Nice to hear that finally someone agrees with me on this one . Let me add to this story that the people I was discussing this with only use their home-grown container in the most basic scenario possible. When I briefly explained to them some more nice features of Castle Windsor, the only thing they were considering is expanding the features of their own container.

I have been on several LOB (line of business) architecture teams that made the choice to roll their own solutions patterning them off various open source projects (picking and choosing through the source code). And a year later they had a nice clean solution . The irony is, that after the custom solution was in production for a while features had to be added back in to meet those edge case needs. Before long their custom solution closely resembled the original project they patterned their custom solution after. Additionally they could not get the benefits of the enhancements to open source community had been making over that year of time. The short answer, they wasted a year of their customers time, and they wonder why the executives are always looking at these cheaper off-shoring alternatives. We do it to ourselves.

I tend to live by the old 80% rule. If there is a solution that can get us 80% where we need to be then the team better have a darn convincing arguments to deviate and roll their own solution.