This one I would have to 100% disagree
-integrate FourQ package directly into the Core code base
During the Breeze presentation we've discussed this swiftly, however I'll re mention it again. I needed a quick metadata hook and bait system a few months back that didn't need the full FarCry stack, but it used fourq pretty well. Now while I was under that cycle, I had some time to port it over to three additional databases.
SAP, Firebird, and CloudScape. I also have a few unit tests for it, so I could verify that compat was kept as well as possible.
Then I go looking towards farcry_core and I must applaud those who built the PG and MySQL support, because the wide spread nature of the sql throughout the core was immense. Something I didn't have time to dive into. Although, I would love to get into it, but it'll just add bloat in the way it has to be done.
The biggest issue with refactoring is ensuring backward compatability. The first steps toward this are developing comprehensive unit tests for the API across areas to be refactored. We've been trying to work up the courage to tackle this task.
It's not something we can realistically achieve at Daemon in the Glamour timeframe.
If you've made any sort of inroads into building a test harness for farcry -- say in something like CFCUNIT -- we'd be very interested in collaborating and getting a module set up in the repository to house it.
It's certainly a project with tremendous merit that could be run in parallel with the Glamour build. Perhaps in expectation of scheduling some degree of refactoring in the 3.0 timeframe.
-- geoff http://www.daemon.com.au/
--- You are currently subscribed to farcry-dev as: [email protected] To unsubscribe send a blank email to [EMAIL PROTECTED] Aussie Macromedia Developers: http://lists.daemon.com.au/
