Due to a reform in the Israeli CS high school curricula, in-service teachers who studied and taught procedural programming are now required to cope with the new paradigm of OOP. In this paper we describe a case study in which we traced the difficulties of one in-service teacher as she taught the topic of inheritance and polymorphism for the first time. We noticed a growth in her knowledge in the laboratory problem-session, whereas in a similar pen-and-paper session no substantial knowledge growth took place. A qualitative analysis enabled us to discern how the combined factors of the teacher's experience, the presence of the computer, and students' expectations for workable solutions, pressured the teacher to seek to resolve difficulties through trial and error, which facilitated knowledge growth. We argue that all these factors have to be fulfilled in order for teacher learning to happen.