Today while 16 participants were simultaneously running the same experiment on MediaLab, 4 of the PCs threw a runtime error 70 and exited MediaLab. The experiment file runs from a server to which the PCs are all linked. Some of the PCs are set up to save to the same folder/data file, but not all of the ones that had the error. We were thinking that maybe multiple PCs were attempting to access the data file at the same time, and so getting a locked file message resulting in the runtime error. Is there something about simultaneously running 20 PCs from an experiment residing on a server that we need to be aware of? Does running the experiment from the same server folder/file cause problems? Attempts to recreate the error in a controlled way have failed so far (e.g., running the experiment on two PCs and both operators hitting the keys at the same time), but this afternoon while running 10 participants we again had runtime errors on several of the machines - even though each was set up to access a different data file to store participants responses... Any ideas?