Hello FPC-Pascal,
Tuesday, May 24, 2011, 8:25:07 PM, you wrote:
MG> IMO it is simply good manner to ask Tim.
MG> He spent quite some time in testing and tuning and explained the
MG> algorithm in detail, so I guess he will be happy to see his fame
MG> spreading.
Do you know the email address of T
Hello FPC-Pascal,
Tuesday, May 24, 2011, 8:25:07 PM, you wrote:
MG> If you do only the insertion+mergesort, then you have only implemented
MG> a long known improved mergesort variant and you can not call it
MG> TimSort. Of course then you don't need to ask Tim for permission.
Well, in first "try
2011/5/24 José Mejuto :
> Hello FPC-Pascal,
>
> Tuesday, May 24, 2011, 1:06:43 PM, you wrote:
>
> >> Why is TimSort specially interesting to you ?
> MG> I need a fast stable sort, so multiple sorts work as expected (contrary
> to
> MG> QuickSort).
> MG> TimSort is a candidate.
>
> That's exactl
"José Mejuto" hat am 24. Mai 2011 um 18:07 geschrieben:
> Hello FPC-Pascal,
>
> Tuesday, May 24, 2011, 1:06:43 PM, you wrote:
>
> >> Why is TimSort specially interesting to you ?
> MG> I need a fast stable sort, so multiple sorts work as expected (contrary
> to
> MG> QuickSort).
Hello FPC-Pascal,
Tuesday, May 24, 2011, 1:06:43 PM, you wrote:
>> Why is TimSort specially interesting to you ?
MG> I need a fast stable sort, so multiple sorts work as expected (contrary to
MG> QuickSort).
MG> TimSort is a candidate.
That's exactly the answer I was looking for, the stabilit
