Errata

"The file system file is the electronic..."

shouldn't this be "The flat system file is the electronic..." or "The
flat file system is the electronic..."?

"We covered an approach to creating such a model which is in-line
with the type and complexity of databases you are likely to encounter
in developing MySQL or mSQL databases."

I assumed "good" databases used foreign keys.

Also on page 26 item 4 under paragraph 4:

4. Relationships are modeled as foreign keys. We will cover this later.

I never saw anything on foreign keys until Chapter 15 where it states:
page 234 paragraph 1:

"The FOREIGN KEY operation is currently not implemented in MySQL.
While the syntax is there, attempting and action on a FOREIGN KEY
will do nothing."

The Question is, how should a proper MySQL database be designed if the
FOREIGN KEY feature does not work. Or if it does and I have missed that
part in the book where can I find it?

"primary key" should be deleted from the table.

errors too numerous to identify here. The example should be replaced with
the following:

Example 2-1. An Example Script for Creating the CD Database in MySQL

CREATE TABLE CD (CD_ID INT NOT NULL,
RECORD_LABEL_ID INT,
CD_TITLE TEXT,
PRIMARY KEY (CD_ID))
CREATE TABLE Artist (ARTIST_ID INT NOT NULL,
ARTIST_NAME TEXT,
PRIMARY KEY (ARTIST_ID))
CREATE TABLE Song (SONG_ID INT NOT NULL,
CD_ID INT,
ARTIST_ID INT,
SONG_NAME TEXT,
PRIMARY KEY (SONG_ID))
CREATE TABLE RecordLabel (RECORD_LABEL_ID INT NOT NULL,
RECORD_LABEL_NAME TEXT,
PRIMARY KEY (RECORD_LABEL_ID))

errors too numerous to identify here. The example should be replaced with
the following:

Example 2-2. An Example Script for Creating the CD Database in mSQL

CREATE TABLE CD (CD_ID INT NOT NULL,
RECORD_LABEL_ID INT,
CD_TITLE TEXT(50))
CREATE UNIQUE INDEX CD_IDX ON CD (CD_ID)
CREATE TABLE Artist (ARTIST_ID INT NOT NULL,
ARTIST_NAME TEXT(50))
CREATE UNIQUE INDEX Artist_IDX ON Artist (ARTIST_ID)
CREATE TABLE Song (SONG_ID INT NOT NULL,
CD_ID INT,
ARTIST_ID INT,
SONG_NAME TEXT(50))
CREATE UNIQUE INDEX Song_IDX ON Song (SONG_ID)
CREATE TABLE RecordLabel (RECORD_LABEL_ID INT NOT NULL,
RECORD_LABEL_NAME TEXT(50))
CREATE UNIQUE INDEX RecordLabel_IDX ON RecordLabel (RECORD_LABEL_ID)

gunzip -c mysql-x.xx.xx.tar.gz | tar xvf -

The first step is to gzip.

mysql_install_db script " ..will also start the server daemon"

I've installed mysql Ver 9.38 Distrib 3.22.32, for pc-solaris2.7 (i386) from
source and the install db script doesn't start the daemon you need to run
safe_mysqld & to do that.

this might be worth checking to save grief for others.

ALSO - though I can't see it really having any place in the text, users
installing it on Solaris need to use gtar (GNU tar) rather than the Solaris tar
when unpacking mySQL - don't know why but its on the mysql web page. Solaris
tar reports checksum errors unpacking the source and binary distributions.

The first step is to gzip !!!

b.) Some relevant information about the MySQL version I'm using:
C:mysqlin>mysqladmin version
mysqladmin Ver 6.8 Distrib 3.21.29a-gamma, for Win95 on i586
TCX Datakonsult AB, by Monty

Server version 3.21.29a-gamma-debug

the mysqlaccess utility.

The text currently reads:

For example, using the command mysqlaccess nobody isp.com
mydata might report the following output:

It should read:

For example, using the command mysqlaccess isp.com nobody mydata
might report the following output:

This is backed by the following information:

Upon issuing the mysqlaccess command with no syntax:
raptor /home/baphomet# mysqlaccess --help
mysqlaccess Version 2.04-alpha2, 09 Feb 2000
By RUG-AIV, by Yves Carlier (Yves.Carlier@rug.ac.be)
Changes by Steve Harvey (sgh@vex.net)
This software comes with ABSOLUTELY NO WARRANTY.

You have to supply a userid.
You have to supply the name of a database.

Usage: mysqlaccess [host [user [db]]] OPTIONS

Version number of MySQL Daemon for further clarrification:
raptor /home/baphomet# mysql -V
mysql Ver 9.38 Distrib 3.22.32, for pc-linux-gnu (i686)

Should the text:
"... CREATE TABLE people (CHAR(10)) ..."
be instead:
"... CREATE TABLE people (name CHAR(10)) ..."
as it refers to the example on the previous page?

should be:

DROP TABLE IF EXISTS table_name

The first example reads:

"CREATE TABLE cities (id INT NOT NULL PRIMARY KEY AUTO_INCREMENT,"

This causes a syntax error on version 3.21. It should read:

"CREATE TABLE cities (id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,"

The salt is two random characters stored at the beginning of the encrypted password
string, and has no correlation to the beginning of the actual string being
encrypted, so the query provided would only return a result on the off chance that
the salt is the first two characters of the name. The correct query should be:

SELECT name
FROM people
WHERE password = ENCRYPT( name, LEFT(password, 2))

and lowercasing ending tags as a convention. I've worked and conversed
with hundreds of HTML authors, HTML editor programmers, HTML validator
programmers, IETF representatives, read many HTML authoring books and
web pages, and never seen this before. Obviously, it isn't incorrect, but
you might want to do something more conventional. The next HTML spec will
require lowercase starting and closing tags (higher density of lc
letters compresses better than using uc) so perhaps standardizing to all lc in
the next printing might be a good idea... and passing the idea onto
editors of other books might help as well.

The author mentions the IMAGE style as usually being incompatible with "hearing-
impaired accessible browsers"... I have friends in both the hearing-impaired and
visually impaired communities, and am fairly sure the author meant "visually-impaired
accessible browsers"

with "</body></html>".

"} elsif ( not $output->param('start') ) {"

to:

"} elseif ( not $output->param('start') ) {"

of a comment ("opens a") on the subsequent line, uncommented.

under use strict: neither $table or $server are declared with 'my'
anywhere.

my @fields = @{$table_data->NAME};

and

my @types = @{$table_data->TYPE};

should be

my @fields = @{$table_data->{NAME}};

and

my @types = @{$table_data->{TYPE}};

since NAME and TYPE are properties of the statement handle object. Without
the {}, the compiler looks for NAME and TYPE methods, generating a compiler
error when they are not found.

Splendid book!

But it should have been:

$query = "select * from sharks";

Because in the latter part a check is performed
and the where part is added!

In the section on PHP, there are several errors.

Line 11 of the example code, which reads:

$query = "select * from sharks where ";

should read:

$query = "select * from sharks ";

The code later tests for search terms, adding "where " to the string if search terms
are present.

Also, on lines 13, 18, and 22, the += operator is used instead of the .= operator.

(.= is used for string concatenation, += is used for addition)

"myql_create_db()"

should be:

"mysql_create_db()

connection=mysql_real_connect(&mysql,
"athens.imaginary.com",
0,0,
"db_test",0,0);

is wrong.
mysql_real_connect requires 8 arguments.

So

mysql_real_connect(&mysql,"atehns.imaginary.com",
0,0,"db_test",0,"",0);

is better.

and...line 3:

#include<mysql.h>

I don't know other Linux distribution.
But In RedHat based linux distribution, I think <mysql/mysql.h> is
more exact than <mysql.h>.

So...I think the next version of the book should discuss the
difference of distribution.

gcc -L/usr/local/mysql/lib -I/usr/local/mysql/include -o select
select.c-lmysql -lnsl -lsocket

should read:

gcc -L/usr/local/mysql/lib -I/usr/local/mysql/include -o select
select.c-lmysqlclient -lnsl -lsocket
^^^^^^^^^^^^^
The "-lsocket" command line switch isn't required under Linux (in fact,
it will cause an error if present) Another one of those trial and error things.

Great book, despite the annoying typos. I've yet to come across a bad
O'Reilly book, come to think of it.

balue..." should be "/* the return value..."

mysql_real_connect needs 8 parameters and not 7
so it should look like:

connection = mysql_real_connect (&mysql, host, uid, pw, db, 0, "", 0);

or some how else ...

INSERT INTO old_data ( id, date, field ) SELECT ( id, date, field)
FROM data WHERE date < 87459300;

should read:

INSERT INTO old_data ( id, date, field ) SELECT id, date, field
FROM data WHERE date < 87459300;

ON clause, reads:

SELECT * FROM people, homes LEFT JOIN people, homes ON people.id=homes.owner

It should read:

SELECT * FROM people LEFT JOIN homes ON people.id=homes.owner

The e.g. currently reads:

(e.g., ENCRYPT('mypass', '3a') should return "3afi4004idgv").

It should read:
(e.g., ENCRYPT('mypass', '3a') will return "3aMC/FUHSlicQ").

Unless ENCRYPT won't always give the same result from a password/salt
combination.

BOOK EXTRACT :-

DAYOFWEEK(date)/WEEKDAY(date)
Returns the number of the day of the week (1 is Sunday) for the given date
(e.g. DAY_OF_WEEK('1998-08-22') returns 7).

ERROR :-

This section is misleading, it gives the impression these 2 functions do
the same thing and return 1 for Sunday, but they dont.
DAYOFWEEK does return 1 for Sunday to 7 for Saturday as stated, but
WEEKDAY is different it returns 0 for Monday to 6 for Sunday.
Also 'e.g. DAY_OF_WEEK' shouldn't have underscores in it.

The last "paragraph" of code in the example calls

$statement_handle->execute("J%");

which should be:

$statement_handle2->execute("J%");

The example code for this command is incorrect. The incorrect example
code, as given is:

07: my %row1 = $mytable_ouput->fetchrow_hashref;
08: my @field_names = keys %row1;
11: my @row1 = values %row1;

In line 7: $mytable_ouput should be $mytable_output (missing "t" in output)
In line 7: %row1 should be $row1 (a variable instead of a hash since a hash
reference is returned)
in line 8: keys %row1 should be keys %{$row1}
in line 11: values %row1 should be values %{$row1}

Correct code segment, therefore should be:

07: my $row1 = $mytable_output->fetchrow_hashref;
08: my @field_names = keys %{$row1};
11: my @row1 = values %{$row1};

The entire example, corrected should be:

01: use DBI;
02: my $db = DBI->connect('DBI:mSQL:mydata',undef,undef);
03: my $query = "SELECT * FROM mytable";
04: my $mytable_output = $db->prepare($query);
05: $mytable_output->execute;
06:
07: my $row1 = $mytable_output->fetchrow_hashref;
08: my @field_names = keys %{$row1};
09: # @field_names now contains the names of all of the fields in the
query.
10: # This needs to be set only once. All future rows will have the
same fields.
11: my @row1 = values %{$row1};

Mysql::errno. To do this the, following line on p. 413 has to be changed.

The line currently reads:
print "Error " . $output->errno . ": " . $output->errmsg ."

It should read:
print "Error " . $db->errno . ": " . $db->errmsg ."

"DATE datatype, 72, 95, 236, 270"
to:

"DATE datatype, 71, 95, 236, 270" ... ?