Lossless TLM has a firm theoretical basis and has a niche applications area in which it is superior to other techniques. On the other hand TLM for diffusion modelling has had ‘psychological’ problems. These are largely due to questions of its identity, its pedigree and its utility. Conventional lossy treatments start with the Telegraphers’ equation and immediately ignore the existence of a major component while trying to treat problems that have long been the realm of finite difference methods.