Michael Glavassevich <mrgla...@ca.ibm.com> wrote on 06/27/2011 12:44:36 AM: > If the intention is to contribute the code to Xalan, in my opinion > it would be better for the effort to be here. Folks interested in > contributing can become committers.
+1000. It isn't hard to work with the Apache development environment, and doing so preserves a lot of history/knowledge about what's been done in the past and what still needs to be done. And, frankly, the odds of getting a sufficient base of people to maintain it going forward aren't as good if the actual development occurs elsewhere. That's a large part of how we got into this situation: IBM was sponsoring most of the development effort (Sun came in long enough to contribute the compiler framework, then dropped out again), most folks sat back and let that happen, and when the IBMers moved to other tasks there weren't enough active volunteers to keep it moving forward. Xalan needs ongoing participation to occur here, in the Xalan community, if Xalan per se is to remain viable. We might want a separate development tree for the 2.0 effort, since that's a fairly drastic rewrite... but I think doing the development elsewhere risks another stagnation after one turn of the crank. It would be nice to know what Oracle plans on doing w/r/t updating the XPath/XSLT support in the Sun JVM, which is based on an archaic version of Xalan and has a bunch of known bugs. (I've spent several months gradually convincing a customer that the W3C, rather than Sun, owns the definition of XSLT and the fact that the Sun processor tolerates something it shouldn't is a bug in that code rather than in ours. Sigh.)
