Abstract
Model transformations (MT), as with any other software artifact, may contain quality flaws. Even if a transformation is functionally correct, such flaws will impair maintenance activities such as enhancement and porting. The concept of technical debt (TD) models the impact of such flaws as a burden carried by the software which must either be settled in a 'lump sum' to eradicate the flaw, or paid in the ongoing additional costs of maintaining the software with the flaw. In this paper we investigate the characteristics of technical debt in model transformations, analysing a range of MT cases in different MT languages, and using measures of quality flaws or 'bad smells' for MT, adapted from code measures.
Based on these measures we identify significant differences in the level and kinds of technical debt in different MT languages, and we propose ways in which TD can be reduced.