An addition to the timber vs. steel subjected to fire discussion: At my alma mater, the University of Wyoming, the roof system of Arena Auditorium (they're so creative with naming buildings at UW, aren't they?) is a geodesic dome utilizing LVL beams, rather than steel, because the wood beams would not buckle. The wood beams would be expected to char no more than 1" deep under the expected temperatures during a fire, while steel beams would have softened to the point of buckling under the compression load, collapsing the roof.
On a sort of off-topic funny observation. I was at an indoor aquatics center (pool, lazy river, sprayers, etc.) last weekend, and I happened to look up and noticed they had a fire sprinkler system. As I looked around, and it's about 90% covered in water and the rest is concrete. Hmm...
At least someone complied with the building codes which should give you some confidence that they were probably observed where they would actually be more critical.
