A metadata repository is a database. More specifically it is a database that stores data about other data. Websites use metadata to provide descriptions about the types of data they contain. Metadata may contain things like the browser agents (Mozilla Firefox or Internet Explorer) as well as other pieces of information that describe the items on the page(s).
The metadata repository is a way to ensure consistency of metadata. Allowing this data to be stored in a database and referenced by the applications or pages that need it keeps the data consistent. If the metadata is referenced as needed, from a metadata repository, the storage space used to store these descriptions of a sites data will be less as well.
A metadata repository, when used appropriately, can allow the data stored about specific web sites, such as those maintained within a single organization, to be consistently searched and made available when the information is needed. Since there is no single metadata repository, the information is not very easily used across companies. As the technology grows and more Internet organizations find a need for these massive quantities of data about data, there will likely be an attempt to centralize the data contained within metadata repositories.
Metadata is different across platforms, devices, and applications. Devices such as an iPod contain music and video data. The metadata for the data stored on the device would be things like the artist name, album name, and track information. Through the use of software, the user of the device and add or edit the metadata of any song or video anytime. Digital cameras also record metadata about the photographs they take, such as image size, camera type, date and time information, and other details. Websites like Flickr can read this information from the photos uploaded to their servers and allow users to look for pictures taken by a certain camera type or in a certain format.
Microsoft Windows also stores metadata for things like word documents. The data is easier to edit or access from Windows Vista than it has been in previous versions of Windows, but data such as author name, access date, date created, file size and many other pieces of information are stored for each file. Windows will not ask for the user or creator of the file to populate many of the items it is capable of collecting. Some things are populated only if the user wishes to populate them, others are recorded when certain actions happen — making them invisible to the user.
A global metadata repository is likely in the future somewhere, and there will be some who do not want to share all of this information, but much of it is used by services that are growing in popularity everyday and sharing this information only stands to improve these services.