Nobody, including Phil Plait of Bad Astronomy, has been able to explain to me how they can ever form. My problem: as matter approaches the even horizon, where the escape velocity reaches the speed of light, from OUR point of view it falls at an asymptotically slower rate... in fact, from our point of view, nothing ever reaches it. The same would appear to be true of the matter on the surface of the collapsing star.
When you say, "from OUR point of view" it takes forever, that's not really correct. Black holes actually form in a finite amount of time from anyone's point of view. People get very confused about this, but it's really not that mysterious. You only come to the infinite-time conclusion if you measure time by the time it would take for a light signal to reach you from something falling through the horizon. That time, naturally, goes to infinity at the horizon (by definition), but that does not mean the black hole hasn't already formed. Moreover it's not really true that you would never see something cross the horizon. What you actually see is that things get darker and darker, and the light from them gets redder and redder, as they approach a surface. Rather rapidly the light gets so dim and redshifted that you cannot see the object (even if it's emitting its own signals). It's true that the last signal you see will always have been emitted from a point slightly outside where the horizon was at that time, but the difference between that and seeing it "actually cross the horizon" is essentially zero. Remember, you cannot ever see it
after it's crossed the horizon - by definition - so what else would you expect?
Let me make an analogy: suppose you had a river flowing down over a waterfall. As the water approaches the waterfall the current gets stronger and stronger, and the water flows faster and faster. Now supposed you were a fish living in that river above the falls, and you had no eyes - your only means of navigating was sonar. So you send your friend down the river to investigate, and she talks to you or sends little pulses of sound back to you as she swims around, telling you what she is observing.
Now if there is a point near the waterfall where the water is flowing faster than the speed of sound, that's a sonic horizon. As your friend approaches that point, the signals she sends (or sonar pulses you send which bounce off her) will take longer and longer to reach you (because the speed of sound
relative to you, or to the banks of the river, goes to zero at the horizon). If she crosses it, you will never hear from her again (in fact unless she is capable of supersonic swimming she is going to fall off the waterfall), but you may continue to receive sound pulses from her for a long time, each arriving at increasingly longer and longer intervals (those are the pulses she emitted as she was about to cross the sonic horizon).
But from hearing those fading sounds you would certainly not conclude that it took her infinite time to cross the horizon, right? That would be wrong. You'd just understand that sound takes longer and longer to escape from a horizon, and that's why her signals took longer and longer to reach you.
For black holes the conclusion that it took infinite time is not
as wrong - it's actually consistent, so long as the black hole doesn't evaporate - but you don't
have to use that particular choice of time, and it's much more useful to think about black holes using other time coordinates. The only major difference with the river is that no signal can ever propagate faster than light, which means it's consistent to totally forget about the interior of the black hole, which means it's OK to use a time coordinate which goes to infinity on the horizon (but again, only if you ignore that fact that black holes eventually evaporate and disappear).
We had a long thread about this - I'll dig up the link to it.
EDIT - OK,
here it is. The discussion starts with DrBaltar's post 226 of that thread (where he stated essentially the same confusion you have), in case that link doesn't take you there.
