Erno de Weerd

Trying to get a bank account…

Erno de Weerd — Sat, 05 Sep 2009 01:35:39 PDT

This article got me wondering in what way, if it’s the computer that is posing this limit, the limit has implemented.

First, I don’t think that this is an actual business rule. Why allow 300 Li Juns and not 301? So this might be a ‘rule’ that has been introduced by the IT team. What might have triggered the designers or implementers to invent this rule?

From a database designer or implementer view there is no immediate reason to limit a column combination of first name and last name. Wondering about this made me think about an every day scenario: you enter the bank to deposit some cash. After waiting in line for a while the account manager asks you for your name en he enters the name or a part of it into the application and presses Find to find your account details. The application responds with a list of people that have, more or less, the same name as you do. The account manager asks for another detail (hometown, date of birth, postal code) and picks the right account from the list.

Most UI designers/implementers limit the number of results to prevent excessive database and network load so the list box containing the search results might have a fixed maximum of 300. A number that is not too small and not too large; 300 entries can easily be scanned by someone. Anyone objecting to this limitation will soon find that the designers/implementer will tell them to provide better or more input to get less results back.

Entering Li Jun can not be more specific. So from over 300 Li Juns only the first 300 are reachable in a system like this. Not that it is a good thing but I understand how this came to be. It is quite easy to fix.

I am just guessing of course but this is a perfect example of how each technical decision has its functional effects.

What do you think, are there any other ways how this rule came into place?

This post in in honor of Emil-Ming, may his life be as special to those around him as his name is unique.

Pretty Pictures on High DPI and the WPF/Silverlight promise

Erno de Weerd — Wed, 19 Aug 2009 05:41:21 PDT

Now that Microsoft tries to get us to use our displays to their full potential (read: maximum resolution), we as developers have to know how to write programs that are aware of different DPI settings. Most posts and articles I read about this show nice APIs for Win32 and MFC developers and a short remark that building DPI aware applications comes for free with WPF. I don’t believe this. Let me show you why.

I made two images with one of my favorite paint tools: Paint.NET

While drawing these images I switched off any anti aliasing and smoothing so when we see any aliasing, smoothing or smearing it will not be the image.

Next up: a WPF application.

        1: <Window x:Class="HighDPITest.MainWindow"

       2:     xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation"

       3:     xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml"

       4:     Title="High DPI Test" Height="400" Width="400">

       5:     <Grid>

       6:         <Grid.RowDefinitions>

       7:             <RowDefinition/>

       8:             <RowDefinition/>

       9:             <RowDefinition/>

      10:         </Grid.RowDefinitions>

      11:         <Rectangle Grid.Row="0" Grid.Column="0" 

      12:                              Width="250" Height="50"

      13:                              Stroke="Black" StrokeThickness="1"/>

      14:         <TextBlock Grid.Row="0" Grid.Column="0"

      15:                              HorizontalAlignment="Center" VerticalAlignment="Center"

      16:                              Text="250x50" />

      17:         <Image Grid.Row="1" Grid.Column="0"

      18:                      Source="Picture1.png" Stretch="None"/>

      19:         <Image Grid.Row="2" Grid.Column="0" 

      20:                      Source="Picture2.png" Stretch="None"/>

      21:     </Grid>

      22: </Window>

Result:

As you can see WPF takes the DPI setting of the image into consideration. The 96 DPI image is displayed exactly the way it was drawn. It matches perfectly on my 96 DPI desktop as each dot of the image is mapped exactly on a pixel of the desktop. The second image is smaller and is a bit blurry. It is blurred because WPF skips some dots in the image data; it only needs 96 dots per inch whereas the image contains 120 dots per inch. So when WPF determines the width of the image it calculates 250 / 120 = 2.08 inch ( x 96 = 200 pixels on screen) as opposed to the first image: 250 / 96 = 2.60 inch ( x 96 = 250 pixels on screen)

Let’s change my desktop’s DPI to 120 DPI and see what happens:

First let me explain that this screenshot has been manipulated by me. When you make a screenshot on a 120 DPI screen you end up with a bigger image because there are now 120 dots per inch instead of 96 so in this post I resized the image to make it as big as the first screenshot. The original screenshot can be seen by clicking on the image. (Are you still with me? ;-) )

So what changed? Well it is quite obvious, this time the 120 DPI image is crisp and the 96 DPI image is blurred. The blurring is caused by WPF making up pixels for the desktop because it needs more dots to fill the image on screen than are available in the image data.

Imagine a graphics designer making a logo for you to use in your application and let’s assume he (or she) creates a bitmap image. How would you use the logo in your WPF application?

My guess is that the designer would not like WPF to skip pixels or invent extra pixels but at the same time you would want your users to be able to switch their display’s DPI. You have a couple of options:

Have the designer create the logo at several DPI levels and load the image that matches the display’s DPI. To retrieve the current DPI (yes, a display might have a horizontal DPI that differs from the vertical DPI (to make thing even worse)):

         1: void MainWindowOnLoaded(object sender, RoutedEventArgs e)

         2: {

         3:     Matrix m = PresentationSource.FromVisual(this).CompositionTarget.TransformToDevice;

         4:     double horizontalDPI = m.M11 * 96.0;

         5:     double verticalDPI = m.M22 * 96.0;

         6: }

Write a special control that always renders an image one dot per pixel. Like this. The good is that it will be pixel perfect but its size will be dependent on the display’s DPI: higher DPI –> smaller image and thus the exact opposite of what the user wanted. Users up their DPI to get bigger text and images. This might also be an option for situations where you need pixel perfect images and do not want any anti aliasing or stretching. (Imagine your doctor examining your MRI Scan looking for a tiny blockade in an artery and he can’t be sure whether that pixel is showing a barrier in your brain or it has been interpolated by WPF)
Use vector drawings. They scale quite nicely (don’t overdo it) but you will not always be able to do this because some images look more like photographs and you can’t easily redraw a photograph in vector format.

The browser is another bump in the road. This article gives a nice overview and even explains a bit about Silverlight and Flash.

In WPF it is quite easy to write a DPI aware application as long as you are sticking to vectors (controls and images) but when you need to display bitmaps, you need to plan ahead and do some extra thinking.

Light Up the Web

Erno de Weerd — Thu, 09 Jul 2009 13:37:41 PDT

Version 3 is here!

Microsoft Silverlight 3

Do you play?

Erno de Weerd — Tue, 30 Jun 2009 13:43:41 PDT

A few years ago I advocated the addition of a Minister of IT to our government in a column. One of the main reasons for supporting this (or even being up for the position myself) is the apparent lack of thorough knowledge of information technology amongst our leaders. I still think that IT is so fundamental to our current society that there is no way of justifying having a Minister of Health, Welfare and Sports and not a Minister of IT.

But the times they are a-changing! (read that text when you don’t know it by heart)

According to this article the UK has a member of parliament who at least know a bit about what he is talking about. Besides this being the ultimate excuse for playing games this also made me think about applying this attitude in my own life and work.

When designing and implementing business software we talk to end users to gather feedback on our interpretation of the problem and solution. By doing this we slowly start to understand the end-user and become one ourselves. Especially during manual tests we play (pun intended) ‘end user’.

But aren’t we kidding ourselves this way? Did we ever use the software for real? Did we ever have to do the job of the end user? Did we ever try a software solution created by a third party?

Are we still using the same patterns, tools, libraries, languages, compilers, platforms, operating systems as we did a couple of years ago? If so, WHY? Did we examine new alternatives while being open to switch to these alternatives or are we old dogs with old tricks?

Game on!

Silverlight, Windows Live Writer and WordPress

Erno de Weerd — Sun, 21 Jun 2009 10:04:00 PDT

Update! Please download version 2 It fixes some relative/absolute url problems.

You might have been wondering why on earth I haven’t included any Silverlight demos in this blog. The reason is as simple as it is frustrating: the software that runs this blog has no support for embedding Silverlight applications. Of course we might build this in due time but so far we haven’t.

So I decided to have a look what is needed to allow Silverlight applications to be embedded in blog posts. Here is the short list:

The blogging software/engine must be able to render the correct HTML so the SL application will be in the post when readers download the post. (This raises an interesting issue: are the mainstream blog readers able to display SL apps?!)
The internet server that hosts the SL applications has have the correct mime types registered so the SL application is displayed in the post instead of the user being prompted whether he wants to downloaded the XAP file.
I use Windows Live Writer (WLW) to write my blog posts, I want to be able to insert and upload the SL application using WLW.

Here is what I did on my personal blog (not this one, as it is the corporate site and I am not allowed to mess with that) which uses WordPress. Please do realize that what I did can only be done when you have nearly full access to the server that hosts the blog. You can NOT follow these steps when your blog is hosted by some 3rd party such as WordPress.com.

Adding support for SL to WordPress was easy as pie. Tim made a very nice plugin for WordPress. Installing it was a breeze. Thanks Tim! Tweaks I made:
1. Redirected the plug-in to my site so Tim’s site doesn’t get all the requests for the image that is being displayed when the reader hasn’t got the correct version of Silverlight installed. (It would be nice if Microsoft hosted some cool images for things like this.
2. Not a real tweak but a true hack: I decided to add an extra parameter to the plug-in WITHOUT changing the plug-in’s code. See step 3.
My blog’s host uses a Configuration Panel that allows me quite some advanced tinkering with my website. Therefore adding correct MIME Type support for Silverlight was a breeze.
This is the most complex step so I’ll break it up in chewable bites, bytes or bits:
- Using the experience I gathered while writing a plug-in for WLW and the Community Server Gallery, some good posts on the web and the, at times pathetic, documentation on the MSDN I wrote a plug-in for WLW from scratch. My goal was a plug-in that makes embedding a Silverlight application into a post as easy as inserting an image. I think I got quite close. Here is the UI I designed (I might make it a bit cooler but this is OK for now):
  
  Some quirks I ran into:
  - WLW will only automagically upload a file (the .xap in this case) when the file is referred to in the post by an image tag (<img src=”Bar.xap”/>) or by a link (<a href=”Foo.xap”/>). The marker that is parsed by Tim’s plug-in does not include such a link so the xap didn’t get uploaded. My first hack was to add a hidden hyperlink to the post but that is just plain ugly. So I decided to trick WLW into uploading the xap by adding a hyperlink to the marker. Tim’s plug-in expects a marker that looks like this: [silverlight: URL, WIDTH, HEIGHT, MINVER] e.g. [silverlight: app.xap, 400, 300, foo=bar#second=true#test=1, 2.0.31005.0] So I made the WLW plugin to generate such a marker with a little extra in red: e.g. [silverlight: app.xap, 400, 300, foo=bar#second=true#test=1, 2.0.31005.0, <a style=”display:none” href=”local path to xap”>Fake Link</a>] Tim’s plug-in ignores this so in the post that is downloaded to my dear readers (yes, you are one of them) there is no pollution.
  - WordPress doesn’t like to have xap files uploaded to it by default. So you have to instruct WordPress to accept xap files. While it is not hard to do so, it requires access to the php code of WordPress on your site. Here is what I did: I opened the file named www.yoursitehere.com/wp-includes/functions.php and searched for this function:
    function wp_check_filetype( $filename, $mimes = null ) in my version it was on line 2034 and added one line (in blue):
    'gz|gzip' => 'application/x-gzip',
    'exe' => 'application/x-msdownload',
    'xap' => 'application/x-silverlight-app',
    This step can also be taken by using the mime config plugin for WordPress: just add the xap extension.
  - I haven’t found any way to force WLW to replace a Silverlight app other than by renaming the xap on the client, removing the old Silverlight app from the post and inserting the new version. The only other way I can think of is by logging in to WordPress as an admin and replace the xap file by hand. I might need this functionality because every now and then I make mistakes and need to update my code…

I am quite content with this solution but it would be nice to have some standard way of doing this for other blogging engines and blog post writing software.

You can download the WLW plug-in source code here. I am not going to put it up for download at the WLW plug-in gallery as I do not like their MSI requirement (it is way too complicated and last time I had to use WiX to make one) which is overkill for simply copying a dll. Use the plug-in at your own risk, use the sources anyway you like; when something goes wrong blame anyone but not me. Install it by dropping it in the \Program Files\Windows Live Writer\Plugins\ folder and restart WLW.

Remember: you can only use this out-of-the-box when you are using WordPress and Windows Live Writer, you must be able to install Tim’s plug-in for WordPress, be able to add the MIME Types of your web server (or ask the provider nicely) and finally be able to edit the functions.php of your WordPress installation. Again: all at your own risk. Make backups before adjusting/trashing your blog!

On the good side, this WLW plug-in could easily be customized into generating markers for other blogging server software. Feel free to do so.

Sampling

Erno de Weerd — Thu, 28 May 2009 00:07:29 PDT

Whenever you test it is vital that you pick correct test data. Suppose you need to verify that you implemented an Add method correctly and you test it by passing 2 and 2 to be added. There are many implementations that would respond with 4. Here are a couple:

     return num1 + num2;

    return 4;

    return num1 + num1;

    return num1 * num2;

    return Math.Pow(num1, num2);

    return Math.Pow(num1, 2);

These might be obvious but most methods we implement aren’t that easy to grasp in the blink of an eye.

Now have a look at this:

Does it look familiar? When spending a lot of time coding in Visual Studio the font you use is more important than you might think. There are a lot of coding fonts and it is worth it to put some time in selecting your favorite. So how do you select a coding font? Well, the obvious thing to do is to test them. But what is a valid test?

The easiest way is to switch to the font and start coding and see if the font works for you. Less work is to examine the font a bit better before switching so that is why Visual Studio has a tiny sample in the dialog above:

Unfortunately most font-aware coders don’t care about A, a, B, b, X, x, Y, y, Z and z. Most care about 1, i, I, L, o, O, 0, B, 8, : and ; spotting the difference between 1 and i is getting more and more important due to the dynamic languages. (Remember “Option Explicit”?)

Microsoft felt our pain and changed the sample in the dialog in Visual Studio 2010 (Beta 1):

I have been working with Visual Studio 2010 (I am writing a C# 4.0 + Parallel Programming training) for a while now and I like what they did to the editor and I can’t wait to get my hands on the final release.

In my search for fonts for coders I stumbled on this very cool tool called FontStruct. So if you are feeling creative, create your own font!

What is your favorite coding font?

Browsers, 3D, Google and OpenGL

Erno de Weerd — Tue, 21 Apr 2009 23:38:00 PDT

Google published a new browser plug-in that implements the O3D standard. It is a very early version but the videos and the demos make its use and intent pretty clear. This implementation uses OpenGL.

I expect to see Flash, Silverlight and OD3 competing for bringing hardware accelerated 3D content to the browser. Competition is a good thing so I hope they all do their utmost in getting the user the best experience AND the developers the best tooling.

Some random thoughts:

Browsers and hardware acceleration: that seems a contradiction. I expect some friction there.
I hope someone will implement O3D in DirectX
The browser is becoming a strange thing: a host, a platform, a content renderer. One of these days someone will rip it up in pieces or someone else will integrate it al perfectly. The distinction between an OS, browser, platform and plug-in is getting less and less clear.

Merger

Erno de Weerd — Mon, 20 Apr 2009 23:21:00 PDT

One of the key arguments in the battle between the .NET stack and the Java stack is the "one supplier" versus "multiple suppliers" argument.

Now that Oracle is buying Sun I dug into the press releases and found this little gem in the Frequently Asked Questions:

"How is the transaction expected to benefit partners?
Oracle and Sun partners are expected to benefit by working with a single vendor to address customer needs for enterprise systems."

Sounds familiar?

As most technical people I am primarily interested in what the Java technology stack will look like. And by that I do not mean the obligatory "don't you worry, we'll support all of them" in the FAQ but I want to know what will remain after 1 to 5 years. What will happen to JavaFX, MySQL, GlassFish and NetBeans? I do not expect Oracle to drop them but they might decide to discontinue the innovation of these products. We need to know about that so we can advise our customers when selecting products and technologies.

Light!

Erno de Weerd — Wed, 18 Mar 2009 08:58:48 PDT

A while ago I attended a presentation and in the slideshow I noticed a picture with very interesting attributes. The background of the slide was pure white and the picture had a white background too so there was a perfect blending in of the image however, the object in the image was white as well so all you could see was its shadows. I have been programming and playing with raytracers for long enough (more than 20 years) to recognize a raytraced image pretty quickly but this one was different, it used radiosity causing the shadows to be more real and correctly colored.

To make a long story short: I dived once again into the world of POV-Ray, Moray and radiosity. My goal was create a stock of transparent images by rendering shadows only so the images would fit on ANY background. I am not going to give you my images (just come to a presentation I give to see them) but I am going to give you my recipe.

But first, in honor of the first CTP of Silverlight 3 that is to be released later today, (if you look around on the internet you can already find the bits) I have rendered this using the recipe below (except for the blue colors).

Here is what you do in POV-Ray:

Setup radiosity like this:

     radiosity {

        brightness       2.2

        count            100

        error_bound      0.75

        gray_threshold   0.0

        low_error_factor 0.2

        minimum_reuse    0.015

        nearest_count    10

        recursion_limit  1

        #if (version>3.1)

          adc_bailout      0.01

          max_sample       -1.0

          media            off

          normal           off

          always_sample    1

          pretrace_start   0.08

          pretrace_end     0.04

        #end

      }

Radiosity works best when not explicitly using any lights like point lights, ambient lights or area lights in POV-Ray or any raytracer due to the way radiosity works. What you need is some objects made from materials that emit light (that sounds like an ordinary light but to the raytracer there is a very distinct difference). To get the effect I was looking for I created two materials:

    #declare Pure_White_Light = 

       material  

       {   

          texture

          {      

             pigment

             {

                color rgb <1.0, 1.0, 1.0>

             }      

             finish

             {

                ambient 1.0

                specular 0.7266

                roughness 0.006139

             }

          }

       }

    #declare White = 

       material

       {   

          texture

          {      

             pigment

             {

                color rgb <1.0, 1.0, 1.0>

             }      

             finish

             {

                ambient 0.0

                phong 0.1151

                phong_size 25.163333

                specular 0.769767

                roughness 0.0

                metallic 0.5

             }

          }

       }

The magic is in the ‘finish’ as you can see the Pure_White_Light material emits an ambient light.

Assign the objects you want to use as lights in your scene this emitting material and you are set to go.

One final step:

    background { color <1.000,1.000,1.000> }
  

The image that is being rendered will be rendered to a white background. This will make it possible to add a transparency to the image based on the ‘darkness’ of the image in a image processing tool like Paint.NET

So lets render a cube:

On a white background this appears quite good and subtle. (Remember this is all a matter of taste and this is my blog so I am entitled to give my opinion and you do not have to agree ;-) ) Now let’s get this cube ready for an other background.

Start Paint.NET after installing the Alpha Mask plugin and open the image. First invert the colors of the image.

Next, apply the Alpha Mask plugin and select the same image as the alpha mask and invert the mask.

You’ll end up with the transparent version of the image above:

(The checker pattern is the background that Paint.NET shows when there is nothing to see)

This image can be used on top of any color in a slideshow. Of course I am going to write a plugin for Paint.NET that does all of this in a single keystroke but that’s for later ;-)

How experienced are you?

Erno de Weerd — Tue, 03 Feb 2009 13:09:00 PST

Normally I do not like companies that create user interface controls because they:

Make all applications look the same
Invent stuff that should have been in the framework
Don't think out of the box

(and, to be honest, I simply envy them because it's such a cool job to create them)

Infragistics has some nice controls but they now published a catalog of User Experience (UX) patterns. Nice job!

Are you comfy?

Erno de Weerd — Wed, 21 Jan 2009 23:11:00 PST

Reading an article like A Field Guide to Developers by Joel Spolsky makes me wonder:

What part of this story is 'American' and what could/should be implemented right here, right now?
If you answered the previous question with 'all of it is very American', my guess is you are a manager. If you answered 'we need all of that right here and now', I think you are a recruiter or addicted to being pampered. (Or is that my latent manager alter ego?)
How do these 'environmental conditions' MoSCoW (Must have, Should Have, Could have, Would have) for me?
How do you, dear reader, MoSCoW them?

My 'environmental priorities'
(Read the article to understand what these topics mean)

Must (can't do without)

A spot where I can work without getting interrupted when I don't want to be interrupted.
A clean, bright office but not sterile or lamp lit
Top notch chair and table
Smart colleagues
Independence and autonomy
Work on something interesting
Use cool new tools
Identify with my company
Getting paid competitively

Should (can do without but only after being/getting convinced)

Free reign on Amazon.com
No politics
Flying first class
Pick my own projects
Multiple screens

Could (it would be pleasant to have these but I can do without)

Nice office location

Would (good to have but I wouldn't even notice if it is lacking)

Limo pick up
Being treated like a star

The funny thing is that having all must have would make me feel like 'being treated like a star'.

While assembling this list I was rather shocked to find out what my Musts and Shoulds are. I like to think of myself as someone who isn't hooked to a lot of stuff and adapts easily. However, the way I prioritized was by thinking about the effect the 'thing' would have on my work. My guess is that I would be more productive with these around. The fact that those makes me happy, content and comfy is just proof of the fact that I would be more productive and deliver a higher quality.

Oh, and by the way, I am not telling you which are the ones I have and which I don't: I am not allowing this post to turn into a battle of the employers.
Please consider that when responding with your MoSCoWs ;-)

Crocs in Code

Erno de Weerd — Thu, 08 Jan 2009 11:49:53 PST

Today someone asked me why you should NOT do the following:

Subclass and leave an inherited operation empty, not abstract but empty.

Now, there are very good reasons to give but my wife told me a true story that I liked so much I decided to share it:

Steve Irwin, the crocodile hunter, kept crocodiles in his zoo in cages without doors. He reasoned as follows: if the cages have doors, there will come a day that someone leaves a door open. So he made the zoo keepers jump over the fences to get into the cages.

When you subclass and leave the method empty you create a door in your code and some day, someone will open that door.

It might be completely harmless (the croc just ate) but it could also be the start of a wild chase throught the entire zoo hunting a croc that is looking for dinner.

Do not create potential hazards, Murphy lives just around the corner.

Yes, I know, there are many good reasons sending shivers down your spine that indicate there is something wrong with the code:

why would you subclass if the inherited method does not belong in the subclass?
what is the functional specification and justification of an empty method?

but I loved this story.

Observe!

Erno de Weerd — Fri, 12 Dec 2008 02:03:08 PST

I just noticed something new in the MSDN documentation about the ObservableCollection<T>:

You can enumerate over any collection that implements the IEnumerable interface, and this is adequate for one-time data binding in Silverlight. However, to set up dynamic bindings so that insertions or deletions in the collection can update the UI automatically, the collection must implement the INotifyCollectionChanged interface. This interface exposes the CollectionChanged event, an event that should be raised whenever the underlying collection changes.

Silverlight provides the ObservableCollection<T> class, which is a provided base class data collection that implements the INotifyCollectionChanged interface, as well as the INotifyPropertyChanged interface. It also has the expected collection support, defined by deriving from the Collection<T> class.

That last part about Silverlight adding the INotifyPropertyChanged is pretty important. Previous versions of the MSDN and books about WPF mentioned that the ObservableCollection in fact did raise the PropertyChanged event but I soon found out it didn't (and still doesn't).

Now Silverlight adds this behavior. At first glance that is a good thing but thinking about it made me wonder: we now have a class (generic) in WPF/NET 3.5 and Silverlight with the same name and role but different behavior. That might prove to be a bad thing.

Let's hope WPF/Net 3.5 will quickly follow the Silverlight.

Do you Scrum?

Erno de Weerd — Fri, 05 Dec 2008 07:33:31 PST

I have been reading about Scrum and if you know me even a little you will not be surprised by me giving you this link: Top 10 Pitfalls Using Scrum Methodology for Video Game Development

A very interesting read (including the comments) and again I'd like to note that game development isn't that different from 'regular' line of business application development.

Silverlight and the Mouse(Wheel) part II

Erno de Weerd — Fri, 21 Nov 2008 06:03:00 PST

I wrote about Silverlight and the mouse(wheel) before and the responses are pretty good. One question that stuck to my brain and has been asked in forums is how to actually have controls like the ListBox respond to the MouseWheel event.

First some code

After decorating a ComboBox like this:

   1: <m:MouseSupport MouseWheel="ComboBoxOnMouseWheel">   2:     <ComboBox>   3:         ...   4:     </ComboBox>   5: </m:MouseSupport>

It is pretty easy to add this code:

   1: private void ComboBoxOnMouseWheel(object sender, SilverlightMouseSupport.MouseWheelEventArgs e)   2: {   3:     ComboBox comboBox = sender as ComboBox;   4:     if (comboBox != null)   5:     {   6:         if (e.Delta > 0 && comboBox.SelectedIndex > 0)   7:         {   8:             comboBox.SelectedIndex--;   9:         }  10:         else if (e.Delta < 0 && comboBox.SelectedIndex < comboBox.Items.Count -1)  11:         {  12:             comboBox.SelectedIndex++;  13:         }  14:     }  15: }

But as you will notice when testing this code the ComboBox only responds to the wheel when it is collapsed. I tried adding the decorator to the ItemsPanel but that froze the entire Silverlight application. I see no solution for that.

On to the ListBox

To make a control that uses the ScrollViewer to scroll its content you need to access the ScrollViewer that is in the template of the control. For some reason the GetTemplateChild method was made protected so the only way to get it you need to sub class the control and add a little code:

   1: public class ListBox : System.Windows.Controls.ListBox   2: {   3:     public ScrollViewer ScrollViewer   4:     {   5:         get   6:         {   7:             return GetTemplateChild("ScrollViewer") as ScrollViewer;   8:         }   9:     }  10: }

This ListBox now has a ScrollViewer property that can be used in code like this:

   1: private void ListBoxOnMouseWheel(object sender, SilverlightMouseSupport.MouseWheelEventArgs e)   2: {   3:     //note: this is a special ListBox, not a standard Silverlight listbox.   4:     ListBox listBox = sender as ListBox;   5:     double scrollStep = 10;   6:     if (listBox != null)   7:     {   8:         ScrollViewer scrollViewer = listBox.ScrollViewer;   9:         if (scrollViewer != null)  10:         {  11:             if (e.Delta > 0 && scrollViewer.VerticalOffset > 0)  12:             {  13:                 scrollViewer.ScrollToVerticalOffset(scrollViewer.VerticalOffset - e.Delta * scrollStep);  14:             }  15:             else if (e.Delta < 0 && scrollViewer.VerticalOffset < scrollViewer.ScrollableHeight)  16:             {  17:                 scrollViewer.ScrollToVerticalOffset(scrollViewer.VerticalOffset - e.Delta * scrollStep);  18:             }  19:         }  20:     }  21: }

I have tried to encapsulate this behavior into the ListBox class itself but the problem I ran into is that to make it easy to use I need a way of manipulating the template of the ListBox to add the decorator. It is easy to replace an entire template of a control but I wanted to add the decorator to the existing template in runtime. That is not possible as far as I can see.

I updated the download with this code so you can test and try for yourself.

Concluding

It looks like a hack. And it is. To get this working we need the Silverlight Base Control developers to help us out on this.

Download here.