[cc-community] Copyleft And Commerce

Greg London email at greglondon.com
Fri May 26 08:55:55 EDT 2006

> The probem with cultural works rather than software is evaluation.
> If a patch doesn't run, it isn't accepted. If a track has low
> aesthetic or technical quality, it also should not be accepted.
> We then come to how we define aesthetic quality, which is a problem,

For software, functionality is a simple, objective, make/break test.
And good functionality is the key to successful software.

I'm not convinced that aesthetics is something that works as a
good make/break test for accepting or rejecting a track.
The main reason is that I like certain types of music and
I don't like other types of music. So, if I'm the gatekeeper
deciding whether to let a track in or not, you're going to find
a site filled with tracks that fit the sound and feel of
students of Hendrix, Led Zep, Stevie Ray Vaugn, and the like.

And my wife likes to point out, she can't dance to any of that
stuff, which means she wouldn't be interested in that project at all.

This magical software I was talking about, this musical
recognition tool, is basically taking the idea of super-tagging,
super-searching, meta-data, and gearing it specifically for music.
Which I get is simply vaporware, but if a good license is one
enabler, the other is good searching tools.

So maybe the problem is that the meta-data, super-searching
idea is too diffuse for a seed project. Maybe what's needed
is a search tool specifically for music. Can software analyze
tracks and find beats, find keys, figure out if its in a
major or minor key, and all that?

With a good search tool, you could catalogue everything and
nothing has to be rejected by the system. Instead, people
look for the tracks they want. And if they want Hendrix,
they get Hendrix. And if they want Club, they get dance music.

THen you solve the storage problem by offloading tracks to
whoever is contributing them, they store them on their website.
And the central site maybe stores just the more popular ones.

The central site could then also combine the software
analysis of music with the listening patterns of its
users, to get some "taste" metrics. And you've got a
hyper-intelligent radio station that only plays the
music you want.

That's the goal anyway. Anyone who subscribes to
Rhapsody music knows that their "radio stations"
try to do something like this, but it never seems
to meet expectations for me.

for 9 bucks a month, you can listen to their entire
catalogue, so, you start out listening to what you
know. Then when you finally want something new, you
can spend hours following the "people who liked this
also liked that" links and get nothing. I am surprised
that Rhapsody users haven't figured out a way to program
their own "radio station", use rhapsody as the music
source / jukebox, and get people to listen to rhapsody
through their station and make a little bit of money
doing it.

That's the problem with music as compared to something
like wikipedia or something like GPL software.
With wikipedia and software, you can usually do a
simple text search using the basic keywords you're
looking for and you'll probably find something that
works. easy to find and read/use. Easy to find and

With music, it seems to be bloody hard to find new
music if you haven't already been introduced to it
through the radio or some other source. It's hard to
find new stuff that you like, so it's also hard to
find stuff you'd like to contribute to.

So, maybe the seed project is really just a music
search tool. Something that can listen to tracks
and get its tempo, speed, major/minor, stuff.
Maybe combined with a website so that it can track
user's preferences, and allow users to "program"
radio stations for other listeners, which then
also helps music contributers find tracks they
can use.

You would then be taking care of all the main
user types around music: creators, listeners,
and the music geeks who have a knack for finding
new music that fits their tastes. Put all three
together, and you've got a seed that is plausibly
achievable, is a pretty cool long term goal:
a site where people can contribute new music,
other people can create 'radio stations' that
reflect their tastes, and the mass of listeners
can then use all of that to hear the kinds of
music they want to hear.

I guess it goes back to "enabling".
Either you do it all by yourself,
or you figure out how to enable others so
that a million people who want the same
thing that you do will come out and build
it for you.


Barbara Bauer makes SFWA's 20 Worst Literary Agencies list

More information about the cc-community mailing list