On Apr 2, 2007, at 11:38 AM, David Jencks wrote:
Over in geronimo we noticed this text and have responded by:
1. for the schemas that we generate code from (using xmlbeans or
jaxb) we checked the schemas into a non-publically-accessible
repository (with our tck stuff) and generate the code there, and
publish source and binary jars, both without the schemas themselves.
2. We also have an ongoing effort to produce unencumbered schemas.
Apparently (IANAL) the actual schema/dtd is not copyrightable as it
is an interface specification, the part that is copyrightable is
the comments, annotations, etc. So, we have been typing up the
schemas ourselves without any of the descriptive material.
See this JIRA for more details on the effort in Geronimo that David
mentioned.
https://issues.apache.org/jira/browse/GERONIMO-2630
Typing the schemas in by hand is tedious but maybe not so bad if we
divide and conquer. I'll volunteer to help if there's a consensus
around that approach. There is a utility attached to
the JIRA that can be used to compare schemas to make sure they are
equivalent.
Best wishes,
Paul
