Good answer. As it stands right now, Abdera2 is very close to being done. I'll be working on shoring up test cases and documentation later this week. I'm still waiting on a response to the post I made on the developer list about making Abdera2 the trunk, however, so I'm hoping that administrative and process delays won't push out the completion unnecessarily.
On Tue, Dec 13, 2011 at 1:29 PM, Chad Lung <[email protected]> wrote: > My two cents as an end user of Apache Abdera would be: whatever gets Abdera > 2 out the door sooner :-) I'm not to concerned with how Abdera currently > handles discovery of extensions and haven't had any performance issues. If > people really want Guice, then could it be added later and ship Abdera 2 > sooner? > > > > On Tue, Dec 13, 2011 at 3:07 PM, James Snell <[email protected]> wrote: > >> Ok... so there's one final remaining major issue I have with Abdera2.. >> and that's configuration. Currently, Abdera2 uses the same fairly slow >> classpath based configuration and discovery mechanism used by Abdera >> 1.x. There have been a number of performance improvements, but in >> general, discovery of extensions (for instance) is still done by >> crawling the classpath looking for files under META-INF/services and >> initializing things that way. Obviously, this doesn't jive well with >> more modern approaches using injection (guice). That said, however, it >> would be a major update to introduce Guice into Abdera and would >> likely result in a number of extensive changes throughout the code... >> That said, however, there have already been extensive changes >> throughout the code so having another wouldn't hurt. This would likely >> be the final major update if I decide to proceed but I wanted to >> solicit feedback from the community. >>
