CDDB

Certainly, the CDDB (short for Compact Disc Database) is a great idea. Many people use it without even knowing it and probably don't even know what it is, but its history is exemplary when it comes to the relationship between free enterprise, the dissemination of knowledge, and major labels.

So, this is the story.

When the audio CD specifications were drawn up, the original designers, Philips and Sony, didn't bother to include any identifiers on the disc, such as the album title, the names of the artists, and the song titles.

As often happens, they didn't immediately realize the possibilities offered by digital technology, and a new medium was seen solely as a replacement for the old one, which was simply a medium for audio, without any related information. They didn't consider that, since the new format was digital, it would be easy to include a record containing data about the content. It wasn't on vinyl, so why put it on CD?

Consequently, there's nothing written in the CD's digital data that allows us to identify the title, artist, and tracks. All that information is contained in the label printed on the disc.

The problem is that technology is rapidly advancing and creating new ways of enjoying music that are often unrelated to a specific physical medium (the major labels are only now realizing this). Almost immediately, it was realized that, even without doing anything illegal, the content of a legitimately purchased CD could easily be transferred to a hard drive for listening on the go or at work on a laptop, and later, with the arrival of MP3, on portable players.

The point is that, once the content was separated from the physical medium, the information printed on the latter was also lost. There were two solutions. The major labels created the CD-Text specification, a format compatible with audio CDs in which there is a space to place data on the content. CD-Text, however, is not universally used.

The second solution was to create an open archive on the Internet that would allow CDs to be identified. A database was thus created in which each CD was identified by a "disc-id," a string of letters and numbers calculated from its contents, taking into account the length and sequence of the tracks. For example, the disc-id of the Cure CD I'm listening to is "910d120c." This identifier is difficult to duplicate because it's unlikely (though possible) that another CD with tracks of the same length exists. If other data, such as the number of tracks, their sequence, and the length of each track in the sequence, repeatability of the ID becomes very difficult.
Each ID is associated with a record of information in text format containing the album title, artist, song titles, year of release, etc. This allows players to connect to the Internet and retrieve the information.

The CDDB (CD database) was invented by Ti Kan and Steve Scherf.
The source code was released under the GNU General Public License, allowing everyone free access to information made available by many people.
Later, however, the project was sold and the licensing terms were changed, requiring an initial fee to cover server usage and necessary support, which fell to the commercial developers. Furthermore, the license also included some clauses that many programmers considered unacceptable: no other similar database could be accessed as a complement to CDDB, and the database logo had to be displayed during access.

The license change motivated the freedb project, which has the same goals but intends to remain free and open source.

In March 2001, CDDB became the property of Gracenote, which prohibited access to its database by all unlicensed applications. Licenses for CDDB1 (the original version of CDDB) were no longer available when developers were asked to convert to CDDB2 (a new version incompatible with CDDB1, and automatically also with freedb). In practice, if someone wanted to develop a player that accessed the CDDB, they had to pay.

After Gracenote's CDDB was released commercially, many media players switched to using freedb, while retaining "CDDB" as a generic term for its database function.

In October 2006, MAGIX acquired freedb, which remains free and open to the public. However, another project, called MusicBrainz, was launched. It aims to be much more than a simple CD database. Its goal is to create an open-content music encyclopedia. It is an online database of information about recorded music, but it is not a music database. MusicBrainz collects information about artists, their recordings, and the relationships between them. Entries for each musical work include the album title, track titles, and the playing time of each track. These entries are maintained according to a common style guide. Recorded works may also include information about the release date and country, the CD ID, the audio signature of each track, and have an optional field for adding text or annotations. As of June 2006, MusicBrainz contained information on 243,000 artists, 399,000 albums, and 4.8 million tracks.

But take a look at this table of online music databases that are largely free of charge from Wikipedia.

The whole story is exemplary of how the commercial system treats information. Note that we're not discussing copyright on content (songs, writings, etc.) here, but only information about the contents of a CD, the compilation of which, ultimately, is a service.

Back