ProxyWarnLog

[Jan Limpens]

I am getting this:

2009-06-25 16:20:46,030 [8] WARN  NHibernate.Engine.StatefulPersistenceContext.ProxyWarnLog - Narrowing proxy to Fex.Businesslayer.ProductSet - this operation breaks ==

What does this mean?

--
Jan

[Tuna Toksoz]

You have a reference to a base class, but when loaded, it becomes the super class, it is narrowed.

Tuna Toksöz
Eternal sunshine of the open source mind.

http://devlicio.us/blogs/tuna_toksoz
http://tunatoksoz.com
http://twitter.com/tehlike

[Fabio Maulo]

/// <summary> 

/// If the existing proxy is insufficiently "narrow" (derived), instantiate a new proxy

/// and overwrite the registration of the old one. This breaks == and occurs only for

/// "class" proxies rather than "interface" proxies. Also init the proxy to point to

/// the given target implementation if necessary. 

/// </summary>

/// <param name="proxy">The proxy instance to be narrowed. </param>

/// <param name="persister">The persister for the proxied entity. </param>

/// <param name="key">The internal cache key for the proxied entity. </param>

/// <param name="obj">(optional) the actual proxied entity instance. </param>

/// <returns> An appropriately narrowed instance. </returns>

public object NarrowProxy(INHibernateProxy proxy, IEntityPersister persister, EntityKey key, object obj)

2009/6/25 Jan Limpens <jan.l...@gmail.com>

--
Fabio Maulo

[Jan Limpens]

How does this happen? (I just get the warning, but don't actually know what causes it, so I need to understand the context better)
Overall it does not seem as a very terrible thing...

2009/6/25 Fabio Maulo <fabio...@gmail.com>

--
Jan

[Fabio Maulo]

For sure is not terrible (for some reason it is a "warning" and not an "error" ;) ).

To understand when and why we should see which the code running in a opened session and I'm pretty sure that it is not so easy to find where the same entity-proxy is requested for two different classes in the same hierarchy.

2009/6/25 Jan Limpens <jan.l...@gmail.com>

--
Fabio Maulo

[Ocsam]

I ran into this issue because i was used an abstract base class instead of an interface. And after digging into the source, i am wondering if a check for abstract classes should not also be made.

/// <summary>

/// If the existing proxy is insufficiently "narrow" (derived), instantiate a new proxy

/// and overwrite the registration of the old one. This breaks == and occurs only for

/// "class" proxies rather than "interface" proxies. Also init the proxy to point to

/// the given target implementation if necessary.

/// </summary>

/// <param name="proxy">The proxy instance to be narrowed. </param>

/// <param name="persister">The persister for the proxied entity. </param>

/// <param name="key">The internal cache key for the proxied entity. </param>

/// <param name="obj">(optional) the actual proxied entity instance. </param>

/// <returns> An appropriately narrowed instance. </returns>

public object NarrowProxy(INHibernateProxy proxy, IEntityPersister persister, EntityKey key, object obj)

{

            bool alreadyNarrow = persister.GetConcreteProxyClass(session.EntityMode).IsAssignableFrom(proxy.GetType()) && !persister.GetConcreteProxyClass(session.EntityMode).IsAbstract;

what do you think?  

Thanks in advance