http://nhforge.org/doc/nh/en/index.html#mapping-declaration-property

Read the access/naming strategies section. This will tell NHibernate how to access the property.

<property name=”EmployerName” access=”nosetter.camelcase-underscore” />

This will tell NHibernate, don’t use the setter, instead use the field that is the camel-cased version of the property name, prefixed with an underscore.

Make sense? It’s a weird syntax, but it works like a charm.

public virtual string EmployerName
{
get { return _employerName; }
set { _isChanged |= (_employerName != value); _employerName = value; }
}

Mapping of that column is really irrelevant, but here it is for context :

@lt;property type=”string” length=”50″ name=”EmployerName” column=”`EmployerName`” /@gt;

A quick way to see what is happening, is just set a breakpoint in one of your entities public setters and call List@lt;@gt; on it.

Thanks again

I have a followup question. It appears to me that NHibernate is setting “IsChanged” equal to True upon a call to “List”. As you may expect this renders that property useless for change tracking. From my experience with CSLA, there is usually a way to tell if the property is being changed during Load or upon a user of the object actually changing something. “Dirty” was never the case after a “Fetch” from the database, because a local variable was set for each column, not using the public Setter method.

Do you see this same behavior?

Thanks again,
Brian