<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type"/>
<title></title>
</head>
<body>
<p style="margin: 0px;"><span> </span></p>
<p style="margin: 0px;"> </p>
<div style="margin: 5px 0px 5px 0px;">
"José Mejuto" <joshyfun@gmail.com> hat am 24. Mai 2011 um 18:07 geschrieben:<br/>
<br/>
> Hello FPC-Pascal,<br/>
><br/>
> Tuesday, May 24, 2011, 1:06:43 PM, you wrote:<br/>
><br/>
> >> Why is TimSort specially interesting to you ?<br/>
> MG> I need a fast stable sort, so multiple sorts work as expected (contrary to<br/>
> MG> QuickSort).<br/>
> MG> TimSort is a candidate.<br/>
><br/>
> That's exactly the answer I was looking for, the stability need ;)<br/>
> I'll try to write an implementation to expand my sorters collection :)<br/>
> but the algo seems quite "confuse" and plenty of "tricks".
</div><br/>
It uses a lot of tricks, but it seems to me that Tim explained all the important things.<br/>
It would be nice, if thetimsort unit has the same license as the FCL (modified LGPL-2).<br/>
Maybe you can ask him for permission.<br/>
The c implementation uses some ugly macros, but the rest should be easily translated to pascal.<br/>
<br/>
<br/>
Mattias<br/>
<br/>
</body>
</html>