Explain the term Naming Packages in details?

As you saw previously name space conflicts arise if two pieces of code declare the similar name. Java remains track of what belongs to each package internally. Therefore if Sun decides to put a Website class in java.lang in Java 2.0, it won't conflict along with a Website class described in an http package. Since they're in different packages Java can tell them apart. Just as you tell John Smith apart from John Jones through their last names, so too does Java tell two Website classes apart through their package names.

Therefore this scheme breaks down if two different classes share the same package name as well as class name. In a world in that several different packages can be downloaded from many different sites, all landing in the user's CLASSPATH, it's not unthinkable in which two different people might write packages known as http with a class called Website.

To ensure that package names don't conflict along with each other, Sun asks that you prefix all your packages with a reversed domain name. Therefore if your domain name is poly.edu, your package names should starts along with edu.poly. Within that domain you are responsible for making sure in which no one else writes a package that conflicts with yours.

This is primarily of interest to applet writers, not applet users. If you're surfing the net and you load one applet from MIT in which has a http.Website class and another from Polytechnic that has a various http.Website class, Java can still tell them apart since before it runs any package it downloads off the net it prefixes everything along with the site from which it got it. Instead it sees edu.poly.www.http.Website and edu.mit.www.http.Website. It's only while you download a package manually and install it in one directory in your CLASSPATH and install another package elsewhere in your CLASSPATH in which real name conflicts can arise if packages aren't carefully prefixed with a domain or email address.