Computation is increasingly important in all branches of science, but most scientists have no idea how reliable their software is, and cannot reproduce even their own computational results if more than a few weeks have passed. \'Live on the web\' labs and other new ideas promise to change this, but face an uphill battle against entrenched practices and institutional lethargy. This talk looks at how we got into this mess, what could get us out of it, and how likely scientists are to actually embrace change.


Talk Number PIRSA:08090037
Speaker Profile Greg Wilson