Hi Milko,

Thanks for your help.

This is how grid is used:

class CellClass : IConvertible
{
    public String DisplayValue = "";
    public MyClass RealValue = null;

    public CellClass(String displayValue, MyClass realValue)
    {
        DisplayValue = displayValue;
        RealValue = realValue;
    }

    public SetValue(String displayValue)
    {
        RealValue = Parse(displayValue);
        DisplayValue = RealValue.AsString();
    }

    #region IConvertible Members
    public object ToType(Type conversionType, IFormatProvider provider)    {
        return Convert.ChangeType(DisplayValue, conversionType);
    }

    public TypeCode GetTypeCode()    {
        return TypeCode.Object;
    }

    public string ToString(IFormatProvider provider)    {
        return DisplayValue;
    }

    //rest of IConvertible just throw new NotImplementedException();
    #endregion


    internal void Clear()
    {
        DisplayValue = "";
        RealValue = null;
    }
}


void BuildGrid()
{
    _gridSource.Rows.Clear();
    _gridSource.Band.Columns.Clear();
    for (x = ....)
    {
        UltraDataColumn col = _gridSource.Band.Columns.Add(name[x], typeof(Object));
        UltraGridColumn gridCol = _grid.DisplayLayout.Bands[0].Columns[_grid.DisplayLayout.Bands[0].Columns.Count - 1];
        gridCol.Style = Infragistics.Win.UltraWinGrid.ColumnStyle.Edit;
    }
}

void LoadData()
{
    List<object> row = new List<object>();

    foreach (......)
    {
        row.Add(new CellClass ("", new MyClass(...)));
    }

    _gridSource.Rows.Add(row.ToArray());
}
   

The principal is that, we are storing own class in every cell of the grid (Object / CellClass), but displaying them as String.

When user edits cells, we are parsing the new String value, and converting it to CellClass.