extended attribute standardization

[Jos van den Oever]

2006/11/19, Michael Burschik <Michael.Burschik at gmx.de>:
> Do you mean some kind of relational database? If not, I'm not so sure
> that searching the index would be faster than searching extended
> attributes. After all, the file system already is a very efficient
> database, even though it is not optimized for the kind of access
> patterns we are talking about.
Yes, a relational db with indexes or an index optimized for text
search such as lucene is _much_ faster than searching through all
external attributes. That's why it's called an index.

> > There are plenty of libs to get that data out. Also the specs are
> > rather simple.
> The individual specs may be simple, but the sheer number of specs will
> certainly introduce complexity.
True, and that is why it nice to have a program like xmlindexer that
gives you only one output format for all metadata.