Hello all, I have the impression that the roadmap we advertise on our website is slightly out of date.
Refreshing and challenging the items part of it would be beneficial. * * Antora migration* -> I believe this should keep on being on our goal. -> I think we should further split things, with one sub-module per application -> We could ship the documentation along with the executables of the application -> Given that distributed application documentation is ready, we should make it available as soon as possible. Given these strategy changes, I believe we can do quick progress on the documentation front. Regarding the roadmap, keep it! * * Spring upcoming deprecation.* -> What we should really do beforehand is promote Guice artifacts. -> I think it would benefit from being a standalone item. Some work had been done on that topic: -> Have a ZIP packaging for Guice apps -> Use JIB to conveniently build a docker distribution for Guice apps -> Group apps in server/apps folder to ease discovery Hopefully the timeline should look like: -> Start shipping Guice artifacts on the download page as part of 3.7.0 -> (If no issues arise) deprecate Spring artifact as part of 3.8.0 -> Removal of Spring based artifact in 3.9.0 I propose making this timeline explicit on the website. * ** * Gradle* There seem that there was no recent activity on the Gradle front for some time. In my opinion we could re-asses this roadmap item, as this sounds likely not to be delivered on the middle term. * ** * JMAP RFC-8621* This specification is mostly implemented and tested with third party applications. Support is still partial, and upcoming efforts will come to further bridge the possible gaps. I refined the corresponding roadmap entry. * * Blob garbage collector* As part of JAMES-3150 my team recently worked on the topic and proposed an implementation suitable for mid-sized deployments. As I am not expecting further work on the topic soon, I propose to remove this roadmap entry. * * Tools to repair projections* We added such tools for mailboxes, mailbox counters, messages, quotas, ElasticSearch... Remains secondary concern projections, for whom I am not expecting progress to be made. As the state of the arts tools are likely satisfying, I propose to also remove this roadmap entry. * * Sum up* The following pull request puts the proposed changes together: https://github.com/apache/james-project/pull/664 Best regards, Benoit
