<i>
Lars's point about the toolkit being more sizzle than steak, more promise than delivery at this point is true. Yes, many of us have built real sites on top of OpenACS 4 but ... it's still weaker in some regards than OpenACS 3. There's a lot of confusion within packages, some of which truly suck, most of which don't really make proper use o f the core tools (due mostly to a lack of guidance and communication between the core team and package implementors when parallel development was happening at a very rapid pace). </i>
<br><br>
This is pretty much the point, why I am a bit confused about the way to a coherent and stable version...
<br><br>
Disclaimer: My following suggestion might be *unrealistic*...
<br><br>
What about getting some cash together from the community to hire a 4.5 insider who tackles these problems full time? How long would it take him?