|
Where would I place a .config file for a windows serice? Would it only be used when I install the service or would it be used everytime the service was started?
Cheers,
Jim
|
|
|
|
|
The .config is always opened when the corresponding executable tries to access it and will have to be positioned at the same location as the executable.
(I assume you mean the "myProgramName.exe.config" file)
|
|
|
|
|
Yes, that's exactly what I meant. I just don't know where exactly the executable sits when its a service. I suppose it doesn't actually go anywhere and calls the same one anyways? Should I just install the service as per usual?
Cheers,
Jim
|
|
|
|
|
When you install your app as a service the current executable path is used, and the same executable path is used every time the service is started.
mav
|
|
|
|
|
Oh Really? That would mean that i only have to recompile a file after I have installed it, as opposed to uninstalling and reinstalling the file every time.
Thanks,
Jim
|
|
|
|
|
I have a question that I'm sure some of the experts here will be able to answer.
I have a custom business object that, by itself, contains a small amount of data. However, during runtime, that object may create and hold more data depending on how it is used. Basically, I want the object to grow-on-the-fly rather than allocate everything all at once.
The way I've set this up is through a property in the object:
public BigStuff[] HugeData
{
get
{
if (this.hugeData == null)
{
this.hugeData = new HugeData();
}
return this.hugeData;
}
}
This way, hugeData is only allocated when it is needed.
Now to the problem: multithreading. This property has the potential to allocate hugeData multiple times if the get_HugeData method is called asynchronously from multiple threads. What is the best way to get around this? Currently, I've placed a [MethodImpl(MethodImplOptions.Synchronized)] tag on the getter of the property, but I don't know how efficient that is, or whether there's a better way to do this. Any thoughts?
Tech, life, family, faith: Give me a visit.
Judah Himango
|
|
|
|
|
Internally this attribute uses the lock(this) construct which is in my opinion rather bad practice. I always prefer locking against a private member variable. If you do so you also have more control over what is locked against what (sound strage heh? )
Efficiency is the same with both solutions and I dont think this should be a problem (unless you are going to call this hundreds of times per second).
|
|
|
|
|
Whoa - so it actually locks the entire object? Are you sure? How, then, would that attribute work on static classes? Sheesh that's scary, I didn't know it locked the entire object. Yeah, maybe I'll lock my private hugeData then, unless locking isn't allowed on a null variable.
Tech, life, family, faith: Give me a visit.
Judah Himango
|
|
|
|
|
Dont get me wrong. Its locked against everything which also uses this attribute or lock(this) (at least in this class).
And locking doesnt lock against variables, it locks against objects. So nulls are not allowed. Dont be shy - just create a private member variable just to lock this function .
private object _myHugeDataLock = new object();
lock (myHugeDataLock) {
if (this.hugeData == null)
{
this.hugeData = new HugeData();
}
}
return this.hugeData;
|
|
|
|
|
Robert Rohde wrote:
Dont get me wrong. Its locked against everything which also uses this attribute or lock(this) (at least in this class).
Ok, I'm not following you now. Are you saying that
[MethodImpl(MethodImplOptions.Synchronized)]
void Test
{
DoSomething();
}
Is equivalent to:
void Test
{
lock(this)
{
DoSomething();
}
}
? The disassembled IL is very much different, but I haven't tested whether they are functionally identical.
Robert Rohde wrote:
And locking doesnt lock against variables, it locks against objects. So nulls are not allowed.
Ah of course, I knew that, but my mind told me otherwise.
Robert Rohde wrote:
Dont be shy - just create a private member variable just to lock this function .
Allocating an object for every function I want synchronized seems like such a waste! :-p I suppose this is another one of those tradeoffs where memory or performace is sacrificed for the sake of synchronized code.
Tech, life, family, faith: Give me a visit.
Judah Himango
|
|
|
|
|
Ah, after doing some testing, I've found what the Synchronized attribute does (and I think I now understand what you were saying).
[sycn attribute here]
void Test()
{
DoLengthy();
}
[sycn attribute here]
void Test2()
{
}
ThreadStart ts1 = Test;
ThreadStart ts2 = Test2;
ts.BeginInvoke(null, null);
ts.BeginInvoke(null, null);
Test2, despite doing absolutely nothing, will not finish until Test finishes! In other words, all methods with the synchronized attribute will be locked, not just the method being called. <Grrr> that stinks, I wonder what thinking went into that design. Well, I learned something new today.
Thanks for your insight.
Tech, life, family, faith: Give me a visit.
Judah Himango
|
|
|
|
|
Hi,
I should first make it clear I know virtually nothing about c# & .NET (so please speak slowly and use short words )
I'm modifying a c# program someone else wrote which, as part of its process sends a mail using MailMessage.
This works... However when the message arrives some (not all) of the new lines (CR,LF) are messed up.
I noticed that if I open the mail in outlook and go View->Encoding this is set to "Western European (ISO)" if I switch this to "Western European (Windows)" the lines look right again.
So I looked into this and noticed the BodyEncoding member, but still can't get this working.
My code is
<br />
MailMessage mail = new MailMessage();<br />
<br />
mail.To = "me@somewhere.com";<br />
mail.Cc = "";<br />
mail.Bcc = "";<br />
mail.From = "Someone@SomewhereElse.com";<br />
mail.Body = <String read from a file somewhere>;<br />
SmtpMail.SmtpServer = "smtp.MyServer.com";<br />
SmtpMail.Send( mail );<br />
<br />
I have tried adding...
<br />
mail.BodyEncoding = System.Text.Encoding.GetEncoding(1252);<br />
and
<br />
mail.BodyEncoding = System.Text.Encoding.Default;<br />
but still the message is always displayed as "Western European (ISO)", does anyone have any idea how I can fix this?
Cheers
DAve
|
|
|
|
|
I have this really wierd error messgae popping up when I run my application (ASP.NET with C# code behind).
I am having the app create a log file and directing the stream to a share on the server.The wierd thing about this is...the application works fine from my machine but when I try the same on one of my fellow developers...it breaks out with this error message. I am not sure if this is a security issue because all of us have the same security settings.
I also turned impersonate=true and given my username and passsword for users in the web.config. and changed the setting in the local security settings per http://www.dotnet247.com/247reference/msgs/25/127951.aspx.
I cannot figure out what and where the trouble is...beacuse the stack trace is no help!!!
I appreciate anyone what can shed a light on this issue
Menita
|
|
|
|
|
This should be a simple answer, but I'm a bit stumped.
I have a Form with a GroupBox of minimum height acting as a visual seperator
in the form. When the Form loads I want only the upper portion of the window
to be viewable. To accomplish this I'm attempting to set the Height property of the form. My problem is how to I equate the location of my GroupBox seperator to the window height ?
-------------------
| |
| |
| |
| |
----------------- <- This is the Group Box I'm using as a seperator
| |
| |
| |
| |
| |
| |
| |
-------------------
|
|
|
|
|
To reduce the size of the form so that the GroupBox is not visible:
<br />
form1.Height = groupBox1.Top + (form1.Height - form1.ClientSize.Height)<br />
If the group box should be visible just add the height of it:
<br />
form1.Height = groupBox1.Top + (form1.Height - form1.ClientSize.Height) + groupBox1.Height<br />
|
|
|
|
|
Thanks Robert, I tried Height=GB.Top, but I didn't add the client size difference...thanks.
|
|
|
|
|
I am new to C# and I know this may be easy for most but I have a question and it may have to do more with variables between form.cs files and other cs files, but here goes. I know the easiest way to add custom buttons to a messagebox is to use a form and make my own. I am looking for a way to have "yes, yes to all, no, no to all" buttons on a messagebox. My question is, if I have a class not in a form cs file and I need to display this messagebox, how do I do this and make it modal, and how do I catch the values (in the none form cs class) of the button that is clicked by the user.
|
|
|
|
|
Calling ShowDialog of your MessageBoxForm will always show it modally and wait until the MessageBox is closed. As your MessageBox will have more result types than DialogResult can offer you will have to make your own enumeration and create a property where you can read the result form after the MessageBox is closed, e.g.:
MyMessageBox mmb = new MyMessageBox();
mmb.ShowDialog();
switch (mmb.Result) {
case MyMessageBoxResult.Yes:
case MyMessageBoxResult.YesAll:
...
}
To make the handling like the one of the original MessageBox just declare a static function in your MessageBox class which returns the result.
|
|
|
|
|
Hi All,
I've created a class library that interops with COM to allow it to be consumed by a COM application. From my development machine everything works great, the COM app recognizes and executes the library. However, when I try to deploy on a machine other than my own, the COMP application does not recognize the class library at all.
I've tried assigning a Strong name to the assembly, then user REGASM on the dll to no avail. I suspect I may need to register the dll with regsvr32 but I am uncertain about how to prepare my class library (I know I need to use GUIDGEN.exe to assign a GUID in an attribute somewhere in the code)...
Any pointers?
Thanks, Rein
|
|
|
|
|
MSDN has a good article on just that, here.
Tech, life, family, faith: Give me a visit.
Judah Himango
|
|
|
|
|
This seems like such a stupid question.... but how do i do it?
(I'll be kicking myself when someone gives me the answer)
Cheers
Cata
|
|
|
|
|
Elementary my dear Cata:
byteArray = BitConverter.GetBytes(varLong);
Let the kicking begin.
|
|
|
|
|
*kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick* *kick*
*ow*
|
|
|
|
|
I have a Form (MainForm) that loads a usercontrol (UserMenu) as a main menu, I do NOT want to show this as a dialog, but I have no way to fire an event on MainForm from a command button on UserMenu without instantiating a new MainForm.
There are properties set on the MainForm that need to be used. This is why I cannot simply instantiate a new MainForm.
I have considered delegates, but honestly struggling to wrap my mind around the concepts. Also, I know the showDialog is the simplist way, but cannot use it for various other reasons.
Any suggestions would be appreciated.
|
|
|
|
|
The best mechinism is delegates/events. The are pretty simplistic to use but, if you want to wimp out , just pass the form to the UserMenu and make the necessary methods public.
|
|
|
|