Hi, i've been looking at the serialization examples as a method to store transactions in a database field before they are committed to the database. This is to be used for publishing control in a content management system. My idea is that the system creates the new record as normal but does not store it instead serializes it into the database. Later a user publishes those changes, the object is deserialized and then it's save function is called. I know this could cause issues if the database structure is changed. But other than is there any reason why this would not work?

 

I wrote a system that did exactly that along time ago with ADO and it's ability to save in XML. The only problem I can see is that the de-serialization might throw and error if the ES assemblies change. I'll do a test here and see if this is the case, but I need a few days as I'm heavy in the middle of ES2008 changes.

---

EntitySpaces | Twitter | BLOG | Please honor our Software License

Just got back from a meeting, we are going to save to our own xml format. We assumed entityspaces would work as I described, however the inability to change the database structure without wiping out a sites change history has forced us to create xml we can process and then deal with changes with out of date/incomplete information. If you have any idea's on how to get around this issue with entityspaces we would still be interested as we feel the serialization approach would be much quicker to implement.