Hi, In the last few days we've nailed quite a few of the remaining open issues for Jackrabbit 1.4 (special thanks go to Marcel for the synchronization fixes!), and I've just dropped another handful of less essential issues from the critical path to the release. As a result we currently have the following nine issues (sorted by subjective priority) open for the 1.4 release:
[JCR-995] Release the OCM component - Blocker for the release. I'll look into promoting the OCM component from sandbox and preparing it for release in a few days, but would appreciate feedback on the current state of the component before that. [JCR-924] Use the Jackrabbit RMI extensions by default in jackrabbit-webapp [JCR-1193] war missing jcr jar - I think we should have these, and probably some extra facelifting to jackrabbit-webapp. I'll still put some effort on these before branching 1.4. [JCR-1180] DatabaseFileSystem and DatabasePersistenceManager don't allow ... [JCR-1197] Node.restore() may throw InvalidItemStateException [JCR-1277] ConnectionRecoveryManager is created twice in DBDataStore init ... - Nice to have. It would be great to have these fixes in 1.4, but postponing to 1.4.1 isn't that bad. [JCR-1242] Improve serialization of NodeReferences for BundleDB PMs [JCR-1249] Improve updating of references to version storage - A bit risky, but probably worth it if we can have good patches in time for some testing. [JCR-1253] Allow to configure autoCommit mode for BundleDB PM to avoid ... - Quite controversial. I'm inclined to dropping this from 1.4, but wouldn't mind if there's some positive outcome soon enough. Taken together and mixed with the long Christmas holidays, it seems like we'll be able to branch 1.4 earliest around the end of next week. This unfortunately means that the final release date will slide over to January. I'm also getting some backchannel feedback that we should still consider including the proposed database connection pools and some other new features in the 1.4 release. That would further postpone the release date and widen the scope of this release that's already by far the biggest change since our initial 0.9/1.0 releases. I guess it would be better to go with a 1.5 release soon after 1.4, but I'll summarize these thoughts better tomorrow. BR, Jukka Zitting
