Good morning,

I am working through some intricate code and I want to make sure I am made aware of any errors that may pop up at run-time.  How I have done this in the past is to create a class that I throw exceptions at and from there send emails or alerts to let me know what is going on.  As I understand it there is some exception handling that is already going on within the framework via an error provider.  How do I take advantage of that and still provide a streamlined way of handling my exceptions?  Also, is this a proper approach to using transactions within my code?

Here is an example of my code that I am trying to work this through:

private void cmdTransfers_Click(object sender, EventArgs e)
{
    try
    {
        // Wrap this entire process in a transaction
        BusinessLayer.TransactionBegin("MyDataSourceName", IsolationLevel.ReadCommitted);

        // Retrieve the next [snip] . . . lots of stuff happening here . . .

       MyLastProcedureToRun(myParm);

        BusinessLayer.TransactionCommit("MyDataSourceName");
    }
    catch (Exception ex)
    {
        // place cool SF exception handling here, if any

        // this is my current way of doing it
        SPXExceptionHandler.ReportException(ex);  // exceptions are logged, emailed, etc.
        BusinessLayer.TransactionRollback("Dynamics");
    }
}

Thanks!
Bill

Yup...I think we met the same conclusion at about the same time.  Here is what I did:

Set a public variable in the dialog form (Form2):

public String _userinfo;

On the cmdOK_Click event, I set the value of the comments textedit control to the public variable.

Then, on Form1, within the deep inner workings of the unhandled exception routine:

MyExceptionData mNewForm = new MyExceptionData(e.UnhandledException);
mNewForm.ShowDialog();
mUserData = mNewForm._userinfo;

I am passing the exception to the form because I am formatting the exception data on the form for my own (or other savvy users') review.

So far, so good...thanks a ton for your help!!
Bill

Cool!



One thing to note is that I used the using statement. It closes the form when done (when the using code block is done). If you don't use this, then you should close the form after you are done with it. The form is still open otherwise.

Oh, how did you send the email? I've been thinking about doing this also, and am investigating ways of doing it.

.NET actually has some good SMTP classes for sending emails.  Here is a simple example (I went ahead and showed how to authenticate as this is generally the next question.  If you are mailing from an internal server that will relay your message you might not need this, but I thought it would help.  Also, you can format the mail message to support HTML, etc.  I just showed you the basics.

Thanks Trent. I've heard that this can fail because ISPs block mail sent to servers outside of their domain. Is this something you've heard of? I.e. I'm accessing the internet via a wireless connection at a hotel or using the network connection at a client site, attempt to send an email via the SMTP server at my company and the hosting network blocks the email.

Thus the credentials.  If you supply a mail server, port, and credentials...that's email...period   We use this everywhere.  The only thing that you cannot do is pull mail down from a POP3 server...but sending...no problem

Very cool! Thanks!