FetchType.LAZY and detached entities

[epbernard]

It's not possible unless you do byte code enhancement.
Because when objects are reattached, there is no way to know wheter you nullify it to remove the collection, the entity or just because it's the end of the graph.

[bill.burke]

"rjamestaylor" wrote:

"rjamestaylor" wrote:

"rjamestaylor" wrote:

"rjamestaylor" wrote:
Alex,

Can you post the pages to a stable location? It doesn't seem that (a) attchments or (b) links to forum messages have a very long life span around here (just from the volume of posts I've seen referring to both followed by "not found" posts shortly thereafter.

If you need a stable URL, let me know.

Thanks.

[clcantrell]

I fell that the spec should address it, and I would not want to see just an empty collection. This would be misleading. null would be much better to indicate the collection has not yet been initilized.

[bill.burke]

NULL would require special bytecode manipulation. While, the other approach, of supplying an instance of a collection that is managed by the persistence manager, does not require special bytecode manipulation.

Bill