Subsections

2.4 Using the Installer

The script `install.pl' is by far the easiest way to install Scoop. Once you've installed all the dependancies (2.1) and downloaded Scoop, all you have to do is change into the scripts directory and run install.pl. The install script isn't really set up to easily handle multiple Scoop sites, though it can work. Even if you use the install script, you may want to read section 2.5 anyway, just to tweak things a bit.

If you're running Windows, the installer won't work; you'll have to do it by hand, and follow a few special instructions besides. See sections 2.5 and 2.7.

2.4.1 Installing Perl Modules

First, the install script will do some checks, then ask if you want to install Perl modules.

Even if you've already installed all the modules listed, it doesn't hurt. If everything is already up to date, then it will skip over all of them and tell you everything is installed. The script basically follows the instructions in section 2.1.2 for installing Bundle::Scoop.

2.4.2 Configure the Database

If you're upgrading to a new computer, re-configuring your database will drop all the data in it, so unless that's your intention, don't let the script do it. If you're just upgrading the code, you shouldn't be using the install script; go read section 2.6.

For a new install, however, you want to let it continue. You'll need to provide the installer with a username, password, hostname, and port so that it can connect to the database and do the work, though most of the defaults will work fine. Note that the user you provide will need to be able to create a database. At the next step, you'll probably want to choose option 1 (Create a new database), since if you rebuild, it'll just drop a current one.

Unless you have a specific reason to change it, such as if you're planning on running multiple Scoop sites on the same computer, the default database name will work just fine.

After this, however, you'll need to make some choices. First is the path to Scoop on the webserver. What you put here depends on how you want to run Scoop. If you intend to run it on a domain or sub-domain (so that ``http://www.mysite.com/'' will access it), then put ``/'' in for the URL path. However, if you're running it in a separate directory (so that you access it with ``http://www.mysite.com/scoop/''), then put whatever you want the path to be in (which would be ``/scoop'' for our example). The next prompt asks simply for the email address of the admin (probably yours).

Finally, you need to choose a password for the initial user (which will be called ``scoop''). This must be at least 8 characters long. After this, the database will be built (or re-built, depending on which option you chose).

2.4.3 Configure Apache

Once the script has created your new database, it will ask a few questions about what kind of setup you want. Namely, virtual hosted or path based, and what path to use if it's the latter, as well as a few other configuration items.

The script then fills in the configuration file according to what you told it, then saves the configuration file with a name based on the site ID it asked for earlier (httpd-<siteid>.conf). Read this file over to make sure everything is set up properly. Put the file generated by the script in the same directory as your httpd.conf file.

Edit httpd.conf to put an Include directive in the appropriate place, referencing Scoop's configuration file. The exact location of this directive will vary depending on how you have Apache set up, and if you've already done some customizations. Since the generated file contains full <VirtualHost> or <Location> directives, simply adding at the end of your httpd.conf file should work.

The .conf file that Scoop generates for you is not a complete Apache configuration file. It must be included in your main httpd.conf file either via pasting it in its entirety into httpd.conf or by using an Include directive.

Run ``apachectl configtest'' to make sure everything is set up properly. It may not catch some misconfigurations, but it'll catch any syntax errors or missing requirements when it tries to compile Scoop.

If the configtest returns no errors, you can start Apache. If it's already running, stop it first, wait for it to completely halt, then start it. Never use the restart command as that will not properly recompile Scoop.

Now, you can try to load your website, using the URL you told the install script you were using, and start setting it up the rest of the way, as described in section 3.


janra
2004-03-26