Re: invalidate() in winforms - C# Discussion Boards

Robert Rohde21-Jul-06 3:54

Robert Rohde

21-Jul-06 3:54

Hold a Bitmap with the size of your form and draw into it:

using (Graphics g = Graphics.FromImage(myBitmap)) {
   //...
}

Then in OnPaint just draw this bitmap onto your form:

e.Graphics.DrawImageUnscaled(myBitmap, 0, 0);

Generally speaking controls do not 'remember' what has been painted onto them. Each time they are invalidated (either by code or for example by another window being dragged over them) everything has to be repainted.

serimcc21-Jul-06 4:07

serimcc

21-Jul-06 4:07

thank you Robert Rohde
if they do not(or can not) remember what has been painted onto them , what is the function of invalidate()?Why do msdn say use invalidate to update form?Where do we use it?
Thanks again
Good works...

LongRange.Shooter21-Jul-06 4:24

LongRange.Shooter

21-Jul-06 4:24

The process of displaying your form is done through the process known as Invalidate(). This is called for just about any reason:
Your mouse moved on the form
A form moved on top of your form
the user resized your form

With any of these events, the Form does some very basic processes to determine if a control must invalidate or if the entire form invalidates. This controls the depth of paint events that occur.

You cannot just draw a line on the form and expect it to be integrated into the rest of the form paint event. The only way of doing this is by doing your own painting. This can be quite a pain if you have to deal with controls as well as just that one line you want to paint.

So then make your task easier:
Create your own user control and name it StaticLineControl. Have the control defined so that you do not care where your control is or where how big the line needs to be. Just do your own painting of a line centered on the control for the width of the control.

Now as a control, your 'line' exists in the proper z-axis in relationship with everything else. You still have just what you need and the work itself is reduced to an e.Graphic.DrawLine() call in your code instead of whatever you are in pain over now.

Remember the K.I.S.S. principal.

Josh Smith21-Jul-06 4:09

Josh Smith

21-Jul-06 4:09

Robert Rohde wrote:
Generally speaking controls do not 'remember' what has been painted onto them.

On a completely unrelated, yet interesting, tangent...

WPF uses a retained state rendering technique, which means that your control can explain to the system how it should be rendered once, and, in theory, it will never again for the life of the control. It uses vector-based rendering instructions (sort of like metafile) which are cached by the underlying composition engine! Big Grin | :-D

:josh:
My WPF Blog[^]

Judah Gabriel Himango21-Jul-06 6:02

Judah Gabriel Himango

21-Jul-06 6:02

Sounds cool. Cool | :cool:

Tech, life, family, faith: Give me a visit.
I'm currently blogging about: Messianic Instrumentals (with audio)
The apostle Paul, modernly speaking: Epistles of Paul

Judah Himango

Josh Smith21-Jul-06 8:12

Josh Smith

21-Jul-06 8:12

Judah Himango wrote:
Sounds cool.

It's a brave new world, Judah! The whole HWND system is replaced. Everything has changed. Cool | :cool:

:josh:
My WPF Blog[^]

Judah Gabriel Himango21-Jul-06 9:01

Judah Gabriel Himango

21-Jul-06 9:01

Josh Smith wrote:
The whole HWND system is replaced

Is that right? Are you sure it's not lurking there under the hood somewhere? I assume it is since it has to interop with Win32 based technologies like Windows Forms, MFC, etc.

Tech, life, family, faith: Give me a visit.
I'm currently blogging about: Messianic Instrumentals (with audio)
The apostle Paul, modernly speaking: Epistles of Paul

Judah Himango

Josh Smith22-Jul-06 3:36

Josh Smith

22-Jul-06 3:36

Judah Himango wrote:
Is that right? Are you sure it's not lurking there under the hood somewhere?

Positive. The only way that HWNDs are involved is in interop scenarios. Other than that, the HWND has no role in WPF. In fact, if you host an HWND-based control in a WPF application, it is not affected by any transforms, animations, etc. It just sits there like a dork in a Lamborghini. Smile | :)

:josh:
My WPF Blog[^]

Judah Gabriel Himango22-Jul-06 6:49

Judah Gabriel Himango

22-Jul-06 6:49

Josh Smith wrote:
In fact, if you host an HWND-based control in a WPF application, it is not affected by any transforms, animations, etc.

Very interesting. Thanks.

Josh Smith wrote:
It just sits there like a dork in a Lamborghini.

LOL!

Tech, life, family, faith: Give me a visit.
I'm currently blogging about: Goof around music jam with my brothers (with video)
The apostle Paul, modernly speaking: Epistles of Paul

Judah Himango

Decoupling components

WillemM21-Jul-06 2:37

C# async Socket howto read nutil buffer is empty

21-Jul-06 2:37

I have always been wondering about this subject.
What's the best way to decouple components from different layers?

I have been thinking about a method like the following:

Data access: Components implement interfaces from the Common namespace
Businesslogic: Components refer to the interfaces from the Common namespace
Webservice: Links data access components to the Business logic by assigning them to the right properties (which are typed using the common interfaces).

When not in a distributed environment, you can replace Webservice by a normal servicing layer. It even provides options to make a proxy construction in which the client can choose between the webservice and the local service Smile | :)

Anyone have interesting thoughts on this?

WM.

What about weapons of mass-construction?

TUX2K21-Jul-06 2:12

Re: C# async Socket howto read nutil buffer is empty [modified]

21-Jul-06 2:12

I have been searching the net, the code project en google.
But I just cant find an answer to my question:
I create a ansync sokcet connection adn use BeginRead te read data from the socket.
like this:

/// <summary><br />
/// Tell the stream to recieve data<br />
/// </summary><br />
private void Receive(){<br />
  _Stream.BeginRead(_Buffer, _Offset, _Count, new AsyncCallback( RecieveCallBack ), _Stream);<br />
}<br />
		<br />
/// <summary><br />
/// The steam has received data<br />
/// </summary><br />
/// <param name="asyncResult"></param><br />
private void RecieveCallBack( IAsyncResult asyncResult ){<br />
	try{<br />
		int _ReceivedBytes;<br />
		_ReceivedBytes =  _Stream.EndRead( asyncResult); <br />
		if( _RecievedBytes > 0 )<br />
		{<br />
			string data = System.Text.Encoding.UTF8.GetString( _Buffer, _Offset, _RecievedBytes);<br />
			<br />
			//Lets pass the data on to the OnRecive event<br />
			FireOnReceive(data);<br />
			// Call recieve to setup the beginread<br />
			if ( IsConnected )<br />
				Receive();<br />
		}<br />
	} catch(ObjectDisposedException) {<br />
		// The object is already Disposed<br />
		// there is nothging more to do<br />
		return;<br />
	} catch(System.IO.IOException e) {<br />
		FireOnError( new ConnectionException(e.Message , ConnectionErrorType.ReceiveError ) );<br />
		}<br />
}

The problem is that it reads only as much as the _Offset allows, this is set 1024 byte.
But if the received is bigger than the offset the socket blocks and i only get half the data i need.
How do I get the complete data form the socket, I looked at some examples on code project bud they all us other method of getting the data form the socket object.

-- modified at 8:29 Friday 21st July, 2006

WillemM21-Jul-06 2:45

Re: C# async Socket howto read nutil buffer is empty

21-Jul-06 2:45

Here's a simplified sample for the received callback.
You forgot the resume reading from the stream. This causes the application to stop receiving data. Also you use the offset, this isn't necessary since you simply fill the receive buffer again from the start.

private void dataReceived(IAsyncResult asyncResult) {<br />
  //Stop reading<br />
  _receivedbytes = _stream.EndRead(asyncResult);<br />
  <br />
  //Process received data<br />
  string data = parseData(_buffer,0,_receivedbytes);<br />
  <br />
  //Start reading again<br />
  _stream.BeginRead(_buffer,0,_bufferlength,new AsyncCallback(dataReceived));<br />
}

WM.

What about weapons of mass-construction?

-- modified at 8:45 Friday 21st July, 2006

TUX2K21-Jul-06 3:08

Re: C# async Socket howto read nutil buffer is empty

21-Jul-06 3:08

Greate this will work.
But still i have a question, the _Buffer is byteArray and has en fixed size.
So that buffer does nothing?.
I gues that this is still the array that holds the bytes, but how to makes this one dynamic?

WillemM21-Jul-06 3:11

Re: C# async Socket howto read nutil buffer is empty

21-Jul-06 3:11

That buffer functions as a temporary place to store the received data.
You can then decode it using a StringBuilder or write it to a memorystream.
This is much more efficient than making the buffer dynamic, array resizing is expensive Smile | :)

WM.

What about weapons of mass-construction?

TUX2K21-Jul-06 3:51

Re: C# async Socket howto read nutil buffer is empty

21-Jul-06 3:51

I get that the buffer is only temporary.
But when you call the beginread fuction the buffer needs a size.
Like in you example:

//Start reading again<br />
_stream.BeginRead(_buffer,0,_bufferlength,new AsyncCallback(dataReceived));

bufferlength is a int, but how big is it, 1024? But I want te read till the end!

WillemM21-Jul-06 4:28

Re: C# async Socket howto read nutil buffer is empty

21-Jul-06 4:28

To what end do you want to read? A network stream potentially has no end.
So this would mean read until you get an error stating the connection was closed.

The method does just that, it will throw an exception when the connection is closed. Normally I would check for a special character during the decoding of the received data to check if I need to close the connection and stop reading.

WM.

What about weapons of mass-construction?

TUX2K21-Jul-06 4:53