|
I care, C# plz 
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
|
|
|
|
|
If you can wait a couple of days, I'll knock something up this weekend. I'm a touch slammed right now.
|
|
|
|
|
Even a list of links would be great, thanks!
|
|
|
|
|
I'll probably pop it on my blog.
|
|
|
|
|
I think I've solved my issue  Took me a while to figure it out, but I now have a working solution. I'll post it on my own blog when I have the time to do it. Posted here.[^]
What my problem was, I was using MarshalByRef on the objects I tried to transmit, at least I'm guessing this was the culprit. The working solution for me was to replace MarshalByRef with [Serializable] and use the tcp-channel instead of the http-channel. In addition, in the client I replaced the code which registered the server object as WellKnownClientType with the Activator.GetObject() routine. The server object is still using MarshalByRef and is in the Singleton mode. Now I have a working client/server. I'm not sure if my solution is "legal", but it gets the job done
-edit with blog link-
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
modified on Sunday, April 17, 2011 5:33 PM
|
|
|
|
|
I've put up a demo project with my implementation here:
Clicky![^]
Hope it helps.
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
|
|
|
|
|
Helfdane wrote: 1) no webserver present (or other software, just the .net runtime)
Presumably you mean that you only want to use .Net to create the server. Because per your original requirements there is a server and a client. And if there is no server then the requirement does not exist.
Helfdane wrote: 2) need to change port & protocol at runtime dynamically.
What specific business need requires this?
Note that there is a signficant different between being able to run a server using a configured port and changing that same port (and handling clients as well) while the server is running.
|
|
|
|
|
The user should be able to change the port number in a console, not while the server is running. It happens often enough that some complementary software from another supplier takes the port we configured. So the sysadmin at the client site should be able to adjust it. Config files are a bridge too far and they get lost in them.
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
|
|
|
|
|
First what is "often enough"?
Once day or once a year?
Second that point still doesn't mean that the port needs to be changed while the application is running. One can provide a GUI to modify a config file and then provide a management app to bounce the server.
Changing the port while running means that you MUST deal with the complication of txns in flight. And also deal with providing a timely notification to clients.
|
|
|
|
|
read my post again
The service is not necessarily running when the port is changed.
And about often, if for every issue I need to send out a consultant to fix it, then even once a year is too often. If I could control everything from A to Z, then this would not be necessary, but I always try make sure the software is as robust as possible.
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
|
|
|
|
|
Helfdane wrote: read my post again
I suggest you read your own post again....
"port & protocol at runtime dynamically."
|
|
|
|
|
First what is "often enough"?
Once day or once a year?
[quote]And about often, if for every issue I need to send out a consultant to fix it, then even once a year is too often. If I could control everything from A to Z, then this would not be necessary, but I always try make sure the software is as robust as possible.[/quote]
As stated above.
Second that point still doesn't mean that the port needs to be changed while the application is running. One can provide a GUI to modify a config file and then provide a management app to bounce the server.Changing the port while running means that you MUST deal with the complication of txns in flight. And also deal with providing a timely notification to clients.
[quote]The user should be able to change the port number in a console, not while the server is running.[/quote]
As stated above.
What did I miss?
A good programmer is someone who always looks both ways before crossing a one-way street. (Doug Linder)
|
|
|
|
|
You said:
"The service is not necessarily running when the port is changed."
You also said, previous to that post:
"port & protocol at runtime dynamically."
|
|
|
|
|
Helfdane wrote: 2) need to change port & protocol at runtime dynamically.
You have a server which processing something and you wish to remotely manage it.
Your possibilities are the following
1. Provide functionality in the application itself which allows you to manage all parts (that you want to administer.) This is limited to the extent that you cannot start the application, but you can stop it.
2. Used a freeware or commercial product specifically designed to manage other processes. An example of this is Microsoft Operations Manager (MOM) and the Windows Management Instrumentation API. Generic term for that seems to be "web based enterprise management" (or at least that is good enough.)
3. Write your own management application.
4. Manually manage it, say via SSH.
I have used all of those except 2. I tried using MOM and decided it was too complex. Simplicity is important with the third since otherwise one ends up with a chicken and egg problem in that one must administer that app as well.
The problem with the first is that some problems will require a restart. Additionally certainly some business situations it is difficult to handle requests that are in flight via sockets when the port changes.
I prefer 3 but the downside is that it must be kept simple. The simplest functional model is to have it do nothing but start the app. Then one can use 1 with a non-dynamic model in use (modify the config file then restart to pick up the changes.) A more complex version is to have it verify that the target app is up and start it if it isn't. And it can also stop and start the app.
|
|
|
|
|
Hello everyone
Since Inlines (CollectionInlines) arnt a dependencyproperty i made a TextBox with dependency Inlines, the Code:
public InlineCollection InlineCollection
{
get
{
return (InlineCollection)GetValue(InlineCollectionProperty);
}
set
{
SetValue(InlineCollectionProperty, value);
}
}
public static readonly DependencyProperty InlineCollectionProperty = DependencyProperty.Register(
"InlineCollection",
typeof(InlineCollection),
typeof(InlineTextBlock),
new UIPropertyMetadata((PropertyChangedCallback)((sender, args) =>
{
InlineTextBlock textBlock = sender as InlineTextBlock;
if (textBlock != null)
{
textBlock.Inlines.Clear();
InlineCollection inlines = args.NewValue as InlineCollection;
InlineCollection oldinlines = args.OldValue as InlineCollection;
if (inlines != null)
{
textBlock.Inlines.AddRange(inlines.ToList());
}
else
textBlock.Inlines.AddRange(oldinlines.ToList());
}
})));
The TextBlock has normally some text after that it gets new inlines, if the Inlines it gets are empty it should get its first(default) text again. Now the Problem:
if it had a the first a text via Inlines then everything is ok coz i can get them again with the oldvalue, if it has a the first normal text with no inlines then the textblock doesnt show nothing because of: textBlock.Inlines.AddRange(oldinlines.ToList()); (no oldvalue of the Inlines)
how can i set the first text at the ende if in this case? any ideas?
Thanks in Advance.
Cheers
Def.
modified on Tuesday, April 12, 2011 9:02 AM
|
|
|
|
|
Hello,
Using Visual C# 2008, I have created a typed DataSet that connects to a SQL Server database. I would like to change it to connect to an ODBC data source (which in fact connects to the same SQL Server database). No need to be dynamic here. Is there a simple way to do that, or do I have to re-create everything from scratch ?
Thanks,
J-L
|
|
|
|
|
Recreate everything from scratch. The classes involved are going to be different and there is stuff that the SQL Server classes supports that the ODBC classes do not.
By using ODBC instead of the SQL Server dedicated classes you are also giving up the flexibility and performance benefits of the dedicated classes.
|
|
|
|
|
Hi
I'm converting a C# code which was using java.io to use system.io.
I need to change
java.io.FileOutputStream file= new java.io.FileOutputStream(path);
&
java.io.InputStream s to corresponding system.io calls.
I cannot instantiate system.io.stream object as its abstract. So how to decide between choosing System.io.memorystream, bufferedstream, streamreader,streamwriter etc?
Thanks
priya
|
|
|
|
|
You're trying to write a file out? Try using System.IO.FileStream.
|
|
|
|
|
Hi All,
I have 2 main Form classes - ParentForm and ChildForm in the same exe.
ChildForm has a bunch of numeric up-down controls on it.
I create an instance of ChildForm, set TopLevel = true and do SetParent with
the ParentForm as the parent handle (doing this enables the non client area of
ChildForm to stay highlighted when its selected).
Now the problem is I need to know when the mouse has left the ChildForm entirely.
The Leave event for the ChildForm is never fired.
If we listen to the MouseLeave event it gets fired if the mouse hovers
over any of the numeric controls and also when the mouse goes to the parent form. Is there any way I can differentiate between these 2 cases ?
Also note that we need to cover the case where some other external application
(say notepad or Task Manager with always on top as true) is positioned exactly above the child form and is clicked
EDIT UPDATE : Thanks to all for the replies ! Have asked the same question on the MSDN Forums - http://social.msdn.microsoft.com/Forums/en-US/winforms/thread/51e040a5-77f9-4b43-a462-1b5e3a13b90a/?prof=required[^]
TIA
modified on Monday, April 11, 2011 12:00 PM
|
|
|
|
|
private void Form1_MouseLeave(object sender, EventArgs e)
{
Point point = this.PointToClient(Cursor.Position);
if (!this.ClientRectangle.Contains(point))
MessageBox.Show("Ding");
}
Hope that helps
"You get that on the big jobs."
|
|
|
|
|
Hi Rob,
Thanks for the code snippet - it works perfectly when the mouse leaves the ChildForm and moves into the ParentForm !
Howeverm for the second casse i mentioned - "some other external application (say Task Manager with always on top as true) is positioned exactly above the child form and is clicked" , the Mouse Leave event on the ChildForm is not triggered. Is there some other event i need to be listening to for this ?
|
|
|
|
|
I think it is hard to solve. And then your mouse could walk into a dropped-down menu or context menu too.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Hey Luc,
Regarding the context menus and tool strip/main menus we can listen to their Activated events, so i don`t think that should be an issue
The tricky part is to differentiate between my applications client area and other windows that may be place directly above it.
In that case, MosueLeave gets fired and the even if the user hovers over the other window, the Cursor is still within the client area boundaries !
One would expect WinForms to provide some notifications for such a basic event but i haven`t (yet) found anything helpful in the Win32 API layer as well.
|
|
|
|
|
Hi TIA
You could start playing around with the form's Deactivate event but it probably wont completely resolve your issue. Also I'm sure you, or someone else has a good argument to implement this but it is unusual behaviour and may confuse the user.
Regards
Rob
"You get that on the big jobs."
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
