On December 17, 1997, Michael Robertson and Greg Flores launched mp3.com. Flores had noticed that MP3 was a popular search term, so he and Robertson purchased the domain from the owner and went into business.
What MP3.com was
Initially mp3.com provided a way for musicians to promote themselves and provide original music to their audiences, acting as an early streaming service. It was immediately successful, receiving 18,000 unique users in its first 24 hours, and its first advertising purchase within 18 hours of launch.
It had 25 million registered users and an average of 800,000 users per day, delivering 4 million MP3 audio files per day. An algorithm paid artists based on the number of streams and downloads of their songs.
Like many dotcoms, MP3.com ran on Red Hat Linux, specifically, versions 5.2 through 7.2. It consisted of 1,500 servers in load balanced clusters. MP3.com went public July 21, 1999 and its IPO raised $370 million.
My.MP3.com
On January 12, 2000, MP3.com launched my.mp3.com, a new service to provide access to music its users already had legitimately purchased on CD.
The idea with my.mp3.com was it would give you access to the music you already owned. When you logged into its web interface, you could insert a disc in your CD-ROM drive and it would scan it. Assuming the CD was in its library, it then gave you access to the MP3 content.
The problem in 2000 was that ripping CDs was still time consuming. Software existed to make it reasonably convenient, but CPUs were not fast enough yet to perform the encoding quickly. I can rip a CD in 5 minutes today on a computer that isn’t particularly new. In 2000, my computer wasn’t especially old, but it would take a couple of hours to rip a CD.
Some of the stuff I downloaded off Napster was stuff I already owned. It was faster, or at least more convenient, for me to download it rather than ripping it myself. I could queue up several downloads and not have to bother switching discs.
Robertson wanted to provide similar convenience and believed that making users prove they possessed the CD would differentiate themselves enough from Napster and other file sharing services to be legal.
The downfall of mp3.com
The downfall for mp3.com was that possession of the CD was not sufficient proof of ownership. There was a small chance that the CD was a copy rather than an original or you might have borrowed the CD from a friend or from the library. mp3.com didn’t have a way of knowing the difference.
UMG recordings sued mp3.com out of business. They settled for $53.4 million, but that didn’t leave mp3.com enough revenue to continue operations. In May 2001, Vivendi Universal purchased MP3.com for $372 million in cash and stock. Vivendi Universal in turn sold it to CNET in 2003.
Today, streaming services collect subscriptions, stream music to you whether you own it or not, and frequently don’t pay the artists. But since it is a corporation performing the theft rather than an individual citizen potentially committing theft, they get away with it. Justice is for those who can afford it.
After MP3.com, Michael Roberstson went on to form Lindows, which later became Linspire.
Amazingly, after all the legal troubles surrounding 1990s efforts to monetize the MP3 file format, when the last MP3-related patent expired, it seemed like nobody really noticed.
David Farquhar is a computer security professional, entrepreneur, and author. He has written professionally about computers since 1991, so he was writing about retro computers when they were still new. He has been working in IT professionally since 1994 and has specialized in vulnerability management since 2013. He holds Security+ and CISSP certifications. Today he blogs five times a week, mostly about retro computers and retro gaming covering the time period from 1975 to 2000.
I think the real downfall of MP3.com was that the courts didn’t understand technology. What they were doing was deduplication of data — standard practice in data centers nowadays, but the word didn’t even exist back in 2000. Making even more cutting edge was that they were doing DISTRIBUTED deduplication, using file signatures for matching.
The court ruled that MP3.com’s service was illegal because they were streaming THEIR files to you, not yours. They would have ruled in favor of a service where you uploaded YOUR rip, and they stored it and streamed to you. (Of course, the costs of doing that would have been prohibitive; storage was still expensive back then. I remember buying my first 1 GB hard drive in that era; it cost as much (in actual dollars, not adjusted for inflation) as a 16 TB drive does now.) They simply couldn’t wrap their heads around the notion that they were the SAME file. The fact that they were made from different copies of the CD was irrelevant; the bits were the same, and therefore the company only needed one copy.