I just got my brand new copy of Barton Zwiebach's A First Course in String Theory and it's sure different from Polchinski or GSW! Here's the info. It's aimed at undergraduates; it was developed from a course given to MIT sophomores. Now you can assume that MIT sophs are better prepared than sophs at most other instituions, but still.. All he requires is "exposed to special relativity, basic quantum mechanics, electromagnetism, introductory statistical physics. Some familiarity with Lagrangian mechanics is useful but not indispensable." I think that many members of this forum could swing that. His aim is to bring the students to the cutting edge as quickly and honestly as possible; this constrains him. He introduces and uses light-cone coordinates and quantization, and postpones covariant quantization till much later. He starts off with relativity in the L-C framework, pulls in multiple dimensions, and then goes into "manifestly relativistic electrodynamics". All this before the first string chapter. He does bosonic strings, often motivating with a prior chapter on the corresponding point particle. His math is appropriate to his audience, he doesn't use anything deeper than basic multivariable calculus without introducing, motivating, and explaining it first. But be aware, he does go the true route and if you're not prepared to do the excercises you'll miss the value of the book. Part 2 is developments. He does D-branes, string charge, string thermodynamics and black holes, T-duality, EM fields on D-branes, nonlinear electrodynamics, finally the covariant quantization, strings and Riemann surfaces, and loop amplitudes in string theory. This is going to be not only a popular textbook, but a lifeline for autodidacts. If you've got any questions about the book I'll try to answer them on this thread. Remember, though, I've just received the book, and haven't read it in depth.