Even with trusted sources - sometimes (often?) we can get it wrong, too. It's easy to misinterpret a 140-character question, or leave out a crucial detail.

Another thing I like about #SQLhelp is that it's bite-sized learning. Joey D'Antoni and I were talking with a few user group attendees last night after the Philly meeting, and one of them asked an exact question that he and I had answered that day on #SQLhelp! It was funny - anybody who watched our exchange on Twitter would have been able to answer that same question too, just based on seeing the #SQLhelp thread.