Re: books

Tom Allison wrote:
> I have the book, "Postgres Introduction and Concepts" by Bruce Momjian
> dated 2001 (version 7.0.2) and O'Reilly's "Pracitcal PostgreSQL" dated
> 2002 (version 7.1.x).
>
> I already noticed a change in the pg_hba.conf files to include an extra
> column from the books to my current version 7.4.2

Just my personal opinion ... I do not have any PostgreSQL
books, and I think the online documentation is excellent.
In fact it is better than most other documentations of open
source projects that I've seen (which is quite a lot).
If someone asked me for a good PostgreSQL book, I wouldn't
hesitate to answer "Use the online docs". Even if you're
a complete SQL newbie, they get you going pretty quickly,
and they probably contain everything you ever need to know.
I find myself using the reference sections on SQL commands
and data types quite often.

In fact, I guess if you just take the online docs and make
a book from it, it would be perfect. Except that it would
be outdated a few months later, while the online docs are
always up-to-date.

> I am trying to assess the need to keep using these books. I am assuming
> that probably 90% of the books are valid until I get into vacuum and
> some administration goodies.
>
> Without reading all the changelogs yet, am I close?
> I know I'll have to keep hitting the docs themselves to be sure, but I
> would hate to toss these before they get a chance to get really worn out.

A lot of details have changed since 7.0.x and 7.1.x. And
it is often those details which can cause subtle problems
when you don't expect them. For example, I have recently
stumbled across the fact that the "serial" datatype does
not imply UNIQUE anymore (it did in PostgreSQL < 7.3).

Just my two Cents, and YMMV.

Regards
Oliver

--
Oliver Fromme, secnetix GmbH & Co KG, Oettingenstr. 2, 80538 München
Any opinions expressed in this message may be personal to the author
and may not necessarily reflect the opinions of secnetix in any way.

I suggested holding a "Python Object Oriented Programming Seminar",
but the acronym was unpopular.
-- Joseph Strout