Question

Say I have a Customer entity type with Name and Phone string properties.

If I do:

objectSet.Load(key, "Root{Customer{Name}}", ...)

versus:

objectSet.Load(key, "Root{Customer{Name, Phone}}", ...)

It generates the same query:

Select [root].[ID], [root].[VersionID], [root].[TypeID], [root].[Permissions], [root].[FastLoadData] from [dbo].[doDataObject-View] [root] where ((([ID] in (106)) and (([ID]=106 and [VersionID]<>1))))

Is this because DO uses the FastLoadData column, which contains the data for both properties?

If so, here's an interesting side effect. Say I use the first fill descriptor (which omits the "Phone" property). If I then access the customer.Phone property, why does DO do a round-trip to the DB? The property's value is already contained in the FastLoadData? This actually isn't a big deal, because the FillDescriptor should include the Phone property if I'm going to consume it. But I wanted to know if there's anything I'm missing.

Accepted Answer

You are absolutely right. In this case FastLoadData cotains both Name and Phone proprties. So resulting query is the same and resulting online DataObject instance is also the same.

At the same time FillDescriptor controls what properties must be transferred to offline instance. While it does not affect queries it allows to affect memory size ObjectSet consumes. And of course don't forget about [LoadOnDemand] properties like e.g. DataObjectCollections. These properties are loaded via separate queries and fill descriptor allows to exclude them.

Subscription:

Related questions