Boost logo

Boost :

From: Herve Bronnimann (hbr_at_[hidden])
Date: 2002-06-20 09:13:07


On Wed, Jun 19, 2002 at 05:05:14PM -0500, Jeremy Siek wrote:
> That stuff is really just "detail" stuff, though some of it should become
> "public", but I haven't yet found the time to move them through formal
> review. The ones that would be good to have "public" would be:
>
> disjoint_sets
> iterator_tests
> fibonacci_heap
> mutable_heap/mutable_queue
> queue
> bucket_sorter
>
> If anyone is interested in helping to get these polished up and through
> review, let me know.
 
As part of my plans with the boost::heaps library, I'll have to take a
close look at the fibonacci_heap, and in particular compare with
Dietmar's implementation. At least I'm thinking I should see what Knuth
has to say!

Also I'll take a look at mutable_heap and mutable_queue. I think the
challenge is to keep the mushrooming of the code by independent parties
while still providing stuff that works for everyone. Otherwise, boost
will end up with three(!) fibonacci heaps, and what else.

Regards,

-- 
Hervé
PS: my plans for boost::heaps is to clear it up and sponsor the review,
but it's not high in my priority list, maybe will have to wait for the
summer. See the previous thread about a month ago on this mailing list.

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk