Hi Matt, Matt Benson <[EMAIL PROTECTED]> wrote on 01/04/2008 12:40:06 PM:
> Hi Xerces guys- > I have attempted to do some homework by searching > the archives, asking around, etc. and what I seem to > find is that the licensing approach used by Xerces-J > is to depend on the APIs published by XML Commons > (which it appears was voted in 2006 to move to > Xerces?). That's right. XML Commons is now a subproject of the Xerces TLP. Releases of Xerces-J bundle the XML APIs and resolver components from XML Commons. > These APIs are of course ASL licensed. I > am trying to run down precisely what requirements must > be met to implement parts of these APIs in an Apache > project, specifically the JXPath subproject of the > Apache (formerly Jakarta) Commons. Is there any > difference between the source and/or binary > representation of the JAXP APIs as published by XML > Commons vs. the underlying ideas expressed > therein?--as best I understand it there is no > hindrance to implementing these APIs despite the fact > that they were originally developed at Sun. Is this > simply by virtue of Sun's having donated the sources > of the APIs to Apache XML, or does any further > documentation exist that expressly states that the > APIs can be freely implemented? There are terms and conditions in the spec which allow for independent implementations. I think the main difference between a donation from Sun and authoring the API classes within the ASF is that we get Sun's javadoc vs. having to write original javadoc ourselves. > Next, I have found statements to the effect that it a > component must pass the JAXP TCK in order to claim to > be "an implementation." Is this indeed the case? I > have it on authority that there is no such thing as > partial TCK certification; does this group concur that > a "partial implementation" (e.g. of XPath only) will > never pass the TCK? I'd agree with that. Passing a TCK implies a full implementation (or at least enough of an implementation to pass all the tests). > Is it possible to distribute a > library as "non-certified", "non-compliant", or > similar? If you're looking for a legal opinion [EMAIL PROTECTED] [1] is probably a better place to ask. I'm not a lawyer. Don't want to mislead you with my perception of how things are or how they ought to be. I don't have a clear understanding myself. > I'd also appreciate information on getting > access to the ASF's JAXP TCK so that, even if Commons > JXPath can never be certified and thus officially > compliant, the portions it does implement can be > verified to be as good as possible. I believe you need to ask for that on the jcp-open mailing list. There are instructions on this page [2] which describe how one goes about requesting it and any other TCK. > Another > possibility might be for a "full" JAXP implementation > to fall back to e.g. Xerces for other than XPath > functionality. I think this ought to be considered > compliant. > > Thanks, > Matt > > ____________________________________________________________________________________ > Looking for last minute shopping deals? > Find them fast with Yahoo! Search. http://tools.search.yahoo. > com/newsearch/category.php?category=shopping > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] [1] http://mail-archives.apache.org/mod_mbox/www-legal-discuss/ [2] http://www.apache.org/jcp/ Michael Glavassevich XML Parser Development IBM Toronto Lab E-mail: [EMAIL PROTECTED] E-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
