I use something like this everyday:

awk '!($0 in a);a[$0]; print}'

I rarely if ever use uniq to remove duplicates. Sorting is expensive.