Zeno: you make the assumption that the chain is rigid for both observers. Think about how this might be a problem given both ends are being moved in the same direction and it's a chain made out of links, find out about rigid bodies and relativistic physics.
Another error is your assumption that both machines can be made to move simultaneously; this isn't as easy to do in practice as it might seem, but go find out why, ok?
I was just now thinking about what your setup is like trying to do.
You can pull an ordinary chain, say a ship's chain at either end so it resembles a rigid body, if it's laying along the ground, which is flat; if this chain's length isn't too great that the curvature of the earth has to be factored in, then it's also "straight".
But if you then lift either or both ends, i.e. apply forces transversally at the ends, it won't be straight. Or rigid, since the links at each end will be loosened. In fact a freely hanging chain is curved by gravity. If you do what you think can be done with a chain out in space or in the absence of gravity (i,e, in an SR setting), the same thing will happen and the chain will curve. This is because the forces applied at the ends by your machines are not "transmitted" to the rest of the chain immediately.