The http lib can be a bit low level sometimes, so why not make it simpler. Use like so:

var requestBuilder = require('./requestBuilder');

requestBuilder.withHeaders(headers)
.withData(data)
.withPort(80)
.makeRequest('http://google.com', '/', function(body, response){
console.write(body);
});

You have probably already used the Hover intent plugin for JQuery. Maybe you want to do the same thing for Silverlight? Well using the reactive extensions you can quite easily. Of course we are using MVVM so we can capture the elements mouse enter event and bind a command to it on the view model.

You can then bind in the XAML like so:
Button x:Name="Eyecon" commands:MouseEnterIntent.Command="{Binding Open}" commands:MouseEnterIntent.DueTime="50"


The MouseEnterIntent class creates the attached behavior using dependency properties. The MouseEnterIntentCommandBehavior class bind the events to the command.

To migrate the live data into the new structure we used the Entity Framework to generate our old domain from the old database and the new domain from the new database structure. They were imported on different namespaces to determine which model was being dealt with. It was a doddle because it meant mapping the old domain object against the new domain object, then inserting the live data into the new structure.

Saved a lot of time and we weren’t writing SQL queries. What I also liked about this, is the project uses Fluent Nhibernate!

Download

The good news is that there will be full UML support in Visual Studio 2010! Yippee! However for me that will be too late

</cc:CSSCombine>

Download it download

This analogy can be compared to for example, a set of Traffic Lights. You may have seen it sometimes when traffic lights are not functioning correctly, or sometimes you’re in a queue and you’re saying “for god sake, the lights don’t stay green long enough.” So it may stay green for about 5 seconds, which is just enough time for you to pull down the hand brake, accelerate and go. In that time you may get about 5 cars though for example. Now lets say the lights go green for 10 seconds. Now you may expect to get 10 cars through, but you may get 15 through. The time to negotiate the hand brake, accelerating and stopping is less frequent, so we are maximizing throughput. Read on: http://sumo.sourceforge.net/docs/dkrajzew_TRAF9_AgentbasedTLS.pdf

This is the same for negotiating each request in our case. I have seen a lot of different ways of people achieving script combining. Most commonly and a good way is to use a Handler, that returns the combined result. I have a handler that works in a similar way, but it can be implemented in a more friendly manor using a Custom Server Control.

The control has other configurable properties, such as caching, minifying and compressing. They are all switched on by default. The code below will look familiar to the ASP.NET AJAX Script Manager, script combining, but the control I wrote gives you more control over caching, minifying and compressing (using GZip).

It can be implimented like so.



</cc:ScriptCombine>

You can download an example project with the source code: Download

You can script combine in your project by taking the ScriptCombine.ashx and the CustomControls.dll and reference accordingly like the example.

Switch off dynamic compression within IIS. Fortunately, there is a way to enable/disable dynamic compression for certain directories and pages. You can add the below to the <system.webServer> section in the web.config.

<urlCompression doStaticCompression=”true” doDynamicCompression=”true” />

This will enable compression for the entire application. To switch it off just for the web service you can add a location section like so…

  <location path=”MyService.asmx”>

    <system.webServer>

      <urlCompression doStaticCompression=”true” doDynamicCompression=”false” />

    </system.webServer>

  </location>