I was testing this WPF application on different operating systems and I have noticed that on Windows 2003 the access key letters in labels were always shown as opposed to the other operating systems where only by pressing Alt you could see the shortcut keys.

Apparently unrelated, there was another difference between Windows 2003 and other operating systems: clicking on elements would show their focus styling, while on any other, one needed to navigate using the keyboard Tab. This until I noticed that pressing Alt made the focus style appear on controls that I had clicked on.

Yes, the two are related and are linked to an operating system option that apparently cannot be changed on an application level. In Windows XP go to en empty area of the desktop, click Properties, go to the Appearance tab and click on the Effects button. Both the behaviors described above can be enabled or disabled by unchecking or checking Hide underlined letters for keyboard navigation until I press the Alt key. Here is the Microsoft link with instructions on how to do it.

In WPF, removing the visual focus to a control is as simple as setting the FocusVisualStyle property to {x:Null}, however there doesn't seem to be a way to enable or disable access key underscore.

Others have noticed this behavior, and people have suggested using the WM_CHANGEUISTATE Message on the window to set this functionality, by setting/unsetting the UISF_HIDEFOCUS flag. I have not tried it, though.

One can also try to change the setting in the registry. It seems the key is HKEY_CURRENT_USER\Control Panel\Desktop\UserPreferencesMask and if you want to hide the key focus and the label underline until you press alt you should unset the flag 20H from the first byte in the key.

In simple words, it seems there is none. After doing a query like SELECT stuff FROM (SELECT * FROM Table1 UNION ALL SELECT * FROM Table2) x WHERE condition, where Table1 and Table2 are identical in signatures, I noticed that it took 30 seconds on a combined 1 million rows. So I thought I would try to move the condition on each of the UNIONed tables: SELECT stuff FROM (SELECT * FROM Table1 WHERE condition UNION ALL SELECT * FROM Table2 WHERE condition) and it took 2 seconds.

So it seems as the server performed the union on all the rows, perhaps moving them in a temporary table, then filtered on the condition.

Coming from the world of Microsoft Sql Server which carefully creates a query execution tree and routinely solves queries similar to there two in an identical and optimised fashion, I was a bit surprised. Then again, being a Linux MySQL, maybe it was just not compiled right or didn't have an obscure option set up or something like that. Anyway, I was dissapointed.

I haven't posted a music entry in quite some time, but this will compensate. Here are three female singers and some very good songs:

Cosmic Love from Florence and the Machine. You might also want to listen to The Drumming Song



The Girl you Lost to Cocaine from Sia. You might also want to listen to Buttons, with a fun video.



Hollywood from Marina and the Diamonds. She is a very prolific song writer and I like many of her songs. Not to mention she has a voice I love and she's cute as well. You might also want to listen to Mowgli's Road

I started my WPF application and I got this weird exception: Internal error: internal WPF code tried to reactivate a BindingExpression that was already marked as detached. At the time of this entry, googling for this yields about 3 results, only one saying something about how internal WPF errors are usually multithreaded exceptions and, if you have it, try to remove IsAsync="True".

Needless to say, I didn't have that. The StackTrace of the exception was showing me, kind of obliquely, that the error was coming from a binding, but not why. Luckily, I looked in the Visual Studio Output window while getting the error. And there it was! A completely different error, this time telling me what the problem was.

Incidently, the error was caused by a style that had a BasedOn="{x:Type Something}" property setting. I know it is supposed to work, but in this case, it didn't, and it needed the more conservative yet safer BasedOn="{StaticResource {x:Type Something}}".

The cause of the error is only a detail, though, as the lesson here is to always look in the Output window while debugging WPF applications, even if you have an exception thrown. The messages there are more verbose and are actually telling you what bindings do, not what internal code is executed.

There are cases when pages need to use the same session, even if they are started from different contexts. One example is when trying to open a new window from within a WebBrowser control, or maybe issues with the ReportViewer control, or even some browsers who choose to open frames and new windows on different threads, like FireFox did for me a while ago. One might even imagine a situation where two different browsers open the same site and you want to use the same session. You have a SessionID, you are on the same server, so you should be able to use the session you want!

Here is how you do it:

var sessionID=(string)Request.QueryString["SessionIdentifier"];
if (Request.Cookies["ASP.NET_SessionId"] == null 
        && sessionID != null)
{
   Request.Cookies.Add(new HttpCookie("ASP.NET_SessionId", sessionID);
}

This piece of code must be added in the Global.asax.cs file (create a Global.asax file for your web site if you don't have one) in the void Global_PostMapRequestHandler(object sender, EventArgs e) handler and the sessionID must be given in the URL parameter SessionIdentifier.

Unfortunately you can't do it anywhere else. I've seen attempts to abandon the session in page load or page init and then do this, but it doesn't work. Basically, this post describes a horrible hack that replaces the default cookie where ASP.Net saves the SessionID value just before it is read.

As it can be a security risk, you should also add some validation logic so that the session hijacking is done only on certain pages that are likely to be opened in different application threads.

The fifth book in the Steven Erikson's Malazan Book of the Fallen series, Midnight Tides is separated by the previous four in location, characters and, I would say, quality. We are witness to a battle between a lost enclave of the Tiste Edur and a lost enclave of The First Empire. From the perspective of the Malazans (which have no involvement at all in this book) both peoples would have been seen as ignorant savages, their conflict merely a petty squabble. The only characters we can recognize are the Crippled God, who is indirectly manipulating things, and Trull Sengar. Trull is almost the main character in the story, explaining his tortured past, although little connects this story with his freeing from the fragment of the Shadow warren in the forth book.

The end, another convergence of characters and stories and gods and magical powers, only opens avenues for further development, rather than actually explaining things. There are some interesting parts to the story, mostly the description of the Letherii culture, so much alike the Western culture today, which Erikson is criticising at every opportunity. He has similar ideas in House of Chains, but he really lets himself free in this one.

Aside from that and from the history of Trull Sengar which is surely to have an impact in the next books, the story was not really that captivating compared to previous chapters in the saga, almost like it all was a prop to describe Trull's way of thinking and to berate capitalism; like one of those TV show episodes that happen in the past so that we can understand what the character will do in the next episode that happens today. Still a good book, though.

Dear God of the Internet, please grant me this one gift, the perfect way to cancel out the world around me and concentrate on Your work!

I had this scenario where a property was attached to objects but I wanted different default values for different objects. The code made it easy to just check if the property was set or not, then set the default myself. In order to do that, use the DependencyObject.ReadLocalValue instead of DependencyObject.GetValue and then check if the result equals DependencyProperty.UnsetValue. Like this:

return element.ReadLocalValue(MyProperty) == DependencyProperty.UnsetValue;

Of course, this can be used as an extension method for any element and/or property.

I am not a Linux guru, but sometimes I need to use Linux systems and for that I use SSH. It is customary nowadays that one doesn't login remotely using the root account, but rather use another user, then use the command su (super user) to gain root priviledges. I've done it tens of times, most of the time checking if I can log in and then su with the new user.

Well, a few days ago I did not and, to my horror, I noticed that I couldn't use su from my new user, as a permission denied error message popped up. Plus, I had already locked the user I had previously logged in with. Add to this that the device in question had no keyboard/monitor jacks, it was remote, it only had a serial connection, my laptop did not and that the serial cable I tried to use with a borrowed desktop computer was not good enough for this damn device and you can understand the hell I was in.

Enough said, the idea is that some Linux distributions (like the ones based on BSD. Gentoo, for example) use what is known as the wheel group or the group that permits users to use su. Use usermod -G wheel myNewUser to add your user to the wheel group and always ALWAYS check if you have enough permissions for login users before you log off from root.

Short answer: you can't! The color of the text selection (not of the background) is given by an (internal) class System.Windows.Document.CaretElement, inheriting from Adorner. Here is the bit of crappy code that makes this impossible:

protected override void OnRender(DrawingContext drawingContext)
{
  if (this._selectionGeometry != null)
  {
      Brush brush = new SolidColorBrush(SystemColors.HighlightColor);
      brush.Opacity = 0.4;
      brush.Freeze();
      Pen pen = null;
      drawingContext.DrawGeometry(brush, pen, this._selectionGeometry);
  }
}

And you might be thinking that it is a simple template change issue and if you don't add the bloody caret to the template it will work. Wrong! The Caret element is instantiated via code in classes as TextSelection, as an implementation of the ITextSelection interface which only has a getter for the Caret property. The TextSelection class is instantiated in all the controls that support text selection as internal properties or fields. Yay!

So unless you recreate the entire selection functionality, you are stuck with the SystemColors.HighlightColor color for selection. As Dr. WPF himself says: "It is entirely impossible in the current .NET releases (3.0 & 3.5 beta). The control is hardcoded to use the system setting... it doesn't look at the control template at all."

Better luck in WPF 4.0... maybe.

WPF is great for styling. They even included different default styles for each Windows operating system so that the controls look "native". So making an application look as if native to the operating system is not a problem. But how can one do the opposite: force the application to always display as if on a specific operating system?

The solution is to include the default theme for that OS in the resources of your application, window, user control, etc, so one can even style different parts for different operating systems, if need be.

Well, this is how you can make you application use the Vista/Windows7 Aero interface. In you App.xaml add an Application.Resources block and a ResourceDictionary with the source /PresentationFramework.Aero, Version=3.0.0.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35, ProcessorArchitecture=MSIL;component/themes/aero.normalcolor.xaml. If you have already a style in your application resources, add the resource dictionary in a ResourceDictionary.MergedDictionaries block. Like this:

<Application.Resources>
    <ResourceDictionary>
        <ResourceDictionary.MergedDictionaries>
            <ResourceDictionary 
                Source="/PresentationFramework.Aero, 
                        Version=3.0.0.0, Culture=neutral, 
                        PublicKeyToken=31bf3856ad364e35, 
                        ProcessorArchitecture=MSIL;component/themes/aero.normalcolor.xaml" />
            <ResourceDictionary Source="/My.Namespace;component/Themes/MyApplication.xaml" />
        </ResourceDictionary.MergedDictionaries>
    </ResourceDictionary>
</Application.Resources>

The operating system theme files that you can choose from are:

Assembly	Theme file
PresentationFramework.Aero	aero.normalcolor.xaml
PresentationFramework.Classic	classic.xaml
PresentationFramework.Luna	luna.homestead.xaml
PresentationFramework.Luna	luna.metallic.xaml
PresentationFramework.Luna	luna.normalcolor.xaml
PresentationFramework.Royale	royale.normalcolor.xaml

Of course, the method of loading the theme might be through code, and so this can be used to programatically and dynamically change the theme:

protected override void OnStartup(StartupEventArgs e)
{
      Uri themeUri = new Uri(myThemeUriString, UriKind.Relative);
      ResourceDictionary theme = (ResourceDictionary)Application.LoadComponent(themeUri);
      Resources.MergedDictionaries.Add(theme);
   }
   base.OnStartup(e);
}

I was working on a small ASP.Net control that inherited from Button, so nothing special. I went into Visual Studio in design mode and I noticed that the control did not display any resize handles so I could only move the control, not resize it using the mouse.

After intensive Googling, trying custom ControlDesigner classes with AllowResize set to true and all other kinds of strange things, I realised that in the Render method I would write a style tag while in DesignMode before rendering the button. The method was described in a previous blog entry to fix the problem with CSS files with WebResource entries inside and PerformSubstitution.

It appears that Visual Studio, by default, only resizes the controls that render an HTML element that has size. A style tag does not! The solution was to add a div around the control (and the style block) while in DesignMode. The only problem was that the div is a block element, while the input of type button is an inline element. That means that two div elements would place themselves one under the other, not next to the other like two buttons would. So I fixed the problem by also adding a float style attribute and setting it to left.

Here is a small code sample:

protected override void Render(HtmlTextWriter writer)
{
if (DesignMode)
{
  writer.AddStyleAttribute("float", "left");
  writer.RenderBeginTag("div");
  // this renders an embedded CSS file as a style block
  writer.Write(this.RenderDesignTimeCss("MyButton.css"));
}
base.Render(writer);
if (DesignMode)
{
  writer.RenderEndTag();
}
}

Another great Malazan book, the fourth in the series, House of Chains starts with the personal history of Karsa, the Teblor, previously known to us as Shai'k's Toblakai guardian. His tale is both brutal and inspiring, as he evolves from a mindless brute to a ... well... slightly minded brute. At the end of the book there is another battle, like I have been already accustomed by previous reads, only it must be the weirdest one yet. You have to read it to believe it.

Since it started with the singular story of Karsa and because of the many characters that were both introduced, developed from the previous stories or simply clashing at the end, the book felt fragmented (like Raraku's warren :) ). That wasn't so bad, however it opened to many avenues that need to be closed in following books.

At this point it is obvious to me that the Malazan Book of the Fallen is not really a series, but a humongous single story with many interlocking threads and characters. Like chains dragging ghosts of books read, I feel the pressure to end the series so I will probably start hacking away at the fifth book this week.

I was trying to make a web control that would be completely styled by CSS rules. That was hard enough with the browsers all thinking differently about what I was saying, but then I had to also make it look decent in the Visual Studio 2008 designer.

You see, in order to use in the CSS rules (or in Javascript resources, for that matter) files that have been stored as embedded resources in the control assembly, a special construct that looks like <%=WebResource("complete.resource.name")%> is required.

The complete resource name is obtained by appending the path of the embedded resource inside the project to the namespace of the assembly. So if you have a namespace like MyNamespace.Controls and you have an image called image.gif in the Resources/Images folder in your project, the rule for a background image using it is background-image:url(<%=WebResource("MyNamespace.Controls.Resources.Images.Image.gif")%>);.

Also, in order to access the CSS file you need something like [assembly:
WebResource("MyNamespace.Controls.Resources.MyControl.css",
"text/css", PerformSubstitution = true)] somewhere outside a class (I usually put them in AssemblyInfo.cs). Take notice of the PerformSubstitution = true part, that looks for WebResource constructs and interprets them.

Great! Only that PerformSubstitution is ignored in design mode in Visual Studio!!. The only solution to render the control correctly is to render the CSS file yourself and handle the WebResource tokens yourself. Here is a method to do just that:

private static readonly Regex sRegexWebResource =
new Regex(@"\<%\s*=\s*WebResource\(""(?<resourceName>.+?)""\)\s*%\>",
          RegexOptions.ExplicitCapture | RegexOptions.Compiled);

public static string RenderDesignTimeCss2(Control control, string cssResource)
{
  var type = control.GetType(); 
  /*or a type in the assembly containing the CSS file*/


  Assembly executingAssembly = type.Assembly;
  var stream=executingAssembly.GetManifestResourceStream(cssResource);

  string content;
  using (stream)
  {
    StreamReader reader = new StreamReader(stream);
    using (reader)
    {
      content = reader.ReadToEnd();
    }
  }
  content = sRegexWebResource
      .Replace(content,
               new MatchEvaluator(m =>
               {
                 return
                     control.GetWebResourceUrl(
                         m.Groups["resourceName"].Value);
               }));
  return string.Format("<style>{0}</style>", content);
}

All you have to do is then override the Render method of the control and add:

if (DesignMode) {
   writer.Write(
     RenderDesignTimeCss2(
         this,"MyNamespace.Controls.Resources.MyControl.css"
     );
}

Update:
It seems that rendering a style block before the control disables the resizing of the control in the Visual Studio designer. The problem and the solution are described here.

My friend Meaflux mentioned a strange concept called polyphasic sleep that would supposedly allow me to spend less time sleeping, thus maximizing my waking time. I usually love sleep, I can sleep half a day if you let me and I am very cranky when forcefully waken up... as in every day when going to work, doh! Also, I enjoy dreaming and even nightmares. Sure, I get scared and lose rest and there are probably underlying reasons for the horrors I experience at night sometimes, but they are cool! Better than any Hollywood horror, that's for sure. My brain's got budget :)

Anyway, as I get older I understand more and more the value of time, so a method that would give me an extra of 2 to 6 hours a day sounds magical and makes me reminisce of the good times of my childhood when I had time for anything! Just that instead of skipping school I would skip sleep. But does it work?

A quick Google shows some very favourable articles, including one called How to Hack your Brain and the one on Wikipedia, which is ridiculously short and undocumented. A further search reveals some strong criticism as well, such as this very long and seemingly documented article called Polyphasic Sleep: Facts and Myths. Then again, there are people that criticise the critic like in An attack on polyphasic sleep. Perhaps the most interesting information comes from blog comments from people who have tried it and either failed miserably or are extremely happy with it. Some warn about the importance of the sleep cycles that the polyphasic sleep skips over, like this Stage 4 Sleep Deprivation article.

Given the strongly conflicting evidence, my only option is to try it out, see what I get. At least if I suddenly stop writing the blog you know you should not try it and lives will be saved :) Ok, so let's summarise what this all is about, just in case you ignored all the links above.

Most people are monophasic sleepers, a fancy name for people who sleep once a day for about 8 hours (more or less, depending on how draconic your work schedule and responsibilities are). Many are biphasic, that means they sleep a little during the afternoon. This apparently is highly appreciated by "creative people", which I think means people that are self employed and doing well, so they can afford the nap. I know many retired people have a biphasic sleep cycle at least and probably children. Research shows that people normally feel they need to sleep most at around 2:00 and 14:00, which accounts for the sleepiness we feel after lunch. The mid day sleep is also called Siesta.

Now, poliphasic sleep means you reduce your sleep (which in the fancy terminology is called core sleep) and then compensate by having short sleep bursts of around 20 minutes of sleep at as fixed intervals as possible called naps. This supposedly "fixes" your brain with REM sleep, which is the first in the sleep lifecycle, however it is a contested theory. The only sure thing seems to come from an italian researcher called Claudio Stampi who did a lot of actual research and who clearly stated that sleeping many short naps is better than sleeping only once at the same number of hours of sleep. So in other words six 20 minutes naps are better than one 3 hour sleep.

Personally, I believe there is some truth to the method, as many people are actually using it, but with some caveats. Extreme versions like the Uberman (six naps a day, resulting in 2 hours of actual sleep) probably take their toll physiologically, even if they might work for the mental fitness. Also, probably some people are better suited than others for this type of customised sleep cycles. And, of course, it is difficult for a working man to actually find the place and time to nap during the afternoon, although I hear that it has become a fashion of sorts in some major world cities to go to Power nap places and sleep for 20 minutes in special chairs. No wonder New Yorkers are neurotic :) On a more serious yet paranoid note: what if this works and then employers make it mandatory? :-SS

So, in the interest of science, I will attempt this for a while, see if it works. My plan is to sleep 5 hours for the core, preferably from 1:00 to 6:00, then have two naps, one when I get back from work (haven't decided if before or after dinner, as there are people recommending not napping an hour after eating) and another close to 8:30 when I go to work. So far I have been doing it for three days, but it seems all this needs at least a few weeks of adjustment.

Now, with 5 hours and 40 minutes of sleep instead of 7 I only gain 1.33 hours a day, but that means an extra TV show, programming a small utility, reading a lot and maybe even writing... so wish me luck!

Update: I did try it, but I didn't get the support I needed from the wife, so I had to give it up. My experience was that, if you find the way to fall asleep in about 5 minutes, the method works. I didn't feel sleepy, quite the contrary, I felt energized, although that may be from the feeling of accomplishment that the thing actually works :) Besides, I only employed the method during the work week and slept as much as I needed in the weekend. I actually saved about 40 hours a month, which I could use for anything I wanted. If one works during that time, it means an increase in revenue to up to 25%. That's pretty neat.