I think it is caused by the changes in tm v0.6 from tm v0.5-10. I haven't played much with tm v0.6, and will investigate it later. Current too busy to improve the code for tm v0.6.
A quick and simple way is to use tm v0.5-10, based on which all original code can run successfully without any changes.
Please post your futurtion questions to the RDataMining group on Linkedin at http://group.rdatamining.com/, instead of here. Thanks.