ASP.NET

Four aspects when evaluating feedback

Tue, 17 Jan 2012 00:03:36 GMT

Feedback is an incredible tool to use for changing something ( a product, a behavior, an repeated action etc.). If not interpreted correctly though, the emerging change might not be always for the best, even if the quality of feedback is great. When we (at Parking Defenders) try to evaluate feedback, we are working from four different angles.

The four aspects

From our prospective, feedback has four (4) different views:

What they are saying
What are you hearing
What are they trying to say
What do you want to hear

An Example

To make things more clear I will use an example. Let’s pretend that I construct pins and one day I decide to change the color of the pins – from White to Dark Gray. I communicate the idea and wait for feedback. Finally I received the following input: Great color! It ’s darker than I expected.

The evaluation

Is it clear? Possibly no. So, I will evaluate it by looking it from the four different angles (that can be converted to questions).

They are saying: Great color! It’s darker than I expected.
I am hearing : Great color!
They are trying to say: It’s should have lighter colors.
I want to hear: Great color!

Is “What I am hearing” the same with “What I want to hear”? Then probably I need to re-evaluate feedback. If that’s not necessary, the product is ready!

Answering the 4 questions is a process that requires both sides (givers and receivers) so it’s essential to try to get answers right after you get the initial feedback. At that time the user is most eager to help you and remembers the best why he gave that feedback at the first place.

My 2 cents!
Cheers!

ASP.NET MVC 3 - A RadioButtonList for Enum Properties

Wed, 04 Jan 2012 08:54:23 GMT

ASP.NET MVC3 doesn’t have a built-in RadioButtonList helper and there are times that I want to display an Enum property. I can always do that using a dropdown list but if there are 2 or 3 values then it’s more user-friendly if the user sees all the options.

So in a simple yes/no question (which I want to force the user to answer so checkbox isn’t really an option) a prefer this:

over this:

So what I did was to create a Enum_RadioButtonList.cshtml file inside Views/Shared/EditorTemplates/ with the following code:

@model Enum
@{
    var listItems = Enum.GetValues(Model.GetType()).OfType<Enum>().Select(e =>
    new SelectListItem()
    {
        Text = e.ToString(),
        Value = e.ToString(),
        Selected = e.Equals(Model)
    });
    string prefix = ViewData.TemplateInfo.HtmlFieldPrefix;
    int index = 0;
    ViewData.TemplateInfo.HtmlFieldPrefix = string.Empty;
    foreach (var li in listItems)
    {
    string fieldName = string.Format(System.Globalization.CultureInfo.InvariantCulture, "{0}_{1}", prefix, index++);
        @Html.RadioButton(prefix, li.Value, li.Selected, new { @id = fieldName })         @Html.Label(fieldName, li.Text)    
    }
    ViewData.TemplateInfo.HtmlFieldPrefix = prefix;
}

And then call it from a page like :

@Html.EditorFor(model => model.EnumProperty, "Enum_RadioButtonList")

So if our Enum was DayOfWeek then we should see something like that:

Of course there are other ways to achive that like using UIHint attribute in your Enum or renaming your file to Enum.cshtml so that every Enum renders that way or even creating an HtmlHelper Extension method.

Please do let me know if the above code is not working in a case.

Cheers!

Edit1: You can check out an improved version by Jason Gable.

Edit2: You can check out an event better version by Jeffrey Tummers that supports .resx files

RazorJS - Source code available!

Wed, 04 Jan 2012 08:53:07 GMT

As of today you can find RazorJS’s source code available on BitBucket. Feel free to download the code and change it to fit your needs. After that you can push any changes you think are useful to this project.

Why BitBucket?

BitBucket uses Mercurial. It has support for issue tracking and wiki but most important (because this is how Mercurial works) it has the ability for a user, to download the source and create a local repository with local-versioning. That means that you can modify the project, have full source-control support and if you do not feel comfortable never push changes to the online repository.
Also BitBucket supports Unlimited Private Repositories which I think is awesome for my projects (as long as I don’t want to collaborate with more that 5 users)!

WP7 Extended ASCII Characters

Wed, 04 Jan 2012 08:52:33 GMT

If you want to display a Greek string that you receive from the Web but all you get are characters like %32, %B6 etc you can do the following

Add this class to your project:

using System;
using System.Linq;
using System.Net;
using System.Reflection;
using System.Text;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Documents;
using System.Windows.Ink;
using System.Windows.Input;
using System.Windows.Media;
using System.Windows.Media.Animation;
using System.Windows.Shapes;
using System.Xml.Linq;

namespace AiaWp7.UI {
    public static class EncoderHelper {
        private static readonly XElement _document = null;
        static EncoderHelper() {
            string path = "Resources/EncodingMap.xml";
            Uri uri = new Uri(path, UriKind.Relative);
            _document = XDocument.Load(Application.GetResourceStream(uri).Stream).Root;
        }

        public static string Decode(string input) {
            if(string.IsNullOrEmpty(input))
                return input;
            StringBuilder sb = new StringBuilder();
            input = input.Replace("%28", "(");
            input = input.Replace("%29", ")");
            input = input.Replace("+", "%+");
            var chars = input.Split(new[] { '%' }, StringSplitOptions.RemoveEmptyEntries);
            foreach (var c in chars) {
                var element = _document.Descendants("Char").FirstOrDefault(x => {
                    var xAttribute = x.Attribute("Encoded");
                    return xAttribute != null && xAttribute.Value == c.Trim(new[] { ' ', '.' });
                });
                if (element == null) {
                    sb.Append(c);
                    continue;
                }
                var attribute = element.Attribute("Value");
                if (attribute != null) sb.Append(c.Replace(c.Trim(), attribute.Value));
            }

            return sb.Replace("+", " ").ToString();
        }
    }
}

Add this file to your Resources folder named EncodedMap.xml (If you don’t have one, create)

<?xml version="1.0" encoding="utf-8" ?>
<Encodings>
  <Char Encoded="B6" Value="Α" />
  <Char Encoded="B8" Value="Ε" />
  <Char Encoded="B9" Value="Η" />
  <Char Encoded="BA" Value="Ι" />
  <Char Encoded="BC" Value="Ο" />
  <Char Encoded="BE" Value="Υ" />
  <Char Encoded="BF" Value="Ω" />
  
  <Char Encoded="C1" Value="Α" />
  <Char Encoded="C2" Value="Β" />
  <Char Encoded="C3" Value="Γ" />
  <Char Encoded="C4" Value="Δ" />
  <Char Encoded="C5" Value="Ε" />
  <Char Encoded="C6" Value="Ζ" />
  <Char Encoded="C7" Value="Η" />
  <Char Encoded="C8" Value="Θ" />
  <Char Encoded="C9" Value="Ι" />
  <Char Encoded="CA" Value="Κ" />
  <Char Encoded="CB" Value="Λ" />
  <Char Encoded="CC" Value="Μ" />
  <Char Encoded="CD" Value="Ν" />
  <Char Encoded="CE" Value="Ξ" />
  <Char Encoded="CF" Value="Ο" />
  <Char Encoded="D0" Value="Π" />
  <Char Encoded="D1" Value="Ρ" />
  <Char Encoded="D3" Value="Σ" />
  <Char Encoded="D4" Value="Τ" />
  <Char Encoded="D5" Value="Υ" />
  <Char Encoded="D6" Value="Φ" />
  <Char Encoded="D7" Value="Χ" />
  <Char Encoded="D8" Value="Ψ" />
  <Char Encoded="D9" Value="Ω" />
  <Char Encoded="DA" Value="Ϊ" />
  <Char Encoded="DB" Value="Ϋ" />

  <Char Encoded="DC" Value="ά" />
  <Char Encoded="DD" Value="έ" />  
  <Char Encoded="DE" Value="ή" />
  <Char Encoded="DF" Value="ί" />
  
  


  <Char Encoded="E1" Value="α" />
  <Char Encoded="E2" Value="β" />
  <Char Encoded="E3" Value="γ" />
  <Char Encoded="E4" Value="δ" />
  <Char Encoded="E5" Value="ε" />
  <Char Encoded="E6" Value="ζ" />
  <Char Encoded="E7" Value="η" />
  <Char Encoded="E8" Value="θ" />
  <Char Encoded="E9" Value="ι" />
  <Char Encoded="EA" Value="κ" />
  <Char Encoded="EB" Value="λ" />
  <Char Encoded="EC" Value="μ" />
  <Char Encoded="ED" Value="ν" />
  <Char Encoded="EE" Value="ξ" />
  <Char Encoded="EF" Value="ο" />
  <Char Encoded="F0" Value="π" />
  <Char Encoded="F1" Value="ρ" />
  <Char Encoded="F2" Value="ς" />
  <Char Encoded="F3" Value="σ" />
  <Char Encoded="F4" Value="τ" />
  <Char Encoded="F5" Value="υ" />
  <Char Encoded="F6" Value="φ" />
  <Char Encoded="F7" Value="χ" />
  <Char Encoded="F8" Value="ψ" />
  <Char Encoded="F9" Value="ω" />


  <Char Encoded="FA" Value="ϊ" />
  <Char Encoded="FB" Value="ϋ" />
  <Char Encoded="FC" Value="ό" />
  <Char Encoded="FD" Value="ύ" />
  <Char Encoded="FE" Value="ώ" />



  <Char Encoded="2F" Value="/" />
  <Char Encoded="30" Value="0" />
  <Char Encoded="31" Value="1" />
  <Char Encoded="32" Value="2" />
  <Char Encoded="33" Value="3" />
  <Char Encoded="34" Value="4" />
  <Char Encoded="35" Value="5" />
  <Char Encoded="36" Value="6" />
  <Char Encoded="37" Value="7" />
  <Char Encoded="38" Value="8" />
  <Char Encoded="39" Value="9" />

  <Char Encoded="28" Value="(" />
  <Char Encoded="29" Value=")" />
  <Char Encoded="+" Value=" " />
</Encodings>

And you are all set!

Android-Binding and GreenDroid

Wed, 04 Jan 2012 07:49:48 GMT

If you want to use android-binding along with GreenDroid’s ActionBarActivity (GDActivity, GDMapActivity etc.) you might stumble into some issues. I know I did.

That’s why I created a Bind method that can be used instead of Binder.setAndBindContentView

Here it is (put it inside a static class):

public static void bind(ActionBarActivity activity, int layoutId, Object viewModel) {
	if (activity == null)
		return;
	FrameLayout contentView = activity.getContentView();
	try {
		Context context = contentView.getContext();
		InflateResult result = Binder.inflateView(context, layoutId, contentView, true);
		Binder.bindView(context, result, viewModel);
	} catch (Exception e) {
		//handle ex
	}
}

Cheers!

Android - Draw a path (array of points) in MapView

Tue, 03 Jan 2012 18:37:27 GMT

Although this post differs from my blog routine I am writing because I haven’t found an efficient way of displaying a list of GeoPoints to a MapView. This is often useful when you want to display directions from Google service inside your Android application.

Key features:

A custom overlay that extends com.google.android.maps.Overlay
Efficient way of displaying a path. The path is drawn over the same overlay.
Smooth Lines, useful when there are a lot of corners like the example below
Easy to use
Easy to extend - Just under 90 lines of code
Available on Bitbucket

Here is a sample result of using my overlay to display a list of points (The green marker is not included).

You can use it like this

List<GeoPoint> path = new ArrayList<GeoPoint>();
//add your points somehow...
mapView.getOverlays().add(new RoutePathOverlay(path));

Here is the complete RoutePathOverlay.java (it's also available on Bitbucket)

import java.util.List;
import android.graphics.Canvas;
import android.graphics.Color;
import android.graphics.Paint;
import android.graphics.Path;
import android.graphics.Point;
import android.graphics.RectF;
import com.google.android.maps.GeoPoint;
import com.google.android.maps.MapView;
import com.google.android.maps.Overlay;
import com.google.android.maps.Projection;

public class RoutePathOverlay extends Overlay {

        private int _pathColor;
        private final List<GeoPoint> _points;
        private boolean _drawStartEnd;

        public RoutePathOverlay(List<GeoPoint> points) {
                this(points, Color.RED, true);
        }

        public RoutePathOverlay(List<GeoPoint> points, int pathColor, boolean drawStartEnd) {
                _points = points;
                _pathColor = pathColor;
                _drawStartEnd = drawStartEnd;
        }

        private void drawOval(Canvas canvas, Paint paint, Point point) {
                Paint ovalPaint = new Paint(paint);
                ovalPaint.setStyle(Paint.Style.FILL_AND_STROKE);
                ovalPaint.setStrokeWidth(2);
                int _radius = 6;
                RectF oval = new RectF(point.x - _radius, point.y - _radius, point.x + _radius, point.y + _radius);
                canvas.drawOval(oval, ovalPaint);               
        }

        public boolean draw(Canvas canvas, MapView mapView, boolean shadow, long when) {
                Projection projection = mapView.getProjection();
                if (shadow == false && _points != null) {
                        Point startPoint = null, endPoint = null;
                        Path path = new Path();
                        //We are creating the path
                        for (int i = 0; i < _points.size(); i++) {
                                GeoPoint gPointA = _points.get(i);
                                Point pointA = new Point();
                                projection.toPixels(gPointA, pointA);
                                if (i == 0) { //This is the start point
                                        startPoint = pointA;
                                        path.moveTo(pointA.x, pointA.y);
                                } else {
                                        if (i == _points.size() - 1)//This is the end point
                                                endPoint = pointA;
                                        path.lineTo(pointA.x, pointA.y);
                                }
                        }

                        Paint paint = new Paint();
                        paint.setAntiAlias(true);
                        paint.setColor(_pathColor);
                        paint.setStyle(Paint.Style.STROKE);
                        paint.setStrokeWidth(5);
                        paint.setAlpha(90);
                        if (getDrawStartEnd()) {
                                if (startPoint != null) {
                                        drawOval(canvas, paint, startPoint);
                                }
                                if (endPoint != null) {
                                        drawOval(canvas, paint, endPoint);
                                }
                        }
                        if (!path.isEmpty())
                                canvas.drawPath(path, paint);
                }
                return super.draw(canvas, mapView, shadow, when);
        }

        public boolean getDrawStartEnd() {
                return _drawStartEnd;
        }

        public void setDrawStartEnd(boolean markStartEnd) {
                _drawStartEnd = markStartEnd;
        }
}

I hope you find it useful!
Cheers!

LinkedIn OAuth Library Update

Wed, 10 Aug 2011 15:21:08 GMT

Version 0.3 of Linked OAuth Library is available via Nuget.

This release has a dependency in Hammock Nuget package.
I have also uploaded the source code at Bitbucket and it’s available to fork it!

Cheers!

RazorJS - Write Razor inside your Javascript files

Tue, 14 Jun 2011 08:52:39 GMT

Update: Full Source code is available at BitBucket

Many times I have a Javascript file that contains a link and in order to be more generic I would prefer writing something like ~/image.png rather than /image.png or /myapp/image.png.

Also when one of my Views in an MVC project interacts with 3rd party components and services (like Google Maps, UserVoice etc.) contains a lot of Javascript code along with html/server-side code. The minute that this page reaches a stable version I tend to put the javascript code in a separate file to keep things cleaner. But if that code depends on my Model and is being rendered by Razor I can’t do that. That means this code must stay on that view and I may end up with duplicating it to other pages with similar functionality.

Struggling with the above I ran into a very good Nuget Package called RazorEngine and thought to create something using that to be able to have more flexibility in my Javascript Code.

And the name of this is RazorJS.

Let’s see a few examples:

Javascript Code in an external file

The JS File

The Razor Code

The Rendered HTML

Javascript Code with Model dependency

The JavaScript File

The Code

The HTML

What about performance?

The JS files are being cached just like Razor files with a CacheDependency so if a file changes the Razor template is being updated instantly. Every file is compiled just like Razor Pages and is being compiled the first time is requested or when a change is detected.

So the time to render each file the second time (and every time after) is really really small.

Known Issues

For now the Nuget package doesn’t have a dependency to RazorEngine because it needs a later version to perform better which isn’t yet available. Once RazorEngine 2.2 is released I will update the package.
For now the only method available to call is Href and the only property is only two properties are Model and Url. Please let me know other methods or properties that could be useful when writing Javascript.

That’s all for now! You can find it at the Nuget Feed!

Cheers!

Native Greek Keyboard for Windows Phone 7

Sun, 05 Jun 2011 08:38:46 GMT

Windows Phone 7 don’t have support for Greek. There is a way to add support for not only Greek but for many more languages using a standalone wp7 app that does some tweaking.

------------------------------------------------------------
READ THIS FIRST:
In order to do that you should be able install standalone .xap files to your phone. If your device is developer unlocked you can do that by using the Application Deployment (if you have installed Windows Phone Tools just search for it in the start menu). If you don’t have a clue on how to do that discard the entire post and wait for the Mango update coming this fall which will add support for Greek.

When installing xap files that do not come from Marketplace a lot of bad things could happen. So it might be necessary to hard reset your phone or worse. You are entirely responsible for destroying your phone or making it unusable as a result of following this post’s directions.

------------------------------------------------------------

Now that we got some things out of the way let’s get started!

In XDA-Developer there is an app that adds keyboard support for many languages (including Greek). Native Keyboard for WP7.

The limitation with this app is that you can’t send an SMS with Greek letters with more that 70 characters.

So after a few tweaks of the app a created a version of Native Keyboard that allows you to send an SMS with 160 characters per message (which is the default size).

Download Native Keyboard with 160-chars SMS support

Known Issues: There is no support for lowercase letters

And here are some screenshots:

PetaPoco and Spatial Data Types

Sat, 14 May 2011 11:15:20 GMT

PetaPoco is a micro-ORM for .NET and Mono. If you haven’t see it yet I suggest you to check it out. Alternatives to PetaPoco are Massive and Dapper. (If you know more projects let me know).

After this brief introduction let me begin by explaining the problem. I need it to build a simple web app with a fairly simple data model that was using Spatial Data. The app would use SQL Server 2008.

Let’s consider a table with a bunch of columns and one of them is a geometry column. I needed to handle that. Entity Framework doesn’t have support for spatial types (it’s coming though so give your feedback) so it was out of the question. Since I was dealing with a simple data model I decided to use a Micro-ORM like PetaPoco.

I fired-up Visual Studio, created a new Web Application and Installed PetaPoco via Nuget.

PetaPoco package has also a few T4 templates that helps you creating classes. There is also a package named PataPoco.Core which contains only the one file that is needed with no T4 templates.

When I tried to create my classes using the T4 template it worked like a charm! For my geometry column an equivalent Microsoft.SqlServer.Types.SqlGeography was created!

Next, I tried to do a query like this:

var location = SqlGeography.Point(lat, lng, 4326);
var ids = db.Query<int>(@"SELECT Id FROM Points WHERE @0.STIntersects(Location) = 1", location.STBuffer(100));

This is were I got the following error:

UdtTypeName property must be set for UDT parameters.

The solution was simple.

Go to PetaPoco.cs
Find the AddParam method
Just before the final else (somewhere around line 406) add the following code:
```
else if (item.GetType() == typeof(bool) && _dbType != DBType.PostgreSQL)
{
    p.Value = ((bool)item) ? 1 : 0;
}/*Already there*/
else if (item.GetType() == typeof(Microsoft.SqlServer.Types.SqlGeography))
{
    p.Value = item;
    ((System.Data.SqlClient.SqlParameter)p).UdtTypeName = "Geography";
}
else if (item.GetType() == typeof(Microsoft.SqlServer.Types.SqlGeometry))
{
    p.Value = item;
    ((System.Data.SqlClient.SqlParameter)p).UdtTypeName = "Geometry";
}
else/*Already there*/
{
    p.Value = item;
}
```

That’s it! Now you can create queries with spatial parameters.

Enjoy!

PS: I hope in the future there will be some kind of extensibility so this fix could be at a separate file.

Why I Still L.O.V.E. ASP.NET WebForms

Fri, 29 Apr 2011 14:00:00 GMT

Objectives

Build Contact Page (a page that interacts with a data-storage)
Build a Contact List Page (a page with a gird)
Do it using ASP.NET WebForms
Create Unit Tests

We will use the MVP pattern and the open-source project WebFormsMVP to accomplish that! Can’t wait? Download the sample!

Ready?

First we open Visual Studio 2010 and create a new ASP.NET Web Application Project like the screenshot.

Then we add a new project to our solution which contains our Application Logic like the screenshot.

Next we fire-up Package Manager Console, set as Default Project .Web and type

Install-Package WebFormsMvp

Repeat the above setting as Default Project .Logic

Now we are ready to create our Database and our Database Access Layer.

Because in this post this is the less important part our DB is a static List and has a table which is represented by the Contact class. Our DAL is the following:

Contact.cs

public class Contact
{
    public string Name { get; set; }
    public string Email { get; set; }
    public string Message { get; set; }
    DateTime? _CreatedAt = null;
    public DateTime CreatedAt
    {
        get
        {
            if (!_CreatedAt.HasValue)
                _CreatedAt = DateTime.Now;
            return _CreatedAt.Value;
        }
    }
}

IContactRepository.cs

public interface IContactRepository
{
    List<Contact> FindAll();
    void Save(Contact entity);
}

ContactRepository.cs

public class ContactRepository : IContactRepository
{
    private static List<Contact> Db;
    static ContactRepository() { Db = new List<Contact>(); }
    public List<Domain.Contact> FindAll() { return Db; }
    public void Save(Domain.Contact entity) { Db.Add(entity); }
}

And a screenshot:

Next let’s start designing and implementing our contact page.

The page needs to handle some data (a model). Let’s create a ContactModel class

public class ContactModel
{
    public string Name { get; set; }
    public string Email { get; set; }
    public string Message { get; set; }
    public enProcessState ProcessState { get; set; }
}

The enProcessState is the following self-explained enum

public enum enProcessState
{
    Pending,
    Saved
}

Remember we are using the MVP pattern and we have still to implement the View and the Presenter.

Let’s create our view. Because a view is possible to be implemented in various ways (desktop, mobile client etc.) we will first create an interface or a contract that we want to code against. Our interface looks like this:

public interface IContactView : WebFormsMvp.IView<ContactModel>
{
    event EventHandler<EventArgs> Contacting;        
}

The view inherits from WebFormsMvp.IView and defines an event. It also defines a model with the type of ContactModel.

Last but not least let’s implement our Presenter. All the above take place inside .Logic project and not inside Web.

Our presenter looks like this:

public class ContactPresenter : WebFormsMvp.Presenter<IContactView>
{
    private readonly IContactRepository _repository;
    public ContactPresenter(IContactView view)
        : this(view, new ContactRepository()) { }
    public ContactPresenter(IContactView view, IContactRepository repository)
        : base(view)
    {
        _repository = repository;
        View.Contacting += new EventHandler<EventArgs>(Contacting);
    }

    void Contacting(object sender, EventArgs e)
    {
        if (View.Model == null || View.Model.ProcessState == Views.Models.enProcessState.Saved)
            return;
        _repository.Save(View.Model.ToDomainModel());
        View.Model.ProcessState = Views.Models.enProcessState.Saved;
    }
}

The presenter inherits from WebFormsMvp.Presenter and defines the type of view that is handling. It also defines two constructors (for now). To explain in a single sentence the main purpose of the presenter I will say that some times is called also a Controller.

Everything looks right except we don’t have an implementation for our View! Let’s move to our .Web project to build one.

We will create a user control named ContactControl.ascx. The controls code-behind looks like this:

public partial class ContactControl : WebFormsMvp.Web.MvpUserControl<ContactModel>, 
    LoveForWebForms.Logic.Views.IContactView
{
    public event EventHandler<EventArgs> Contacting;
    private void OnContacting()
    {
        if (Contacting != null)
            Contacting(this, EventArgs.Empty);
    }
    protected void btnContactUs_Click(object sender, EventArgs e)
    {
        Model.Name = txtName.Text;
        Model.Email = txtEmail.Text;
        Model.Message = txtMessage.Text;
        OnContacting();
        View activeView = Model.ProcessState == enProcessState.Saved ? vSaved : vContact;
        mv.SetActiveView(activeView);
    }
}

and mark-up looks like this

<asp:MultiView runat="server" ID="mv" ActiveViewIndex="0">
    <asp:View runat="server" ID="vContact">
        <fieldset>
            <legend>Contact Us</legend>
            <table>
                <tr>
                    <td>
                        Name
                    </td>
                    <td>
                        <asp:TextBox runat="server" ID="txtName" />
                        <asp:RequiredFieldValidator ErrorMessage="*" ControlToValidate="txtName" runat="server" />
                    </td>
                </tr>
                <tr>
                    <td>
                        Email
                    </td>
                    <td>
                        <asp:TextBox runat="server" ID="txtEmail" />
                    </td>
                </tr>
                <tr>
                    <td>
                        Message
                    </td>
                    <td>
                        <asp:TextBox runat="server" ID="txtMessage" TextMode="MultiLine" Rows="10" />
                    </td>
                </tr>
                <tr>
                    <td colspan="2">
                        <asp:Button Text="Contact Us" runat="server" OnClick="btnContactUs_Click" />
                    </td>
                </tr>
            </table>
        </fieldset>
    </asp:View>
    <asp:View runat="server" ID="vSaved">
        Contact request was Saved! <a href="ContactList.aspx">View all</a> contact requests.
    </asp:View>
</asp:MultiView>

Our markup is pretty much standard. In our code-behind file when the user clicks the button we prepare our Model and raise the Contacting Event. If everything went OK the ProcessState property of our model becomes Saved and the appropriate view of MultiView control is displayed.

All we wave to do is to create a page and put our UserControl inside. Just a simple page.

But I didn’t declared anywhere the presenter!! How is ContactPresenter is selected? This is where MvpWebForms projects comes!

Pay attention to names. ContactControl.ascx, ContactPresenter,IContactView. When you follow these conventions (just like MVC has it’s own) no further actions are required. If something like this isn’t possible you can use attributes like

[WebFormsMvp.PresenterBinding(typeof(LoveForWebForms.Logic.Presenters.ContactPresenter))]

in the user control to indicate which presenter to use.

Dependency Injection

Let’s say that I am a WebForms developer. When I ‘m around MVC developers I want to be able to say things like: “Yeah Dependency injection rocks!” or “When you say DI god kills a kitten, but still using it in my projects! DI DI DI…”

Can I do it? Is it helpful in any way? Yes and yes!

Go to Package Manager Console target the Web Project and type

Install-Package WebForms.Mvp.Autofac

You might notice that WebFormsMvp has packages for Autofac, Unity and Castle. Let’s use Autofac for now.

After the package is being installed we make the following changes to global.asax

public class Global : System.Web.HttpApplication, IContainerProviderAccessor
{
    protected void Application_Start(object sender, EventArgs e)
    {
        var builder = new ContainerBuilder();
        builder.RegisterType<ContactRepository>().As<IContactRepository>();
        _containerProvider = new ContainerProvider(builder.Build());            
        PresenterBinder.Factory = new AutofacPresenterFactory(_containerProvider.ApplicationContainer);
    }
    static IContainerProvider _containerProvider;        
    public IContainerProvider ContainerProvider
    {
        get { return _containerProvider; }
    }
}

Implement IContainerProviderAccessor and add the code above to Application_Start.

Now let’s change our ContactPresenter and remove the first constructor. Now our presenter looks like

public class ContactPresenter : Presenter<IContactView>
{
    private readonly IContactRepository _repository;
    public ContactPresenter(IContactView view, IContactRepository repository)
        : base(view)
    {
        _repository = repository;
        view.Contacting += new EventHandler(view_Contacting);
    }

    void view_Contacting(object sender, EventArgs e)
    {
        if (View.Model.ProcessState == Views.Models.enProcessState.Saved)
            return;

        _repository.Save(View.Model.ToDomainModel());
        View.Model.ProcessState = Views.Models.enProcessState.Saved;
    }
}

The instantiation of IContactRepository is taken care by Autofac.

What about creating a grid page with paging that also uses an ObjectDataSource?

What about validation?

Because this post is really long those scenarios are implemented in the source code! So download and see for your self! But really quick I am posting a view screenshots to get an idea. When you download the project don’t forget to run the Install-Package command to get dependencies from Nuget as they are not included by default.

What about testing?

I 've written a detailed article on DotNetSlackers about testing!

I think it’s time for refactor!

Download Sample

When you download the project don’t forget to run the Install-Package command to get dependencies from Nuget as they are not included by default. Commands are

Install-Package WebFormsMvp for .Logic project
Install-Package WebFormsMvp.Autofac for .Web project

Filter Tweets by Language

Tue, 26 Apr 2011 08:16:02 GMT

Twitter is great but is missing a feature that is badly wanted (at least from me). This is the ability to filter tweets based on the language of the tweet.

Although some clients support automatic recognition of the language such an implementation would be a nightmare if you want to implemented bug free.

So I am suggesting the following “workaround” which can be easily supported both by Twitter and 3rd party clients.

Let’s suppose that you want to tweet Hello Word in four different languages. English, Greek, German and French.

An English tweet would have no difference

Hello World

But the other three tweets would have three (3) more chars. Two (2) chars that represent the country code (as seen here) and the special char “|” as a delimiter. And those chars would be at the beginning of the tweet so it’s easy to implement the filtering mechanism. Last but not least if you use the “new” re-tweet feature the it’s possible to filter more tweets!

Here are the example tweets:

gr|Γειά σου Κόσμε

fr|Bonjour monde

de|Hallo Welt

So the above is just a suggestion to Twitter and 3rd party apps.

As a user I don’t mind those extra three (3) characters because I consider the ability to filter tweets more valuable.

As a developer I think it’s rather trivial to implement a feature like that. It’s also easy to create a UI that fills that kind of info without having the user to type them.

The power of Twitter lies on it’s users! So, if this is something that you think it could be useful make a buzz about it and I believe we can push it forward.

Cheers!

ASP.NET MVC - Passing a Flag Enum to an Action as Parameter

Fri, 22 Apr 2011 14:02:03 GMT

A flag Enum is an Enumeration that can hold multiple values per instance An example of such Enum is:

[Flags]
public enum enDays
{
    None = 0,
    Monday = 1,
    Tuesday = 2,
    Wednesday = 4,
    Thursday = 8,
    Friday = 16,
    Saturday = 32,
    Sunday = 64
}

And we could use it like this:

enDays days = enDays.Monday | enDays.Friday;

So our “days” field would have both Monday and Friday

If we want to create a controller and inside it an action that takes as an input the following enum we would write something like:

public partial class HomeController
{
    public virtual ActionResult DoSomething(enDays days)
    {
        //TODO
    }
}

Then we could navigate to /Home/DoSomething?days=Monday and give value enDays.Monday to “days” argument.

But if we write /Home/DoSomething?days=Monday,Friday it’s the same as enDays.Monday | enDays.Friday!

We can also write /Home/DoSomething?days=17 which is the same as above!

Comma-separated Enum values or the equivalent numeric value, your choice!

Cheers!

ASP.NET Membership - Change password without asking the old (WITH Question and Answer)

Fri, 22 Apr 2011 14:01:39 GMT

I have received many comments and questions about how you can do what is described in this post when you site requires question and answer. (Better solution at the bottom of this post)

The solution is definitely not the best and should be used with EXTREME caution because in a high traffic website can cause problems but I write it down anyway.

We will use reflection in order to solve our problem.

And this is the code

MembershipUser user = Membership.GetUser();
string newPassword = "newPass";
string tempPassword = string.Empty;
if (Membership.Provider.RequiresQuestionAndAnswer)
{
    var _requiresQA = Membership.Provider.GetType().GetField("_RequiresQuestionAndAnswer",
        System.Reflection.BindingFlags.Instance | System.Reflection.BindingFlags.NonPublic);
    //change the value in the private field
    _requiresQA.SetValue(Membership.Provider, false);
    //do the reset
    tempPassword = user.ResetPassword();
    //set it's original value
    _requiresQA.SetValue(Membership.Provider, true);
}
else
{
    tempPassword = user.ResetPassword();
}

//do the Change
user.ChangePassword(tempPassword, newPassword);

But this code changes the only instance of MembershipProvider meaning if you access somewhere else from your application the property RequiresQuestionAndAnswer until you set back it’s original value you will get false instead of true.

So again be VERY careful.

OR

You can ignore the above and use the brilliant and simple solution that RichardD suggested in comments.

Add this to your web.config

The concept is to use one provider for password reset and one provider for every other function in your site.

And the code becomes like this

Beautiful isn’t it?

Hope you find it useful!

Note: The above solution is only when you store your passwords in a Hashed way. In any other occasion the above solution is pointless.

ASP.NET Membership - Change password without asking the old

Fri, 22 Apr 2011 14:01:08 GMT

Recently I was in a situation where a user was required to change the password upon first login.

But MembershipUser’s ChangePassword requires 2 arguments. Old and new password. In my case the password was hashed and I couldn’t retrieve it unless the user enter it.

So I wanted to change the password without asking the old one. The solution was :

Reset the password
Keep the generated password in a variable
Call MembershipUser.ChangePassword using the generated password as the old one!

It does an extra query to the database but does the job!

If the password was stored in Clear or Encrypted form it could be retrieved by calling user.GetPassword().

Linked-In OAuth Library

Fri, 22 Apr 2011 13:58:31 GMT

I uploaded a package in Nuget Gallery that simplifies the interaction with linked-in. Uses a modified version of Hammock (because latest version has a bug) so be careful if you want to add it to your project. When the bug is fixed the correct dependency will be added to the package.

This package is in early development stage so a lot are subject to change. This post will be updated accordingly.

Version 0.0.1

Code that begins the Authorization process:

var session = LinkedInSession.RetrieveFromUserSession();
if (session == null)
{
    var token = OAuthManager.Current.CreateToken(callback: "http://localhost:1134/Account/LogOn");
    //The user is being redirected
    OAuthManager.Current.BeginAuth(token, endResponse: true, displayAllowDenyScreen: false);
}

Code that completes the Authorization process and retrieves user profile:

var token = OAuthManager.Current.GetTokenInCallback();
var session = OAuthManager.Current.CompleteAuth(token);
if (session.IsAuthorized())
{
    session.StoreInUserSession();
    var p = session.GetProfile();    
}

That’s all for now. There are a lot TODOs and many improvements to be done so stay tuned!

Where is the bug? - C#

Fri, 22 Apr 2011 13:57:37 GMT

Below are a few lines of code written in C#. Is the code bug-free? What will this print in the console? Can you do it without VS?

static void AVeryCoolMethod()
{
    var numbers = new[] { 1, 2, 3, 4, 5, 6 };

    var ngt5 = numbers.Where(n => n > 5);

    var n = ngt5.First().ToString();

    Console.WriteLine(n, numbers);
}

Cheers!

Detect if Firebug is enabled and deny access

Wed, 09 Mar 2011 20:49:09 GMT

Firebug is an essential tool for every web developer. It allows to do many things such as script debugging, DOM manipulation etc. What happens though when it get’s to the hand of an experienced attacker? Then it could be used to exploit your site very easily!

How can you prevent that? Well, you can’t actually (or you can if you write secure code) because the source is already downloaded to the browser and anyone can do anything with that. But there is a way to make that process a little bit more trickier (attention: it’s still overridable but I won’t say how ☺).

Just put the following code in your page and try to active firebug’s console

$(window).load(function() {setInterval(chkh, 5000);});
function chkh() {
    if (window.console && window.console.firebug) {
            var path = window.location.pathname;
            if(path!='/DenyAccess.html')
                window.location = '/DenyAccess.html';
    }
}

What the above script does is that it checks every five (5) seconds and if Firebug’s console is enabled, redirects the user to a page that perhaps prompts the user to disable Firebug’s console or what ever you think. Checking if a variable is defined every five or ten seconds isn’t a heavy process so the overhead is not big.

Do you think the above code can prevent anything? Then use it wisely!

I ‘m saying again that this code checks only for Firebug’s console. Other tabs like HTML, Dom, Script are active and functional.

Hope Driven Development - Are you doing it?

Tue, 08 Mar 2011 07:58:52 GMT

HDD (and it don’t mean Hard Disk Drive ☺) is intuitively a bad approach on Software Development. If someone tells you that many times is doing Hope Driven Development, only by name you might say “Oh that’s a bad thing. Stop it now!”

But everyone is doing it… Not always, not even most of the times, but sometimes it happens. Why is that?

First of all let me give you my definition of HDD and please do comment on that.

Hope Driven Development is an approach to Software Development that encourages the developer to write code, hoping that this code will run always as expected and ensuing that belief by praying. HDD also encourages the developer to overlook the fact that this code might not work as expected and points out that it might work. Last but not least, when someone is doing HDD it’s possible (yet, less likely) to believe that when the time comes and everything goes wrong a magic unicorn will come and save the day.

Now that we have a definition (or at least we think we have) we must figure a way to tell when we are doing HDD.

So by definition, when we use phrases like: “should/might work”, “even if it breaks no big deal” (deep down we know it will but hoping that this day will never come), “it got to work” etc then we probably doing HDD. And I ‘m not talking about the entire process of Development. I ‘m taking about that tiny piece of code that we are referring to.
When we don’t have unit tests (or we have an empty project that we call “Tests”).
When we are worrying about performance too much.
ex: I expect a Person argument in this function and I shouldn’t validate it cause I already validating it three calls behind. So I believe and hope that I will always get my argument validated.
You are writing “spaghetti” code and no-one else in the team wants to hear about that code (hoping they will never have to find out)
You are writing “spaghetti” code and you your self want to forget that you wrote that piece of code (hoping that you will never have to change it again)
There is also a possibility when you write spaghetti code and you love it plus the team want to know more about how this could be more complex. In that case you are not doing HDD. You are just weird

How can you change your HDD approach to a project?

STOP doing the above

This post is more of a note to myself because there are times that the pressure is huge and wrong seems right…

Comment and share if you feel the need to…

JavaScript is disabled - Should I Care?

Sat, 05 Mar 2011 07:39:23 GMT

Many times a Web Developer faces the question “This won’t work if JavaScript is disabled. Should I change it?” If it’s something simple like a submit button instead of a link it’s easy and we don’t put too much thought in the decision. What happens though, if we have something more complicated like an interface using Drag ‘n Drop?

First of all, if you don’t plan to make the site 100% accesible to users with disabled JavaScript then don’t bother at all. There is no point in making a registration page that works with no JavaScript and then other pages (like the login page or the user page) require JavaScript. It’s the same thing like requiring IE6+ on registration and IE7+ on the user page.

Let me tell you when I think a Web Developer should care. A web dev should care if the answer is yes to one of the following questions:

Am I targeting users with disabled JS?
Do I have the time to do that? (Or is everything else done and nothing needs improvement?)
Did I make a promise to the client?
Is the client paying separately that requirement? (Or Do I have the money to do that?)
Will the website traffic be reduced by a significant amount if I restrict users with disabled JavaScript?
Will this requirement take less than or equal to 1/20 of the total time spent in developing the website?

If you answered yes to at least one of the above questions then a fully functional JS-disabled website requirement is in your next project!

For me trying to develop sites with the above functionality is like trying to develop websites that work with IE6. If you target countries like China with more that 30% of IE6 users then it probably makes sense. If you target European users with 2-3% of the users having IE6 then is a lot less important if this browser is supported.

So making a decision about supporting or not this feature is something that needs to be done in every project and the decision should affect the entire project.

I have always been a great supporter of browsers with expiration date. It’s a feature that will save the next generation of Web Developers (because our generation is doomed to ask again and again the same questions about browsers compatibilities along with backward compatibilities in web standards)

An important note is that I do believe and I always encourage anyone who asks me to test their website with the lowest-enabled possible JavaScript settings. A user that worries about security, so reduces the things that JavaScript can do for him (like minimize,or resize windows etc.) is a healthy user that deserves to be treated equally as the other users! A user that worries about security, so disables JavaScript is a user who follows a Greek saying that goes like this: “If the head hurts, cut the head”.

In the end, I believe that JS-disabled websites should becoming fewer as the time passes and the need to ask the question “Should I care?“ should eventually go away.

Thoughts and comments always welcomed!