Good book choice?

[rzed]

"Rob Andrews" <rob at diespammerdieuselesspython.com> wrote in message
news:3D873E8F.7060501 at diespammerdieuselesspython.com...
> > <vent>
> > This isn't directed at the book mentioned (which is why I didn't
quote
> > its name in this rant), rather at all books which come with CDs.
I hate
> > it when books come with CDs.  I'm perfectly capable of downloading
the
> > source code used in the book, and in fact I prefer to hand-type it
while
> > reading from the book anyway, to make sure I see every part of the
> > code.  When a book comes with a CD, it is cumbersome to read
portably,
> > which is the whole advantage of books, unless you take the CD out.
And
> > if I take the CD out, then there is no doubt that it will get
lost, or I
> > will be unable to return the book if dissatisfied, etc.  Plus, you
can't
> > tell me that the inclusion of a CD and the envelope/publishing
used to
> > include the CD doesn't add a few units of currency to the book's
cost.
> >
> > Poke this opinion full of holes, but when I buy a book, I want a
book,
> > not a software package.  The book should be good enough to stand
on its
> > own without the "bonus" material on CD.  I think twice about
buying a
> > book that comes with a CD, to tell you the truth.
> > </vent>
>
> All perfectly valid sentiments, of course. But for the village in
Nepal
> lacking bandwidth because they are still six months away from having
a
> phone line (not as far-fetched an example as it may seem), an
entirely
> different set of considerations may be found.
>
> I find myself on the middle path on this issue, since I rarely have
> personal use for the CDs, but find that they do make it easier to
> distribute Python in my fits of zealous advocacy. I have bandwidth,
but
> no burner, and some are more interested in hearing me out if they
don't
> have to face a long (in some cases, expensive) download before they
even
> get to see if they would be interested.
>

The further aspect to it is that code written in a book may be
dependent on the environment the author is working with. Some or all
of it may still be valid a couple of releases later, but if things are
not working, it is useful to be able to set up the *same* environment
the author had to verify that the code is not the problem. Otherwise
you're left translating to the current version of things, and then
debugging code that doesn't correspond to what the author wrote to
begin with.

--
rzed