my.net

SQL Server Database Restore for Dummies – like me

2012-01-30T18:05:00.001+01:00

I got a SQL server (express) database backup (.bak file) from a customer to do a test against. In the past, I just restored from my own, previously existing databases, so I never came across the issue of restoring one from a complete different machine. Using SSMS I got the following error messages

The backup set holds a backup of a database other than the existing database – or

File 'Db' cannot be restored to 'C:\Program Files\Microsoft SQL Server\MSSQL10_50.CUSTOMER\MSSQL\DATA\Db.mdf'. Use WITH MOVE to identify a valid location for the file.

The path does not exist on my machine. After browsing the web I found out the following:

create a new and empty database named as the one from the customer (in this case ‘Db’)
use a script to run
RESTORE DATABASE [Db]
FROM DISK = 'c:\theDbBackup.bak'
WITH REPLACE

Excellence in Software Engineering Conference 2012 - in Zurich

2012-01-29T22:02:00.001+01:00

This year the ESE conference 2012 will take place in Zurich April 24. and 25. – hosted by Zühlke Engineering AG.

That’s where software experts from around the world meet their peers and prominent speakers. The conference targets software architects, project managers, IT manager, process experts and all kinds of techies. Interactive discussion and talks are a major part of the two days.

There are three trendy tracks: Agile, Cloud and Mobile. See this for more details.

Struggling with Azure SDK Samples –> use Microsoft.ServiceBus.dll Version 1.6! or still 1.5?

2012-01-16T10:47:00.001+01:00

Got some issues recently with the Azure SDK samples (e.g. RelayedConnectionSample with NetTcp, MulticastSample, etc.). SecurityExceptions were thrown, etc.

Until I noticed that the samples still reference the Microsoft.ServiceBus.dll in

version 1.5.0.0 under c:\Program Files\Windows Azure AppFabric SDK\V1.5\Assemblies\NET4.0\ instead of the latest one
version 1.6.0.0 which is now installed under C:\Program Files\Windows Azure SDK\v1.6\ServiceBus\ref

So you need to

change assembly references to Microsoft.ServiceBus 1.6.0.0
change .config settings to 1.6.0.0 to:

    <extensions>
      <bindingExtensions>
        <add name="netEventRelayBinding" type="Microsoft.ServiceBus.Configuration.NetEventRelayBindingCollectionElement, Microsoft.ServiceBus, Version=1.6.0.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35" />
      </bindingExtensions>
    </extensions>
  </system.serviceModel>
 
  <runtime>
    <assemblyBinding xmlns="urn:schemas-microsoft-com:asm.v1">
      <dependentAssembly>
        <assemblyIdentity name="Microsoft.ServiceBus"
                          publicKeyToken="31bf3856ad364e35"
                          culture="neutral" />
        <bindingRedirect oldVersion="1.0.0.0"
                         newVersion="1.6.0.0"/>
      </dependentAssembly>
    </assemblyBinding>
  </runtime>

and it works again.

However, some of the samples did not with 1.6 but only with 1.5. Maybe I messed up with the latest SDK installation.

Azure Connect

2012-01-16T10:42:00.001+01:00

Using Windows “Azure Connect” - a service from the Windows Azure platform – one can establish a protected network connection between enterprise computers and the cloud roles (E2C). This virtual IPsec network allows for:

easier migration of tiers of an applications to the cloud
tunnel firewalls, proxies, NATs
hybrid (cloud to on-premise) apps
various higher-level communication protocols
apps where the protocol might change
have high number of communicating endpoints
domain joins (some ActiveDirectory; logins)
remote administration, debugging and troubleshooting
DNS
etc.

Setting up Connect is a 3 step process. There is a tutorial showing all the details.

1. Get an activation token from the Azure portal and configure your Azure roles to use it.

Copy the token into your role’s properties / virtual network settings

Now you publish your Azure roles to the staging or production environment in the cloud. Wait until the service is ready and check that it appears in the portals “Virtual network” –> “Groups and Roles” and “Activated Endpoints”.

2. Enable local machines to connect to roles in the cloud. Install the Connect Agent on local machines. Copy the link URL to a browser, download the configured executable and run it. A system tray icon like the following should appear

3. Pair the local machines and Azure roles in a “Group” (network policy). In the portal click Create Group

The system tray icon should change to connected now.

Please note that ping will not work until you enable it in your role using a startup task script. From my limited home network I experienced some significant (first-hit) network delays

Pinging RD00155D3A622D [2a01:111:f102:121:858:cf20:a9dd:5639] with 32 bytes of data:
Reply from 2a01:111:f102:121:858:cf20:a9dd:5639: time=299ms
Reply from 2a01:111:f102:121:858:cf20:a9dd:5639: time=258ms
Reply from 2a01:111:f102:121:858:cf20:a9dd:5639: time=359ms
Reply from 2a01:111:f102:121:858:cf20:a9dd:5639: time=368ms

Ping statistics for 2a01:111:f102:121:858:cf20:a9dd:5639:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 258ms, Maximum = 368ms, Average = 321ms

Microsoft Azure SQL Data Sync

2012-01-16T10:36:00.001+01:00

Azure Data Sync let’s you synchronize enterprise (on-premise) SQL Server (2005 and above) databases with SQL Azure databases – or several SQL Azure databases. There is a local agent and a cloud sync group service necessary for it to work.

This is the first preview from Nov 14 2011; let’s have a try.

In the management portal navigation tree go to Data Sync, select a subscription, a data center region,

You now have two options:

Let’s synchronize on-prem and a cloud SQL Azure database. The wizard leads you through the configuration of a sync group

Download and install the so called Sync Agent for the local server.

Then start the Data Sync Agent Preview from the start menu.

Now generate an agent key from the management portal and copy it into the local agent wizard.

Then register a SQL Server

to get

Back in the portal we add the agent to the Sync Group

Now we are half way through. Now let’s add an SQL Azure database:

Just creating a new SQL Azure database for this demo – and selecting it in Step 3 above.

In the following step (4) we configure the scheduling, and the conflict resolution (client or hub wins)

Select the on-premise SQL database

Click deploy now

Done!

Now, let’s see that it worked correctly. In the local Northwind database there are 2 people from Switzerland, add a row filter for Country = ‘Switzerland’. Then go the SSMS, connect the SQL Azure DB. There is a new dbo.Customers table now. Please also note the added tables for the change tracking. There are also some triggers and stored procedures added to your database. This may impact your runtime performance.

There are the 2 rows as expected:

That’s it again in the portal

This was actually super simple to setup. There are some best practices out there.

Negative point: there does not seem to exist an API or .NET library for scripted configuration and operation.

Fun Project: DotNetNuke module views

2011-11-25T20:38:00.001+01:00

Last time I showed you how to create, build, package, upload and use a simple DNN module, developed in C# using Visual Studio
2010. Let’s get started by thinking (and writing down ) some requirements.

Requirements

Non-functional

R-NFR1 DotNetNuke 05.06
R-NFR2 ASP.NET 4
R-NFR3 local testability without DNN installation

Use Cases

As a client I want to:

R-U1 – Create an online account on the site
R-U2 – Log in and out
R-U3 – Choose a treatment
R-U4 – Choose a time/date
R-U5 – Book a massage
R-U6 – Got sent an email confirmation
R-U6.1 with an outlook appointment attached to it
R-U7 – get shown my booked treatments in detail

As an administrator I want to:

R-A1 – Define treatments (duration, title, description)
R-A2 – Define time windows (opening or working hours)
R-A3 – Show (todays, etc.) bookings
R-A4 – Got sent an email with a client booking information
R-A5 – Manage accounts (list, add, edit, delete)

New Web Project

To separate DotNetNuke module stuff from booking web functionality I added a new ASP.NET web project to the solution

remove the generated pages, sub folders and strip down the web.config file to a minimum.

Now add a web user control to the BookingViews project

Telerik

Writing an outlook style schedule/calendar control in ASP.NET seems like a huge effort. Let’s get a commercial control library suite including a schedule control. As DNN already contains Telerik controls the evaluation was short. They have an excellent reputation when talking to colleagues.

Data Model

First let’s show the data access layer where the massage bookings and treatment offerings are persisted. Using the Entity Framework 4.0 model first approach, I create the following model:

The BookingEntry and Treatment entities were added in the model edmx in VS design view. However, the User entity was imported from the existing DotNetNuke database (table first). Adding some relations and I can take advantage of the DNN register/login/user functionality in my bookings.

View

Learning and configuring the module view took more effort. This is the resulting aspx:

<%@ Page Language="C#" AutoEventWireup="true" CodeBehind="BookingForm.aspx.cs" Inherits="BookingViews.BookingForm" %>
 
<%@ Register Assembly="Telerik.Web.UI" Namespace="Telerik.Web.UI" TagPrefix="telerik" %><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head runat="server"><title></title></head><body><form id="form1" runat="server">
    <asp:ScriptManager ID="ScriptManager1" runat="server">
    </asp:ScriptManager>
    <div style="height: 527px; width: 827px">
        <telerik:RadScheduler ID="RadScheduler1" runat="server" Culture="de-CH" DayEndTime="19:00:00"
            DayStartTime="07:00:00" FirstDayOfWeek="Monday" Height="558px" SelectedView="WeekView"
            ShowViewTabs="False" Skin="Forest" Width="800px" WorkDayEndTime="19:00:00" WorkDayStartTime="07:00:00"
            SelectedDate="2011-11-01" DataEndField="End" DataKeyField="Id" 
            DataStartField="Start" DataSubjectField="Subject" 
            onappointmentcommand="RadScheduler1_AppointmentCommand" 
            ontimeslotcreated="RadScheduler1_TimeSlotCreated" 
            DataSourceID="BookingEntryDataSource" DataDescriptionField="Remarks" 
            EnableDescriptionField="True">
            <ResourceTypes>
                <telerik:ResourceType DataSourceID="TreatmentDataSource" 
                    ForeignKeyField="TreatmentId" KeyField="Id" Name="TreatmentResource" 
                    TextField="Name" />
            </ResourceTypes>
            <WeekView DayEndTime="19:00:00" DayStartTime="07:00:00" HeaderDateFormat="dd.MM.yyyy"
                WorkDayEndTime="19:00:00" WorkDayStartTime="07:00:00" />
            <AppointmentTemplate>
                <span style        ="font-weight: bold; font-size: small">
                    <%# Eval("Subject") %>
                </span>
                <br />
                <asp:ImageButton ID="ImageButton1" runat="server" ImageUrl="~/Resources/Calendar.png" OnClick="ExportOutlook" CommandName="ExportOutlook" />
            </AppointmentTemplate>
        </telerik:RadScheduler>
    </div>
    <telerik:RadAjaxManager runat="server">
        <AjaxSettings>
            <telerik:AjaxSetting AjaxControlID="RadScheduler1">
                <UpdatedControls>
                    <telerik:AjaxUpdatedControl ControlID="RadScheduler1" />
                </UpdatedControls>
            </telerik:AjaxSetting>
        </AjaxSettings>
    </telerik:RadAjaxManager>
    <asp:ObjectDataSource ID="BookingEntryDataSource" runat="server" 
        DataObjectTypeName="My.DotNetNuke.Modules.BookingModule.BookingEntry" 
        DeleteMethod="Delete" InsertMethod="Add" SelectMethod="GetAll" 
        TypeName="My.DotNetNuke.Modules.BookingModule.Data.BookingEntryDataRepository" 
        UpdateMethod="Update"></asp:ObjectDataSource>
    <asp:ObjectDataSource ID="TreatmentDataSource" runat="server" 
        DataObjectTypeName="My.DotNetNuke.Modules.BookingModule.Treatment" 
        DeleteMethod="Delete" InsertMethod="Add" SelectMethod="GetAll" 
        TypeName="My.DotNetNuke.Modules.BookingModule.Data.TreatmentDataRepository" 
        UpdateMethod="Update"></asp:ObjectDataSource>
    </form>
</body>
</html>

The control’s data sources are of type ObjectDataSource. BookingEntryDataSource and TreatmentDataSource both implement the IRepository<T> interface. This hides the EF stuff behind a façade and allowed better testability than the EntityDataSource:

    /// <summary>
    /// Interface to decouple entities from data access.
    /// </summary>
    public interface IRepository<T> where T: class
    {
        void Add(T entity);
        void Delete(T entity);
        ICollection<T> GetAll();
        T GetById(T entryWithKeyOnly);
        void Update(T entity);
    }

Note that I added the telerik rad scheduler, configured a week view, added some appointment resources and DataSources to be bound in the scheduler control.

And this is the final result:

Double clicking shows or edits the details of the appointment

I hope this 3-part series of my (hobby project) development “minutes” gave you some insight around DotNetNuke module development in Visual Studio using C#.

Azure AppFabric Caching ASP.NET Session State Provider – not working with 2 Instances in the local Compute Emulator

2011-11-25T09:35:00.001+01:00

A customer asked me for help about this.

Introduction

Windows Azure AppFabric Cache is a subset of the Windows Server 2008 R2 AppFabric distributed in-memory cache (aka “Velocity”) and typically provides it’s services for ASP.NET session provider and output caching.

Enable Cache

To speed up your web apps performance you first need to enable this Azure feature on the management portal

Click New Service Namespace –> select Cache –> select your Azure subscription (for the billing) –> the region of the data center –> and a name.

Configure Web App to use the cache

In the management portal Properties copy your host name and authentication token

to your web.config <Configuration> section:

<configSections>
  <section name="dataCacheClients" type="Microsoft.ApplicationServer.Caching.DataCacheClientsSection, Microsoft.ApplicationServer.Caching.Core" allowLocation="true" allowDefinition="Everywhere"/>
</configSections>
 
<dataCacheClients>
  <dataCacheClient name="default">
    <hosts>
      <host name="[SERVICE-HOST-NAME]" cachePort="22233" />
    </hosts>
    <securityProperties mode="Message">
      <messageSecurity
        authorizationInfo="[AUTHORIZATION INFO]">
      </messageSecurity>
    </securityProperties>
  </dataCacheClient>
</dataCacheClients>
 
<system.web>
 
  <sessionState mode="Custom" customProvider="AppFabricCacheSessionStoreProvider">
    <providers>
      <add name="AppFabricCacheSessionStoreProvider"
            type="Microsoft.Web.DistributedCache.DistributedCacheSessionStateStoreProvider, Microsoft.Web.DistributedCache"
            cacheName="default"
            useBlobMode="true"
            dataCacheClientName="default" />
    </providers>
  </sessionState>
 

Add assembly references to the Azure libraries

Microsoft.ApplicationServer.Caching.Client.dll
Microsoft.ApplicationServer.Caching.Core.dll
Microsoft.Web.DistributedCache.dll
Microsoft.WindowsFabric.Common
Microsoft.WindowsFabric.Data.Common

2 Instances

Configure your web role to use 2 instances. Otherwise you won’t notice the problem.

Attention

Make sure, you do not confuse with the Windows Server 2008 R2 AppFabric caching libraries. Use the ones under

C:\Program Files\Windows Azure AppFabric SDK\V1.5\Assemblies\NET4.0\Cache

Access the cache from code

In the following example from the AppFabric Lab Samples, store some shopping cart entries into the session

List<string> cart = this.Session["Cart"] as List<string> ?? new List<string>();
cart.Add(selectedItem);
Session["Cart"] = cart;

and read back

var itemsInSession = this.Session["Cart"] as List<string> ?? new List<string>();

Test it

Now when I start the app and add a product to the cart, stop the compute emulator, do an iisreset.exe and refresh the page – then I still see the product. This now comes from the Azure AppFabric Cache:

on instance 1. From the debugger:

Checkout on instance deployment16(44).AzureStoreService.MVCAzureStore_IN_1

Pressing F5 to refresh the page again in the browser I suddenly don’t see the product anymore

on instance 0. From debugger

Checkout on instance deployment16(44).AzureStoreService.MVCAzureStore_IN_0

Pressing IEs F5 again and again, shows or don’t shows the product …. ?!?

Problem

So the findings are, that the session cache basically works, but not in a (load-balanced) multi-instance environment on the local development fabric compute emulator! Depending on the web role instance (0 or 1) a different session cache is used.

It looks like each web role instance is still using default ASP.NET in-proc session state, meaning you'd have different session state on each instance.

However, the session id is the same

I made sure the default in-proc session state provider is removed by commenting it and <clear/>ing the providers list:

<!--<sessionState mode="InProc" customProvider="DefaultSessionProvider">
  <providers>
    <add name="DefaultSessionProvider" type="System.Web.Providers.DefaultSessionStateProvider, System.Web.Providers, Version=1.0.0.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35" connectionStringName="DefaultConnection" applicationName="/" />
  </providers>
</sessionState>-->
<!-- If session state needs to be saved in AppFabric Caching service, add the following to web.config inside system.web. If SSL is required, then change dataCacheClientName to "SslEndpoint". -->
<sessionState mode="Custom" customProvider="AppFabricCacheSessionStoreProvider">
  <providers>
    <clear/>
    <add name="AppFabricCacheSessionStoreProvider"
          type="Microsoft.Web.DistributedCache.DistributedCacheSessionStateStoreProvider, Microsoft.Web.DistributedCache"
          cacheName="default"
          useBlobMode="true"
          dataCacheClientName="default" />
  </providers>
</sessionState>

All over again – it works in the Cloud

Not sure I did something wrong in the Lab sample – I started all over again from scratch with a simple app that stores the text box value into the Session[] state and reads it back using the buttons. Value/InstanceId/SessionID are written to the label below.

Same problem on the local fabric – BUT …

… storing from IE’s first tab the value “m5” to web role instance 0 in session 5WJJ:

Reading back from IE’s second tab the value “m5” from web role instance 1 in session 5WJJ:

gives me what I expected.

Help

As I understand this should not be the case? Seems to be an issue in the local compute emulator using the AppFabric cache service! It works in the cloud – but still makes local testing hard. Any ideas?

Fun Project: DotNetNuke module development

2011-11-23T17:10:00.001+01:00

This post shows you how to create a DNN module in C# using Visual Studio 2010 – from beginner to beginner – to make expectations clear.

Requirements

In time with the site going live, new requirements popped up:

online bookings of treatments by clients
manage massage treatments through my wife
booking selected treatments using a kind of schedule/calendar view (clients) in predefined time windows / opening hours (admin).
email confirmations (client and admin)

I found no module for these .

Visual Studio Project Template

As Visual Studio and C# are my friends, I downloaded the DNN template from dotnetnuke.codeplex.com and added a new project to my VS 2010 solution:

which added the following structure:

changed to .NET framework to 4.0 and added my custom namespace

Performing a Release build

Packaging

Right clicking the module project in the solution explorer and choosing “Build Deployment Package” creates the necessary CMS upload package BookingModule_00.00.01_Source.zip in the packages sub folder.

The packing script error:

Error 1 The target "Package" does not exist in the project.

can probably be ignored. It indicates the build script references the solution file rather than the csproj project file.

Note the packaging scripts are included in your DNN module project folder “BuildScripts”

Host Login

When I created the DNN site within my hoster’s portal, I got asked to create two login account credentials:

Administrator: create pages, add installed modules to pages, change skins and styles, etc.
Host: same as the administrator, but with additional rights, one of these is installing new modules

So you must log in with your host account

to see the additional site menu items “Admin” and “Host”. Clicking host shows the Extensions icon:

to see the DNN framework (version), Authentication System, Container, Library, Module, Provider, Skin, Skin Objects and in the bottom right pane the command to install a new module Install Extension Wizard:

selecting and uploading the previously built release package, next, booom!

Install Extension

Upload Results
If you have reached this page it is because the installer needs to gather some more information, before proceeding.

Error reading the zip package - see below

StartJob
Reading Installation Manifest file

Info
Reading Package Manifest - Module - BookingModule

Failure
Install aborted - The module being installed is not compatible with your host application version ( an upgrade is required )

I spent some time googling for answers, found some useful hints, saying the module is not compatible with my DNN version. But my problem remains: I cannot update my CMS with my hoster and I’ve got the latest Visual Studio template!? Hm. The solution was in the VS project DNN manifest file (project.dnn):

<dotnetnuke type="Package" version="5.0">
  <packages>
    <package name="BookingModule" type="Module" version="00.00.01">
      <…/>
      <dependencies>
       <dependency type="CoreVersion">06.00.00</dependency>
      </dependencies>

You need to delete the dependencies node, rebuild, repackage, upload again:

Yeah.

You can now add it to you page / pane as host –> edit mode –> common task –> add module

and when the module has the same rights as the page, then you can see it. There’s a context menu added

that let’s you switch to an “Edit Module” view or “Settings” view. Do you remember the three .ascx asp.net user controls in the above module project?

I hope this blog post gave you some overview of DotNetNuke module developments to get started with, including some pitfalls (I have fallen into).

Next time, I plan to show you the development of the web stuff, including the database model, data access, UI.

Fun Project: DotNetNuke – Failing with WebMatrix/ Orchard

2011-11-23T17:04:00.001+01:00

No, I’ve not been in jail – but I’ve been very busy in a project recently and could not share my .NET insights with you. So I start again with a fun project, in a field I never work or accept customer mandates so far: web development, CMS, homepages.

It all started in an evening effort to create a web site for my wife’s new hobby (ehm, business): wellness, wellbeing, massage.

Microsoft WebMatrix / Orchard

I started with the latest and smallest .NET (what else) open source content management system (CMS) out there: Microsoft Orchard using WebMatrix. Learned the tool basics, got into the CMS, created some content, browsed through some free skins, modified CSS (with blood and sweat), had fun and insights …

… but failed due to hosting restrictions from my shared web hoster (Microsoft Orchard 1.2.41 / ASP.NET 4.0 / SqlCompact 4) with a security exception (Exception Details: System.Security.SecurityException: That assembly does not allow partially trusted callers.). As I did not want to update my subscription to a more expensive one, I went back to “field 1” to start all over again.

Learning: Orchard needs full trust shared hosting.

DotNetNuke

But my hoster supports DotNetNuke (DNN) 05.06 Community Edition! Another new experience for me.

Installation

It took a couple of web forms filling (SQL connection, accounts, etc.) and 10 minutes waiting time for the installation on my hoster’s management portal. A no brainer.

Customization

After log-in to the deployed CMS you can start adding pages, adding modules to pages, adding content to modules.

Modules are added to panes of a page. Panes are typically the following areas:

Blood and sweat again: there just the so called “MinimalExtropy” skin to start with – or get a bunch of commercial skins or create your own using commercial tools. But I wanted to migrate my Orchard CSS to MinimalExtropy, which took me another (long) evening.

@import url("css/menu.css");

/* CSS Document */
#Body {background: url(images/body_bg.png) #c6e29f fixed no-repeat center bottom; height:100%; margin:0; padding:0; }
 
/* ControlPanel style */
.ControlPanel{background:transparent url(images/cpanel_center.png) repeat-x bottom left; border:none;padding-bottom:10px;}
.cpanel_left {background: url(images/cpanel_left.png) no-repeat bottom left;padding-left:21px;}
.cpanel_right {background: url(images/cpanel_right.png) no-repeat bottom right;padding-right:33px;}
 
/*-------- Default Style --------*/
body,th,td,table,h1,h2,h3,h4,h5,h6,.Normal,.NormalDisabled,.Head,.SubHead,.SubSubHead, a:link, a:visited, a:hover, input, .CommandButton{color:#2E3D47; font-family:Verdana, Arial, Helvetica, sans-serif;}
body,th,td,table,.Normal,.NormalDisabled,.Head,.SubHead,.SubSubHead, a:link, a:visited, a:hover{font-size:12px;}
h1, h2, h3, h4, h5, h6, h2 input{color:#375162;font-weight:bold;margin:1ex 0;}
h1{font-size:18px;margin:10px 0;}
h2, h2 input{font-size:14px;}
h3{font-size:13px;}
h4{font-size:12px;}
h5{font-size:11px;}
h6{font-size:10px;}

… and the like.

There are a couple of pre-installed modules for DNN and an ecosystem of free and commercial modules out there. A module is a kind of a web part like blogging, picture gallery, etc.

Windows Mobile Device Center crashing with "Services is not a valid partnership ID"

2011-07-19T15:36:00.001+01:00

Almost drove me crazy, this error on my Windows 7 64bit Ultimate, connecting Outlook 2010 with my Windows Mobile 6.5 HTC Touch Diamond 2. From one day to the other synchronization did not work any more because the WMDC process crashed. Un- and re-install did not help.

The solution was to delete all sub keys (partnerships) from the following registry node:

HKEY_CURRENT_USER\Software\Microsoft\Windows CE Services\Partners

ESE Software Engineering Conference in Zürich – Call for Papers

2010-11-02T09:50:00.001+01:00

The international Software Engineering Conference ESE Conference 2011 (12.-14. April) opens the Call for Papers.

So, if you are an expert in a technical software topic you can submit your talk until 12. December 2010.

The conference allows the discussion and exchange of experts and participants.

Silverlight 3 Trap : OneWayToSource

2010-08-11T10:54:00.001+01:00

Silverlight is missing some details you may be using in WPF. One of them is the missing data binding mode OneWayToSource. Another one is the missing SelectionChanged command of a ComboBox. Both created a funny behavior in my current project.

I have a ViewModel property

public string Selected
{
    get
    {
        return this._selected;
    }
    set
    {
        this._selected = value;
        this.RaisePropertyChanged("Selected");
    }
}

and two ComboBox TwoWay (OneWayToSource does not exist) binding on it with SelectedItem

The funny thing was, that when I made a selection on combo box c1 the command handler of my c2 fired (too) – due to the this.RaisePropertyChanged("Selected") above:

- c1 copies the SelectedItem from the target to the source

- the source property raises the property changed event and

- copies the Selected property value to the other c2 target

- c2 fires the SelectionChanged command

Et voilà, your in the Guatemala City trap…it would be better to have more fine-grained data binding control with OneWayToSource in the first place.

Silverlight 3 Debugging in Visual Studio 2008: Breakpoint not hit

2010-07-02T11:27:00.001+01:00

Today I hit the problem, that I could not any longer debug Silverlight client applications (it worked yesterday).

A lot of solutions have been posted out there, which did not help for me. The most common cause with other people was that the web project had the Silverlight Debugger option unticked. So enable it again.

Going back to the source code from yesterday and doing it again, showed me the following screen (“Do you want to enable Silverlight debugging for this project? …” – event when this tick was set!

After some time struggling around I found out, that I had to set the “Set as Start Page” again on my .htm page. I looks like this fixes something (btw, even when the start project was set correctly before!).

Entity Framework 4 : Cascaded Delete in Parent-Child Relationship?

2010-05-21T15:55:00.001+01:00

Extending the previous post, let’s look at cascaded delete

[TestMethod]
[ExpectedException(typeof(System.Data.ObjectNotFoundException))]
public void DeleteParentEntity_WithOneToOneChildEntityRelationshipConstraint_ChildIsAutomaticallyDeleted()
{
    // save parent and child (1:1)
    FinanceMinister financeMinister = new FinanceMinister
    {
        Name = "Giulio Tremonti",
        Email = "gt@finanze.it"
    };
    Country country = new Country
    {
        Name = "Italy",
        GDP = 2118.264M, // billions ? 
        ForeignDebt = 0M,
        FinanceMinister = financeMinister
    };
    this.target.Add(country);
    this.target.Save();
    int financeMinisterId = financeMinister.Id;
 
    // check if that was done right
    Country reloadedCountry = this.target.GetById<Country>(new Country { Id = country.Id });
    Assert.AreEqual<string>(country.FinanceMinister.Name, reloadedCountry.FinanceMinister.Name);
 
    // delete Country and check that treasurer is also gone
    this.target.Delete(country);
    this.target.Save();
    this.target.GetById<FinanceMinister>(new FinanceMinister { Id = financeMinisterId });
}

throws

Now setting the relationship to “cascade”

Runs OK, deleting the country and the corresponding finance minister in one go:

Entity Framework 4 : Delete child from ParentChild Relationship – throws or what?

2010-05-21T15:00:00.001+01:00

[topic from a discussion; addressing a colleague]

Imagine the very common situation where you have a parent-child relationship like the following, and you want to delete the child (Currency; or better just remove the relation by setting the parent’s child navigation property to null):

Having code (show in an intergration test) like this

[TestMethod]
public void RemoveChildEntityNavigationFromParentEntityProperty_WithManyToOneConstraint_Throws()
{
    // save parent and child with a 1:* relationship constraint
    int dbGeneratedCityId = -1;
    Currency chf = new Currency { Symbol = "CHF", Name = "Schweizer Franken" };
    City city = new City { Code = "ZRH", Name = "Zürich", Currency = chf };
    this.target.Add(city);
    this.target.Save();
    dbGeneratedCityId = city.Id;
 
    // get the parent again from the database and get rid of the child
    City reloadedCity = this.target.GetById<City>(new City { Id = dbGeneratedCityId });
    Assert.IsNotNull(reloadedCity.Currency);
    reloadedCity.Currency = null;
    this.target.Save();
 
    // assert that the child was properly deleted
    City reloadedUpdatedCity = this.target.GetById<City>(new City { Id = dbGeneratedCityId });
    Assert.IsNull(reloadedUpdatedCity.Currency);
}
 

Using SQLProfiler shows that the following queries are emitted for the first code block

Calling Save() in the second block throws the following exception:

Aha, sure, this is due to the “1” in the “1:*” relationship – no cities without a default local currency (not even south european EU contries … sorry, no, not funny).

Let’s change this to “0..1”

Running the same test again shows “green”

emitting the Update statement.

Please note:

depending on the constraint, there may be no Delete(chf) necessary
relationship constraints violations throw exceptions
simply set the parent’s child navigation property to null
loosening the constraint to “0..1” leaves the child entity in the database

In comparison we can do the following (with “0..1” relationship):

[TestMethod]
public void DeleteChildEntityFromParent_WithManyToOneConstraint_Throws()
{
    // save parent and child with a 0..1:* relationship constraint
    int dbGeneratedCityId = -1;
    Currency euro = new Currency { Symbol = "EUR", Name = "Euro" };
    City city = new City { Code = "ATH", Name = "Athens", Currency = euro };
    this.target.Add(city);
    this.target.Save();
    dbGeneratedCityId = city.Id;
 
    // get the parent again from the database and get rid of the child
    City reloadedCity = this.target.GetById<City>(new City { Id = dbGeneratedCityId });
    Assert.IsNotNull(reloadedCity.Currency);
    this.target.Delete(reloadedCity.Currency);  // Athens and no more Euro
    this.target.Save();
 
    // assert that the child was properly deleted
    City reloadedUpdatedCity = this.target.GetById<City>(new City { Id = dbGeneratedCityId });
    Assert.IsNull(reloadedUpdatedCity.Currency);
}

and get

Setting the constraint back to “1:*” throws too:

DeleteChildEntityFromParent_WithManyToOneConstraint_Throws threw exception:
System.Data.UpdateException: A relationship from the 'CityCurrency' AssociationSet is in the 'Deleted' state. Given multiplicity constraints, a corresponding 'City' must also in the 'Deleted' state.

Guessing the same behavior for NHibernate. Can anybody confirm this?

Public Interfaces of WCF in Silverlight 4

2010-05-17T13:15:00.001+01:00

Microsoft seems to have added some more interfaces to System.ServiceModel in Silverlight 4 (came across a blog post from Dominick Baier). These can be used as WCF extensibility points.

Here is a list of all the public interfaces I found reflecting the assembly:

System.ServiceModel.IDefaultCommunicationTimeouts
System.ServiceModel.ICommunicationObject
System.ServiceModel.Channels.IChannel
System.ServiceModel.Channels.IChannelFactory
System.ServiceModel.Channels.IChannelFactory`1[TChannel]
System.ServiceModel.Channels.IInputChannel
System.ServiceModel.Channels.IOutputChannel
System.ServiceModel.Channels.IDuplexChannel
System.ServiceModel.Channels.ISession
System.ServiceModel.Channels.IInputSession
System.ServiceModel.Channels.IOutputSession
System.ServiceModel.Channels.IDuplexSession
System.ServiceModel.Channels.ISessionChannel`1[TSession]
System.ServiceModel.Channels.IDuplexSessionChannel
System.ServiceModel.Channels.IInputSessionChannel
System.ServiceModel.Channels.IOutputSessionChannel
System.ServiceModel.Channels.IRequestChannel
System.ServiceModel.Channels.IRequestSessionChannel
System.ServiceModel.Channels.IHttpCookieContainerManager
System.ServiceModel.Description.IEndpointBehavior
System.ServiceModel.IExtensionCollection`1[T]
System.ServiceModel.Description.IContractBehavior
System.ServiceModel.IExtensibleObject`1[T]
System.ServiceModel.IExtension`1[T]
System.ServiceModel.Description.IOperationBehavior
System.ServiceModel.IContextChannel
System.ServiceModel.IClientChannel
System.ServiceModel.Dispatcher.IClientOperationSelector
System.ServiceModel.Dispatcher.IParameterInspector
System.ServiceModel.Dispatcher.IClientMessageFormatter
System.ServiceModel.Dispatcher.IClientMessageInspector
System.ServiceModel.Dispatcher.IDispatchMessageFormatter
System.ServiceModel.Channels.IMessageProperty

generated (without Reflector) using

IList<Type> types = typeof(IClientMessageInspector).Assembly
    .GetTypes()
    .Where(t => t.IsInterface && t.IsPublic)
    .ToList();
 
I have not tried them, but the bold ones are interesting for WCF extensibility.

Exclude from Code Coverage (Visual Studio 2010)

2010-05-03T19:45:00.001+01:00

I was working on a library that adds some customer specific functionality to web services and WCF. I decided to have it very highly covered by unit tests. Unfortunately there is some static functionality in WCF (OperationContext) that is hard to be used in unit tests (it throws). So I wrapped this code behind a custom C# interface with a production implementation calling OperationContext and an (independant and fast) unit test implementation/mock. The small and simple production implementation had some negative impact on Visual Studio (instruction) code coverage (being below 100%). I figured out how to exclude this piece of code.

The way to do this is to add the [ExcludeFromCodeCoverage] attribute to class, property, method or event:

namespace MyTrials
{
    // [ExcludeFromCodeCoverage]
    public class NotCoveredClass
    {
        [ExcludeFromCodeCoverage]
        public string NotCoveredProperty { get; set; }
        public string CoveredProperty { get; set; }
 
        [ExcludeFromCodeCoverage]
        public EventHandler<EventArgs> OnNotCoveredEvent { get; set; }
        public EventHandler<EventArgs> OnCoveredEvent { get; set; }
 
        [ExcludeFromCodeCoverage]
        public void NotCoveredMethod()
        {
        }
 
        public void CoveredMethod()
        {
        }
    }
 
    public class CoveredClass
    {
    }
}

Write some more useful unit tests than I did for this post:

[TestMethod]

public void CodeCoverage_ExcludeWith_ExcludeFromCodeCoverage_FromClassMethodPropertyEvent()

{

CoveredClass target = new CoveredClass();

NotCoveredClass target2 = new NotCoveredClass();

target2.CoveredMethod();

}

to have it excluded:

Other ways I’ve seem were to add [System.Diagnostics.DebuggerHidden] or [System.Diagnostics.DebuggerNonUserCode] to methods – but with some side effects!

Learning Tests

2010-04-12T22:30:00.001+01:00

Robert C. Martin brought it into my world in his book “Clean Code – A Handbook of Agile Software Craftmanship” - “Learning tests”.

“Learning tests” (by K. Beck?)

are unit tests
to explore (learn) a new API or framework
that builds a set of tests for regression when a new version of the component is used to validate that your expectations of the API is still fulfilled

For an upcoming training I wanted to use AutoMapper (by J. Bogard) for entity to/from DTO transformations. There is not a lot of documentation around for this open source library and my first trial with the API ended in a strange exception.

So I decided to write some unit tests first (getting them red was rather easy ;-) and investigate on the common use of AutoMapper (www.google.com/codesearch).

Here are some of my tests (please note the method names which should give you an idea about my expectations):

/// <summary>
/// Learning tests for automapper by Jimmy Bogard from http://automapper.codeplex.com/ 
/// </summary>
[TestClass]
public class AutoMapperTests
{
    [TestInitialize]
    public void TestInitialize()
    {
        Mapper.Reset();
    }
 
    [TestMethod]
    [ExpectedException(typeof(AutoMapperMappingException))]
    public void Map_NoMappingExists_Throws()
    {
        DestinationType destination = Mapper.Map<SourceType, DestinationType>(SourceType.CreateSimple());
    }
 
    [TestMethod]
    public void Map_CreateMap_ReturnsMappingExpression()
    {
        IMappingExpression<SourceType, DestinationType> mappingExpression = Mapper.CreateMap<SourceType, DestinationType>();
 
        DestinationType destination = Mapper.Map<SourceType, DestinationType>(SourceType.CreateSimple());
 
        Assert.IsInstanceOfType(destination, typeof(DestinationType), "must return instance of destination type");
    }
 
    [TestMethod]
    public void Map_CreateMapForOneMember_ReturnsDestinationWithOneMember()
    {
        Mapper
            .CreateMap<SourceType, DestinationType>()
            .ForMember(d => d.Text, c => c.MapFrom<string>(s => s.Text));
        SourceType source = SourceType.CreateSimple();
 
        DestinationType destination = Mapper.Map<SourceType, DestinationType>(source);
 
        Assert.AreEqual<string>(source.Text, destination.Text, "member must be mapped");
    }
 
    [TestMethod]
    public void Map_CreateMapForMembersIgnoreOne_ReturnsDestinationWithoutMember()
    {
        Mapper
            .CreateMap<SourceType, DestinationType>()
            .ForMember(d => d.Text, c => c.MapFrom<string>(s => s.Text))
            .ForMember(d => d.Floating, c => c.Ignore())
            .ForMember(d => d.Number, c => c.MapFrom<int>(s => s.Number));
        SourceType source = SourceType.CreateSimple();
 
        DestinationType destination = Mapper.Map<SourceType, DestinationType>(source);
 
        Assert.AreNotEqual<double>(source.Floating, destination.Floating, "ignored member must not be mapped");
    }
 
    [TestMethod]
    public void Map_CreateMapForChildMembers_ReturnDestinationWithChildMembers()
    {
        Mapper
            .CreateMap<SourceType, DestinationType>()
            .ForMember(d => d.Child, c => c.MapFrom<SourceType>(s => s.Child));
        SourceType source = SourceType.CreateSimple();
        source.Child = SourceType.CreateSimple();
 
        DestinationType destination = Mapper.Map<SourceType, DestinationType>(source);
 
        Assert.AreEqual<string>(source.Child.Text, destination.Child.Text, "child member (reference type) not mapped");
    }
 
    [TestMethod]
    [ExpectedException(typeof(AutoMapperMappingException))]
    public void Map_CreateMapWithoutRecursiveMaps_Throws()
    {
        // map "outer" (root) type 
        Mapper.CreateMap<TopSourceType, TopDestinationType>()
            .ForMember(d => d.DestinationType, c => c.MapFrom<SourceType>(s => s.SourceType));
 
        TopSourceType source = TopSourceType.CreateSimple();
        TopDestinationType destination = Mapper.Map<TopSourceType, TopDestinationType>(source);
 
        Assert.Fail("inner (reference type) map not done. Automapper must throw");
    }
 
    [TestMethod]
    public void Map_CreateMapWithRecursiveMaps_ReturnsTreeDestination()
    {
        // map "inner" type
        Mapper.CreateMap<SourceType, DestinationType>()
            .ForMember(d => d.Text, c => c.MapFrom<string>(s => s.Number.ToString(CultureInfo.CurrentCulture)));
 
        // map "outer" (root) type 
        Mapper.CreateMap<TopSourceType, TopDestinationType>()
            .ForMember(d => d.DestinationType, c => c.MapFrom<SourceType>(s => s.SourceType));
 
        TopSourceType source = TopSourceType.CreateSimple();
        TopDestinationType destination = Mapper.Map<TopSourceType, TopDestinationType>(source);
 
        Assert.AreEqual<string>(
            source.SourceType.Number.ToString(CultureInfo.CurrentCulture), 
            destination.DestinationType.Text,
            "child member (of different reference types) not mapped");
    }
 
    // … and more
}

And here are my tests classes:

/// <summary>
/// Tests-only source type for mapping tests.
/// </summary>
internal class SourceType
{
    public int Number { get; set; }
    public string Text { get; set; }
    public double Floating { get; set; }
 
    public SourceType Child { get; set; }
 
    public static SourceType CreateSimple()
    {
        return new SourceType { Floating = 1.1, Number = 13, Text = Guid.NewGuid().ToString() };
    }
}
 
internal class TopSourceType
{
    public int? Id { get; set; }
    public SourceType SourceType { get; set; }
 
    public static TopSourceType CreateSimple()
    {
        return new TopSourceType { Id = 13, SourceType = SourceType.CreateSimple() };
    }
}
 
/// <summary>
/// Tests-only destination type for mapping tests.
/// </summary>
internal class DestinationType
{
    public int Number { get; set; }
    public string Text { get; set; }
    public double Floating { get; set; }
 
    public DestinationType Child { get; set; }
 
    public override string ToString()
    {
        return string.Format(CultureInfo.CurrentCulture, "Number = {0}, Text = {1}, Floating = {2}", this.Number, this.Text, this.Floating);
    }
}
 
internal class TopDestinationType
{
    public string Id { get; set; }
    public DestinationType DestinationType { get; set; }
}

Specification Pattern with Func<>

2010-03-16T14:44:00.001+01:00

Specification is a pattern (Evans/Fowler) that can be used to apply rules for your domain entities:

Create a specification that is able to tell if a candidate object matches some criteria. The specification has a method isSatisfiedBy (anObject) : Boolean that returns "true" if all criteria are met by anObject.

It defines an interface like

public interface ISpecification
{
    bool IsSatisfiedBy<T>(T candidate);
 
    ISpecification And(ISpecification other);
 
    ISpecification Or(ISpecification other);
 
    ISpecification Not();
}

where IsSatisfiedBy() wraps the business logic (not contained in the entity as OO would suggest).

That allows to write business logic code like

price.And(seats).And(slow.Not())

My car specification:

public class CarSpecification : Specifications.CompositeSpecification
{
    public Func<Car, bool> IsSatisfiedFunction { get; set; }
 
    public CarSpecification(Func<Car, bool> isSatisfiedFunction)
    {
        this.IsSatisfiedFunction = isSatisfiedFunction;
    }
 
    public override bool IsSatisfiedBy<T>(T candidate)
    {
        if (!(candidate is Car))
        {
            throw new ArgumentException("must pass Car entity", "candidate");
        }
 
        if (this.IsSatisfiedFunction != null)
        {
            return this.IsSatisfiedFunction(candidate as Car);            
        }
        return false;
    }
}

The whole code:

public partial class MainWindow : Window
{
    private IList<Car> cars;
    private CarSpecification price = new CarSpecification(c => c.Price <= 40000);
    private CarSpecification seats = new CarSpecification(c => c.Seats >= 7);
    private CarSpecification slow = new CarSpecification(c => c.HorsePower < 140);
 
    public MainWindow()
    {
        InitializeComponent();
 
        this.cars = ApplyRules(Cars.Get());
        this.DataContext = this.cars;
    }
 
    private IList<Car> ApplyRules(IList<Car> cars)
    {
        ISpecification myCarSpecification = price.And(seats).And(slow.Not());
        return cars.Where(c => myCarSpecification.IsSatisfiedBy<Car>(c)).ToList();
    }
}

Definitively this simple example could be written in other ways. More importantly you evaluate your new car ….

Linq in a Generic Data Access Layer with EF4

2010-03-16T11:16:00.001+01:00

Introduction

Decoupling a data access layer (DAL) from business or service logic has become mainstream for some time now – not only due to testing (mocking) reasons but also because replacing your ageing data access technology or your database (manufacturer) may be a topic.

The Interface

I’ve created a simple repository pattern (CRUD) kind of interface over the last year or two. The methods of the interface are generic

int Add<T>(T entity);
int Delete<T>(T entity);
int Update<T>(T entity);
IList<T> GetAll<T>();
// and more

The implementation was done with NHibernate 2.1 and recently with Entity Framework 4.0 (EF4).

The Domain Model

I had several domain models and object-relational mapping (ORM) definitions. Here’s a simple model to show things:

A blog or facebook like system that stores Entry objects allows for people to add Comments or just say they like it (Liker).

Using Linq

Linq is a powerful language feature that can be used to submit queries from the layers on top of the DAL. It is data access and database independent because it works on the domain model – so no coupling to a particular technology.

For this reason I provided a Linq entry point to the interface. Shown here is the EF4 implementation:

public IQueryable<T> GetLinq<T>()
{
    ObjectContext context = GetObjectContext();
    return context.CreateObjectSet<T>() as IQueryable<T>;
} 

Or even easier with NHibernate and the NHibernate Linq provider:

public IQueryable<T> GetLinq<T>()
{
    return this.Session.Linq<T>();
}

Now you can program statements like

l = databaseContext.GetLinq<Entry>().Where(entry => entry.Title.Length > 5).ToList();

emitting SQLExpress 2008 T-SQL:

Or more complex

var result = from x in this.databaseContext.GetLinq<Entry>()
             join y in this.databaseContext.GetLinq<Liker>()
             on x.Id equals y.Entry_FK
             select new { x.Title, y.Posted };

producing

Clean and optimized! The above seems to work due to lazy evaluation and loading.

Note that you could use GetAll<T>().Where() kind of code, but this would return all (10000+) records and only filter on the result set – so a performance killer.

Create your own Web Casts with Microsoft Expression Encoder Screen Capture

2010-01-15T16:36:00.001+01:00

Have you ever explained a certain topic to a peer or a live audience and wanted to

record your computer screen
record the audio
save this to a file

for later reuse or another audience?

Then the Microsoft Expression Encoder (3) Screen Capture is an application to look at. Screen Capture is included in Expression Studio 3, Expression Web 3 or the Expression Professional Subscription.

Start the application, enable the microphone (and web cam; no tried yet myself) and press the red dot button.

Select the screen area to capture

Press the red button again to confirm your selection

Pausing recording is done with Ctrl+Shift+F11, stopping with Ctrl+Shift+F12.

An .xesc file is now create. To encode this to a more common format (like mp4 or wmv) Microsoft Expression Encoder 3 can be used:

S’Zockerpapier het (ned) abgschlage

2009-11-16T18:37:00.001+01:00

Don’t quote me on this, colleagues at home, readers, or anybody out there … because I do not even know the source of this Swiss-German saying, but the meaning is the following:

something is not as sweet as it has been once

You can apply this to this years PDC give-away:

The traditional bag is missing some potency and this years drinking bottle was made from recyclable plastic only.

Let’s hope this is not foreboding of the upcoming technologies, products and services from Microsoft ;-)

--- Updated ---

Steven Sinofsky must have read my post before his day 2 keynote, where he announced the unbelievable compensation

a Win7, multi-touch, tablet subnote for 5000 attendees. So

things are even sweeter than ever before …

What a way to promote multi-touch adoption with developers!

Windows HPC Server 2008 and how to run your jobs

2009-11-16T16:08:00.001+01:00

A twelve hours flight to Los Angeles is a very long time that can be used in a couple of way. One is sleeping, using the on board entertainment system (see on the right; so not to sleep) or reading.

One job you come across every couple of years (or even month as did I lately) as a Windows developer is running computational work that last for a couple of minutes, hours or event up to days. Microsoft Windows HPC Server may look promising to replace a lot of your distributed runtime infrastructure code. It contains as a major component a job scheduler I will make some notes about here.

Overview

HPC 2008 brings high performance computing in a cluster environment to the Windows platform. A so-called job may contain several tasks, which basically are single executables (a simple sequential program or an already parallel one) you plan to run. There are a lot of ways to (1.) submit your job:

Job management console
CLI & PowerShell
SOA APIs (WCF)
COM APIs
.NET APIs or
WS-HPC Basic Profile (Web service)

Job scheduling (2.) can get quite sophisticated using First-come-first-serve, exclusive scheduling, resource optimizing. Job execution (3.) goes through the states Queued, Running, Finished, Failed or Cancelled. The scheduler decides according to defined requirements which job to run. Failed jobs can get automatically re-run. Jobs can run in parallel on several nodes or on several local CPUs. Tasks are usually not designed to communicate (I may come back to this in another post). Jobs may even get preempted (killed or slowed down to let other jobs be run before).

The head note (which can also be a compute node) acts as the central job management point. Jobs are stored in an underlying SQL-Server (Express?) database and execute under the submitting users account. There exists only one queue for a cluster.

(source: Microsoft)

Job Submission

Jobs get priorities, allocated resources (nodes, CPU and run time), dependency information or node exclusivity upon submission. A job can be submitted from the command line like this:

job new [options] [/f:jobdescriptionxmlfile] [/scheduler:host]

The two primary client user interface tools to use are the Job Management Console and the Administration Console. Jobs defined with the new job wizard can be saved to a job description XML file (that is intended to be used or even automatically generated). Message-Passing-Interface (MPI) executables must be prepended with mpiexec.

(source: Microsoft)

Job input (data) and output (results) can be on the local node or on a file share. Large and static data should be copied to the nodes whereas small and dynamically updated data should be placed centrally. Data should be cut into pieces (fragmented) and run in so-called “sweep”, which is some kind of (for int i = 0; i < 100; i++) – loop index.

Jobs can also be submitted using the .NET API and the Microsoft.Hpc.Server namespace classes Scheduler (to connect and submit), ISchedulerJob job = scheduler.CreateJob(), job.AddTask(t), etc.

Another way is to call the scheduler using Web Services and the HPCPBClient class (hpcbp.CreateAcitivity(jdsl), hpcbp.GetActivityStatuses(), etc.).

Job Scheduling

The most simple scheduling is a FIFO queue. “Backfill” is a method where time windows are used for small jobs to run according to the definitions of the large submitted jobs. The “resource matchmaking” method is done by the scheduler according to compute, network and application requirements. “Job templates” are administrator or system owner predefined jobs that can be used by high performance computing client users.

Microsoft seems to have done a good job when it comes down to security. As mentioned above, jobs are executed under the submitting users account. Credentials are passed during submission and safely (encrypted, etc.) stored to be used when the job needs to run. Credentials are passed by secured .NET remoting channels to the nodes and deleted after the job run.

Picture: The fail-over topology from Microsoft

Please note the HPC 2008 Server supports the MPI Standard by it’s own MS-MPI implementation.

HPC 2008 Server looks reasonable priced at around 500$ per node (as I found out after a quick search). You are getting some Quality-of-Services you probably cannot code yourself for this money. So have a look yourself next time defining IJob, IJobScheduling interfaces in your project.

IE8 vs. Google Codesearch

2009-11-06T15:53:00.001+01:00

Google codesearch (www.google.com/codesearch) is great to search for source code. Since installing Internet Explorer 8 recently there is an error on the page. Who’s the culprit:

Webpage error details

User Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.0; Trident/4.0; GTB6; SLCC1; .NET CLR 2.0.50727; InfoPath.1; .NET CLR 3.5.21022; .NET CLR 3.5.30729; .NET CLR 3.0.30729)
Timestamp: Fri, 6 Nov 2009 14:45:24 UTC

Message: Not implemented

Line: 339
Char: 257
Code: 0
URI: http://www.google.com/codesearch/js/CachedFile/F1A2CB189D0FCB1FF201C42BF6A5447C.cache.js

I found out IE8 compatibility view does the trick:

Google states they are working on it http://groups.google.com/group/google-code-search/browse_thread/thread/587e5f2e521b384d

F# – sharp what?

2009-10-27T16:24:00.001+01:00

I’m really happy with C# as an expressive, powerful and readable language (after recently being somewhat unfaithful with VB.NET). For a customer project doing heavy mathematics I had a look on the suitability of F# for concurrent table- and matrix-oriented math.

(logo from whattofix.com; not found one from Microsoft!?)

Overview

On the Microsoft F# Developer Center I got a first overview (e.g. F# will be part of Visual Studio 2010).

F# is a functional, compiled but object-oriented language. An interactive console allows for initial coding:

There is no autocompletion, syntax highlighting or intellisense. Errors are quite understandable however.

Or you can create a new Visual Studio project:

There is useful little video on channel9 that gives a first idea of the language. Since Cobol/Lisp I have not felt so “strange” with a new language …

Downloads

F# October 2009 CTP
F# samples (V 0.3) on code.msdn

Installation

The installation for VS2008 took some minutes:

Learning

In the samples there is a solution called Samples101 that you can build and run:

Concurrency

On the development center site there a book chapter about asynchronous and concurrent programming (Apress). Unfortunately this book is based on an older version. I found some slides from Matthew Podwysocki.

The technique I am interested in is called “asynchronous workflows” (a powerful set of techniques for structuring asynchronous programs in a natural way).

// mleder.blogspot.com : my.net
// my first f# concurrent program
 
#light 
 
// import or open namespaces
open System.IO
open Microsoft.FSharp.Control.CommonExtensions
 
// define array of key/value pairs
let people = ["Markus", "F# is cool";
               "Jonas", "eff what?";
               "Aaron", "(doing something different)";
               "Elias", "bla bla bla"]
 
// define a function to be executed asynchonously                       
let myAsync(name:string, comment:string) =
    async {
        System.Console.WriteLine("Waiting {0} ... {1} / {2}", System.Threading.Thread.CurrentThread.ManagedThreadId, name, comment)
        System.Threading.Thread.Sleep 1000        
        System.Console.WriteLine("Waited! {0}", name)
    }
 
// change the number of threads in the pool
// let ok = System.Threading.ThreadPool.SetMinThreads(10,10)
 
// start function on thread pool threads
for nm,url in people do
    Async.Start(myAsync(nm,url))
 
// wait for termination (using the "pass-forward" or "pipe-forward" operator to chain calls)
System.Console.ReadKey() |> ignore

Which seems to run on 2 pool threads (instead of 2 per CPU which would mean 4) !?:

Uncommenting the SetMinThread line above gives a different behavior

F# Summary

a language you have to get used to
steep learning curve (for a C# guy like me)
not widely used yet. Check maintainability and people skills.
suited for certain problem domains (like mathematics, analytics, simulation, …); to be preferred to other CLR languages
building on existing .NET framework assets
well integrated into Visual Studio (even if my often used “go to definition”/F12 does not work)
concurrent and asynchronous operation is an integrated part of the language

ASP.NET MVC 2 Resources

2009-10-14T09:03:00.001+01:00

I’ve been asked about MVC tutorials, blogs, how to learn. Here are some useful links to get started for all of you:

The official Microsoft ASP.NET MVC web site is here under http://www.asp.net/mvc/
As MVC is Open Source Software (OSS) see also codeplex under http://aspnet.codeplex.com/Wiki/View.aspx?title=MVC
There’s the blog of ScottGu (“who lives in Seattle and builds a few products for Microsoft” :-) under http://weblogs.asp.net/Scottgu/
Upcoming features of MVC 2 on dotnetslackers.com from Ben Scheirman here http://dotnetslackers.com/articles/aspnet/A-First-Look-at-ASP-NET-MVC-2.aspx
Sample chapter from Wiley “Professional ASP.NET 3.5” here http://media.wiley.com/assets/1539/15/professionalaspnet35mvc_chapter13.pdf
Great cheat sheets from SADev.co.za under http://www.sadev.co.za/content/aspnet-mvc-cheat-sheets

Have fun!

Unit Testing – A Book you have been waiting for

2009-09-23T07:31:00.001+01:00

Does your project have low test coverage, unreadable test code, high investments in maintaining unit tests? Do you feel writing unit test being awkward and time consuming, questioning the overall benefit, not trusting some of the unit tests? Ever had discussions about fake/stubs/mocks, what is better Record-Replay or Arrange/Act/Assert (triple A) syntax, where to set up your fakes, differences between unit and integration tests, testing best practices, different tools and test frameworks?

No? Working as a hair dresser or what? Then you can easily forget about the following …

I just finished reading Roy Osherove’s (300 pages) book “The Art of Unit Testing” which is an easy but MUST-READ book, I’ve been waiting for a long time now. You do not have to agree with all of his views, but definitively opens up your mind concerning unit testing.

Go an get it. Or if you are a team lead or manager go and get a stack of it. Your investment will pay back soon!

VS MSTEST Test Run Deployment Item Error – DLL not trusted

2009-09-22T16:32:00.001+01:00

I remember having this error for a couple of times already during the last few years – so I write a post of my own about the solution to the following issue:

Running Visual Studio 2008 (MSTEST) unit tests fails with:

Failed to queue test run '2009-09-22 17:11:11': Test Run deployment issue: The location of the file or directory 'sqlite3.dll' is not trusted.

There were some hints about using the caspol tool to add full trust to the assembly which DID NOT WORK FOR ME:

C:\Program Files\Microsoft Visual Studio 9.0\VC>caspol -m -ag 1.2 -url \\localhost\c$\project\tools\* FullTrust

Microsoft (R) .NET Framework CasPol 2.0.50727.4016
Copyright (c) Microsoft Corporation. All rights reserved.

The operation you are performing will alter security policy.
Are you sure you want to perform this operation? (yes/no)
yes
Added union code group with "-url" membership condition to the Machine level.
Success

A better approach was:

1. Remove the reference to the dll from VS
2. Shut down VS
3. Right click the assembly (here the sqlite3.dll) und click the UNBLOCK button.
4. Restart Visual Studio
5. Re-add the reference to the file

(thanks to markrob35)

WLW Syntax Highlighting with CopyAsHtml

2009-09-15T12:44:00.001+01:00

Some of my fellow blogger colleagues at my company hinted using SyntaxHighlighter to post source code. As I am using CopyAsHtml (Visual Studio plugin) for quite some time now and found out about the HTML view in Windows Live Writer

I keep sticking to it. Looks much better again:

   21 private void timer1_Tick(object sender, EventArgs e)
   22 {
   23     Color c = ColorUnderCursor.Get();
   24     this.label1.Text = "Color under cursor = " + c.ToArgb().ToString("x8") + " [hex]";
   25     this.label2.Text = c.ToArgb().ToString() + " [int]";
   26 }

Color under Cursor in C#

2009-09-14T12:46:00.001+01:00

To easily “steal” some styles and colors from a given web app, I wrote the following little application to find out the color under the mouse pointer.

And here’s the code for the major parts:

using System;
using System.Drawing;
using System.Runtime.InteropServices;

namespace ColorUnderPixel
{
    public class ColorUnderCursor
    {
        [DllImport("gdi32")]
        public static extern uint GetPixel(IntPtr hDC, int XPos, int YPos);

[DllImport("user32.dll", CharSet = CharSet.Auto)]
public static extern bool GetCursorPos(out POINT pt);

[DllImport("User32.dll", CharSet = CharSet.Auto)]
public static extern IntPtr GetWindowDC(IntPtr hWnd);

        /// <summary>
        /// Gets the System.Drawing.Color from under the mouse cursor.
        /// </summary>
        /// <returns>The color value.</returns>
        public static Color Get()
        {
            IntPtr dc = GetWindowDC(IntPtr.Zero);

POINT p;
GetCursorPos(out p);

            long color = GetPixel(dc, p.X, p.Y);
            return Color.FromArgb((int)color);
        }
    }

    [StructLayout(LayoutKind.Sequential)]
    public struct POINT
    {
        public int X;
        public int Y;
        public POINT(int x, int y)
        {
            X = x;
            Y = y;
        }
    }
}

[Update] Please note that the hex values must be reversed to represent a RGB value like in a CSS

background-color: #3b5998;

Human Interaction Workflow

2009-09-14T09:32:00.001+01:00

Windows Workflow Foundation 4.0 does not provide explicit framework classes or tools for workflow human interaction (e.g. waiting for a real actor of a workflow to make a decision how to continue).

Approach

Therefore we gave the following approach a try:

Create a workflow with a SQL persistence extension that represent an order confirmation process
Send email to actor, including a link to a ASP.NET application, a URL with request argument (the workflow instance id)
Add a pick activity with two pickbranches, each including a aec.CreateNamedBookmark(<bm>)
Create an ASP.NET application page to confirm a process, loading the workflow instance and calling ResumeNamedBookmark(<bm>)

Here a some details.

Bookmarks

To block the workflow and wait for human interaction we used the concept of “Bookmarks”. A bookmark is a named point in your workflow to unload and later resume. To create a bookmark and unload we created a simple custom activity:

public class CreateBookmark : NativeActivity
{
public InArgument<string> Bookmark { get; set; }

    protected override void Execute(ActivityExecutionContext context)
    {
        context.CreateNamedBookmark(Bookmark.Get(context), new BookmarkCallback(BookmarkCallback));
    }

    private void BookmarkCallback(ActivityExecutionContext executionContext, Bookmark bookmark, object value)
    {
        executionContext.RemoveAllBookmarks();
    }
}

Sending Email

[Credits for this code to Alex]

Before we can unload our workflow we need to send an URL to the human. The link requests a page that resumes our workflow later on.

public class SendMailActivity : CodeActivity
{
    public InArgument<string> From { get; set; }
    public InArgument<string> To { get; set; }
    public InArgument<string> Subject { get; set; }
    public InArgument<string> Body { get; set; }

    protected override void Execute(CodeActivityContext context)
    {
        var fromAddress = new MailAddress(From.Get(context));
        var toAddress = new MailAddress(To.Get(context));

var addresses = new MailAddressCollection();
addresses.Add(toAddress);

        using (var message = new MailMessage(fromAddress, toAddress))
        {
            message.Subject = Subject.Get(context);
            message.Body = Body.Get(context);

            var mail = new SmtpClient("localhost", 25);
            mail.Send(message);
        }
    }
}

Get Workflow Instance Id

We did not find a simple way to get the workflow instance identifier in our workflow. To work around this we created another simple activity:

public class WorkflowIdActivity : CodeActivity<Guid>
{
    protected override void Execute(CodeActivityContext context)
    {
       Result.Set(context, context.WorkflowInstanceId);
    }
}

Pick and Pickbranch

To summarize: we have a workflow that sends a link to a human, persists and unloads. The question is now (the answer has cost me a beer) what WF4 construct to use to continue our workflow; resuming on two parallel branches. We used the pick and pickbranch activities and included a CreateBookmark trigger activity:

<p:Pick>
<p:PickBranch>
    <p:PickBranch.Trigger>
      <w:CreateBookmark Bookmark="[‘Accepted’]" />
    </p:PickBranch.Trigger>
    <e:SendMailActivity Body="[‘Your order has been accepted’]" From="[‘goodboss@email.com’]" Subject="[‘Laptop order’]" To="[‘goodboss@email.com’]" />
</p:PickBranch>
<p:PickBranch>
    <p:PickBranch.Trigger>
      <w:CreateBookmark Bookmark="[‘Rejected’]" />
    </p:PickBranch.Trigger>
    <e:SendMailActivity Body="[‘Your order has been rejected’]" From="[‘badboss@email.com’]" Subject="[‘Laptop order’]" To="[‘badboss@email.com’]" />
</p:PickBranch>
</p:Pick>

The Web Page

Last step is to build a simple web page, including a green and a red button. Note how we parsed the URL argument to get the workflow identifier.

public partial class OrderConfirmationForm : System.Web.UI.Page
{
    static SqlPersistenceProviderFactory persistenceProviderFactory;
    static AutoResetEvent instanceUnloaded = new AutoResetEvent(false);
    private Guid workflowId;

    protected void Page_Load(object sender, EventArgs e)
    {
       string id = this.Context.Request.Params["id"];
         workflowId = new Guid(id);
    }

protected void btnAccept_Click(object sender, EventArgs e)
    {
        ResumeWorkflow("Accepted");
    }

   protected void btnDecline_Click(object sender, EventArgs e)
    {
        ResumeWorkflow("Rejected");
    }

    private void ResumeWorkflow(string bookmark)
    {
        SetupPersistence();

        PersistenceProvider persistenceProvider = persistenceProviderFactory.CreateProvider(workflowId);
        WorkflowInstance instance = WorkflowInstance.Load(new OrderProcessing(), persistenceProvider);
        instance.Extensions.Add(persistenceProvider);

instance.OnUnloaded = () => instanceUnloaded.Set();

instance.ResumeBookmark(bookmark, null);
instanceUnloaded.WaitOne();

ClosePersistence();
}

    static void SetupPersistence()
    {
        persistenceProviderFactory = new SqlPersistenceProviderFactory(@"Database=Instances;Integrated Security=True", false, false, TimeSpan.FromSeconds(60));
        persistenceProviderFactory.Open();
    }

    static void ClosePersistence()
    {
        persistenceProviderFactory.Close();
    }
}

WIF (aka Geneva) Glossary

2009-09-14T09:06:00.001+01:00

Federated claims-based security as implemented in Microsoft Windows Identity Foundation brings a set of terms that are important to understand. Because I did not find a glossary “out there” I put together a list of terms annotated with a definition (copied from other sources like K. Brown or written myself).

[Sorry: partly in German]

Term	Description
Claim	a bit of identity information such as name, email address, age, membership in the Sales role, and so on
Token	A security token is a serialized set of claims that is digitally signed by the issuing authority
STS	builds, signs, and issues security tokens according to the interoperable protocols
RP	an application that relies on claims
R-STS	the .NET Access Control Service features a Resource-STS: that R-STS can be configured for accepting tokens from specific IP-STSs, and perform various claim transformation including emitting authorization claims
IP-STS	Identity provider STS, like Live ID/passport, ADFS, Sun Metro, CA SiteMinder, IBM, Oracle
Issuer	Issuing authority like Domain Controller, Certificate Authority, Ausweisbüro der Gemeinde/Kanton
Rule	Claims transformation rule (input -> output)
Federate	“verbünden”, trust, establish base for SSO
Claims transformation	Values of input claims are transferred to output claim values (copied, constants, from claim type to other claim type). Mapping.
Scope	Container for rules and token settings (like token renewal, etc.).
Claim type	A URN string describing the meaning of a claim. There are standard claim types; you can add your own.
Action claim	Claim representing an ACS right (e.g. servicebus send/listen/etc.)
ClaimsAuthorizationManager	Modul/Handler used by an RP to implement special authorization. Also possible to do other claims transformation/mapping.
ACS	Azure .NET Services component for Identity und Access Control
Service Bus	Transparent, routed WCF web service through the cloud, providing quality of services.

Code Contracts in Visual Studio 2010

2009-09-09T14:59:00.001+01:00

Currently still a Microsoft DevLabs Project (planned to be included in 4.0/VS2010) – Code Contracts is a very powerful tool “to sue your code”.

Definitions

Preconditions: Required state for a method to run successfully. Contract.Requires();
Postconditions: Guarantees condition that will always be true upon completion of a method. Contract.Ensures(Contract.Result<int>() != 0);
Assertions: Contract.Assert(); Contract.Assumer();
Object Invariants: guarantees about conditions that will always be true for the objects. [ContractInvariantMethod]protected void ObjectInvariant()

Code Contracts are inherited (seems correct decision according to OO substitution principle), but preconditions cannot be added, postconditions can. Interface contracts can be applied with attributes, passing a class to specifiy contract.

A basic concept is that your IL code is re-written with CC code (like it or not). Code can be statically checked or at runtime. In Visual Studio there is a new project settings tab “Code Contract”.

Unit Test Timeout

2009-09-09T13:57:00.001+01:00

A nice feature we were just shown is that a mstest [TestMethod] can get a [Timeout(100)] attribute to allow long-running (hanging) tests from blocking your unit test run for too long:

WF 4.0 Persistence

2009-09-09T09:14:00.001+01:00

Day 3 starts with a recap of what we have done so far with WF4, including the

Testing practices

we did. Important here:

isolate dependecies (e.g. web service send/receive activities)
design workflow for testability (like creating testable sub workflows and putting all the untestable plumbing activities into a top-level workflow

Today we will have a look at the new workflow persistence concepts in WF4.0. There are samples to look at.

Workflow persistence

is a must for long-running workflows. An important concept here are bookmarks which are a kind of marker for unloaded/persisted workflow instances that can be resumed. What was formerly known as workflow services are now extensions. A persistence provider creates a provider to persist your instances. There is one provider for SQL Server at the moment as far as we’ve seen.

static SqlPersistenceProviderFactory persistenceProviderFactory;
persistenceProviderFactory = new SqlPersistenceProviderFactory(@"Data Source=.\SQLEXPRESS; AttachDbFilename=|DataDirectory|\SampleInstanceStore.mdf; Integrated Security=True", false, false, TimeSpan.FromSeconds(60));
persistenceProviderFactory.Open();
…
// start, unload, run
WorkflowInstance instance = new WorkflowInstance(myWorkflow);
id = instance.Id;
instance.Extensions.Add(persistenceProviderFactory.CreateProvider(id));
instance.OnIdle = () => IdleAction.Unload;
instance.OnUnloaded = () => instanceUnloaded.Set();
instance.Persist();
instance.Run();
…
// load, resume
PersistenceProvider persistenceProvider = persistenceProviderFactory.CreateProvider(id);
WorkflowInstance instance = WorkflowInstance.Load(activity, persistenceProvider);
instance.Extensions.Add(persistenceProvider);
instance.OnUnloaded = () => instanceUnloaded.Set();
instance.ResumeBookmark(readLineBookmark, input);
…
persistenceProviderFactory.Close()

Code Generation with T4

2009-09-09T07:26:00.001+01:00

The text templating engine (tt or t4) is part of the Visual Studio environment for quite a few version now. It can be used to generate static wrapper classes for workflow in and out arguments.

Static code (text) sections are mixed with dynamic C# code to emit the target file. My 3 minute sample here:

Syntax highlighting comes from tangible engineering VS plugin (a free edition exists).

HelloWorld Workflow Exercise

2009-09-08T10:54:00.001+01:00

The task is creating a workflow that requests for a name (2 to 3 chars long) from the console and prints it, saying hello.

Creating a new VS2010 workflow Flowchart console application crashes first time, but restarting allows continuing.

We create a FlowChart workflow with a DoWhile activity, create a ReadlineActivity

assign the result of the code activity (string) to workflow variable (id):

press “F9” to add breakpoint in workflow designer and run it with “F5”:

Enter name:
l
Enter name:
leder
Enter name:
lem
Hello lem from WF4.0
Press any key to continue . . .

Now, we want to implement best practices and don’t block in an activity, so let’s refactor it to run asynchronous.

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Activities;
using System.Threading.Tasks;

namespace HelloWorldWorkflow
{
    /// <summary>
    /// Asynchronous code activity that does the bookmark plumbing code correctly.
    /// </summary>
    /// <typeparam name="T">Type of the return value.</typeparam>
    public abstract class AsyncCodeActivity<T> : CodeActivity<T>
    {
        /// <summary>
        /// Activity execution method.
        /// </summary>
        /// <param name="context">The code activity context.</param>
        protected override void Execute(CodeActivityContext context)
        {
            var asyncContext = context.SetupAsyncOperationBlock();
            var task = new Task((ac) =>
            {
                var asyncCtx = (AsyncOperationContext)ac;

T ret = InternalExecute(context);

                asyncCtx.CompleteOperation(new BookmarkCallback(InternalCompleteAsyncWork), ret);
            }, asyncContext);
            task.Start();
        }

        /// <summary>
        /// Internal activity execution method.
        /// </summary>
        /// <param name="context"></param>
        /// <returns></returns>
        protected abstract T InternalExecute(CodeActivityContext context);

        /// <summary>
        /// Internal async code activity completion method.
        /// </summary>
        /// <param name="ctx"></param>
        /// <param name="bm"></param>
        /// <param name="ret"></param>
        protected virtual void InternalCompleteAsyncWork(ActivityExecutionContext ctx, Bookmark bm, object ret)
        {
            Result.Set(ctx, ret);
        }
    }
}

and my ReadlineActivity would look like this:

public class ReadlineActivity : AsyncCodeActivity<string>
{
    protected override string InternalExecute(System.Activities.CodeActivityContext context)
    {
        Console.Write("Enter name: ");
        string id = Console.ReadLine();
        return id;
    }
}

Much easier and less error prone!

Workflow Foundation 4.0

2009-09-08T09:08:00.001+01:00

Now getting more technical with Alex, we start learning about the new framework and tools.

But first we do an excursion to other workflow tools existing: K2 blackpearl, skelta BPM.NET (looks interesting, 100% embeddable, designer), Aspouse.Workflow for .NET, AddFlow for .NET (looks depricated?!), Nintex Workflow 2007 (runs only in SharePoint).

(source: msdn.microsoft.com)

What’s new: new designer, new runtime, clearer data flow, new activities (no state-chart style yet, but new FlowChart), more activities, XAML only (no XOML), better WCF integration (WorkflowServiceHost, xamlx files for IIS activated workflows, threading based on ThreadPool, Rules and Expressions (as activities, no stand-alone engine yet, no forward-chaining (yet?), no vocabulary support), different programming models (WorkflowElement and WorkflowElement<T> as base type, with return value), bookmarks (set workflow to named idle state: aec.CreateNamedBookmarkus(“bm”, resumeCallback) and workflowInstance.ResumeBookmark(“bm”);), asynchronous operation: aec.SetupAsyncOperationBlock(); t = new Task(…), t.Start(); asyncContext.CompleteOperation(cb).

Backwards compatiblity: WF3 runs under .NET 4.0 (WF3.0), old activities incompatible with WF4.

Now let’s go hands-on.

Business Modeling

2009-09-08T08:11:00.001+01:00

Day 2 is starting at 8 am with a very short introduction from Manuel into the modeling of business process. This should give all the foundations to dig into Windows Workflow Foundation 4.0 later today.

“BM is the creation of a model of an abstract but real business”. Like our software models, business models only represent a certain view onto a business. There is no just 1 correct model. “Modellieren = Hervorheben + Weglassen”. Models for an IT architect are typically much more detailed than for the business.

Types of diagrams: Business Use Case (BUCD), Activity, Business Analysis (BAM).

Process: Model –> Document –> Verify (iterate). BM Hamburger: Strategy –> Operations –> IT (business goals) and backwards (verification).

A business process: has a goal, input, output, needs resources, has impact.

Model diagram: shows business actors and business workers, swim lanes, activities, sequences, parallelism/joins, decision nodes, start and end node(s)

BPMN example (one possible notation; quite complex and very detailed; not suitable for business people, this is a model diagram for IT specialists):

(source: ebmpl.org)

BPEL example: very exact; executable model

.NET ServiceBus Exceptions

2009-09-07T16:38:00.000+01:00

I would like to collect some common exceptions from the service bus here in this post.

Wrong solution name:

Unhandled Exception: System.ServiceModel.EndpointNotFoundException: No DNS entri
es exist for host mlede.servicebus.windows.net. ---> System.Net.Sockets.SocketEx
ception: No such host is known
at System.Net.Dns.GetAddrInfo(String name)

Wrong password:

Unhandled Exception: System.ServiceModel.FaultException: authN failed: 'mleder'
of PasswordCredential (#<aGuid>)
at System.ServiceModel.Security.IssuanceTokenProviderBase`1.DoNegotiation(Tim
eSpan timeout)
at System.ServiceModel.Security.IssuanceTokenProviderBase`1.GetTokenCore(Time
Span timeout)

Timeout (happens quite often as the CTP cloud hosting infrastructure seems to be quite light weight):

Unhandled Exception: System.TimeoutException: The open operation did not complet
e within the allotted timeout of 00:01:00. The time allotted to this operation m
ay have been a portion of a longer timeout. ---> System.TimeoutException: The so
cket transfer timed out after 00:00:59.9098704. You have exceeded the timeout se
t on your binding. The time allotted to this operation may have been a portion o
f a longer timeout. ---> System.Net.Sockets.SocketException: A connection attemp
t failed because the connected party did not properly respond after a period of
time, or established connection failed because connected host has failed to resp
ond
at System.Net.Sockets.Socket.Receive(Byte[] buffer, Int32 offset, Int32 size,
SocketFlags socketFlags)

Problem with added ConnectionStatusBehavior:

“ConnectionStatusBehavior requires an IChannelListener that exposes IConnectionStatus”

Trying to emit service meta data:

System.ServiceModel.EndpointNotFoundException was unhandled
Message="Could not connect to http://servicebus.accesscontrol.windows.net/sts/username_for_certificate. TCP error code 10060: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond 65.55.54.15:80. "
Source="System.ServiceModel"
StackTrace:
    Server stack trace:
       at System.ServiceModel.Channels.HttpOutput.WebRequestHttpOutput.GetOutputStream()
       at System.ServiceModel.Channels.HttpOutput.Send(TimeSpan timeout)

Cloud not reachable at the moment:

System.ServiceModel.CommunicationException was unhandled
Message="Unable to reach servicebus.windows.net via TCP (818, 828) or HTTP (80, 443)"
Source="Microsoft.ServiceBus"
StackTrace:
at Microsoft.ServiceBus.NetworkDetector.ChooseConnectivityModeForAutoDetect(Uri uri)

Service Bus Meta Data

2009-09-07T16:07:00.001+01:00

The question came up how to emit WCF service meta data for possible clients. According to hhaaii this is not possible with wsHttRelayBinding but only with netTcpRelayBinding. netTcp did not work with svcutil so I tried the following:

<?xml version="1.0" encoding="utf-8" ?>
<configuration>
<system.serviceModel>
    <behaviors>
      <endpointBehaviors>
        <behavior name="EndpointBehavior">
          <transportClientEndpointBehavior credentialType="UserNamePassword">
            <clientCredentials>
              <userNamePassword userName="<solution>" password="<password>" />
            </clientCredentials>
          </transportClientEndpointBehavior>
        </behavior>
      </endpointBehaviors>
      <serviceBehaviors>
        <behavior name="ServiceBehavior">
         <serviceMetadata httpGetEnabled="true" httpGetUrl="meta" />
        </behavior>
      </serviceBehaviors>
    </behaviors>
    <services>
      <service behaviorConfiguration="ServiceBehavior" name="Microsoft.ServiceBus.Samples.EchoService">
        <endpoint behaviorConfiguration="EndpointBehavior" binding="wsHttpRelayBinding"
          contract="Microsoft.ServiceBus.Samples.IEchoContract" />
        <endpoint address="mex" binding="wsHttpRelayBinding" bindingConfiguration=""
          name="MexEndpoint" contract="IMetadataExchange" />
      </service>
    </services>
</system.serviceModel>
</configuration>

The service started. Going to the service bus registry, clicking the service asked for my credentials

but then a Fault was generated:

Giving up, sorry.

Service Bus Registry: public discovery

2009-09-07T14:49:00.001+01:00

One of the components in a SOA environment is the so-called service registry. A registry lists running services. In .NET Services there is a ATOM feed that shows your current publicly listed services:

The echoservice is only listed when a ServiceRegistrySettings behavior is added to the service endpoint:

IEndpointBehavior serviceRegistrySettings = new ServiceRegistrySettings(DiscoveryType.Public);
endpoint.Behaviors.Add(serviceRegistrySettings);

Otherwise it is not publicly visible!

Unfortunately I could not add this registry behavior using app.config configuration. The WCF config editor tool does not even reflect on the behavior in the Microsoft.ServiceBus.dll. I’m not sure if I’m doing something wrong or this is not possible yet with July CTP.

Microsoft Azure .NET Services

2009-09-07T11:15:00.001+01:00

The upcoming series of posts are dedicated to “Herr Hannes”.
Starting today with the Microsoft Azure platform is Mario, 7.9.09.

The main drivers to go into the Cloud are: Operation, Maintenance, Quality-of-Services, Security.

Windows Azure Platform (source blogs.msdn.com)

Introduction

Azure

Microsoft Windows Azure (Platform; compute/store/manage; open for 3rd party hosters)
vs. Azure Services (Live Services, .NET Services, SQL Services, SharePoint Services, Dynamics CRM Services and more)
for application and service hosting (in the cloud)
Pricing & Licensing (2000 VM hours compute, 50 GB storage, 20 GB network throughput / day), then
compute 0.12$/hour, storage = 0.15$, SQL Azure 1 GB 10$/month or 100$/month,
.NET Services 0.15$/100k messages, 0.10$ bandwidth/GB

.NET Services

for .NET, Java, Ruby; REST, SOAP, RSS, ATOM. Focus on application integration and access control
Quality of Services, flexible scale out

Service Bus

Patterns: Service Registry, Connectivity (relay or direct), pub/sub
"abge-space-ter ESB", NAT/firewall traversal, HTTP and TCP (using ports 808, 818, 819, 828)
Addressing: [http|sb]://solution.servicebus.windows.net/project/serviceXy , new schema prefix "sb" for TCP
One-way connection relays,

Service Connectivity (source: msdn.microsoft.com)

Azure is building on existing assets such as WCF. WCF is enhanced in two areas:

bindings (see below) and
behaviors

WCF Bindings and new service bus Relay bindings:

BasicHttpBinding (BasicHttpRelayBinding)
WebHttpBinding (WebHttpRelayBinding)
WSHttpBinding (WSHttpRelayBinding)
WS2007HttpBinding (WS2007HttpRelayBinding)
WSHttpContextBinding (WSHttpRelayContextBinding)
WS2007HttpFederationBinding (WS2007HttpRelayFederationBinding)
NetTcpBinding (NetTcpRelayBinding)
NetTcpContextBinding (NetTcpRelayContextBinding)
n/a (NetOnewayRelayBinding)
n/a (NetEventRelayBinding)

Resources

So let’s do some hands-on. There’s a sample called EchoService in the SDK.

As always with WCF you create a contract:

[ServiceContract(Name = "IEchoContract", Namespace = "http://samples.microsoft.com/ServiceModel/Relay/")]
public interface IEchoContract
{
[OperationContract]
string Echo(string text);
}

implement the contract:

[ServiceBehavior(Name = "EchoService", Namespace = "http://samples.microsoft.com/ServiceModel/Relay/")]
class EchoService : IEchoContract
{
    public string Echo(string text)
    {
        Console.WriteLine("Echoing: {0}", text);
        return text;
    }
}

Configure the self-hosting console application in app.config:

<?xml version="1.0" encoding="utf-8" ?>
<configuration>
<system.serviceModel>
    <services>
      
      <service name="Microsoft.ServiceBus.Samples.EchoService">
        <endpoint contract="Microsoft.ServiceBus.Samples.IEchoContract"
                  binding="netTcpRelayBinding" />
      </service>
    </services>
</system.serviceModel>
</configuration>

Do the self-hosting:

// Determine the system connectivity mode based on the command line
// arguments: -http, -tcp or -auto (defaults to auto)
ServiceBusEnvironment.SystemConnectivity.Mode = GetConnectivityMode(args);

Console.Write("Your Solution Name: ");
string solutionName = Console.ReadLine();
Console.Write("Your Solution Password: ");
string solutionPassword = ReadPassword();

// create the endpoint address in the solution's namespace
Uri address = ServiceBusEnvironment.CreateServiceUri("sb", solutionName, "EchoService");

// create the credentials object for the endpoint
TransportClientEndpointBehavior userNamePasswordServiceBusCredential = new TransportClientEndpointBehavior();
userNamePasswordServiceBusCredential.CredentialType = TransportClientCredentialType.UserNamePassword;
userNamePasswordServiceBusCredential.Credentials.UserName.UserName = solutionName;
userNamePasswordServiceBusCredential.Credentials.UserName.Password = solutionPassword;

// create the service host reading the configuration
ServiceHost host = new ServiceHost(typeof(EchoService), address);

// create the ServiceRegistrySettings behavior for the endpoint
IEndpointBehavior serviceRegistrySettings = new ServiceRegistrySettings(DiscoveryType.Public);

// add the Service Bus credentials to all endpoints specified in configuration
foreach (ServiceEndpoint endpoint in host.Description.Endpoints)
{
endpoint.Behaviors.Add(serviceRegistrySettings);
endpoint.Behaviors.Add(userNamePasswordServiceBusCredential);
}

// open the service
host.Open();

Console.WriteLine("Service address: " + address);
Console.WriteLine("Press [Enter] to exit");
Console.ReadLine();

// close the service
host.Close();

On the client side we are doing a similar thing using the ChannelFactory:

// create the service URI based on the solution name
Uri serviceUri = ServiceBusEnvironment.CreateServiceUri("sb", solutionName, "EchoService");

// create the channel factory loading the configuration
ChannelFactory<IEchoChannel> channelFactory = new ChannelFactory<IEchoChannel>("RelayEndpoint", new EndpointAddress(serviceUri));

// apply the Service Bus credentials
channelFactory.Endpoint.Behaviors.Add(userNamePasswordServiceBusCredential);

// create and open the client channel
IEchoChannel channel = channelFactory.CreateChannel();
channel.Open();

So what was added to WCF for Azure:

the binding (NetTcpRelayBinding)
a cloud rendez-vouz address (“sb://mleder.servicebus.windows.net/EchoService”)
a behavior (TransportClientEndpointBehavior) to pass Azure solution credentials
a behavior for the registry (ServiceRegistrySettings)

Behaviors and credentials above can also be done using config files and the Microsoft Service Configuration Editor tool:

However, the new TransportClientEndpointBehavior has only one property (CrendentialType) that can be edited. Not very useful; hope this will be improved until release.

Blogging with Windows Live Writer

2009-09-03T07:49:00.001+01:00

Not using one of Scot Hanselman’s The Big Ten Life and Work-Changing Utilities is rather provoking. So I decided to give Live Writer a try for blogging (which is infrequent I admit; because time or projects not permitting).

It’s free, it’s WYSIWYG, can insert (well) text,

pictures

maps

columns

Tool	Impression
Live Writer	nice and slim

Finally: new WCF 4 Configuration

2009-08-25T07:26:00.002+01:00

According to Dr. Nick some long awaited improvements concerning WCF configuration come with .NET 4.0:

- short type names for configuration extensions : have you ever used strong naming and assembly versioning in your project (sure you have)? have you ever implemented a WCF extension (maybe you have)? did you ever have to always change your config files with every new project (assembly) version, because you had to use the assembly qualified type name - including version and token - in the WCF configuration (I did)?! Therefore you probably also added your extension in code only. That's why I love this new one.

- Channel factories with custom configuration: up to now you had two options for your WCF configuration (1. config file 2. create proxy the hard way in code). Now there's a third one (3. write your own serializable/persistable config OBJECT and pass it to channel factory).

p&p Acceptance Test Engineering Guide

2009-07-02T12:53:00.002+01:00

Patterns & Practices J.D. Meier and his team just released a beta of a 209 pages paper about Acceptance testing ("...acceptance testing is the planned evaluation of a system by customers and customer proxies to assess to what degree it satisfies their expectations"). Very important topic.

See here.

Windows Mobile 6 with C# (compass control)

2009-07-01T15:53:00.004+01:00

Actually the textual representation did not look to great. So I decided to look for a free .NET mobile compass control - which I did not find.

Let's hack some dirty user control drawing code:

/// <summary>
/// Paints the control.
/// </summary>
/// <param name="e"></param>
protected override void OnPaintBackground(PaintEventArgs e)
{
    base.OnPaintBackground(e);
 
    int penWidth = 5;
    using (Graphics g = this.CreateGraphics())
    using (Pen penBlackThin = new Pen(Color.Black, penWidth - 3))
    using (Pen penBlack = new Pen(Color.Black, penWidth))
    using (Pen penDarkGray = new Pen(Color.DarkGray, penWidth))
    using (Pen penRed = new Pen(Color.Red, penWidth + 3))
    using (Pen penWhite = new Pen(Color.WhiteSmoke, penWidth + 3))
    using (Brush brushDarkGray = new SolidBrush(Color.DarkGray))
    using (Brush brushWhite = new SolidBrush(Color.White))
    using (Brush brushLightGray = new SolidBrush(Color.LightGray))
    using (Brush brushRed = new SolidBrush(Color.Red))
    {
        double radius = (this.ClientRectangle.Width - 50) / 2;
        double radiusInner = (this.ClientRectangle.Width - 70) / 2;
        int xcenter = this.ClientRectangle.Width / 2;
        int ycenter = this.ClientRectangle.Height / 2;
        float halfFont = this.Font.Size / 2;
 
        // g.Clear(this.BackColor);
 
        // draw circles
        g.FillEllipse(brushLightGray, penWidth, penWidth, this.ClientRectangle.Width - 2 * penWidth, this.ClientRectangle.Height - 2 * penWidth);
        g.DrawEllipse(penDarkGray, penWidth, penWidth, this.ClientRectangle.Width - 2 * penWidth, this.ClientRectangle.Height - 2 * penWidth);
        g.DrawEllipse(penBlack, 2 * penWidth, 2 * penWidth, this.ClientRectangle.Width - 4 * penWidth, this.ClientRectangle.Height - 4 * penWidth);
 
        // draw separator lines
        for (int l = 0; l < 16; l++)
        {
            double angle = l * 22.5 + this.Heading;
            double xout = Math.Sin(this.ToRadian(angle)) * radiusInner;
            double yout = Math.Cos(this.ToRadian(angle)) * radiusInner;
            double xin = Math.Sin(this.ToRadian(angle)) * (radiusInner - 2 * this.Font.Size);
            double yin = Math.Cos(this.ToRadian(angle)) * (radiusInner - 2 * this.Font.Size);
            g.DrawLine(penBlackThin, (int)xout + xcenter, (int)yout + ycenter, (int)xin + xcenter, (int)yin + ycenter);
        }
 
        // draw N, W, E, S
        double x = Math.Sin(this.ToRadian(this.Heading)) * radius;
        double y = Math.Cos(this.ToRadian(this.Heading)) * radius;
        g.DrawString("N", this.Font, brushRed, (float)(xcenter - x) - halfFont, (float)(ycenter - y) - halfFont);
        g.DrawString("S", this.Font, brushRed, (float)(xcenter + x) - halfFont, (float)(ycenter + y) - halfFont);
        g.DrawString("W", this.Font, brushRed, (float)(xcenter - y) - halfFont, (float)(ycenter + x) - halfFont);
        g.DrawString("E", this.Font, brushRed, (float)(xcenter + y) - halfFont, (float)(ycenter - x) - halfFont);
 
        // draw needle
        g.DrawLine(penWhite, xcenter, ycenter, xcenter, this.ClientRectangle.Height - 50);
        g.DrawLine(penRed, xcenter, 50, xcenter, ycenter);
 
        // draw digital heading
        g.FillRectangle(brushWhite, (int)(xcenter - 11 * halfFont), (int)(ycenter - 5 * halfFont), (int)(22 * halfFont), (int)(11 * halfFont));
        g.DrawString(string.Format("{0,3:000}°", this.Heading), this.Font, brushRed, xcenter - 4 * halfFont, ycenter - 5 * halfFont);
 
        // draw height above sea level
        g.DrawString(string.Format("{0,4:0000.0}m", this.Height), this.Font, brushRed, xcenter - 10 * halfFont, ycenter);
    }
}

which looks about this on the device:

Needs some adjustments, especially to look great on other screen resolutions too. Maybe later ...

Windows Mobile 6 with C# (GPS)

2009-07-01T15:27:00.002+01:00

It's been some time since my last fun project. So after getting my brand new HTC Touch Diamond 2 I thought about ending this period and digging somewhat into C#.NET mobile device development.

As I am new to this platform I first browsed through the Mobile SDK samples to get an idea about the slighty restricted framework capabilities. An initial MyApp application was easily created using the older Visual Studio 2005. Changing the main form to my special 800x480 screen resolution did not work properly, so the designer does not represent my device screen.

The fact that my HTC has a built-in GPS device fascinated me very much. So I had a closer look at the GPS application GPS sample and copied some classes into MyApp. Instantiating goes like this:

   55 this.gps = new Gps();
   56 gps.DeviceStateChanged += new DeviceStateChangedEventHandler(gps_DeviceStateChanged);
   57 gps.LocationChanged += new LocationChangedEventHandler(gps_LocationChanged);

getting a GpsPosition class instance with the following major fields:

//** Position + heading related
internal double dblLatitude = 0.0;            // Degrees latitude.  North is positive
internal double dblLongitude = 0.0;           // Degrees longitude.  East is positive
internal float flSpeed = 0.0f;                // Speed in knots
internal float flHeading = 0.0f;              // Degrees heading (course made good).  True North=0
internal double dblMagneticVariation = 0.0;   // Magnetic variation.  East is positive
internal float flAltitudeWRTSeaLevel = 0.0f;  // Altitute with regards to sea level, in meters
internal float flAltitudeWRTEllipsoid = 0.0f; // Altitude with regards to ellipsoid, in meters

After a couple of seconds running the application I got my first position. Output formatted as string was fine ... for the first day ;-)

Team Foundation Error: You are not logged in to Windows Live Messenger

2009-06-12T07:22:00.002+01:00

You may know BizTalk or Sharepoint or PerformancePoint Server to be a cacophony of technologies.

Team Foundation Server seems to be no better, or how do you read this:

;-)

WPF Validation

2009-02-17T11:15:00.004+01:00

Up until recently the mechanisms that Windows Presentation Foundation provided to implement validation were quite unsatisfying.

Starting with .NET 3.0 there was a way to do simple field validation:

<Binding Path="SemanticId" UpdateSourceTrigger="PropertyChanged">
    <Binding.ValidationRules>
        <XyValidationRule/>
    </Binding.ValidationRules>
</Binding>

For most applications this is not enough, however. Usually several fields from an entity are bound to several elements on a form and they have to be validated as a whole to be consistent.

A new approach came with .NET 3.5 where an POCO entity could implement an interface to be used by :

public class Contact : IDataErrorInfo

What I don't like is messing up my entities with validation code. I would prefer a separated (aspect-oriented? and) configurable way.

Rather unnoticed yet another approach came with 3.5 SP1, where you can group several element bindings to one or more entities and define validation rules for this named group:

<Grid.BindingGroup>
    <BindingGroup Name="TrainValidationGroup">
        <BindingGroup.ValidationRules>
            <my:TrainValidationRule ValidationStep="CommittedValue"/>
        </BindingGroup.ValidationRules>
    </BindingGroup>
</Grid.BindingGroup>

It is used in each binding

<TextBox Text="{Binding Path=SemanticId, BindingGroupName=TrainValidationGroup, ValidatesOnDataErrors=true}"/>

Not to forget to explicitely calling

RootElement.BindingGroup.CommitEdit();

Check out the excellent posts I had as source:
MSDN blog
Vincent Sibal

WCF behaviors on callback channel

2009-02-02T08:41:00.002+01:00

Adding behaviors to a WCF channel is widely used and easy.

For an IEndpointBehavior the following code can be used:

public void ApplyDispatchBehavior(ServiceEndpoint endpoint, EndpointDispatcher endpointDispatcher)
{
    // add endpoint behavior to service side
    endpointDispatcher.DispatchRuntime.MessageInspectors.Add(this);
}

A less documented way is to add the same behavior to the callback of a duplex channel:

public void ApplyClientBehavior(ServiceEndpoint endpoint, ClientRuntime clientRuntime)
{
    // add endpoint behavior to callback
    clientRuntime.CallbackDispatchRuntime.MessageInspectors.Add(this);
}

You just need to know ...

Expression evaluation, thinking again

2009-01-30T11:21:00.003+01:00

I recently discussed .NET runtime expression evaluation. Integrating some sample code into your productive applications may have some smells; so this may not be your prefered way to solve simple business logic expression evaluation issues.

Reading a blog from Brad Abrams about Avoiding Custom Delegates explaining

Expression<…> represents function definitions that can be compiled and subsequently invoked at runtime but can also be serialized and passed to remote processes. Continuing with our example:

            Expression<Func<int, int, double>> expression = (x, y) => (double)x / (double)y;
            Func<int, int, double> divide = expression.Compile();
            Console.WriteLine(divide(2, 3));

Microsoft .NET Services (Azure .NET Service Bus) samples and relay bindings

2009-01-30T10:32:00.003+01:00

After some initial issues getting a Microsoft .NET Services invitation code and creating a solution on partially non-working servers I finally wanted to run some samples from the SDK.

Being behind a NAT/firewall infrastructure (no open custom ports other than 80/443) of our organisation (yes, indeed the net.tcp samples all failed starting the service) I tried to get wsHttpRelayBinding sample working.

The service is quite simple:

        static void Main(string[] args)

        {

            string serviceBusSolutionName = GetServiceBusSolutionName();

            Uri address = new Uri(String.Format("http://{0}/services/{1}/EchoService/", ServiceBusEnvironment.DefaultRelayHostName, serviceBusSolutionName));

            ServiceHost host = new ServiceHost(typeof(EchoService), address);

            host.Open();

            Console.WriteLine("Service address: " + address);

            Console.WriteLine("Press [Enter] to exit");

            Console.ReadLine();

            host.Close();

        }

with the following username/password configuration:

<?xml version="1.0" encoding="utf-8" ?>

<configuration>

  <system.serviceModel>

    <behaviors>

      <endpointBehaviors>

        <behavior name="UserNamePasswordCredentials">

          <transportClientEndpointBehavior credentialType="UserNamePassword">

            <clientCredentials>

              <userNamePassword userName="solution" password="password" />  

            </clientCredentials>

          </transportClientEndpointBehavior>

        </behavior>

      </endpointBehaviors>

    </behaviors>

    <bindings>

      <!-- Application Binding -->

      <wsHttpRelayBinding>

        <binding name="default">

          <security mode="None"/>

        </binding>

      </wsHttpRelayBinding>

    </bindings>

    <services>

      <!-- Application Service -->

      <service name="Microsoft.ServiceBus.Samples.EchoService">

        <endpoint name="RelayEndpoint"

                  contract="Microsoft.ServiceBus.Samples.IEchoContract"

                  binding="wsHttpRelayBinding"

                  bindingConfiguration="default" 

                  behaviorConfiguration="UserNamePasswordCredentials"

                  address="" />

      </service>

    </services>

  </system.serviceModel>

</configuration>

Checking the code looks like HTTP communication to me, that should pass our security perimeter infrastructure. Getting the following exception mentioning about net.tcp gives me some thinking.

System.ServiceModel.EndpointNotFoundException was unhandled
Message="Could not connect to net.tcp://servicebus.windows.net:828/services/solution/UserNameAuthenticationService/. The connection attempt lasted for a time span of 00:00:21.0021000. TCP error code 10060: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond 65.55.54.16:828. "
Source="Microsoft.ServiceBus"
StackTrace:
at Microsoft.ServiceBus.RelayedOnewayTcpClient.Connect()
at Microsoft.ServiceBus.RelayedOnewayTcpClient.EnsureChannel()
at Microsoft.ServiceBus.RelayedOnewayTcpClient.OnOpen(TimeSpan timeout)
at Microsoft.ServiceBus.Channels.CommunicationObject.Open(TimeSpan timeout)
at Microsoft.ServiceBus.Channels.CommunicationObject.Open()
at Microsoft.ServiceBus.RelayedOnewayTcpListener.OnOpen(TimeSpan timeout)
...

Using TCP port 828...

Is there some probing going on, even when specifying a HTTP binding? Trying again from home.

C# runtime expression evaluation

2009-01-23T15:32:00.007+01:00

I recently found out about a XAML declarative value converter for WPF which is based on a Visual Studio Linq lambda expression sample called DynamicQuery (Dynamic.cs is bundled with VS2008 as part of \Samples\1033\CSharpSamples.zip).

For a rule evaluation (engine) investigation I created a simple application that accepts a list of expression parameters and calculates the result of a formula:

Takes a couple of [ms] the first time but performs great after the initial hit.

WebBrowser in WPF

2009-01-19T14:01:00.004+01:00

Sketching a very simple RSS reader for a demo today, I just faced the question on how to display HTML in the Windows Forms WebBrowser control in a WPF application.

You just need to wrap it into a WindowsFormsHost element:

<DockPanel Grid.Column="1" Margin="10" TextBlock.FontFamily="Verdana" >

    <TextBlock FontSize="16" FontWeight="Bold" x:Name="RssTitle" DockPanel.Dock="Top"/>

    <WindowsFormsHost>

        <wf:WebBrowser x:Name="RssBrowser" />

    </WindowsFormsHost>

</DockPanel>

Setting the namespace for wf to

xmlns:wf="clr-namespace:System.Windows.Forms;assembly=System.Windows.Forms"

and then call

this.RssBrowser.DocumentText = syndicationItem.Summary.Text;

from code. Does anybode know how to set the font family, by the way?

[ThreadStatic] and ThreadPool threads

2008-12-01T14:49:00.002+01:00

You can use [ThreadStatic] on static fields of a class to store state - per thread.

Please note that using the CLR ThreadPool may re-use a thread and the state is still sticking to that thread.