Why add a constraint in one to many relationship entity framework code first

Question

I'm trying to understand the difference in my one-to-many relationship after I add a virtual call back to my main entity 'space'. All it does, looking at the tables is create a constraint. Can someone please explain the difference between having

public virtual Space Space { get; set; }

and leaving it out?

Here is my main entity 'space' and 'spaceimage' that has the virtual call back to 'space'

public class Space
{
    public int SpaceId { get; set; }
    public virtual ICollection<SpaceImage> Images { get; set; }  
}

public class SpaceImage
{
    public int SpaceImageId { get; set; }
    public byte[] Image { get; set; }
    public byte[] ImageThumbnail { get; set; }
    public string ContentType { get; set; }
    public bool IsMain { get; set; }

    public virtual Space Space { get; set; } // adds a constraint in sql server
}

Answer 1

the difference is you can make a call like this :

  var thisSpace = _context.Space.Include( x => x.Images).Where( blah == blah);

ok, so know you got your space from the database , now you don't need any other query at all to get images, you would just do.

  thisSpace.Images // now this is filled with a collection of images 

You do not need a join , you don't need to do anything at all - see how awesome Entity Framework is. The down side... is the is some overhead in a framework that can do stuff like this.

UPDATE:

you asked about Lazy Loading , you still have control of lazy loading or not.

See if you were to do :

   var thisSpace = _context.Space.Include( x => x.Images).Where( blah == blah).ToList();

then the query will execute right away and the Images will be attached. or..

you could do something like

   var thisSpace = _context.Space.Include( x => x.Images).Where( blah == blah);
   foreach( var image in thisSpace.Images){
   // when you hit this in code the query will actually execute
   // if lazy loading
   }