (null != web) {
    // Get the Site Collection root path to get the master page gallery.
    string siteCollectionRoot = web.Site.RootWeb.Url;

    // Set both Master Pages to v4.master
    var siteMaster = new Uri(siteCollectionRoot +
                             "/_catalogs/masterpage/" +
                             "v4.master");
    web.CustomMasterUrl = siteMaster.AbsolutePath;
    web.MasterUrl = systemMaster.AbsolutePath;

    // Clear the Alternate CSS
    web.AlternateCssUrl = string.Empty;

    // Save the changes back to the web
    web.Update();
  }
}

All fairly simple so far, and once you've deployed this feature, you can activate it and it will apply your custom master pages to you team sites. As you've hopefully already guessed, these are based on the "System Master Page" - so you really need to ensure that your template is left aligned and fluid - there's a whole other post hidden in that statement I'm sure.

Automatically assigning Custom Master pages on new sites as they are created

Create another feature in your project, and this time scope it to at least "Site - Activates a Feature for all web sites in a site collection", however your milage may vary, and will depend on the setup of your Farm - if you have multiple Web Applications that don't all have this branding applied to them, then you don't want to be scoping this at the Farm level. If the majority of your Web Applications share the same branding scoping it to WebApplication and deactivating it on those Web Applications that don't use the branding might suffice.

Next, add a new "Empty Element" to the project:

First, open the CustomMasterPages feature, and switch to the "Manifest" view to find the id of the feature:

Then in the Elements.xml add the details of the features you want to activate:

<Elements xmlns="http://schemas.microsoft.com/sharepoint/">
  <FeatureSiteTemplateAssociation Id="e5f2c515-d45f-4096-a0b1-7d486a4d011a"
                        TemplateName="GLOBAL" />
</Elements>

As you can see, no Publishing Features were harmed in the making of this feature. I've set this to apply my custom Master Pages to every site template so that the users have a nice consistent branding throughout their Intranet - however you can pick and choose which ones you want here, or you could do it in code during the feature activation if (for example) you want to apply a different Master Page to the Blog Site template.

Then in the stapling feature, add the Element:

Save, deploy and test. You should now have your custom master pages in action every time you create a new site within SharePoint.

Filed under: SharePoint

While I don't do anywhere as much with 3ds Max as I should, I do like to keep learning about it and playing with it every now and again.

One of the most useful resources I've come across in a long time is Jamie Gwilliam's Blog: Jamie's Jewels. Jamie's recently been running a series of "Quick Tips" on his blog which I've found to be incredibly useful, and he also occasionally runs competitions where you can win the odd bit of Autodesk/3ds Max branded or related gear.

His last one was just a call for shots of 3ds Max jewel cases or software boxes, and luckly I had a few very old ones sitting around in the loft, having switched to digital distribution years ago. I only got around to sending a shot of the boxes to the day before the deadline, so was very happy to recieve a very nice iPhone case this week

Anyway, follow @3dsJamiesJewels over on twitter now to find out all sorts of things about this great app.

While it's clearly perfectly understandable to me how Doodle Grouper actually works, I'm occasionally reminded that it's not as intuitive for others as I'd like, so here's a little "How-To" guide, that covers most of the features that currently exist in 1.0.0.7.

Adding Users to a List

The easiest way to add a user to a list is to use the "Add to list..." action the plugin adds to the actions on the User Icon:

This brings up the "Add User To List" Dialog:

Here you can add the user to an existing List by selecting it from the list, or create a new List, press "Add a new list" and then select it and press OK.

The Lists you create with this plugin appear in the User Lists tab of Seesmic:

You can display or hide the columns by clicking on them - the order they appear in the main Seesmic space is determined by the order you show them.

Modifying a List

Once you've displayed a list, you can modify it through the column settings, selecting "Modify this list..."

Then on the "Modify List" dialog, you can rename the list, add more users manually (it's Case Sensitive so be careful doing it this way), delete the entire list or remove a user by selecting them from the list and pressing "Update":

The remove function only removes one user at a time at the moment.

Advanced Functions

On the Settings pane for the Doodle Grouper plugin, you can import and export your lists for example for use on a different computer. While you can export the file to pretty much any location you like, note that you can only import a file from within your "My Documents" folder.

The format of the xml is as follows:

<?xml version="1.0" encoding="utf-8"?>
<ArrayOfUserList xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
                 xmlns:xsd="http://www.w3.org/2001/XMLSchema">
  <UserList>
    <Name>List Name 1</Name>
    <Users>
      <string>User Name 1</string>
      <string>User Name 2</string>
    </Users>
  </UserList>
  <UserList>
    <Name>List Name 2</Name>
    <Users>
      <string>User Name 3</string>
      <string>User Name 4</string>
    </Users>
  </UserList>
</ArrayOfUserList>

I tend to leave Seesmic running, and have encountered memory issues when running with too many lists, so I added the ability to throttle the number of updates that the plugin keeps in each List.

If you "Enable logging", the plugin will add more verbose details into the Seesmic log file to tell you what it's doing.

At the time of writing (1.0.0.7) the "Enable grouping of retweets" doesn't currently work.

Filed under: Plug-ins, Seesmic

It's finally here: After 12 months of other people defining, and committing to support it, one of the latest StackExchange sites to enter public beta is LEGO Answers - Bricks.StackExchange.com^*, the place to ask all your LEGO and compatible brick building questions.

This is the first time I've been really actively involved in the private beta of one of these sites - I was in the WebMasters one enough to fulfil my obligations and earn the Beta badge, but not as much as this time, so here are my thoughts on the process.

Remote collaboration is hard

Working with a group of others remotely, on a site that doesn't encourage "chatter" is hard - Chat is there, but it's not that heavily used at the moment - when what you really want to do is sit down with a few drinks and discuss some of the issues: I had to remind myself on a few occasions not to take things personally, especially as there is a large group of users for whom English isn't a first language - they're doing a lot better than I would do if the site were in German or French , I just hope I didn't cause any offence either...

Moderation requires effort - let us help!

One of the features of StackExchange sites is that as users gain reputation, they gain access to tools and features on the site that give them more say in the running of the site - as is pointed out on the FAQs:

At the high end of this reputation spectrum there is little difference between users with high reputation and ♦ moderators. That is very much intentional. We don’t run LEGO® - Stack Exchange. The community does.

But also, during the beta phase, the reputation requirements are lowered so that the actual users rather than CHAOS moderators do most of the day to day running of the site.

During Private Beta, the reputation requirements are very low, around 1000 rep gets you access to moderation tools - by the end of private beta, we had three or four users who were just over this threshold and helping out Dori and Robert (the two mod's I've really seen around, no disrespect to Geoff, Emmett or Rebecca). Suddenly the site goes into Public Beta, and the threshold's instantly raised to 2000 rep, and a load of new users join the site (that's a good thing ) but suddenly, it's left to Dori and Robert to manage the influx (it probably wasn't that big) of flags, suggested edits, etc., and my edits suddenly need approval again - I mean, I've proved myself already haven't I?

I guess part of this would have been resolved if the influx of users had actually been asking questions that people could answer and vote on - then the reputation would have risen fast enough to counter this - as it is, after nearly two weeks in public beta, I've nearly regained access to the moderation tools .

Conclusion

We need your help - come and ask questions - come and give better answers - and don't be put off by thinking your question is silly, or someone's edited your answer - we're here to help, and craft the best resource for LEGO Q&A's there is - we really need to increase our daily question rate for example.

As some examples, I've had some great answers to building graveyard rooms for Heroica games and building a roof with a 60º pitch, others have been trying to work out how to invert the direction of bricks so that the studs are touching.

* The name may well change - The LEGO Group understandably don't want "LEGO" in the title. Vote for your choice over on the meta question: Site may not be described as LEGO Answers - call for alternative titles

Filed under: Lego, StackExhange, Websites

There are a number of useful blog posts out there on the web that talk you through setting up SharePoint's People search correctly however the following were increadibly useful to me:

Corey Roth's "How to set up People Search in SharePoint 2010" was the most helpful for me - however I think there's something different about the instance of SharePoint I'm running:

When Corey suggests giving the Default Content Access account rights in the User Profile Service Application, I didn't have the option of "Retrieve People Data for Search Crawlers", however giving it access to "User Profile Application" and "User Profile Service Application Proxy" seems to work quite nicely*.

Also, as all my browsers are set up to be British English for preference over US English, or just English, we weren't seeing any of the Nickname or Phonetic search features that are included in SharePoint 2010.

Thanks to Kristof Kowalski's post "SharePoint Server 2010 Phonetic and Nickname Search" and the results of his support request, we were able to get this working for all our British speaking users , and now searching for "Benjamin" will bring me back as "Ben".

* Having finally installed Service Pack 1, and the June, August and October Cumulative Updates the permissions list now matches Corey's blog.

Filed under: Search, SharePoint

This is one of those jobs that looks so simple on the outside, but can lead you down a massive rabbit hole if you're not too careful:

We're in the process of changing an instance of SharePoint from a true "Single Server" setup to having the databases on a separate SQL Server, and there's plenty of documentation about that, starting with: Move all databases

We'd also followed the recommendations in Harden SQL Server for SharePoint Environments but after we'd moved the Content Database to the new server and attached it in SQL, all attempts to "Add a content database" in Central Admin resulted in errors such as:

Cannot connect to database master at SQL server at SqlServer2. The database might not exist, or the current user does not have permission to connect to it.

Digging into the ULS logs we found the following Critical errors:

Unknown SQL Exception 53 occurred. Additional error information from SQL Server is included below.  A network-related or instance-specific error occurred while establishing a connection to SQL Server. The server was not found or was not accessible. Verify that the instance name is correct and that SQL Server is configured to allow remote connections. (provider: Named Pipes Provider, error: 40 - Could not open a connection to SQL Server)

It turns out that sometimes the SQL Configuration Manager needs a little bit of a helping hand, and you need start the Old School "SQL Server Client Network Utility" with the command (note the 8.3 format of the name):

cliconfg.exe

On the Alias tab of this app, add the Aliases you added in the SQL Configuration Manager, and you should be good to go.

Filed under: Fixes, SharePoint

I'm drafting this post in WordPad*, on the basis that I don't have much else installed on Windows 8 - I could have written it directly into the browser, but this feels like a better place to start - I've got a lot to say, and may need to make a small mini-series out of it.

Setting the scene a bit, I was encouraged to write this after I pulled Scott Hanselman into a G+ discussion about other peoples views on the Windows 8 Developer Preview. We were challenged by him to stop wasting our precious keystrokes and blog about it, and others had previously been riled by my statement that we're not "typical" users any more, and wanted to know my thoughts… All the while I'm trying to "remember that this is a Developer Preview, it's not Beta, or even Alpha".

Whilst I've not yet spent all that long playing with it, I've read and watched a few things around the web this week, and so have been forming thoughts for a little while, and here's my opinion, and my guesses as to some possible justifications - note that I am guessing here, and your guesses might be different and more accurate.

One of the big discussions we'd been having was "Who is Metro aimed at?". It's clearly a "Touch First" UI, and indeed preferring a tablet form factor over anything else - using a mouse in there really does feel like a poor runners up choice. My stance was that these days more "typical" users are using tablet devices (mostly with an i at the front) and MS are clearly hoping to get in on that action, and they are using them for content consumption rather than content creation (whatever that content may be, from graphics, music or video through to word documents, spreadsheets, and applications). In that sort of situation, were a user is consuming some media, perhaps along side watching the telly, a full screen app held at not quite arms length on a touchable interface makes a lot of sense. For those of us still creating content, generally this will require the more formal interfaces of mouse and keyboard**.

Now that I've been using it, my initial thoughts are somewhat confused - but I hope that this is because of the early nature of things - a lot of the apps behave inconsistently - for example when I first saw the Metro start screen I thought: "I hope that scrolls left and right as I scroll the mouse wheel up and down", and it does, so I am happy navigating around in there. But so many of the other apps don't do this at the moment (Socialite, Tweet@rama and News/RSS to name three) - which worries me: is this something that Devs are going to have to implement in every app? Shouldn't this be the "native" behaviour of the "Grid Application"?

I know Metro is targeting the Tablet form factor, and I don't yet know how much the frameworks are hiding the fact that the user is using a mouse or a finger for navigating around, but there seems to be a lot of "mystery meat" navigation or buttons, and the usual trick of hovering over them doesn't work, because nothing pops up to tell you what the name of the button is - presumably because you can't hover with a finger.

Take Metro IE for example - what does the button on the right do:

Open a new tab? Open a new URL? Apparently it allows you to search for content, or switch to Desktop IE:

One of the things I've been dabbling with over the last few years is XNA - this isn't in the Windows 8 Dev Preview, and hasn't been mentioned on the technology plans (other than in the negative) - I believe the reason for this is the move to DirectX 11 - XNA at the moment targets the implementations of DirectX on the XBox and Windows Phone platforms (9) - and there are at least two factors at work here:

1. Different Hardware - Historically, new versions of DirectX have needed new hardware to fully support them, this isn't an option with the XBox or Phones, and adding a new version of XNA doesn't work on the XBox will cause confusion I'm sure.
2. Different Teams and Timescales - XNA has often lagged the framework updates a little bit.

This was exacerbated for me at least with the demo at //build/ where they show building a controller aware application, and to do this, they had to build a C++ wrapper - this raises a much higher barrier to entry than just adding a Controller object to your class.

This is a shame, because I've always held the opinion that the easiest way to get kids interested in programming is to get them writing games - it's how I started, and how a lot of my friends started. But then along came the consoles, and they were hard to develop for unless you had the funds for a dev kit ($10,000 for the original XBox I believe); XNA changed all that, and it would be a shame to lose it again.

It's possible that these wrappers will move into the main framework (or sit officially in the Microsoft namespace rather than System) at some point in the not too distant future.

Overall, I think I'll like it, the system is generally snappy and responsive (but this hardware is no slouch), and I'm putting the glitches there down to early drivers and odd hardware rather than anything else, but I'm not sure how much time I'll spend in Metro.

There's more to come, but that will do for now, as I was up too late last night playing with Windows 8.

* That was a very bad choice, at this point in time, on this computer, the rendering of the text in the document, and more importantly, the location of the caret in relation to it is way off - I've not used it in so long I can't say if this is a feature of the app or the OS.

** Which now neatly sets the scene for Voice Recognition to come of age, and we'll all be plugging a Kinect into our PCs to enable the touch features even without a touch screen. April 2012 at the latest, along with my flying car.

Filed under: Opinion, Windows 8

I noticed an interesting thing today while attempting to make some parameters on a PowerShell script required.

As I'm just starting to get the hang of PowerShell, you'll have to bear with my script-kiddie tendencies, however, here's the background:

I'm creating a script that will take the following parameters (can you tell I'm still working with TFS?):

• [string] WorkItemType
• [string] ProjectsToUse
• [switch] ExportWitd

I wanted to make the WorkItemType parameter required, and searching around the internet led me to the following code:

[string] $WorkItemType = $(throw 'workItemType is required')

Which does indeed "work" if you're happy with exceptions being thrown if you don't set it:

workItemType is required
At C:\Scripts\UpdateTFSWorItems.ps1:17 char:37
+     [string] $WorkItemType = $(throw <<<<  "workItemType is required"),
    + CategoryInfo          : OperationStopped: (workItemType is required:String) [], RuntimeException
    + FullyQualifiedErrorId : workItemType is required

And it also doesn't help if more than one parameter is required (you only get the first error), and get-help doesn't help:

PS C:\Scripts> get-help .\UpdateTFSWorItems.ps1
UpdateTFSWorItems.ps1 [[-WorkItemType] <String>] [[-ProjectsToUse] <String>] [-ExportWitd]

As you can see, get-help thinks that WorkItemType is optional (the square brackets around the name and the type).

The actual answer came in the form of the parameter attribute:

[parameter(Mandatory = $true)]
[string] $WorkItemType

Now when I run the script without the WorkItemType parameter I get prompted to supply the required parameters as a read prompt, rather than nasty exceptions:

cmdlet UpdateTFSWorItems.ps1 at command pipeline position 1
Supply values for the following parameters:
WorkItemType:

However, this also has some fairly major consequences:

It converts your script into an "Advanced Function", as some of the documentation on TechNet for this states:

[T]o be recognized as an advanced function (rather than a simple function), a function must have either the CmdletBinding attribute or the Parameter attribute, or both.

There doesn't appear to be a way to apply one or other of these attributes and not be recognised as an advanced function.

Because your script is now "advanced", get-help is a bit different:

PS C:\Scripts> get-help .\UpdateTFSWorItems.ps1
UpdateTFSWorItems.ps1 [-WorkItemType] <String> [[-ProjectsToUse] <String>] [-ExportWitd]
[-Verbose] [-Debug] [-ErrorAction <ActionPreference>] [-WarningAction <ActionPreference>]
[-ErrorVariable <String>] [-WarningVariable <String>] [-OutVariable <String>]
[-OutBuffer <Int32>]

Where did all those extra parameters come from? Your function now supports all the "Common Parameters", this is more clearly stated if you add some help documentation to your script:

<#
.Parameter WorkItemType
    Required. The name of the work item type you want to update in TFS.
.Parameter ProjectsToUse
    New or Legacy, defaults to New.
#>

Calling get-help now results in the following syntax line being auto-generated:

SYNTAX
    C:\Scripts\UpdateTFSWorItems.ps1 [-WorkItemType] <String> [[-ProjectsToUse] <String>]
    [-ExportWitd] [<CommonParameters>]

In general though I think this is a bonus, as it allows me to call the write-verbose and write-debug methods to get additional output and breakpoints.

If you're going to go to the effort of adding parameter documentation, you might as well also supply the HelpMessage parameter of the Parameter attribute:

[parameter(Mandatory = $true, HelpMessage = "The name of the work item type you want to update in TFS.")]
[string] $WorkItemType

Which then allows the user to get some help as they are prompted:

cmdlet UpdateTFSWorItems.ps1 at command pipeline position 1
Supply values for the following parameters:
(Type !? for Help.)
WorkItemType: !?
The name of the work item type you want to update in TFS
WorkItemType:

Gotcha with Write-Verbose

Write-host appears to take an array of objects and write a string representation of them out to the screen, which can be (ab)used in "interesting" ways:

write-host "Exporting WIT Definition from TFS Project" $selectedProjects[$i]

Results in output like:

Exporting WIT Definition from TFS Project Generic

Write-verbose (and indeed write-debug) explicitly only take a single string however, and calling them in the same way results in a nasty exception to that effect, so make sure you concatenate them:

write-verbose ("Project folder " + $selectedProjects[$i] + " already exists")

Which is slightly tedious.

Filed under: PowerShell, TFS

I'm a massive fan of LINQPad - it's much more lightweight than SQL Management Studio if you want to muck around with your data, and allows me to work against SQL with C#, rather than having to remember how to use cursors and the like when I've got some otherwise tedious data re-mapping to do.

That said, I'd never really pointed it at an Entity Framework EDM before, but I'd been wondering about the best way to tidy up the code for the tag clouds in the albums section of this site, so thought that I should see what I could do with the entity model as it stood, without resorting to stored procs.

Firing up the "Add Connection" wizard, I selected "Entity Framework" from the "Use typed data context from your own assembly", selected the website library, selected the (only) EDM from the library, and confirmed the provider.

LINQPad then created the data context correctly, so I pulled up a query to start working against it, only to get the following error message:

An assembly with the same identity 'System.Data.Entity, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089' has already been imported. Try removing one of the duplicate references.

This was slightly confusing - obviously both my project and LINQPad are referencing the System.Data.Entity namespace, and if I remove the reference from my project, I won't be able to build it. How do I remove it from LINQPad then?

It turns out that I don't have to do either. The clue (for me) lay in the notes section of the "Using LINQPad with Entity Framework" page:

If your assembly references other assemblies, LINQPad will pick up these, too, providing they live in the same folder.

Note the bit in bold from the original page - it would appear that my reference to the System.Data.Entity library was set to "Copy Local". Setting that to false and restarting LINQPad removed the error message, allowing me to clean up the tag code somewhat.

Filed under: Entity Framework, Fixes