Just a quick update on the status of ongoing FarCry development. We've had our heads down at Daemon coding, but i thought I ought to break radio silence. I've outlined versioning changes, major functional milestones, and release dates - the rubbery kind.
In light of changes to (Gonzales) version numbering (see below), we're looking at pushing some new enhancements available in trunk through to the p300 maintenance branch for FarCry 3.0.x users. And leaving only the "form tool" technology and associated architecture changes in the 4.0 release.
FarCry 3.0.2 (due any day)
Currently in release candidate; includes minor bug fixes, significant changes to Verity configuration and Oracle support.
We're hoping to release this as a milestone as soon as we're collectively happy with the build. I have to admit that daemon is not in a great position (resource wise) to put this build through its paces for new users eg. the installer. However, we're running it in production for a number of large sites and it's running fine.
FarCry 3.0.3 (due end of September)
We're looking at pushing the fourq object broker technology into this milestone release. Currently this is only available in trunk. The object broker technology provides a significant performance boost to any farcry site by holding the COAPI in memory and circumventing the multitude of content object lookups to the database.
FarCry 3.0.4 (due end of October)
Overhaul of search components. It's become apparent in the attempted re-engineering of the Verity components in 3.0.2 that there are a variety of limiting factors in the current design. We want to rip out and replace the Verity componentry and replace it with a more appropriate and technology agnostic solution.
(public beta due as soon as we can get a decent installer in place)
The major contribution to the code base this time round is "form tools" technology, and significant re-factoring of the underlying FarCry framework. These advances, headed up by Mat Bryant, really transforms the way you build things in FarCry. This is a very significant change to life as a FarCry developer.
We want to make life easier for ourselves and remove the need to provide backward compatibility to CF6.1. It's hard enough providing a seamless upgrade path for FarCry projects and Linux/Windows support without having to worry about the foibles of our code changes running on CF6.1.
The upshot of all this is that we're looking at changing the next version to be a 4.0 release rather than a 3.1 release. We think the functionality change is significant enough to warrant the change in release number marketing wise. But perhaps more significantly, if we choose to go to CF7 as a base level requirement then it compels us to change version number under our current version numbering regime (ie. break in backward compatibility).