Remove redundant vec implementations traits

[erickt]

vec.rs has a couple redundant implementations of various traits that are needed until #4148 is fixed.

[nikomatsakis]

Not critical for 0.6; removing milestone

[emberian]

@erickt which redundant implements are you thinking of? I do not see any that are obvious, but I don't really understand #4148

[erickt]

@cmr: I'm having trouble remembering, to be honest. I think this was about the fact that we had to copy-paste implementations for ~[T], &[T], and @[T]. For example:

    impl<'self,T:Eq> Eq for &'self [T] {
        fn eq(&self, other: & &'self [T]) -> bool {
            self.len() == other.len() &&
                self.iter().zip(other.iter()).all(|(s,o)| *s == *o)
        }
        #[inline]
        fn ne(&self, other: & &'self [T]) -> bool { !self.eq(other) }
    }

    impl<T:Eq> Eq for ~[T] {
        #[inline]
        fn eq(&self, other: &~[T]) -> bool { self.as_slice() == *other }
        #[inline]
        fn ne(&self, other: &~[T]) -> bool { !self.eq(other) }
    }

    impl<T:Eq> Eq for @[T] {
        #[inline]
        fn eq(&self, other: &@[T]) -> bool { self.as_slice() == *other }
        #[inline]
        fn ne(&self, other: &@[T]) -> bool { !self.eq(other) }
    }

The reason we have to do this is that our handling of generic functions cannot automatically coerce a generic argument of type ~[T] to &[T], which means we need to have these multiple implementations. @nikomatsakis, do you remember if this is what we could clean up once #4148 is implemented?

[brson]

Sounds like this is probably obsolete.

[thestinger]

Agreed, this stuff will need to be redone as a whole. I have an issue open about generalizing it for Vec<T> and I think it will cover any redundancy like this.

[thestinger]

#11876