Can you totally trust the books?

Here is a puzzle from '303 Perplexing Chess Puzzles', a 2004 book compiled by Fred Wilson and Bruce Albertson and endorsed as an official MENSA Puzzle Book. Fairly high recommendation indeed.

This is #57 from the book, a reasonably good looking problem.

Now the problem for you is, is this correct?

Have a go at it now.

I read once before that one should read analysis but never trust it until you have confirmed it as correct. I guess the same applies to books!

ur right! cool!

Chess history is chock-full of "cooked" problems, and any number of chess books (including many pre-computer classics) contain faulty analysis. The challenge is, knowing the "cook" in the problem above, can you repair it to make the idea work?

Nice find dude.  No, you can't always trust the books.  By the way, you can't always trust the computers either. 

Yes, but black will have turned a position of some advantage into a draw, which can't be a positive conclusion. In any event, a Ra2 by white may serve to stop some of the perpetual threats.

@ DimKnight, the fix in this one is easy but I agree with you, sometimes finding a bust and then finding the fix can be a fun thing to do.