ENIAC, the electronic numerical integrator and computer, announced to the public in 1946, is one of the most famous ``first'' electronic general-purpose computers. It is also a transitional machine: unlike our modern concept of a computer, the original ENIAC had no real programming interface. In order to set up a problem on the machine it had to be manually wired: this was considered as one of two major bottlenecks (the second being memory) that was in need of a solution (and which was partially solved after the rewiring). The machine, before its rewiring, was also very modern as compared to the EDVAC design because of its highly parallel character, a property which made it even more challenging to ``program'' the machine. Even though access to the ENIAC was highly restricted, quite a different number of researchers worked with it and had their problems run on it. The physical contact/meeting between this special and difficult machine and different types of researchers resulted in quite different views on computing. The aim of this talk is to discuss the ENIAC experience from the point of view of three of its users: John von Neumann (mathematician and ``homo universalis''), Derrick H. Lehmer (number theorist) and Haskell B. Curry (logician). Each of these researchers had quite explicit and different views on what coding and programming could be like. It is shown how their meeting with ENIAC played a significant role in the formation of those views. The fact that the experience with the same machine by different people results in so different views not only illustrates the significance of the particularity of a concrete human-machine interaction, but also supports the view of the multi-faceted and general-purpose character of ENIAC.