Seed

In the computer world, a seed may refer to three different things: 1) A random seed, 2) seed data, or 3) a client on a peer-to-peer network.

1. Random Seed

A random seed is a value used to generate random data. Since computers are programmed to follow instructions based on specific input, creating random output is more difficult than it might seem. In order to generate a random value, another value called a "random seed" must be provided as input. This may be a timestamp (which can include milliseconds), a system value (such as a GUID), a hardware serial number, or another unique value.

A random seed may be generated automatically (e.g., a timestamp) or may be explicitly passed to a randomizing function as a parameter. The function uses the seed value to produce random output, which is typically a number.

2. Seed Data

Seed data is information that is loaded to enable a function or program to work correctly. If a function queries an empty database, for example, it will not produce useful output. If the database is "seeded" with data, the function will generate meaningful results. Seed data is often used for testing purposes. It may be created using an automated process or can be entered manually.

3. P2P Seed

A peer-to-peer (P2P) seed is a or computer (or "peer") that uploads one or more files on a file sharing network, such as BitTorrent. Once a user downloads a complete file, he or she can share the file with other users. Peers that provide files for other torrent users to download are called seeds. Peers that download files but rarely or never upload them are called leeches.