Re: [boost] [sorting] Is there still interest?

From: Corrado Zoccolo (czoccolo_at_[hidden])
Date: 2007-03-21 04:20:58

Hi Sam,
I see sorting problems and sorted data structures as two faces of the
same medal.
In STL, for example, we have std::sort and std::set that both work
with a two way comparison functor (std::less).
I see a great potential in your work on operator[] with trie data structures.
I think that an organic proposal of radix_sort with trie data
structures (mimicking std::set and std::map), that both use the same
operator[] approach would be very interesting.

Corrado

On 3/21/07, Stjepan Rajko <stipe_at_[hidden]> wrote:
> Hi Sam,
>
> On 3/20/07, Sam Schetterer <samthecppman_at_[hidden]> wrote:
> >  Hi. I am just wondering if there is still interest in the library, because
> > I have not recieved any comments on the library in the last two days, no
> > posts at all, while I have made multiple posts, some with very important
> > ideas in them.
>
> I've been sort of following the discussions on sorting, so I'll try to
> share some thoughts on what I noticed.  To me, the main themes that
> stood out were that
>
>  1) there is quite a bit of good sorting functionality available in
> the standard libraries
>  2) the benefits of added / alternative implementations are questionable.
>
> I think it's great that you are putting effort into trying to develop
> something for the community, and if that "something" has to be sorting
>  then it might be crucial to put time into clearly showing how your
> planned contribution addresses the above two items.  If a lot of the
> previous discussion contained things like "I am skeptical that an
> alternative implementation would be faster, simpler to use, etc.",
> then chances are that people of that opinion won't be inclined to
> discuss an alternative implementation unless you can clearly show that
> there is, in fact, some potential in it.
>
> So, if you take into account all the discussion points that have been
> posted so far, take some time to address them, and come back with "I
> have implemented xyz-sort and tested it on both killer and random
> sequences and found it faster than the standard implementation by 30%
> on average on two different platforms", or "I have found a fully
> portable way to express sorting criteria in an incredibly concise,
> never before imagined way, here's an example", it might peak some more
> interest.
>
> Perhaps you've already done that and I missed it... or perhaps people
> are just busy with other things :-).  Or perhaps it's getting too late
> and I'm just rambling about things that I shouldn't be rambling
> about... on that note, I should call it a night.
>
> Best regards,
>
> Stjepan
> _______________________________________________
> Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
>

-- 
__________________________________________________________________________
dott. Corrado Zoccolo                          mailto:zoccolo_at_[hidden]
PhD - Department of Computer Science - University of Pisa, Italy