WTKXSerializer does not currently have any dependencies on org.apache.pivot.wtk and can technically be used to deserialize any arbitrary bean hierarchy. Moving it to pivot-core will allow developers to take advantage of it on the server where it can be used to facilitate design patterns such as dependency injection and aspect-oriented programming.
There is a question of what prefix to use for the internal processing namespace (currently "wtkx"). Should this continue to be "wtkx" for WTKX files, or should it be something more generic ("bxml", "bx", "jbx", etc.)? This prefix could also be used as a generic file extension for bean markup files when a more appropriate extension is not available.
There is also the question of what to call the @WTKX annotation. This should probably also parallel the namespace prefix.
It may be valid for WTKX files to retain the .wtkx extension, since it clearly articulates what type of bean XML the file contains. This notion is further reinforced by the fact that the default namespace for WTKX files refers to the WTK package, org.apache.pivot.wtk. However, using a dedicated extension for WTK markup may simply promote confusion.