During the last years the HCI research field has been in a flux: topics and approaches have changed and new ones  kept emerging. A particularly interesting point in this context is that certain issues of original Scandinavian PD that were watered down when it immigrated to US have been suddenly re-recognized now, after a 20 years delay. The talk will try to make sense of the situation by analyzing it against a broader frame of reference: the "practice turn" or practice theory discussion in social sciences. An interpretation is made that the change in HCI can be seen as a shift in the unit of analysis from interaction towards "practices" as they are understood in social science. The practice view has been around in HCI since 1980s, but as a fringe issue: it has had a central role only in PD and CSCW. The current situation suggests that this may be changing. Such change would be, however, rather significant, and need a reorientation by the HCI community in the ways how research is done and evaluated. Partially the HCI community could draw from the social sciences practice research experience, but there are areas essential to HCI where social sciences are of little help. In particular, the dynamics of practices and the role of artifacts are such issues where the HCI community might  -- in principle -- be able to contribute back to the practice research discussion in general. Before such contribution certain gaps in our current knowledge should first be filled, however.