Hi,

"Johan Holmberg" <holmberg / iar.se> wrote:

> An elegant language deserves an elegant book. This one isn't.
[...]
> This organisation is quite OK, but ...
[...]
> But the real problems with the book are of another sort.
> I wonder if anyone has bothered to proof-read the text at all.
[...]
> It's a mystery to me that O'Reilly publishes such a
> poorly proof-read book.

While it's easy to dismiss or attack the bearer of uncomfortable news, I for
one want to thank Johan for daring to raise this issue. (Of course producing
a book is enormous work, and of course it's easier to critize than to write,
and of course no one's perfect, but these also easy-to-say
counter-criticisisms are irrelevant.)

Despite my reservations about the book, I still just bought a copy of it
from my neighborhood book store. It's not as elegant as Ruby, but for me it
is easy to read and contains sufficient patches of elegance that I still
wanted to have it anyway.

I think Matz has made a great start (thank you!), and I think the discussion
Johan has started will vastly increase the odds that Matz will get the
editorial support he deserves in order to make the next edition of this book
(and his forthcoming book) into first class works, for the mutual benefit of
the Ruby World. (Meanwhile, at least O'reilly fixed that damn snail cover
before the book went to press!)

Conrad

PS: Maybe this incident will at last drive O'Reilly to adapt the Thomas-Hunt
process of generating testable examples from the book document source for
future books. As a consumer of many other O'Reilly books (and corresponding
eratta links) and having read about similar problems with first (and
sometimes only) editions of their other books on other newsgroups, I for one
would like to see them test-in reliable examples from the start rather than
depend on highly fallible eyeballs to catch problems.

Since O'Reilly has their fingers in some sort of Open Document project,
maybe they could standardize and release the tools needed to make this
practice widespread and quasi-standard. Of course they wouldn't want to
advertize their leadership in such a widespread quality improving venture.
:-)