Through the study of the evolution of computers and of their interfaces, this research seeks to survey the origins of the graphic interface and the motives why they evolved so little in more than twenty years since its introduction. In a second stage, through the analysis of 10 interviews, the techniques used by some professionals of graphic areas to deal with digital information in their machines are discussed - how they get round some of the limits imposed by the graphic interface and how they alter their way of working when they find limits which cannot be transposed. The results presented can serve as subsidies to the development of future interfaces best adequate to the activities of the professionals who live off the manipulation of images.