By using XML as database files, the benefits of XML that can impart to such XML database files are:
- easy for porting and development
- vendor , programming/scripting platform and OS neutral
- easy to read literally
- can be easily extended and scalable
The main rationale of thinking of the XMLDB is simply because the very fact that every database use their own file format and it would be difficult for databases to share database files when necessary and when developers need to write some programs between databases, they need to know the format. XMLDB would be so versatile and easily extendable.
Just imagine XMLDB as a CSV database file but in a more complex , more defined and more structured version.
I am not sure if someone already have or implemented this idea exactly or very similarly to my idea. If there is, my ideas came not from reading your ideas but my ideas just came... so I am not copying anyone but just plain coincidence.
I hope that XMLDB could be adopted and used in many databases like how H2 database is able to not only handle it's own file format but also CSV format and maybe XMLDB may even make it into being a stable client-server , cluster and embedded database ?
I don't wish to patent or license this XMLDB if no patent or license ever existed and if someone, having read and decided to adopt this XMLDB as their patent and/or license themselves or for someone, then may the person's patent or license fail to be approved (this is a literal curse) and if they did it in the name of business, then their business (All) would fail badly and bring them to their knees. If a patent of license already existed before, then there's nothing I can do.
Just to clarify, XMLDB is about the file format and the engine format. I would specify more about the specs of the file format and engine format for those who wish to implement and modify it.