|
Hello,
I manage to make it work.
After connect, read or write I check the _tcp.Client.Poll(...).
|
|
|
|
|
Normally, when no actual data is being exchanged on a TCP connection, no packets will be sent in either direction. If the connection is lost, one may wait forever to receive data without any clue that anything is wrong.
Most TCP stacks, however, offer a 'keepalive' time setting. When this is enabled and the indicated amount of time has elapsed since the last packet exchange, the "bored" machine will pretend that it hasn't received the acknowledgment for the last byte of data it sent, and will thus resend it with an indication that it is a retransmission that needs acknowledgment. The other machine should receive this and acknowledge it appropriately. If the other machine has dropped the connection but is presently running, it will immediately send a reset and kill the connection. If the other machine is not running or is not accessible, the "bored" machine will detect the acknowledgment failure after a minute or so and drop the connection itself.
|
|
|
|
|
how can edit the reports at run time?
bah10z
|
|
|
|
|
Hi,
I am designing an architecture in .NET framework. I use the Three tier model to design my application but I have a concern about how is the way to properly model the access to database layer. I have separate DLL for each layer.
In my business layer I want to access the database, for example to add new registry in a way that the members are not passed each one in a parameter but all in a single object. For example:
namespace myapp.bussines_layer
{
class Point
{
int x;
int y;
MyDatabase db = new MyDatabase();
void InsertInDatabase()
{
db.NewRowInDB(this);
}
void InsertInDatabase()
{
db.NewRowInDB(this.x, this.y);
}
}
}
The thing is that this way I have circular dependency because Point knows MyDatabase; and MyDatabase has to know Point in order to recognize Point as an parameter in the function.
Does my idea fit correctly in the good practices and designs to access databases? Is better design to pass all the parameters as native types (even when the amount os parameter is high)??
I would be very grateful if anyone could help me and/or give me references to good design patterns with the connection between business and database layers.
Thank you very much in advance.
Julen.
|
|
|
|
|
Julen wrote: The thing is that this way I have circular dependency because Point knows MyDatabase; and MyDatabase has to know Point in order to recognize Point as an parameter in the function.
Point is a predefined type; it is declared externally of both DLL's. You can do the same thing for your own classes, to avoid circular references.
Julen wrote: I would be very grateful if anyone could help me and/or give me references to good design patterns with the connection between business and database layers.
The dOOdads[^] architecture was a nice example. It's deprecated by now, but it's clean and readable code. Fastest way to dive into that code is to use the myGeneration application to generate a small testapplication that contains them dOOdads.
I are troll 
|
|
|
|
|
Eddy Vluggen wrote: Point is a predefined type; it is declared externally of both DLL's. You can do the same thing for your own classes, to avoid circular references.
In this case Point is just an example; my class is one of my own that has at least 10 members. If I include this class in both DLL, would'n I be breaking the separation between the business and data layer?
|
|
|
|
|
Create a separate DLL for OwnClass, and reference it from both assemblies. The separation is one of responsibilities, not of classes; you're going to pass data from one DLL to another (or, from one layer to another) - and you'll need some common ground for them to talk to each other.
In pseudocode;
class Person { string Name; }
class DALPerson: Person
{
public DALPerson(string connectionString)
{
Name =
}
}
class BLPerson: DALPerson
{
public override ToString()
{
return "High Lord " + Name;
}
}
We have met the enemy, and he is us
|
|
|
|
|
its very simple 
|
|
|
|
|
Once you know it, yes
I are troll
|
|
|
|
|
Julen wrote: would'n I be breaking the separation between the business and data layer?
Not necessarily.
As mentioned by others, the class could be external to both layers, or it could be defined in the lowest layer that uses it (probably the business layer).
And layers aren't necessarily tiers; don't confuse the two concepts.
|
|
|
|
|
Oh; that's nice to say because I thought both concepts mean the same. What is the difference?
|
|
|
|
|
The logical layers are the conceptual division, the tiers are the physical division.
(Source[^])
I are troll
|
|
|
|
|
Julen wrote: I use the tree tier model
It's "three-tier", as in the number 3, not "tree tier".
modified on Sunday, June 21, 2009 11:48 PM
|
|
|
|
|
Dave Kreskowiak wrote: It's "three-tier", as in the number 3, not "tree tier".
Sorry for the spelling. 
I don't dislike creating a separate DLL to come to a common ground between layers. So, if I have three tier architecture; will I have 5 DLL? Three for the layers themself; and two for the common grounds?
Presentation layer -- Common ground 1 -- Bussiness layer -- Common ground 2 -- Data Layer
Could recommend me an article, reference, matter on this issue? 
Thanks!
|
|
|
|
|
Create a Common library for Data Transfer objects and share it among all layers.
So this would be
(Data Transfer Objects) (Data Transfer Objects)
Data Layer -----------------------> Business Layer ------------------------> UI Layer
|
|
|
|
|
Shukla Rahul wrote: Create a Common library for Data Transfer objects and share it among all layers.
So this would be
(Data Transfer Objects) (Data Transfer Objects)
Data Layer -----------------------> Business Layer ------------------------> UI Layer
To create those Data Transfer Objects, is there a pattern or standard design to create them? Also, regarding the picture; do you mean with the arrows that the data access layer has a reference to business layer instead of reverse?
Thanks!
|
|
|
|
|
You can create a project class library project which holds such classes. Provide the reference of this project in all layers.
Sorry, the arrow shows the flow of the data objects not the reference.
Reference would be as follows
Data Access Layer - Ref of common
Business Layer - Ref of Data Access Layer and Ref of Common
UI - Ref of Business Layer & ref of Common
Download sample project from here.
-Rahul Shukla
|
|
|
|
|
This sounds like a serious case of entanglement. Why should Point know about MyDatabase and why should MyDatabase know about Point? What you might want to look at instead, is using interfaces to control the way things interact, and then code against the interface.
"WPF has many lovers. It's a veritable porn star!" - Josh Smith As Braveheart once said, "You can take our freedom but you'll never take our Hobnobs!" - Martin Hughes.
My blog | My articles | MoXAML PowerToys | Onyx
|
|
|
|
|
Using VB.net...
I've got an owner drawn toolstrip combobox that displays my system fonts. Each item in the drop down list is in its respective font. Works okay, but the top of the combo where the selection is displayed also uses whatever font is selected. How do I get the combobox to use its default font at the top and still have the drop down list display multiple fonts?
Also, the "Cambria Math" font in Vista displays as being about three lines tall, with tiny text in the middle of the line. That's in my text editor, OO.org and MS Works. A friend tried it in Word and it just produced onscreen gibberish. Anyone else encounter this with Cambria Math?
AB
|
|
|
|
|
every time i run the program an i press the send button i get an error " Exception from HRESULT "
Public Class Form1
Dim SendMessage As String
Dim GetMessage As String
Private Sub Form1_Load(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles MyBase.Load
Winsock1.LocalPort = 139
Winsock1.Listen()
End Sub
Private Sub Winsock1_ConnectionRequest(ByVal requestID As Long)
Winsock1.Close()
Winsock1.Accept(requestID)
End Sub
Private Sub Winsock1_DataArrival(ByVal sender As Long)
Winsock1.GetData(GetMessage)
txtMessage.Text = txtMessage.Text & GetMessage
End Sub
Private Sub btnSendMessage_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles btnSendMessage.Click
SendMessage = txtSendMessage.Text
Winsock1.SendData("server" & SendMessage)
End Sub
End Class
|
|
|
|
|
I'm on a project with several components (projects) which are deployed on multiple servers. Several components have their own web.config or app.config files with logging configuration settings, all flat file loggers. Some have config files with no logging configuration. Since all the trace listeners point to the same file, everything manages to get to the destination. This will change. I need to understand how the ELLB deals with multiple configuration files.
|
|
|
|
|
Taliesin99 wrote: This will change.
What will change?
Taliesin99 wrote: I need to understand how the ELLB deals with multiple configuration files.
Does it? I thought each application supplies it's own configuration?
|
|
|
|
|
Sorry for my lack of clarity.
There are several projects withing this solution. Many of them provide their own app.config or web.config files. At the moment, each of the categories in all of configuration files ultimately point to the same flat file, trace.log. So all the entries are written to the same file, at least the same file on a particular server. As requirements are changing, different projects need to point to files named something other than trace.log. There will be multiple log files in use. Also, there will be situations where different projects will still need to write to the same log file. Right now, that means managing multiple configuration files, which I would like to avoid if possible.
There are also projects in the solution which make calls to logger.write, but have no config file defined within that project. All the projects have a reference to what I assume are the same binaries for the Application Logging Block (we are not extending that code in any way). The logging in these projects appears to be working, so my assumption is that they are finding the categories defined in some other project and making to the log file. This may be an incorrect assumption on my part.
I guess what I am really looking for is a way to manage the logging across multiple projects (assemblies) within the same solution using as few configuration files as possible, ideally one. This may not be possible. But I have not yet found any documentation that discusses this situation.
|
|
|
|
|
Taliesin99 wrote: I guess what I am really looking for is a way to manage the logging across multiple projects (assemblies) within the same solution using as few configuration files as possible, ideally one.
I’m not an expert on the logging block but it wouldn’t seem likely that any sort of organization like you describe is part of it. There would be far too many variables to account for in a generalized library. If you want some sort of central management of logging configuration across your suite of applications you will probably have to develop that yourself.
I don’t have your complete requirements (and I don’t want them), however one thing you might consider is that a build system could potentially allow you to develop a centralized configuration that will automate the distribution of information into the project files. I’m pretty sure you could use ANT to accomplish that so I imagine MS Build might work as well.
|
|
|
|
|
How do I go about using filtering on the built in trace listeners, such as System.Diagnostics.DefaultTraceListener and System.Diagnostics.TextWriterTraceListener?
I don't want to have to override write methods and explicitly check the filtering, but I can find no way to attach a level to trace information?
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
