The questions that Andy Heath has posed are challenging and need more time for reflection than is possible here. The questions posed will inform the research as it develops further. However, in the interests of debate we would like to give our initial replies. We agree in general with Andy Heath's assessment of the limitations of the approach we are adopting. We recognise that this approach uses a very limited response to AccessForAll principles: our Transformation Augmentation and Substitution Service (TASS) is localised, not global, and relies on a limited set of learning patterns matched against a small subset of the potentially infinite set of learner profiles. Our intention is certainly not to reproduce the considerable efforts of the IMS AccessForAll or Dublin Core Adaptability working groups, but to interpret their potential impact on the thinking of courseware designers, tutors and students.