On Tue, Oct 20, 2009 at 2:52 PM, Jukka Zitting <jukka.zitt...@gmail.com> wrote: > So I'm thinking of putting JCR-RMI on a maintenance track for old 1.x > releases, and perhaps even dropping the RMI layer entirely from > Jackrabbit 2.0. > WDYT?
I would like to see the RMI access to continue. There may be other transport mechanisms, but in my opinion you also want something as direct as direct calls and other protocols are more overhead and workaround to get to use procedure calls over HTTP. The implementation of the current RMI layer may not be the right one, as the performance is really bad (but I don't see this as fundamental) but one shouldn't influence this kind of decisions based on a current implementation (not having caching has its benefits as pointed out). The RMI access is not just an internal implementation layer, it is important for many projects and is relied upon. Deprecating such a layer should be smeared out over a very long time. But I would opt for the RMI access to be improved and continued and be on par with SPI-WebDAV as an option. I wouldn't mind putting time into it (in fact, if RMI is deprecated soon, I would probably need to). \Berry
