Serialization

It's used when you need to save an object and later to load it, to send object over network and then recreate it on the receiving side and in similar purposes.

For example, if two instances of application communicate over socket by sending messages, message can be an instance of some class. Then you just fill object's properties with values you want to send to the remote instance, then serialize the object into XML and write it to socket. Remote application reads XML from the socket and deserializes it into object of message class. In that way, one appication instance has effectively sent an object to another application instance.

Similarly, you can use binary serialization, but that is dangerous in sense that version of object must be exactly the same - any change in internal fields of the class will cause deserialization failure. Apart from that problem, binary serialization produces much smaller output.

Power of this way of working is that it's simple - serializing and deserializing one object takes only several lines of code, regardless of object complexity.

Drawback of this way of serialization is that you have to write class knowing in advance that it will be serialized. For example, all content must be gettable and settable via properties. Nothing that is set like using AddThis method will not be serialized. Also, circular references cannot be serialized (that would end up in infinitely large XML).

3 Replies

Know the answer? Post it — somebody with the same question will find it here.

ali raza16y ago
Hi satyapriya, read this post it'll help you understanding serialization and its practical implementation in ASP.NET.

http://techgulf.blogspot.com/2010/11/viewstate-error-in-aspnet.html
0
Sam Hobbs16y ago
Serialization is primarily for storing to disk and restoring from disk. It can also be used for other purposes such as what Zoran describes.
0
Zoran Horvat16y ago
It's used when you need to save an object and later to load it, to send object over network and then recreate it on the receiving side and in similar purposes.

For example, if two instances of application communicate over socket by sending messages, message can be an instance of some class. Then you just fill object's properties with values you want to send to the remote instance, then serialize the object into XML and write it to socket. Remote application reads XML from the socket and deserializes it into object of message class. In that way, one appication instance has effectively sent an object to another application instance.

Similarly, you can use binary serialization, but that is dangerous in sense that version of object must be exactly the same - any change in internal fields of the class will cause deserialization failure. Apart from that problem, binary serialization produces much smaller output.

Power of this way of working is that it's simple - serializing and deserializing one object takes only several lines of code, regardless of object complexity.

Drawback of this way of serialization is that you have to write class knowing in advance that it will be serialized. For example, all content must be gettable and settable via properties. Nothing that is set like using AddThis method will not be serialized. Also, circular references cannot be serialized (that would end up in infinitely large XML).
0