var fDesc=new Array();
fDesc[0] = "In the March 1997 issue of Scientific American, Shawn Carlson described a remarkable algorithm for tackling challenging problems, like the Travelling Salesman problem, by adopting the approach &quot;Nature uses&quot; when a metal anneals or when a crystal forms. The algorithm is called &quot;simulated annealing&quot; and was implemented in C.<br /> <br /> EuroSoft converted the C code to Visual Basic and provided an easy-to-use visual interface which makes it extremely straightforward to solve the Travelling Salesman&#039;s problem for visiting anything from 10 to 3000 arbitrary cities, the location of each city being specified by its x- and y- co-ordinates.<br /> <br /> New Zealander Tony Cooper took the version developed by EuroSoft and enhanced it as follows:<br /> <br /> - A graphical interface was added to allow the interactive plotting of points (left-click to place a point at the cursor) and to allow the progress of convergence to be viewed visually. This second capability provides dramatic visual evidence of the effectiveness of the simulated annealing algorithm.<br /> <br /> - A C-based dynamic linked library (Carlson.dll) was developed for the algorithm to significantly improve running times. For example, on an Intel Core Quad Q6600 based system, a 1000-point problem takes around 6 seconds, and a 2000-point problem around 17 seconds.";

function tShowHide(id, show)
{
	var s = document.getElementById("desc");

	if ((s.innerHTML.length<=212 || show==1) && show!=2)
	{
		s.innerHTML = fDesc[id];
		if (document.getElementById('m1'))
			document.getElementById('m1').style.display='none';
		if (document.getElementById('m2'))
			document.getElementById('m2').style.display='none';
		if (document.getElementById('more_txt'))
			document.getElementById('more_txt').style.display='inline';
	}
	else
	{
		s.innerHTML = '';
	}
}