https://github.com/haskell/containers/pull/8
It shows that a more efficient implementation is available, which avoids `splitLookup` and therefore `union`. (Disclaimer: I haven't run any benchmarks to show that my internal implementation is actually faster.)
As I mention in that pull, these functions are useful for very practical things like consistent hashing, etc.
Mike Craig
On Thu, Feb 16, 2012 at 2:37 PM, Johan Tibell
<johan.tibell@gmail.com> wrote:
Hi,
First, thanks for writing such a detailed proposal.
I don't have strong feelings one way or the other. Like always I'm
worried about API growth. How about performance? Can these functions
be implemented efficiently outside the library?
Since I don't feel strongly either way I'm fine with this if Milan and
the community are.
Cheers,
Johan