Re: [GENERAL] Book nearing final form

> At 06:28 PM 07-06-2000 -0400, Bruce Momjian wrote:
> >I have been told by the publisher that most changes to the book must be
> >done by next week.
> >
> >The book is at:
> >
> > http://www.postgresql.org/docs/awbook.html
> >
> >Again, many thanks to those that have supplied suggestions for the book.
>
> More "typical usage" examples on commands would be helpful. For example the
> cluster command may need a typical usage example. Not everyone can parse
> man pages and understand the differences with stuff being between <> and []
> and {} so on... Typical usage examples can help the majority actually get
> things done.

CLUSTER has limitations currently so I don't want people running it
without reading the manual.

>
> Also a LIMITS section/appendix would be very useful. This should list stuff
> like limits on tuples, queries, names, fields, data types, number of tables
> (theoretical and practical), number of indexes, and so on. It can be rather
> misleading to say that "text" has no limit and only when people bump into
> the tuple limit, we tell them of the infamous 8KB row/tuple limit.
> Naturally this would be out of date soon, but that just sets the stage for
> the 2nd edition ;).

That is in the FAQ. That way, then can look at the current version to
see the current values. FYI, 8k limit is going away in 7.1, and the
book will be out after 7.1 is released.

>
> An example for "create user with password" would be good too - since the
> web example was unclear/problematic the last time I checked - I had to do
> some trial an error.

Has to be associated with pg_hba.conf. Too complicated.

>
> Also another thing should be made clear:
>
> Stuff like vacuumdb, createdb etc are basically scripts, whereas the actual
> commands are vacuum, create database and so on. A clear distinction should
> be made between the two, I feel this would clarify things a lot, especially
> when you configure Postgres in a way which some scripts aren't comfortable
> with (documentation or usagewise)- access control for instance.

Not sure.

>
> Should there be mention of turning off fsync for performance reasons? Or is
> that a bit risky/not officially supported? I find there is a dramatic
> performance difference, however I'm still a bit nervous about it :).

Also will be gone in 7.1. Was removed from book.

--
Bruce Momjian | http://www.op.net/~candle
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026