RPM's and Debian packages might still be provided but this is not sure due
to the current lack of security of the code. If they are, find them on:

	http://www.sqlay.org/download/

	or

	http://sourceforge.net/projects/showfiles.php?group_id=69329

Otherwise, simply download the source code from the same locations and
read further.

From that point, you basically need to decide what you're interested in:
all, the C-C++ applications only, the PHP framework only or the PHP
framework with the PHP module (using the C drivers).

If using RPM's or deb's, you may take a look at point 2 which introduces
the C-C++ applications, then skip to last point which will explain the
use of the PHP framework.

If not using RPM's nor deb's and wishing to use the C-C++ stream or the 
PHP module, or PHP code only but with manual installation, you first 
need to compile and install this way:




1. Manual compilation and installation:

The simplest compilation doesn't need any developpement package and should
normally compile on any Linux machine. Some other Un*x flavors may work
partially (i.e. MacOSX, FreeBSD, Solaris). 
Nevertheless, I'm afraid Cygwin support is now broken.

If you wish to build the ncurses binaries, you will need the ncurses-devel 
package (I started to develop using 5.2 but maybe older can work as well), 
and in order to build the Qt binaries, you will need xdvel and qt-devel; 
the Qt schema viewer builds (and runs with most functionalities) with Qt 
from 1.x (I tested from 1.45 until 3.3.4) while you will need QT > 2.x if you also
want the datasheet. If you wish to build the PHP framework, you need the
msgfmt executable, shipped with gettext.

The C database drivers of course need the client and devel packages for their
RDBMS. You can check in the man pages, available also in html format in the
source tree under the doc/ subdirectory, to see with which versions of the
clients the drivers have been successfully compiled.

Those packages are shipped with most Linux distributions, so the easiest way 
to install them is probably from your CD's, otherwise you can visit those
sites: 
- http://www.gnu.org/software/ncurses 
- http://www.trolltech.com/  
- http://www.postgresql.org/ 
- http://www.mysql.com/ 
- http://www.freetds.org/
- http://mdbtools.sf.net/ 
- http://www.sqlite.org/
- http://www.unixodbc.org/
- http://www.iodbc.org/

So, if you have none, some or all of those installed, now run:

	$ ./configure --help

in order to get an idea of the options available, namely --with-php, 
--with-ncurses and --with-qt for the front-ends, and --with-pgsql, 
--with-mysql (by default on), --with-freetds, --with-sqlite, --with-mdbtools 
--with-unixodbc and --with-iodbc (by default off) for the back-ends.

--enable-debug would allow you to run the data sheets with edition enabled
but be aware that it theoretically could damage you data ( GPL + alpha tag
+ small project = for sure no warranty, you should be conscious of that ).

--enable-build-method (hopefully) allows you to circumvent some problems
if building on other than Linux (see below --enable-release-notes).

You may need to adapt the xyz-libs and xyz-includes options if you see 
warnings during the configure phase eventhough it tries to guess as much as 
it can.

You may also try:

	$ ./configure --enable-release-notes
	
for last minute explanations: this just prints some kinda release notes and 
compilation hints then exit.	

Then type:
	$ ./configure --with-the-options-that-you-wish
	$ make
	# make install

If you need to remember the various standard configure options, its standard
INSTALL file is located in the build directory. Note that some of the more
advanced options, automatically added by configure, have not been tested nor
even understood by me :)

Note also that if you have chosen a custom --prefix, or in the (very unlike)
situation where /usr/local/lib is not listed in /etc/ld.so.conf, before
trying to use the software, you must make sure that the directory $prefix/lib
(or $libdir if you choosed to change --libdir) is either listed in
/etc/ld.so.conf, then run ldconfig as root, or, easier for a first try, make
sure that this directory is in the environment variable LD_LIBRARY_PATH.




2. Testing C-C++ applications:


Now, give it a try by typing :

	$ sqlsca
or
	$ sqlscc
or
	$ sqlscq
	
or even just

	$ sqlsc

which is normally a symlink to sqlscc (you may change if you wish).
This should load the dummy (random) driver and show you a schema.

For a real example, you can search for an SQL file on your harddisk or generate 
one by yourself by typing:

	$ pg_dump -s your_database > pg_dump_file.sql
	
if you use Postgres, then try:

	$ sqlsc pg_dump_file.sql

or do online visualisation with:
	
	$ pg_dump -s your_database | sqlsc
or
	$ sqlsc pgsql:
or
	$ sqlsc mysql:
	
and so on ... which would try to connect to your default database on localhost
without password or, i.e.:

	$ sqlsc mysql:yourdb/
	
(note the trailing slash) to connect to a specific db.	
	
You should then see a schema, including pk's indicated by '+' on the 
left side of the column name, fk's indicated by '-' and datatypes 
on the right side.

The data sheets are available with executables called sqlds, so, i.e.

	$ sqldsc mysql:yourdb/histable

should display in ncurses flavour data in read-only mode if you are on a view 
or if you haven't compiled --with-debug.

So, from here, I suggest you to read the man pages if you want to go further.




3. Configuring the PHP framework:

First, You will need to make sure that a subdirectory called sqlay is accessible from 
your include_path, (i.e. SqLay has been installed with --prefix=/usr, the PHP 
framework should be located under /usr/share/sqlay/php/sqlay and /usr/share/sqlay/php 
is in your include_path, then everything is fine). This can be done in several ways: 
inside each page, in the site-wide php.ini or, in Apache (others ?), either in the 
(virtual) server config or in a .htaccess file by using the php_value directive.

Second, you should make SqLay icons accessible in your web server tree, i.e. by
declaring an Alias in the (virtual) server config pointing towards the icons (located
under $prefix/share/sqlay/icons/default). Alternatively, you may copy them anywhere
in your existing web tree.

Then you need to make sure that a file named sqlay_config.php is found by every page 
calling SqLay library functions. This can be done in the same way as explained above.
This sqlay_config.php is to be written by setting some global variables for the site:
you may take inspiration from a sample sqlay_config.php file.

From that point, you should read the man pages (sqlay_php and sqlay_config) to go
further. Note that some simple sample pages are provided and you can start from
modifying them.


4. Final word:

Please, do not expect too much from those tools. I personnaly use the C/C++ tools for
developpement myself with some satisfaction, but for nothing else.

As for the PHP framework, it is able to satisfy end-users in small teams on not too
sensitive data nor complex schemas, but no more. At this stage, it simply doesn't
provide sufficient functionalities to build production level applications, used 
full-time, or by larger teams, or handling quantities of data greater than several 
thousands (or even hundreds, depending of the use) of rows per table.