Thursday, January 21, 2010
Recently I have started using Mercurial (also known as ‘hg’ which is the command line tool) for source control management (also known as version control system) for some of my projects. Most of the projects I deal with use Subversion which is sufficient for most needs. My first exposure to distributed source control management was Linus Torvald’s presentation on git at Google. In a nutshell, he rails on CVS (his initial intention was to design a system as “What Would CVS Never Do” [WWCCVSND] and he considers tarballs and patches as superior to CVS) and also sees Subversion as “the most pointless project ever started.”
Popular Open Source Distributed SCMs / DVCSs
There are several popular open source distributed SCMs/DVCSs include Darcs, Bazaar, Mercurial, and Git. Currently, it seems that Git is the most “popular” of the bunch (at least in the blogs and tweets that I’ve seen) due to the github and projects like the Linux kernel and others. A quick comparison from stackoverflow also verifies popularity in terms of the number of tags related to the respective systems (I listed the main tag only although there were quite a few subtags for Git and Mercurial):
Why Mercurial?
I believe Mercurial, Git, and other DVCSs will work great for most cases. There are several analyses on the web with varying results, but the all of the available open source DVCSs are strengths and weaknesses and are capable. That being said, there are several reasons why Mercurial is a good choice.
1. Portability
Mercurial is written primarily in Python with small portions written in C for speed. This makes it extremely portable to almost every architecture. While Git is available on POSIX systems, there are basically two Windows versions with one requiring Cygwin and the other is called msysGit.
2. TortoiseHG
Another reason is TortoiseHg which is a shell extension primarily for Windows, although there is a Gnome/Nautilus CLI extension and a Mac OS X version is in the works. TortoiseHg is written with gtk/tkinter so that makes it easier to port to other systems than TortoiseSVN or TortoiseGit (both are Windows only) although I wish TortoiseHg used wxPython instead. I’ve also used QGit and GitGUI the experience with Tortoise* is generally better in my opinion.
3. “Smart” HTTP
Mercurial has built-in “smart” HTTP which leads to more efficient transfer when using HTTP. There is discussion of Git implementing smart HTTP. The reason HTTP is important is it is a universal common denominator. Some corporations block the ports used for native git or git+ssh access. Also, I have done work with people in foreign countries where the native protocols didn’t work, or we could use ssh, and HTTP was the only option.
'hg serve' is an extremely easy way to start the Mercurial internal web server. As one person noted:
I chose Mercurial since it feels clean and elegant -- I was put off by the shell/Perl/Ruby scripts I got with Git. Try taking a peek at the git-instaweb.sh file if you want to know what I mean: it is a shell script which generates a Ruby script, which I think runs a webserver. The shell script generates another shell script to launch the first Ruby script. There is also a bit of Perl, for good measure.
4. Google support
Google officially supported Mercurial for project hosting in Google Code on Friday, April 24, 2009. Google also released a report on their internal DVCS analysis with the following summary:
In terms of implementation effort, Mercurial has a clear advantage due to its efficient HTTP transport protocol.
In terms of features, Git is more powerful, but this tends to be offset by it being more complicated to use.
I personally doubt Google’s support of Mercurial will sway the DVCS competition significantly towards Mercurial as I believe Git is powerful, robust, and here-to-stay. Google’s support does give significant credence to Mercurial though. I expect improvement in both Git and Mercurial in the future which will probably blur the lines even more.
Using Mercurial with VMs
The primary benefit of using a DVCS with virtual machines is that the VM does not need to be connected to the outside world but instead only needs access to the host. Most of my VMs use “Host only” networking which allows the VM to connect to the host, but not the outside network or Internet. Thanks to the easy Mercurial built-in web server, it is easy to share a repository from the host to the VM.
Step 1: Create or Clone repository on Host
For example, if we take the 'hello' repository from the Mercurial tutorial, we can clone an existing repository on the host environment:
Step 2: Starting the Mercurial Web Server
You can either start the web server by using 'hg serve' from the repository directory, or use TortoiseHg Web Server:
If you are going to be pushing repository changes from the VM back using HTTP, then don’t forget to enable pushing via HTTP in the host '$(repository)/.hg/hgrc' file:
[web]
allow_push = *
push_ssl = false
Step 3: Cloning the Repository from the VM
This step is similar to Step 1, except change the source to the repository on the host, e.g. :8000/">http://<yourmachinename>:8000/.
Then modify the source as necessary on the VM and commit your changes to the local repository.
Step 4 – Option A: Pushing Changes back to the Host
Once you are ready, you can push your changes back to the repository on the host:
If you get a failure, then check the [web] settings in Step 2 and restart the web server if you change the 'hgrc' file.
Step 4 – Option B: Pulling the Changes back to the Host
Alternatively, you could start the Mercurial Web Server on the VM and pull the changes from the host. This just requires you to start the web server on the VM, and also make sure the specified port, e.g. 8000, is not used by another web server and is open. You can add a sync alias for the VM(s) which makes it easy to switch to different sources/repositories with the Repository Explorer:
Conclusion
It is easy to use Mercurial, or Git for that matter, in combination with development on virtual machines to allow developers to commit and experience a real SCM/VCS experience.
Wednesday, December 30, 2009
At one point in time I found a cool clock graphic and I was looking for it again because it gave a fairly detailed explanation on making a nice looking vector based clock (it was the Codeproject article by Marc Clifton). During my search I found numerous implementations of both SVG and Silverlight clocks and I decided to catalog my findings.
The animated clocks are not necessarily practical to display on web pages, but an interesting exercise and fun as well.
SVG Clocks
by Adobe (unknown date)
A very basic looking clock that uses three <animateTransform type="rotate"> for the hour, minute, and second hands along with a load script to initialize the hands. Note that Adobe has discontinued their SVG Viewer and there are conflicts between SVG and Flash.
by Shelley Powers on November 30, 2007
This is an improvement on a very simple SVG clock by Jason Davis (the original link was not working). This implementation supports displaying time in a different time zone rather than the user’s current time using the statement: setInterval("setClock(calcTime(-6))", 1000);. The code is licensed with LGPL.
I had to download the file directly as it was not shown in FF 3.5. The author has a rather appropriate statement in the article:
Is the clock worth the extra burden on the client's machine? Yes, and no. As a demonstration of what you can do with SVG and simple animation, I think it's a valuable tool. There is a Catch 22 about SVG: we don't use SVG because browser support is incomplete or inefficient; effort to better incorporate SVG is of secondary importance because SVG is little used. The only way to break this cycle is to actually start using the specification, and pushing a bit at the edges while we go about it.
by Alexander Jones & Eric Pierce from Wikipedia on October 14, 2006
This is Wikipedia entry that I thought was animated but it turns out it is not animated. Bummer. It wouldn’t be hard to animate though.
by Kam-Hung Soh
This example uses Javascript at the DOM document level rather than inside the SVG document to rotate the clock hands. It also uses a regular expression to change the rotation angle.
by Tavmjong Bah (unknown date)
This is a really cool example created with Inkscape where every gear is independent and does not interfere with other gears. According to the author, the gears outlines were generated with a Python script by Aaron Spike.
by Stefan Oskamp (unknown date)
A nice looking clock that uses inline SVG along with a Javascript setInterval function to adjust the rotate attributes for the hour, minute, and second hands. The actual background is a 256 x 256 PNG file, rather than a vector based image.
by Damien Dawber (ibuildstuff.net)
This clock is a self generated SVG image using JavaScript and the Raphael JavaScript Vector Library.
by G. Wade Johnson
A clock with a sunburst effect that uses an external script to drive the animation. The author also has a version that runs backwards by simply changing a parameter in the initialization script.
by Steve Pomery
A nice looking clock created using Inkscape. Now at version 2.2.
by Chris W. Johnson on April 29, 2008
An inline base64 encoded SVG clock. The SVG has a version of the Google analytics tracking script "because it's nice to know if people care."
Silverlight Clocks
by Microsoft
A clock sample from the Silverlight 1.0 SDK which uses clock.xaml and clock.js to create the clock. The result is actually very similar to some of the SVG implementations since Silverlight 1.0 did not have codebehind capability.
Version 2.0 of the clock uses codebehind along with a XAP file. There is also a version using IronRuby although the XAP file is significantly larger (11KB vs 1MB!) because it includes several assemblies in the XAP. Then there is the Python version as well which is almost identical to the Ruby version except for the language substitution.
There is also a Microsoft walk-through: Creating a Silverlight Clock by Using Expression Blend or Code, which has a detailed step by step guide for Expression Blend.
by Marc Clifton on April 18, 2004
A very nice looking clock complete with a walkthrough on how to create the look and feel of the clock. Technically this isn’t a Silverlight clock as it is intended for the (now defunct) MyXaml and then improved upon by VG.net. The graphics should be mostly compatible with Silverlight though I haven’t tested it yet.
by Pencho Popadiyn on October 30, 2008
Version 1 introduces the MVP design pattern and the design of the clock using the MVP pattern.
Version 2 - This implementation uses VS2008, .NET 3.5, Silverlight 2 using an MVP pattern. The sample has a very nice looking graphic and improves upon the Clock view.
by Alex Bell on November 7, 2008 (based on SVN timestamps)
A modest clock using Silverlight although the author claims:
Hybrid Clock is, by far, the most sophisticated animated Date/Time visualization control developed for Microsoft Silverlight™ 2.0.
by Michael S. Scherotter on April 24, 2008
This is a gorgeous clock using Silverlight, Expression Blend, and Expression Design. For some reason this is a Silverlight 1.0 applet instead of 2.0 which means it uses some Javascript event triggering for the loading, but it looks great.
Miscellaneous Silverlight Clocks
I was taking way too much time, so here are other implementations without screen shots.
http://www.c-sharpcorner.com/UploadFile/nipuntomar/SilverlightClock08122008132436PM/SilverlightClock.aspx – A good Silverlight clock.
http://pixelplastic.de/2008/01/12/ZoomableUltraHighResolutionSilverlightClock.aspx – A Silverlight clock that allows the user to drag and move the clock. The applet did not work in IE or FF, so I assume it was developed with a pre-release version and has not been updated on the blog article. You can download the source code though. Marcel Hoyer has an older version of the clock as well as other cool Silverlight examples on his blog.
http://weblogs.asp.net/andrewrea/archive/2009/08/12/programmatic-drawing-with-silverlight-2-0-3-0-an-analogue-clock.aspx – A basic looking clock although the clock is mostly generated by code (tick marks, etc.) which is cool. The blog post has a lengthy code listing and downloadable code.
WPF Clocks
by Charles Petzold on April 7, 2006
This is an all XAML clock without any C# or .NET code. It was based on a Microsoft sample and Charles Petzold improved upon that sample by eliminating the C# code, improved the animation technique, and eliminated the repetition in the XAML. The clock graphic is basic and not fancy, but practical.
Friday, January 30, 2009
Recently I needed to display the text that was sent to log4net in a TextBox in a Form. It turns out this is very easy to do using a custom appender and the AppenderSkeleton. The custom appender uses a delegate to allow the host program to define a callback function to handle the log text as necessary.
1. Create a new custom appender class
I named my appender DelegateAppender because it needed to use a delegate to pass the text to the Form.
using log4net.Core;
namespace log4net.Appender
{
public delegate void LogTextAppend(string text);
public class DelegateAppender : log4net.Appender.AppenderSkeleton
{
private LogTextAppend logTextAppend;
public LogTextAppend LogTextMethod
{
get { return logTextAppend; }
set { logTextAppend = value; }
}
public DelegateAppender()
{
logTextAppend = EmptyAppend;
}
private void EmptyAppend(string text)
{
// Do nothing
}
protected override void Append(LoggingEvent loggingEvent)
{
if (logTextAppend != null)
logTextAppend(RenderLoggingEvent(loggingEvent));
}
}
}
2. Create and assign the delegate to the appender
As far as I could tell, there was no easy way to specify the delegate to use in the configuration of the DelegateAppender. Therefore I added a simple method to assign the delegate in the Form to the DelegateAppender.
public void AddStatus(string message)
{
textBoxStatus.AppendText(message);
}
private void InitializeLogging()
{
bool initialized = false;
if (!log.Logger.Repository.Configured)
{
log4net.Config.XmlConfigurator.Configure();
textBoxStatus.AppendText("WARNING: log4net not automatically configured. " +
"Check AssemblyInfo.cs for - " +
"[assembly: log4net.Config.XmlConfigurator(Watch=true)]\r\n");
}
foreach (log4net.Appender.IAppender appender in log.Logger.Repository.GetAppenders())
{
if (appender.GetType() == typeof(log4net.Appender.DelegateAppender))
{
log4net.Appender.DelegateAppender delegateAppender = (log4net.Appender.DelegateAppender) appender;
// .NET 2.0+
delegateAppender.LogTextMethod = this.AddStatus;
// .NET 1.1
//delegateAppender.LogTextMethod = new log4net.Appender.LogTextAppend(this.AddStatus);
initialized = true;
}
}
if (!initialized)
{
textBoxStatus.AppendText("ERROR: Unable to add DelegateAppender to logging!\r\n");
}
}
3. Add the log4net configuration to app.config and AssemblyInfo.cs
The code for AssemblyInfo.cs just tells log4net to configure itself and also watch for changes.
[assembly: log4net.Config.XmlConfigurator(Watch = true)]
The app.config is as follows:
<?xml version="1.0" encoding="utf-8" ?>
<configuration>
<configSections>
<section
name="log4net"
type="log4net.Config.Log4NetConfigurationSectionHandler, log4net"
/>
</configSections>
<log4net>
<logger name="log4netAppender.LogTestForm">
<level value="ALL"/>
</logger>
<root>
<level value="ALL" />
<appender-ref ref="DelegateAppender" />
</root>
<appender
name="DelegateAppender"
type="log4net.Appender.DelegateAppender" >
<layout type="log4net.Layout.PatternLayout">
<param
name="ConversionPattern"
value="%m%n"
/>
</layout>
</appender>
</log4net>
</configuration>
And there is a simple example of how to create a custom appender with log4net and hook it into an application.
Technorati tags:
C#,
log4net
Thursday, November 06, 2008
Series History
Introduction
The goal of these posts is to build a spinning cursor similar to the Mac OS X wait cursor through programmatic means in Silverlight. The cursor is still very rough and will undergo improvements progrressively. One of the reasons to build the cursor programmatically is to have more control over the output such as changing the number of slices or rotation or other parameters.
For this post, I added an animation to the rotation angle of the canvas which causes the slices to spin. I also added a simple navigation to the previous examples.
Step 1: Adding the Animation
In the previous post, I had one canvas for the background image as well as the slices. This time I realized I needed two canvases: one for the background and one for the slices. I needed separate canvases so that I could apply a rotation transform animation to the slices without affecting the background image. I used Expression Blend 2.0 sp1 and selected the "SpinningCanvas" and added an Storyboard named "SpinStoryboard". I then changed the Angle of the RotateTransform from 0 - 360° for the time range 0.0 - 2.0 sec.
This resulted in the following XAML:
<UserControl.Resources>
<Storyboard x:Name="SpinStoryboard" RepeatBehavior="Forever">
<DoubleAnimationUsingKeyFrames
BeginTime="00:00:00"
Storyboard.TargetName="SpinningCanvas"
Storyboard.TargetProperty="(UIElement.RenderTransform).(TransformGroup.Children)[2].(RotateTransform.Angle)">
<SplineDoubleKeyFrame KeyTime="00:00:02" Value="360"/>
</DoubleAnimationUsingKeyFrames>
</Storyboard>
</UserControl.Resources>
Step 2: Starting the Animation
When the control is created or updated, the Update() method is called. I modified the method as follows:
void Update()
{
SpinStoryboard.Stop();
CursorCanvas.Children.Clear();
SpinningCanvas.Children.Clear();
CreateOutlineGuides(CursorCanvas);
CreateBackground(CursorCanvas);
CreateSlicePaths(SpinningCanvas);
SpinStoryboard.Begin();
}
Step 3: Adding the Sample Navigation
I wanted to support future samples, so I used reflection to find all of the UserControls not including App or Page. I used a Dictionary<string, Type> to map the Type.Name with the Type in case I need it for future samples. I then dynamically created a button for each sample.
void Page_Loaded(object sender, RoutedEventArgs e)
{
Assembly assembly = Assembly.GetExecutingAssembly();
Type[] exportedTypes = assembly.GetExportedTypes();
foreach (Type t in exportedTypes)
{
// Testing various inheritance detection methods
bool isNotApp = (t != typeof(App));
bool isNotSelf = (t != this.GetType());
bool isUserControl = t.IsSubclassOf(typeof(UserControl));
if (isNotApp && isNotSelf && isUserControl)
{
_implementedTypes.Add(t.Name, t);
}
}
Button lastButton = null;
// Sort the names of the UserControls to make more sense
List<string> sortedKeys = new List<string>(_implementedTypes.Keys);
sortedKeys.Sort();
for (int index = 0; index < _implementedTypes.Count; ++index)
{
string key = sortedKeys[index];
Type t = _implementedTypes[key];
lastButton = AddButton(t, index);
}
// Use the last UserControl for the initial display
if (lastButton != null)
{
Button_Click(lastButton, null);
}
}
To make it simple, I used name of the UserControl for the Button.Content and in the Button_Click event I used System.Activator to create an instance of the class.
private void Button_Click(object sender, RoutedEventArgs e)
{
UIElement control = null;
try
{
Button button = sender as Button;
string typeName = button.Content as string;
Type type = _implementedTypes[typeName];
object instance = System.Activator.CreateInstance(type);
control = instance as UIElement;
}
catch (Exception ex)
{
System.Diagnostics.Debug.WriteLine(ex.ToString());
}
WorkArea.Children.Clear();
if (control != null)
WorkArea.Children.Add(control);
}
Conclusion
The animation looks nice and adds a lot to the overall effect. For the next part, I plan to dynamically create the animation instead of using Blend and perhaps improve the background shape to more closely resemble the Mac OS X spinning cursor.
Technorati tags:
C#,
Silverlight
Wednesday, October 29, 2008
Series History
In this post, I add curvature to the slices and refactor the code to support upcoming features. The goal of these posts is to build a spinning cursor similar to the Mac OS X wait cursor through programmatic means in Silverlight. One of the reasons to build the cursor programmatically is to create it will different number of slices or rotation or other parameters. At this point, the cursor is still very rough and just beginning to resemble the final result. In upcoming posts, I will animate the cursor and adjust the appearance to more closely resemble the desired cursor.
Step 1: Adjust the cursor properties
I added AlternateSlices and Origin properties, and also DefaultSliceCount and DefaultRotationAngle to the class.
public static readonly int DefaultSliceCount = 10;
public static readonly double DefaultSliceRotationAngle = 360.0 / DefaultSliceCount;
/// <summary>
/// Show alternating slices (true) or all slices (false)
/// </summary>
public bool AlternateSlices { get; set; }
/// <summary>
/// The origin of the ellipse for the spinning cursor
/// </summary>
public Point Origin { get; private set; }
Step 2: Set the default parameter values
I specified the default slice count and the radius and origin of the cursor.
void Page_Loaded(object sender, RoutedEventArgs e)
{
txtVersion.Text = this.GetType().Name;
AlternateSlices = true;
SliceCount = DefaultSliceCount;
txtSliceCount.Text = Convert.ToString(SliceCount);
SliceRotationAngle = DefaultSliceRotationAngle;
txtRotation.Text = Convert.ToString(SliceRotationAngle);
RadiusX = SpinningCanvas.Width / 2.0;
RadiusY = SpinningCanvas.Height / 2.0;
Origin = new Point(RadiusX, RadiusY);
Update();
}
Step 3: Create a background circle (ellipse)
I used a simple linear gradient for now because the final gradient looks much more complicated and will take more time.
void CreateBackground(Canvas cursorCanvas)
{
GradientBrush brush = new LinearGradientBrush();
GradientStop stop1 = new GradientStop();
stop1.Color = Color.FromArgb(255, 255, 0, 0);
stop1.Offset = 0.25;
brush.GradientStops.Add(stop1);
GradientStop stop2 = new GradientStop();
stop2.Color = Color.FromArgb(255, 0, 255, 0);
stop2.Offset = 0.5; // cursorCanvas.Width / 2.0;
brush.GradientStops.Add(stop2);
GradientStop stop3 = new GradientStop();
stop3.Color = Color.FromArgb(255, 0, 0, 255);
stop3.Offset = 0.75; // cursorCanvas.Width;
brush.GradientStops.Add(stop3);
Ellipse ellipse = new Ellipse();
ellipse.Height = cursorCanvas.Height;
ellipse.Width = cursorCanvas.Width;
ellipse.Fill = brush;
cursorCanvas.Children.Add(ellipse);
}
Step 4: Implement the alternating slices
Implementing the alternating slices was simple:
/// <summary>
/// Create an ellipse using rotated slices to build the ellipse
/// </summary>
void CreateSlicePaths(Canvas cursorCanvas)
{
// Create Slices
for (int index = 0; index < SliceCount; ++index)
{
PathFigure pathFigure = CreateSliceFigure();
PathGeometry pathGeometry = new PathGeometry();
pathGeometry.Figures.Add(pathFigure);
Path path = new Path();
path.Stroke = new SolidColorBrush(Color.FromArgb(128, 0, 0, 0));
path.StrokeThickness = 1.0;
path.Fill = new SolidColorBrush(Color.FromArgb(192, 128, 128, 128));
path.Data = pathGeometry;
// Rotate the slice for all slices after the first slice
if (index > 0)
{
RotateTransform t1 = new RotateTransform();
t1.CenterX = RadiusX;
t1.CenterY = RadiusY;
t1.Angle = SliceCenterAngle * index;
TransformGroup transformGroup = new TransformGroup();
transformGroup.Children.Add(t1);
path.RenderTransform = transformGroup;
}
cursorCanvas.Children.Add(path);
if (AlternateSlices)
++index;
}
}
Step 5: Add the curvature to the slices
The curvature isn't perfect, but it looks fine with the default settings (10 slices and 36° rotation). I used a Bezier curve, but perhaps the standard arc would work better in the future.
// Curve control weighting
double curveWeight = 0.75;
// Create the first line
BezierSegment seg1 = new BezierSegment();
seg1.Point1 = point0;
seg1.Point2 = CalculatePointOnEllipse(0.0, RadiusX * curveWeight, RadiusY * curveWeight, Origin);
seg1.Point3 = point1;
pathFigure.Segments.Add(seg1);
// Use an arc for the circular side
ArcSegment seg2 = new ArcSegment();
seg2.Point = point2;
seg2.Size = new Size(RadiusX, RadiusY);
seg2.RotationAngle = SliceCenterAngle;
seg2.IsLargeArc = (SliceCenterAngle > 180.0);
seg2.SweepDirection = SweepDirection.Counterclockwise;
pathFigure.Segments.Add(seg2);
// Close shape by going back to the starting point
BezierSegment seg3 = new BezierSegment();
seg3.Point1 = point2;
seg3.Point2 = CalculatePointOnEllipse(SliceCenterAngle, RadiusX * curveWeight, RadiusY * curveWeight, Origin);
seg3.Point3 = point0;
pathFigure.Segments.Add(seg3);
Conclusion
The spinning cursor now has a dynamic number of slices and rotation angle, and it is starting to look more like the desired result. In the next post, I will animate the cursor to give it the spinning effect.
Technorati tags:
C#,
Silverlight
Monday, October 20, 2008
After my previous spinning wait symbol, I decided to see how difficult it would be to create a Silverlight version of the Mac OSX wait cursor that I referenced in the previous post. The Mac OSX cursor is commonly referred to as the "Spinning Pizza of Death" or the "Marble of Doom" and in fact there is a Marble of Doom web site dedicated to the amount of time spent waiting while watching the spinning cursor. The Marble of Doom web site has a very nice and large version of the cursor using Flash although it doesn't have any vector information but is using video frames (they probably just published the final product and did not include the vector/animation information). The purpose of this post is to programmatically build the cursor and then in later posts to animate it.
Step 1: Decide on the initial interface properties
I realized quickly that I would need a little geometry to programmatically build the cursor, but the first step was to build the interface requirements. The essential properties were:
public int SliceCount { get; set; }
public double SliceCenterAngle { get; private set; }
public double SliceRotationAngle { get; set; }
public double RadiusX { get; set; }
public double RadiusY { get; set; }
The SliceCount determines how many slices or divisions to create, and the SliceCenterAngle is simply 360° / SliceCount. The SliceRotationAngle is the angle to twist or bend the slice. I decided to have a RadiusX and RadiusY to support ellipses in the future as well.
Step 2: Manually create a slice
Before I could programmatically create a slice, I needed to find out how to create a slice using XAML and Blend. The points on the slice would be in the center of the circle, and then two points on the circle determined by the SliceCenterAngle. The biggest question was how to create the arc and maintain the circular appearance. Fortunately, the Geometry Overview on MSDN was very helpful and got me started on the right track with the PathGeometry. I was able to create the simplest scenario with a single slice from a circle with four slices:
<Path Stroke="Black" StrokeThickness="1">
<Path.Data>
<PathGeometry>
<PathGeometry.Figures>
<PathFigure StartPoint="50,50">
<PathFigure.Segments>
<LineSegment Point="0,50" />
<ArcSegment Size="50,50" IsLargeArc="False"
RotationAngle="90" SweepDirection="CounterClockwise" Point="50,100" />
<LineSegment Point="50,50" />
</PathFigure.Segments>
</PathFigure>
</PathGeometry.Figures>
</PathGeometry>
</Path.Data>
</Path>
The next step was to create the same quarter-circle except with two slices:
<Path Stroke="Black" StrokeThickness="1">
<Path.Data>
<PathGeometry>
<PathGeometry.Figures>
<PathFigure StartPoint="50,50">
<PathFigure.Segments>
<LineSegment Point="0,50" />
<ArcSegment
Size="50,50"
Point="14.645,85.355"
RotationAngle="45"
IsLargeArc="False"
SweepDirection="CounterClockwise"
/>
<LineSegment Point="50,50" />
</PathFigure.Segments>
</PathFigure>
<PathFigure StartPoint="50,50">
<PathFigure.Segments>
<LineSegment Point="14.645,85.355" />
<ArcSegment
Size="50,50"
Point="50, 100"
RotationAngle="45"
IsLargeArc="False"
SweepDirection="CounterClockwise"
/>
<LineSegment Point="50,50" />
</PathFigure.Segments>
</PathFigure>
</PathGeometry.Figures>
</PathGeometry>
</Path.Data>
</Path>
Step 3: Create a slice programmatically
The general idea is to create one slice and then rotate the slice around the circle to create the complete circle.
/// <summary>
/// Create an ellipse using rotated slices to build the ellipse
/// </summary>
void CreateSlicePaths(Canvas cursorCanvas)
{
SliceCenterAngle = 360.0 / SliceCount;
// Create Slices
for (int index = 0; index < SliceCount; ++index)
{
PathFigure pathFigure = CreateSliceFigure();
PathGeometry pathGeometry = new PathGeometry();
pathGeometry.Figures.Add(pathFigure);
Path path = new Path();
path.Stroke = new SolidColorBrush(Color.FromArgb(255, 0, 0, 0));
path.StrokeThickness = 1.0;
path.Data = pathGeometry;
// Rotate the slice for all slices after the first slice
if (index > 0)
{
RotateTransform t1 = new RotateTransform();
t1.CenterX = RadiusX;
t1.CenterY = RadiusY;
t1.Angle = SliceCenterAngle * index;
TransformGroup transformGroup = new TransformGroup();
transformGroup.Children.Add(t1);
path.RenderTransform = transformGroup;
}
cursorCanvas.Children.Add(path);
}
}
/// <summary>
/// Create the base shape for the slice
/// </summary>
private PathFigure CreateSliceFigure()
{
// Start at the center of the ellipse
Point point0 = new Point(RadiusX, RadiusY);
// Next point is the left side of the ellipse
//Point point1 = new Point(0.0, RadiusY); // if no rotation
Point point1 = CalculatePointOnEllipse(SliceRotationAngle);
// Calculate the bottom point on the ellipse
Point point2 = CalculatePointOnEllipse(SliceRotationAngle + SliceCenterAngle);
// Starting point
PathFigure pathFigure = new PathFigure();
pathFigure.StartPoint = point0;
// Create the first line
LineSegment seg1 = new LineSegment();
seg1.Point = point1;
pathFigure.Segments.Add(seg1);
// Use an arc for the circular side
ArcSegment seg2 = new ArcSegment();
seg2.Point = point2;
seg2.Size = new Size(RadiusX, RadiusY);
seg2.RotationAngle = SliceCenterAngle;
seg2.IsLargeArc = (SliceCenterAngle > 180.0);
seg2.SweepDirection = SweepDirection.Counterclockwise;
pathFigure.Segments.Add(seg2);
// Close shape by going back to the starting point
LineSegment seg3 = new LineSegment();
seg3.Point = new Point(point0.X, point0.Y);
pathFigure.Segments.Add(seg3);
pathFigure.IsClosed = true;
pathFigure.IsFilled = true;
return pathFigure;
}
/// <summary>
/// Returns a point on the ellipse based on the rotationAngle, using
/// RadiusX/Y as the center (0, 0).
/// </summary>
private Point CalculatePointOnEllipse(double rotationAngle)
{
double angleRadians = rotationAngle * Math.PI / 180.0;
double x = Math.Cos(angleRadians);
x = RadiusX * x;
double y = Math.Sin(angleRadians);
y = RadiusY * y;
Point result = new Point(x, y);
result.X = RadiusX - x;
result.Y = RadiusY + y;
return result;
}
Step 4: Going Forward
Obviously this still needs a lot of improvement before it approaches the appeal of the Marble of Doom, which I will work on in the coming posts. However, the initial effort to create the "pizza" slices has been achieved and it is easier to build upon a base.
I added a grid in the background when I had some difficulty with the path geometry, but it is useful to I added two text boxes for the number of slices and rotation angle so that I could see the shape update dynamically.
Since I just used LineSegments to connect the slice points, the shape does not have the swirl or twist effect yet. Next time I will add the twist as well as rotation.
Technorati tags:
C#,
Silverlight
Wednesday, October 08, 2008
My wife has a Treo with Windows Mobile and I when I was using it I noticed it had a cool rotating wait symbol, so I wondered how difficult it would be to build the symbol in Silverlight. The symbol is similar to the old BeOS wait cursor and has as well as the Mac OS X wait cursor which I've always thought looks nice. At one point in time I created a Windows cursor that duplicated the look of the BeOS cursor but I don't use it anymore. If I found a really nice looking 24-bit cursor then I might use it again.
The first step is to create the four quadrants. I used the path geometry to create each quadrant, for example the first quadrant is:
Data="M0,0 C0,0 100,0 100,100 L0,100 z"
For reference , the path markup syntax is documented at MSDN. The shape starts at (0,0) and creates a curve point with a control point at (100,0) and end point at (100,100). Then there is a LineTo (0,100) and finally the close ("z") marker to complete the shape. While this is fairly easy through a graphical editor such as Expression Blend (blog), I created this manually so that I could have more control over the exact coordinates rather than relying on the graphical editor.
I repeated the shape for each quadrant by adjusting the coordinates as necessary. It is also possible to simply apply a rotation to the shape so you have rotations of 90, 180, and 270 degrees, but remember to set the RenderTransformOrigin to the correct corner such as (0,1) to rotate by the bottom left corner.
Once the quadrants are built, then I placed a rotating quadrant named 'Spinner' over the combined shape. I used an alpha mask on the spinner to differentiate it from the other quadrants. Finally, I added a Storyboard to rotate the 'Spinner' 360 degrees every two seconds:
<Storyboard x:Name="SpinStoryboard" RepeatBehavior="Forever">
<DoubleAnimationUsingKeyFrames BeginTime="00:00:00" Storyboard.TargetName="Spinner"
Storyboard.TargetProperty="(UIElement.RenderTransform).(TransformGroup.Children)[2].(RotateTransform.Angle)">
<SplineDoubleKeyFrame KeyTime="00:00:02" Value="360"/>
</DoubleAnimationUsingKeyFrames>
</Storyboard>
The end result is a nice looking animation that is fairly simple. I'm tempted to try to duplicate the Mac OS X wait cursor, but that will have to wait until I have more time.
Technorati tags:
Silverlight
Sunday, September 14, 2008
Recently I wanted to make a very simple sample in Silverlight that used a little code to animate bouncing balls. The overall effect is fairly simple, but getting the sample down to the basics took a little time. As part of my research, I looked at several old bouncing ball demos using JavaScript and it was an eye-opening reminder of the dark ages of browsers and JavaScript.
For the sample, I wanted to keep everything very simple. I started with a circle (ellipse with the same height and width) in a canvas. In order to move the ball, I chose to position the ball at (0, 0) and use a TranslateTransform to adjust the position. You could easily move the ball around with Left and Top, but I also plan to use RotateTransform in later samples so it makes sense to use a TransformGroup to manipulate the object.
The XAML for the base page is just a Canvas and a rectangle that acts as a border. The ball is placed on the canvas using the common TransformGroup that is created by Expression Blend when you add a transform to an object. Expression Blend also uses the same order for the transforms (ScaleTransform, SkewTransform, RotateTransform, TranslateTransform) and will reorder the transforms back to this order if you adjust any transform using Blend (at least as of Blend 2.5 June 2008 Preview). Here is the XAML for the page:
<UserControl x:Class="BounceTest.Page"
xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation"
xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml"
Width="400" Height="300">
<Canvas x:Name="LayoutRoot">
<Rectangle x:Name="Boundary" Height="300" Width="400"
Canvas.Left="0" Canvas.Top="0"
Fill="#FFFFFFFF" Stroke="#FF000000" />
<Ellipse x:Name="Ball01" Height="25" Width="25"
Canvas.Left="0" Canvas.Top="0"
Fill="#FFEE3131" Stroke="#FF000000"
RenderTransformOrigin="0.5,0.5">
<Ellipse.RenderTransform>
<TransformGroup>
<ScaleTransform/>
<SkewTransform/>
<RotateTransform/>
<TranslateTransform X="50" Y="50"/>
</TransformGroup>
</Ellipse.RenderTransform>
</Ellipse>
</Canvas>
</UserControl>
I created a small class to manage the position and velocity of the ball. The constructor takes the source shape (the ball), the velocity, and the boundary shape. For this scenario, I assume that the ball already has the RotateTransform otherwise it throws an exception. The code could easily create a RotateTransform if it isn't present. The Update method takes the elapsed time since the previous method call. The boundary checking is extremely simple and just changes the direction of the motion when the ball hits a boundary edge. This obviously needs a lot of work for more complex scenarios, but it will do for this simple case.
public class ShapeVelocity
{
public Shape shape;
public Vector velocity;
public TranslateTransform translate;
public Size bounds;
public Size container;
public ShapeVelocity(Shape AShape, Vector AVelocity, Shape BoundsContainer)
{
this.shape = AShape;
this.velocity = AVelocity;
var renderTransform = this.shape.RenderTransform;
if (renderTransform is TransformGroup)
{
TransformGroup transformGroup = (TransformGroup)renderTransform;
foreach (Transform transform in transformGroup.Children)
if (transform is TranslateTransform)
this.translate = (TranslateTransform)transform;
}
if (this.translate == null)
throw new ArgumentException("Shape must have a TranslateTransform in it");
container = new Size(BoundsContainer.Width, BoundsContainer.Height);
bounds = new Size(
this.shape.ActualWidth + this.shape.StrokeThickness,
this.shape.ActualHeight + this.shape.StrokeThickness);
}
public void Update(TimeSpan Interval)
{
Rect pos = new Rect(
translate.X,
translate.Y,
bounds.Width,
bounds.Height);
if ((velocity.X < 0.0) && (pos.Left < 0.0))
velocity.X = -velocity.X;
else if ((velocity.X > 0.0) && (pos.Right > container.Width))
velocity.X = -velocity.X;
if ((velocity.Y < 0.0) && (pos.Top < 0.0))
velocity.Y = -velocity.Y;
else if ((velocity.Y > 0.0) && (pos.Bottom > container.Height))
velocity.Y = -velocity.Y;
translate.X += velocity.X * (double) Interval.Milliseconds / 1000.0;
translate.Y += velocity.Y * (double) Interval.Milliseconds / 1000.0;
}
}
In anticipation of future examples, I used a List<Shape> collection to store the ball shape to updae the position. For the motion, I used the StoryBoard instead of a DispatcherTimer based on the recommendation of Adam Kinney.
public partial class Page : UserControl
{
private DateTime _lastTime = DateTime.MinValue;
private double _initialSpeed = 50.0;
private Storyboard _storyboard;
private List<ShapeVelocity> _shapes;
public Page()
{
InitializeComponent();
_shapes = new List<ShapeVelocity>();
_storyboard = new Storyboard();
_storyboard.Duration = TimeSpan.FromMilliseconds(10);
_storyboard.Completed += new EventHandler(storyboard_Tick);
this.Loaded += new RoutedEventHandler(Page_Loaded);
}
void Page_Loaded(object sender, RoutedEventArgs e)
{
_shapes.Add(new ShapeVelocity(Ball01, new Vector(_initialSpeed, _initialSpeed), Boundary));
_lastTime = System.DateTime.Now;
_storyboard.Begin();
}
void storyboard_Tick(object sender, EventArgs e)
{
DateTime now = System.DateTime.Now;
TimeSpan interval = now - _lastTime;
foreach (ShapeVelocity s in _shapes)
{
s.Update(interval);
}
_lastTime = now;
_storyboard.Begin();
}
}
So that is the very simple bouncing ball sample.
Technorati tags:
C#,
Silverlight
Thursday, July 31, 2008
A co-worker gave a presentation on a project we worked on a long time ago and he asked me to update an feature in the web application. The web app was originally programmed in ASP and I had almost forgotten how painful it was to work in that environment. The new request was to enable live presence information in a web part using the new RTC libraries. Fortunately, another developer had a component to perform all of the calls using the RTC libraries, but I needed to retrieve the group members for the selected group and cross-reference the group members with online presence.
The actual details involved opening "MTSAdmin.Catalog.1" object and looking for a specific COM+ package, enumerating the "RolesInPackage", and then getting the "UsersInRole". There were several layers of filtering going on and I wanted an elegant solution. Unfortunately I kept trying to be too elegant and I was getting hung up on how to make it more elegant. I wanted to have a "before", "item", and "after" callback functions at each of the function depth levels where the "item" function would occur in the iteration loop. This would allow for scenarios where items might need to be collected after iteration, particularly to avoid making multiple unnecessary calls to the lower filtering methods.
Small example from jQuery
ajax: function( s ) {
// Extend the settings, but re-extend 's' so that it can be
// checked again later (in the test suite, specifically)
s = jQuery.extend(true, s, jQuery.extend(true, {}, jQuery.ajaxSettings, s));
// ... snip, snip, snip ...
// Allow custom headers/mimetypes
if ( s.beforeSend && s.beforeSend(xhr, s) === false ) {
// cleanup active request counter
s.global && jQuery.active--;
// close opended socket
xhr.abort();
return false;
}
The example above from jQuery only uses the "before" concept that I mentioned above, but it shows a very nice implementation where the 's' object is based on jQuery.ajaxSettings. Just before the AJAX call is about to be made, jQuery checks for any custom headers and calls the 'beforeSend' function if it exists.
Start Simple and Add Progressive Elegance
I did manage to get the function working correctly, but I had to take a step back and start simple and progressively build additional functionality into the component. Sometimes you have inspiration and you can just design and code and everything is great. Other times it can take a while before the inspiration comes. It is better to start coding and have something you can refine rather than wait and lose valuable time.
To use a baseball analogy: Don't try to hit a homerun every time; sometimes you need to play "small ball" and get people in scoring position and just try to make contact. A similar soccer (futból!) analogy would be don't just try to dribble down the field and score every time, but you need to pass and set up the players around you to maximize the scoring opportunity.
Friday, June 13, 2008
I was working through a Communications sample of connecting SilverLight to POX, then Web Services, and finally WCF and I came across the following instructions:
(Part 1: POX) The Generic handler will process an incoming request using the code declared in the ProcessRequest function. This function should create some new instances of CityData and add them to the myCities list. Here are some examples of cities with longitude and latitude { (London, 51.5, 0), (Stratford-upon-Avon, 52.3, -1.71), (Edinburgh, 55.95, -3.16) }. See if you can write the code to do this.
(And later in the lab...)
In this example your ASHX served up hard-coded data for 1 city. Can you build it so that it can accept parameters on the URI string (i.e. http://localhost:8001/Sample1Web/GetData.ashx?city=whatever)?
(And even further in the lab...)
Write a function that takes in a country and builds a List<CityData> of several cities for that country. Here are some cities and their latitudes and longitudes:
("Paris", 48.87, 2.33);
("Lourdes", 43.1, 0.05);
("Toulouse", 43.6, 1.38);
("London", 51.5, 0);
("Stratford-Upon-Avon", 52.3, -1.71);
("Edinburgh", 55.95, -3.16);
("Berlin", 52.52, 13.42);
("Munich", 48.13, 11.57);
("Hamburg", 53.58, 9.98);
Provided Solution
The sample provided "solution" code at the end of the instructions. The code is basic and unimaginative, but perhaps that is all the is needed of sample code (I would offer that samples are the perfect to push the boundaries -- lead by example).
// File: CityData.cs
public class CityData
{
public string CityName { get; set; }
public double Latitude { get; set; }
public double Longitude { get; set; }
public CityData(string strCityName, double nLatitude, double nLongitude)
{
CityName = strCityName;
Latitude = nLatitude;
Longitude = nLongitude;
}
public CityData()
{
}
}
// File: GetData.ashx
public class GetData : IHttpHandler
{
private List<CityData> getCities(string strCountry)
{
List<CityData> ret = new List<CityData>();
switch (strCountry)
{
case "france":
ret.Add(new CityData("Paris", 48.87, 2.33));
ret.Add(new CityData("Lourdes", 43.1, 0.05));
ret.Add(new CityData("Toulouse", 43.6, 1.38));
break;
case "uk":
ret.Add(new CityData("London", 51.5, 0));
ret.Add(new CityData("Stratford-Upon-Avon", 52.3, -1.71));
ret.Add(new CityData("Edinburgh", 55.95, -3.16));
break;
case "germany":
ret.Add(new CityData("Berlin", 52.52, 13.42));
ret.Add(new CityData("Munich", 48.13, 11.57));
ret.Add(new CityData("Hamburg", 53.58, 9.98));
break;
default:
ret.Add(new CityData("London", 51.5, 0));
ret.Add(new CityData("Stratford-Upon-Avon", 52.3, -1.71));
ret.Add(new CityData("Edinburgh", 55.95, -3.16));
break;
}
return ret;
}
}
Alternative 1
I wanted to explore alternatives to the supplied code, so I decided to initialize the CityData list with the data instead of using "Add". In C++ this was more efficient (it set the initial list size instead of using the default capacity sizing algorithm among other things), but I'm not sure of performance in terms of C#. I would not expect any significant performance difference with this small sample though.
public List<CityData> getCities(string strCountry)
{
List<CityData> ret = new List<CityData>();
switch ( strCountry.ToLower() )
{
case "france":
ret = new List<CityData>() {
new CityData("Paris", strCountry, 48.87, 2.33),
new CityData("Lourdes", strCountry, 43.1, 0.05),
new CityData("Toulouse", strCountry, 43.6, 1.38)
};
break;
case "germany":
ret = new List<CityData>() {
new CityData("Berlin", strCountry, 52.52, 13.42),
new CityData("Munich", strCountry, 48.13, 11.57),
new CityData("Hamburg", strCountry, 53.58, 9.98)
};
break;
case "uk":
default:
strCountry = "UK";
ret = new List<CityData>() {
new CityData("London", strCountry, 51.5, 0),
new CityData("Stratford-Upon-Avon", strCountry, 52.3, -1.71),
new CityData("Edinburgh", strCountry, 55.95, -3.16)
};
break;
}
return ret;
}
Alternative 2
I haven't worked very much with LINQ, so I wanted to compare the same functionality using LINQ. I added a string "Country" to the CityData class and here is the result:
private static List<CityData> cityList = new List<CityData>()
{
// UK
new CityData("Paris", "UK", 48.87, 2.33),
new CityData("Lourdes", "UK", 43.1, 0.05),
new CityData("Toulouse", "UK", 43.6, 1.38),
// France
new CityData("Paris", "France", 48.87, 2.33),
new CityData("Lourdes", "France", 43.1, 0.05),
new CityData("Toulouse", "France", 43.6, 1.38),
// Germany
new CityData("Berlin", "Germany", 52.52, 13.42),
new CityData("Munich", "Germany", 48.13, 11.57),
new CityData("Hamburg", "Germany", 53.58, 9.98)
};
public List<CityData> getCities(string strCountry)
{
IEnumerable<CityData> data = from city in cityList
where city.Country == strCountry
select city;
return new List<CityData>(data);
}
This is still not the most efficient or optimal method, but I think it looks nice and is helps introduce LINQ in an understandable manner.
Conclusion
Samples and labs should be simple and yet it is a perfect time to include new technologies and push the boundaries so to speak. One thing I certainly don't want to see in samples are basic programming inefficiencies (things that make you cringe) due to lack of effort by the sample creator. That was not necessarily the case with this sample, but it would have been better to present alternatives such as the examples I listed.
Technorati tags:
C#,
LINQ
Tuesday, May 27, 2008
I'm in the process of updating an Excel spreadsheet that is failing when it is running inside of Internet Explorer. The issue is related to the ActiveSheet and other global properties having a value of Nothing when the code is assuming they have valid references. As I am going through this spreadsheet, I am noting a wide variety of programming deficiencies and inefficiencies. Here is a list of some of the issues encountered:
If you run spreadsheets under Internet Explorer, use Application.ThisWorkbook and to ensure the browser evaluates the reference correctly.
NOTE: This is not an exhaustive list, nor do I claim that the "Better" examples are the best code example, but rather they are (hopefully) significant improvements over the "Bad" examples. In many cases, I have kept the design of the bad example so that you can see the differences however you should strive to completely refactor the code if possible.
-
Do not abuse the "With" statement in VBA.
Some people like "With" and some people don't and I don't particularly care for it. In particular, do not use nested "With" statements. Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
With ActiveSheet
With TestForm
' ...
End With ' TestForm
End With ' ActiveSheet
' ... BAD EXAMPLE: DO NOT USE ...
-
Avoid using the "Goto" statement in VBA.
"Goto" is almost always a bad idea and a sign of spaghetti code. There are very few cases were it is appropriate to use it -- very few cases!
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
For i = iHere To lRealLastRow Step 2
BuildStr = "A" & i + 1
Range(BuildStr).Select
If Range(BuildStr).Value = "---" Then GoTo DoneWithNames
SelectItemDlg.Items_List.AddItem Range(BuildStr).Text
Next
DoneWithNames:
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code (I still would have designed it differently)
' ...
Dim MyCell As Range
For i = iHere To lRealLastRow Step 2
Set MyCell = Cells(i + 1, 1)
If MyCell.Value = "---" Then
Exit For
End If
SelectItemDlg.Items_List.AddItem MyCell.Text
Next
' ...
-
Avoid using the name of a Form inside the Form code
The current form is implied in the code. It is not necessary to use the form name to reference controls on the form. If you want to differeniate form controls, you can use the 'Me' keyword such as: Me.Hide
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
' This example is doubly bad because it uses the form name (ChartConfig)
' both explicitly and in a With statement
If ChartConfig.Variable1.Text = ChartConfig.Variable2.Text Then
MsgBox "Variable1 cannot be same as Variable2. Please choose another variable type.", vbOKCancel
'Turn off the Change effect on Variable2.
changeCasevar = False
With ChartConfig
.Variable2.Text = ""
.Variable2Desc.Text = ""
.Variable2Uom.Text = ""
End With
'Turn back on the Change effect on Variable2.
changeCasevar = True
Exit Sub
'
End If
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code (I still would have designed it differently)
' ...
If Variable1.Text = Variable2.Text Then
MsgBox "Variable1 cannot be same as Variable2. Please choose another variable type.", vbOKCancel
'Turn off the Change effect on Variable2.
changeCasevar = False
Variable2.Text = ""
Variable2Desc.Text = ""
Variable2Uom.Text = ""
'Turn back on the Change effect on Variable2.
changeCasevar = True
Exit Sub
End If
-
Use Error Handling and Resume to restore Application.ScreenUpdating
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
' There are quite a few global variables and a pass-by-reference
' boolean that should just be the function return value.
Application.ScreenUpdating = False
iOption = "Single"
x = ChartPicker.Input_Val_Text
Call CheckUserRangeInputsAndOutputs(iOption, OkToChart)
iOption = ""
If OkToChart = True Then
With ChartConfig
CreateChartingArrays
End With
End If
Application.ScreenUpdating = True
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code (I still would have designed it differently)
' ...
On Error GoTo ErrorHandler
Application.ScreenUpdating = False
If CheckUserRangeInputsAndOutputs("Single") Then
ChartConfig.CreateChartingArrays
End If
Exit_Handler:
Application.ScreenUpdating = True
Exit Sub
ErrorHandler:
MsgBox "Error " & Err.Number & ": " & Err.Description
Resume Exit_Handler
-
Use If, Else If, and Else correctly.
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
If TextBoxMinimum.Text = "" Then
MsgBox "Must enter Minimum Value before proceeding", vbOK
Result = False
Exit Function
End If
If TextBoxMaximum.Text = "" Then
MsgBox "Must enter Maximum Value before proceeding", vbOK
Result = False
Exit Function
End If
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code
' ...
' For a explicit method, you can directly refer to the controls
Result = False
If Not IsNumeric(TextBoxMinimum.Text) Then
MsgBox "Must enter Minimum Value before proceeding", vbOK
TextBoxMinimum.SetFocus
ElseIf Not IsNumeric(TextBoxMaximum.Text) Then
MsgBox "Must enter Maximum Value before proceeding", vbOK
TextBoxMaximum.SetFocus
Else
Result = True
End If
' If you can use a generic method, you can iterate through all of the controls
' There are obviously better ways of doing this, especially with .NET
Dim ValidateControl
Dim ControlList = Array(TextBox1, TextBox2, TextBox3)
For Each ValidateControl In ControlList
If Input_Val_Text = Null Or Input_Val_Text = "" Then
MsgBox "Must enter Value in " & ValidateControl.Name & " before proceeding", vbOK
Result = False
Exit Function
End If
Next
-
Avoid using the ActiveSheet, Range, Cells, and other global variables.
Rather than assuming a position, it is much better to set a variable to the desired
Workbook, Worksheet, or Range.
In particular, if there is any chance the spreadsheet might run under a browser,
use Application.ThisWorkbook to ensure the browser
evaluates the macro correctly.
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
'NOTE: Arrays start with 1st position as "0", not "1"
'Thus ColLtrs(0) is not used but placed there as a spacer for letter to column alignment:
ColLtrs = Array("", "A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W", "X", "Y", "Z")
With ActiveSheet
Range(ColLtrs(iOutCol) & iOutRow).Select
If UnitSystem = "Metric" Then
iRow = ActiveCell.Row
For i = iRow To iLastRow - 1
currCell = ColLtrs(ActiveCell.Column) & i + 1
If Range(currCell).Value <> "" Then
'capture the value from original calculatated data
Range(currCell).Select
strOrigValueCell = ColLtrs(iResultsCol) & ActiveCell.Row
Result = ConvertUom(Range(strOrigValueCell))
End If
Next i
End If
End With
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code
' ...
Dim mySheet As Worksheet
Dim Value As Variant
Set mySheet = ActiveSheet
For i = iRow To iLastRow - 1
Value = mySheet.Cells(i, iOutCol).Value
If Value <> "" Then
Result = ConvertUom(UnitSystem, Value)
End If
Next i
' To get a column letter/code, do not reinvent the wheel!
' Various functions from:
' http://www.dicks-blog.com/archives/2004/05/21/column-numbers-to-letters/
Function ColLetter(ColNumber As Long) As String
On Error Resume Next
ColLetter = Application.Substitute(Application.ConvertFormula("R1C" & ColNumber, xlR1C1, xlA1, 4), "1", "")
End Function
Function ColumnLetter(ByVal c As Long) As String
Dim p As Long
While c
p = 1 + (c - 1) Mod 26
c = (c - p) \ 26
ColumnLetter = Chr$(64 + p) & ColumnLetter
Wend
End Function
-
Avoid using .Select and moving selection in macros unless absolutely.
Do not change the current sheet or selection unless that is the intention of the macro.
In general, don't mess with the user, flip sheets, or other heinous acts and
try to leave things the way you found them.
Example of Bad Code
' ... BAD EXAMPLE: DO NOT USE ...
For i = 1 To 10
Range("A" & i).Select
ActiveCell.Value = "Row " & i
Next i
' ... BAD EXAMPLE: DO NOT USE ...
Example of Better Code
Dim iColumn As Long
Dim mySheet As Worksheet
Dim myRange As Range
iColumn = 1
Set mySheet = ActiveSheet
For iRow = 1 To 10
Set rng = mySheet.Cells(iRow, iColumn)
rng.Value = "Row " & iRow
Next iRow
Thursday, April 24, 2008
I was building a quick test web site and I was using the aspnet_regsql tool to add membership to a SQLEXPRESS database. At first, I tried:
aspnet_regsql -A all -C "Data Source=.\SQLEXPRESS;Integrated Security=True;User Instance=True" -d "C:\code\Test\APP_DATA\aspnetdb.mdf"
For some reason, the SqlConnection insisted that it try to create the database and disregarded the full path to the database (note the path in the exception).
SQL Exception:
System.Data.SqlClient.SqlException: Directory lookup for the file "C:\Documents and Settings\user\Local Settings\Application Data\Microsoft\Microsoft SQL Server Data\SQLEXPRESS\C:\code\Asp.net\ServerControlTest\App_Data\Database.mdf" failed with the operating system error 123(The filename, directory name, or volume label syntax is incorrect.).
CREATE DATABASE failed. Some file names listed could not be created. Check related errors.
Creating the C:\code\Asp.net\ServerControlTest\App_Data\Database database...
at System.Data.SqlClient.SqlConnection.OnError(SqlException exception, Boolean breakConnection)
at System.Data.SqlClient.SqlInternalConnection.OnError(SqlException exception, Boolean breakConnection)
at System.Data.SqlClient.TdsParser.ThrowExceptionAndWarning(TdsParserStateObject stateObj)
at System.Data.SqlClient.TdsParser.Run(RunBehavior runBehavior, SqlCommand cmdHandler, SqlDataReader dataStream, BulkCopySimpleResultSet bulkCopyHandler, TdsParserStateObject stateObj)
at System.Data.SqlClient.SqlCommand.RunExecuteNonQueryTds(String methodName, Boolean async)
at System.Data.SqlClient.SqlCommand.InternalExecuteNonQuery(DbAsyncResult result, String methodName, Boolean sendToPipe)
at System.Data.SqlClient.SqlCommand.ExecuteNonQuery()
at System.Web.Management.SqlServices.ExecuteFile(String file, String server, String database, String dbFileName, SqlConnection connection, Boolean sessionState, Boolean isInstall, SessionStateType sessionStatetype)
Next I tried the simple commands (credit Scott Mitchell) that worked great:
sqlcmd -S localhost\SQLExpress -Q "EXEC sp_attach_db 'Foobar', N'pathToDBfile'"
aspnet_regsql.exe -S localhost\SQLExpress -d Foobar -E -A all
References
- Working with SQL Server 2005 Express Database
- Using ASPNET_RegSQL.exe with SQL Express databases in APP_DATA
In my SharePoint experiments with form-based authentication (FBA), I have been installing self-signed SSL certificates since I am developing in a virtual machine without a certificate authority. Last night, I shut down my virtual machine instead of the usual Suspend operation. This morning, I started the virtual machine but SharePoint wasn't working and said: Cannot connect to the configuration database.
I then tried to open SQL Server Management Console and tried to connect, only to receive the following message:
I didn't freak out, but I did check the SQL Server Service which was running, and then the SQL Server Log files which looked normal. There was one line which I didn't recognize:
The certificate was successfully loaded for encryption.
This is actually a normal line in the log file, but it was then that I realized that the self-signed certificates I created were interfering with the login process. I searched around for Internet resources, but finally found what I was looking for in the Certificate MMC snap-in configuration. Since I had created multiple SSL certificates, I deleted the unused certificates and checked the encryption and certificate settings in the SQL Server Configuration Manager. After a quick reboot, I was back in business.
References:
- Configuring Certificate for User by SSL
- How to enable SSL encryption for an instance of SQL Server
- SQL Protocols: Troubleshoot Connectivity Issues in SQL Server 2005, Part III
Wednesday, April 23, 2008
I recently set up an isolated, single-server SharePoint site with forms-based authentication (FBA) with the hopes that it would eliminate the need to create unrelated user accounts for all of the SharePoint users. The FBA setup and installation went fine and I used the SharePoint FBA tool on codeplex to administer the accounts and users which is great.
Unfortunately I didn't realize that disabling "Client Integration" in the SharePoint Cental Admin / Application Management / Authentication Providers would have such a dramatic effect on usability. Some of the challenges are using SharePoint Designer, the MySites functionality is impacted, inability to export list data to Excel, as well as inability to integrate with Outlook. The composite screen shot below shows the difference between enabled and disabled client integration.
After spending quite a bit of time configuring and setting up FBA, I don't want to abandon it immediately. I am looking into two options to use the FBA user store (ASPNETDB) with IIS, basic authentication, and SSL. The two likely candidates are MADAM (Mixed Authentication Disposition ASP.NET Module) and Custom Basic Authentication.
I will post updates as available.
Monday, April 21, 2008
I was looking at Douglas Crockford's 360 blog today and came across a hilarious quote:
Do you write regularly for any publications in this field?
Just the blogs.
In researching the book, did you come across any surprising facts, figures, or statistics that the press might be interested in?
Surprisingly, facts have very little to do with web development.
Douglas Crockford is simply amazing and is the original author of expat (one of the first XML parsing libraries), as well as JSLint, JSMin, and JSON (one of the primary components of AJAX).
P.S. Bonus points for any one identifying the blog post title source.