|Perl: the Markov chain saw|
IMRDO (In My Respectfully Differing Opinion ;--) you can _hope_ that people will write proper reviews, but there is not much you can really do to improve what they write.
Especially the "factual" stuff. I can get the author bio and the number of pages from Think Geek or Amazon or my local bookstore. I don't care about the number of illustrations in a book, I'd rather the review writer spend more time telling me what he thought about the book than counting pictures! ;--)
What is really interesting in a book review is whether the reviewer thinks the book is helpful, technically accurate and generally enjoyable. All of which are quite non-factual. And besides it is quite hard for a reviewer to judge both the helpfulness of a book (it has to be someone who needs to learn whats in the book) and its technical accuracy (in which case it helps if the reviewer knows a lot more than whats in the book).
So a review that tells me who the reviewer is and then what he thought about the book is just fine. Other posters with different background can then comment and help the reader get the complete picture.
At least this review showed us a book that I am sure few people here knew.
And now back to some factual information: the book is 1400 pages, paperback, pretty thin paper I've been told, although I don't have the figures here, it was first published in August 1999, it looks like the current edition is still the 1rst (actually an important info, 1rst editions have more typos and errors than subsequent ones, plus publishers usually only prints a second edition if they've sold all of the first one), but there seems to be another one named the "CD-ROM edition" out there, it's listed price is $49.99 but its street price is around $35-$40. Is that enough non-biased info?
Man, I don't think I've ever known so much about a book I will never buy!