# The Tragedy of MP3.com
## January 15, 2020
### From gopher.somnolescent.net, the Gopherhole of the weird lightbulb people
A fallen behemoth that's left not a trace--unless you know where to look.

---

I've been obsessed lately. Up at 4am, scrubbing through the Wayback Machine.
Digging through tarballs. Trying to piece together the story of what once
was, and what you've never heard of--despite the fact that you probably
should have.

What if I told you there was a site that did what iTunes did before iTunes,
what Bandcamp did before Bandcamp, and what Google Play Music did before
Google was even a drip of gayonnaise in Larry Page's briefs? What if I told
you the streaming revolution happened not in the 2010s, but in 1998? And
what if I told you all that remains of this site are a few archived pages
and four screenshots on the visible internet?

Let's talk about MP3.com.

## The short of it

So even if you're a mom, you probably know what an MP3 is. It's a funny
little music format that squishes down to a relatively comfortable size by
sacrificing some of the sound quality--the bright side to that being it's the
bits you likely can't hear. (I don't think you want me to go into
psychoacoustics though >_>)

The lads behind MP3.com didn't start with the intention of making a digital
music site. In the late 90s, MP3s hadn't yet become the juggernaut that
they'd later become, but among computer kids (not the woke kind), decent
quality digital audio in 4MB (or less, if you were willing to put up with
more artifacting) was kind of a big deal in the age of 6.4GB hard drives.

The oft-repeated story is in 1997, Michael Robertson noticed some spikes in
searching for MP3s on his (now utterly obscure) FTP search engine Filez.com.
Thinking it'd drive some good traffic to Filez, he bought the MP3.com domain
for a grand off someone whose online alias was quite literally MP3.
Initially, he just hosted links to MP3 encoders and music news, but that
wasn't enough. People came to MP3.com for MP3s, surprisingly enough. He tried
to court labels into distributing their music through his site, but
trailblazers that they were, they didn't see the point.

Of course, with a thriving indie scene struggling to distribute music over
the internet of the day, Michael went to the bands instead. It worked.

At its sale to Vivendi, MP3.com was shifting four million MP3s a day to
800,000 unique users, about a gigabyte of traffic globally every second, or
about four terabytes monthly. Bands fuckin' adored the place. They could sell
off downloads, made-to-order CD-Rs of their work, and monetize their
streaming traffic in the most soothsayer way possible--stream our tracks and
we get royalties from it.

So MP3.com had a business model that iTunes and later Spotify would take to
the goddamn bank and everyone loved it. How'd it fall?

Well, the labels, of course. You might wonder how the labels had an issue
with indie bands, and the answer is, they didn't. It all came down to a funny
little thing called My.MP3.com. The service, again far too ahead of its time
for its own good, was the late 90s equivalent of Google Play Music or iTunes
Match. You'd register your CDs with it and MP3.com would let you download and
stream them anywhere. Michael's reasoning? To grow MP3.com further, they'd
have to move past simply streaming the music of nobodies and stream the music
of *somebodies*. I mean, they already had Alanis Morrissette behind them; why
not?

Labels got antsy. To them, this was dangerous, copyright-infringing territory
that'd kill their *future business prospects* in this strange new realm, and
UMG personally sat up to smash MP3.com's femurs in court. The case was later
settled and My.MP3.com discontinued, but the dot-com bubble had just popped
and the whole money thing wasn't panning out for startups like MP3.com. As a
result, they sold to Vivendi in 2001, who really couldn't do jack or shit
with the service and sold it again to CNET, who promptly shut it down.

CNET promised replacement services, but that never came to pass. It mostly
sat its fat ass on the name and assets. Trusonic, one of MP3.com's divisions,
partnered with an MP3.com competitor in GarageBand.com to allow some 1.7
million tracks (supposedly "most" of the full library, but finding specifics
is rough) to see the light of day again, but that wasn't in the cards either.
(GarageBand.com has since folded and now Apple owns the domain for their
cut-down Logic clone of the same name.)

And what became of it all? MP3.com is now a cheesy discount music-focused
Buzzfeed, the library is now tied up in acquisition after acquisition if not
outright lost, and the story has all but faded from the internet. As said,
four measly screenshots still exist of the service, aside from my own
featured here.

The same people who crow about the loss of Geocities don't even realize how
many other Geocities existed at the time, and they all died the exact same
death. Don't think your new favorite toy site is safe.

## To archive

Stymieing archival attempts was the fact that MP3.com streamed through a CGI
script and thus spiders couldn't touch any of the actual music. So ignoring
that, what *did* get saved? The Wayback Machine, of course, has a partial
archive of the pages (and a partial archive of their assets), but that's
nothing special, really. Archiveteam's working on collecting all of the
made-to-order CD-Rs, but their records of them aren't great and not every
MP3.com artist used the CD-R service.

As far as the site itself goes, though, there's a skeleton archive that at
least gives us an overview of what was available. Now, it's only artist pages
and no stylesheets or images, but that's better than nothing.

As for some of the details of the service and what artists were capable of on
it, I'm still a little shaky.

- It seems artists had the ability to customize their profile to some
degree--I've seen ones with custom backgrounds, ones with blurbs (complete
with images!), custom colors, the works.
- I imagine the made-to-order CD-Rs were of the same MP3s the artists would
upload through the service--I mean, not like lossless was much of a thing
yet. Seems like it defeats the point of a CD to have mid-fi MP3s on it, but
hey, what the fuck do I know.
- As far as streaming goes, "lo-fi" play streamed in 24kbps and "hi-fi" play
in 128kbps. Downloads were 128kbps. As for the streaming format, I'm not
sure. I know it sounds obvious, but I didn't even really know MP3 as a
streaming format compared to, say, RA or ASF. Apparently it could be used
that way though, so I'm gonna go with MP3 through a CGI script.
- Of note is that MP3.com got a redesign in ~2003 that changed the logo and
made everything a whole lot more yellow. The skeleton archive is of the
later pages.

## Closing up shop

Some people are uniquely pragmatists when it comes to archival. They see
the user-created landfills of MP3.com and any number of other sites as being,
well, landfills. "Archive enough to get the point across and move on,"
they'll probably say. Hell, some will say it's a waste of time to save this
shit at all.

When sites like MP3.com die, millions of manhours are poofed from existence.
Each track holds personal significance to someone; more importantly, they're
an artifact in time. Seeing these user-generated monoliths die time and time
again is enough to blackpill a man. Have a peek at this little eulogy for it,
from a saxophonist who used MP3.com to promote his own music. What he talks
of--the traffic, the back catalog, the statkeeping, the promotion he got
through it--it's no bullshit.

MP3.com served every genre under the sun. I don't listen to jazz. I barely
know any jazz whatsoever. Yet it served them as much as it served indie
weirdos with beat-up guitars and 4-trackers, as much as it served shitty
bedroom trance musicians--MP3.com had a scrappy little spirit, a spirit that
more than likely lives on in a Bandcamp or a ReverbNation to this day.

But there's no denying that we lost history when MP3.com shut down. Given how
tough it is to find anything, *anything* of value on the service to this day
without diving into obscure Internet Archive uploads and secondary sources
from the blogs of the day, it's safe to say we haven't taken that loss
heavily enough, nor have we learned anything from it.

Keep your backups. Save everything you'd like to keep. It might get bought by
CNET one day.