Good comments everyone....
I was "raised" as an engineer to let the fabricator do the work. This made sense on typical connections. We always designed any connection that was unique.
However, over the last 10 years or so, the nature of liability cases in the U.S. has directed the "blame" back at the EOR as the ultimate source of responsibility for the entire structure. This was especially hit hard on me when I sat at a structural conference and listened to the experiences of Jack Gillum who was the EOR of the Hyatt-Regency in Kansas City. The walkway that collapsed did so through a comedy of errors and oversights. He was depending upon a fabricator to design a connection. It didn't happen. What did happen was:
1. His draftsman failed to place the required load on the detail.
2. The fabricator began the connection design, but, getting a bigger project, farmed the Hyatt out to another fabricator.
3. Fabricator No. 2 "thought" that the partially detailed connection was designed and simply completed the detail.
4. Gillum's firm quickly reviewed the shop drawings, thinking that the connection, as detailed, was designed.
5. The connection was built and one year later collapsed, killing 114.
This drives home the thought that if we depend upon a fabricator to design "our" connection, then we'd better have the discipline to review these connections. That moves me to think that I just ought to design the connection in the first place. Most of the repetitive, basic connections are fairly standardized already (AISC single plate shear connections and AISC double angle connections). The judge in the Hyatt case told Mr. Gillum that he could delegate tasks, but he could not delegate responsibility.
Which takes longer, designing the connections, or spending the time to review someone else's design?