Yep, we know about it, and I'm here policing it, since this one seemed to have slipped by somehow.
So, this means the code needs to be removed from CVS yes? -- dIon Gillard, Multitask Consulting Blog: http://blogs.codehaus.org/people/dion/ "Roy T. Fielding" <[EMAIL PROTECTED]> wrote on 09/08/2003 07:14:23 AM: > > As far as I can see, doing an import (which I understand to simply mean > > referring to a class and using its methods) does not create a > > derivative > > We've already gone over this many times. Java doesn't work that way. > It isn't sufficient to read the LGPL (intended for static compilation) > and then do a mental translation to what any sensible person thinks it > should say for a late-bound-by-name language like Java. If you just > read the text as is, linking by name does cause it to be a derived work > covered by section 6 because the module and method names have to be > copied into the executable. The FSF has confirmed that interpretation, > which is consistent with their licenses not being sensible in the > first place. > > Serge, licensing the API using a ASF/BSD/MIT/Artistic-style license > is sufficient to allow ASF java code to import those names. I would > still caution against creating a dependency on the presence of an > LGPL work, but as long as the API is licensed such that others can > create alternative implementations it should be okay. > > There are also ways to use LGPL libraries such that the core code > is invoking a generic API rather than the API covered by the LGPL, > but I assume Maven already knows about that. > > ....Roy > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] >