I thought all photons had spin 1, it was just the direction that changed...
sub-atomic particles can have positive or negative spin. and depending on the type of particle it can be integer spin or half integer spin, ie. -3, -2, -1, +1, +2, +3 or -2.5, -1.5, -0.5, +0.5, +1.5 etc.
The problem with using entanglement to communicate is because, as others have said, the process of measuring the spin collapses the wave function... which means, that before the spin of the particle is measured, we do not know what it is. So particle A has a property of spin, either +1, or -1 for simplicity, and particle B has the opposite spin to particle A. However we do not know which one is +ve and which one is -ve. When we measure particle A it then has a known spin, and we know what the spin of B is.... great - so you'd think we could now change the spin of A, and then that change would be seen in B? Well, not really, since whenever we are NOT measuring the spin, we don't know what the spin is. It doesn't matter if you measured it half a microsecond ago, it could be doing anything now - in fact, according to our best understanding of quantum mechanics, the particle may now be a house in the Yorkshire Dales since we cannot say anything at all about it while we are not measuring it. Measure it again and it's spin may be different just because... because why we don't understand yet.
So although action at a distance is indeed a real phenomena (which has been known about for decades as others have said), there is currently no real viable way of using it for communications...
Now, i'm gonna read that article and see if something new has cropped up to make me eat my words ;P
edit: For the record, anyone wondering how they can do the experiment if it's not yet useful for communications... They made the measurements 18km apart from each other, and then had to use standard methods of communication (whether that's intarwebz or the post office) to compare data and see that the two particles did indeed display this behaviour...