dream system

[Curt Siffert]

Yeah, it was fun to try and pull together all the efforts at
once.

Rob has written before that he feels the moderation system needs
to be expanded, and could offer artists the ability to police
their own data a bit more. I don't feel that any of the current
efforts are counter to that direction, so I don't feel too
concerned about the moderation. Just needs a few more volunteers
to work with it.

As far as the rest of the stuff I looked at, the conclusion I've
basically come to is that there doesn't really need to be yet
ANOTHER from-scratch effort to encompass all this; we've got a
lot of resources to pull from already.

I'm really suprised that a file sharer/trader doesn't yet have
musicbrainz support built in and I'm not sure who would be the most
approachable candidate. Napster has started signaturing with
relatable, which I think is really cool, but I really doubt
they are cross-referencing the information against the musicbrainz
database. Scaling considerations aside, it would be very cool
if they did that.

The only thing I think we're truly missing is a good open-source
collaborative filtering system. I know relatable is doing
something there, but I didn't get the feeling it would be
on the same level as musicbrainz in terms of open-ness; our ability
to interact with and extend the functionality. (I know some
relatable folks are on this list; I'd love more clarification.)
I just think it would be ultra cool to have an external open-source
system out there like musicbrainz that would just be a huge recommendation
engine - in a way that would also be beneficial for new artists by
soliciting ratings for them, too.

I'm not sure of the best way to start efforts on such a thing.
The algorithms are cutting-edge A.I./CS stuff that would probably
make my brain hurt. And the datasets would of course be huge.
BUT, one thing I've found out from reviewing the different
research lists is that the folks that play with the algorithms
LOVE datasets so they can test out their own research. So even
if we didn't have the recommendation service set up, an effort
could probably draw attention to themselves in the community simply
by soliciting ratings. It might be as simple as asking Rob to put
in a "Rate this song (1-10)" doohickey in freeamp for while a song
is playing, and setting up a server to receive the ratings (identified
by audio signature).

(writings at http://www.tangrams.com/music/)

Curt

On Tue, May 01, 2001 at 02:53:31PM +1000, Adrian wrote:
> Hey Curt,
>
> I really like what you're doing. It's such a complex environment to analyse
> I know.
> I feel that the whole 'moderation' thing in Musicbrainz is unfortunate.
> There is a dream system out there, one that the open-source community could
> build where artists upload their own content and write their own metadata
> into Musicbrainz or wherever. The question is. . . . is it just a dream . .
> . . or a vision???
>
> ~Adrian
> _______________________________________________
> Musicbrainz mailing list
> ***@musicbrainz.org
> http://www.musicbrainz.org/mailman/listinfo/musicbrainz

[Jim Carrico]

>I'm really suprised that a file sharer/trader doesn't yet have
>musicbrainz support built in and I'm not sure who would be the most
>approachable candidate. Napster has started signaturing with
>relatable, which I think is really cool, but I really doubt
>they are cross-referencing the information against the musicbrainz
>database. Scaling considerations aside, it would be very cool
>if they did that.


the Espra freenet client dev team were making noises about this not so long
ago...

no time for IRC, so I don't know what's up over there these days :-(


>The only thing I think we're truly missing is a good open-source
>collaborative filtering system. I know relatable is doing
>something there, but I didn't get the feeling it would be
>on the same level as musicbrainz in terms of open-ness; our ability
>to interact with and extend the functionality. (I know some
>relatable folks are on this list; I'd love more clarification.)
>I just think it would be ultra cool to have an external open-source
>system out there like musicbrainz that would just be a huge recommendation
>engine - in a way that would also be beneficial for new artists by
>soliciting ratings for them, too.


I've been thinking about an open-source micropayment systems that could
interface with p2p clients, mp3 or divx players, web browsers, etc. - the
idea being to realize the potential, described in Scott McCloud's
're-inventing comics' [1] and elsewhere, of a direct market for created
works, that takes advantage of the efficiencies of ubiquitous networks to
eliminate unnecessary middlemen from the artist/audience equation.

Pay the artist directly, for all music, from now on. Let the artists decide
how much of this dough the label should get, and management, etc, according
to their contracts and consciences. We have an opportunity right now to
re-write the rules for the music business.

One of the "excuses" for the current top-heavy structure of the music
industry is that we need them to weed out the crap, to do all the hard work
of finding and "developing" talent and packaging it for our edification.
So clearly if we're going to have to dispense with this objection, by
developing a corresponding open-source recommendation system, in which the
fans, the listeners, are cajoled, coerced or convinced into becoming
promoters of their fave tunes.

A problem with rating systems of this type is may be their inherent
spammability - what's to stop anyone from voting over and over, for
instance. The answer may be to attach a cost - small enough not to be
"noticed" in normal use, but which would require significant resources to
abuse. One method for attaching "cost" to a network operation, as a
spam-vaccine, is to require payment in 'hash cash' or other proof-or-work
operation. But why not demand payment in *cash* cash?

It may make more sense to see the micropayment system as a by-product of a
collaborative filtering system, rather than the other way round which is
the way I've been looking at it - the main point being the synergetics of
this relationship.

In the 'penny-per-play' universal jukebox scenario, each payment
constitutes a "vote" for a particular artist and a particular track - in
the sense of 'voting with your wallet'. This seems like a promising
opportunity to bootstrap an alternative economy for artists and creators -
possibly combined with a revenue-sharing system described in Ian Clarke's
"fairshare" document. [2]


-Jim Carrico
http://www.potlatch.net

[1] http://www.scottmccloud.com/
http://www.thecomicreader.com/html/icst/icst-5/icst-5.html

[2] http://freenetproject.org/index.php?page=fairshare

[Curt Siffert]

This post might be inappropriate. Click to display it.

[Adrian Wajsbrem]

Curt:- HOW could the 'dream system' be built? Would it need funding so
people could physically get together in the one place to discuss?

[Curt Siffert]

Heh... well, if you're offering, sure. Otherwise I think it's
a combination of tracking current efforts and pestering them
with suggestions.

If I had funding I'd probably concentrate on developing something
exactly like what I described, but I honestly think it's going to
be created through a combination of open-source tools in the next
year. The one piece that I think would take the most prodding is
the open-source recommender system. I can approach some academic
folks on CF lists for ideas and interest levels. I've been swamped
for the past couple weeks but I want to continue cogitating about
what features it would need and how it could fit in with payment
mechanisms, like how we were just talking about it in regards to
potlatch.

The other key step is integrating musicbrainz into a file-sharer
app but that is inevitable given enough evangelism.

I wonder if anyone has patented "0-click payments". Amazon's
silly patent covers one click that buys/confirms all at once,
but all it does is buy. You could argue that a click that
charges money and *also* delivers media at the same time would
be a 0-click payment. This doesn't have anything to do with
the rest of this letter but it's tangential to possible payment
methods, plus I figured it'd be good to bring it up since it
could be "prior art" in a public mailing list archive for future
patent protection. :-)

Curt

On Fri, May 11, 2001 at 09:50:50PM +1000, Adrian Wajsbrem wrote:
> Curt:- HOW could the 'dream system' be built? Would it need funding so
> people could physically get together in the one place to discuss?
>
> _______________________________________________
> Musicbrainz mailing list
> ***@musicbrainz.org
> http://www.musicbrainz.org/mailman/listinfo/musicbrainz

[Jim Carrico]

Curt Siffert <***@tangrams.com> said:
>If I had funding I'd probably concentrate on developing something
>exactly like what I described, but I honestly think it's going to
>be created through a combination of open-source tools in the next
>year. The one piece that I think would take the most prodding is
>the open-source recommender system. I can approach some academic
>folks on CF lists for ideas and interest levels. I've been swamped
>for the past couple weeks but I want to continue cogitating about
>what features it would need and how it could fit in with payment
>mechanisms, like how we were just talking about it in regards to
>potlatch.



It seems like there's 4 distinct aspects to the "dream system" -
production, distribution, promotion, and payment.

The first will be handled the way it is now, with a tendency toward the
lean & mean over time. The second seems like it's pretty much happening -
gnutella has apparently picked up the slack from napster, and there's
plenty of other systems in beta. That leaves promotion and payment which I
propose to consider together. Ideally, this process should be aggregated
onto a public network and implemented as an open protocol that developers
can plug into any application.

my current thinking goes like this:

consider the initial goal to be a rating system, intended to function as a
parallel "pop chart" that is tunable to the tastes of individual
listeners/end users. (lets assume that these are pseudonymous entities to
protect individual privacy.) To achieve this, listeners are encouraged to
"vote" for what they like, either by conscious acts (eg. clicking buttons),
or by setting automated systems which are built into client/player/browser
software. To prevent spamming, these votes must be backed with a small
payment, large enough to make it difficult to "fix" the charts - let's say
1 vote costs 1 cent. To create incentives for participation, let's add the
"fairshare" concept into the mix. Half the votes=payments for a given
track accrue to the "artist"* and the other half are split proportionally
between the previous voters=payers. In other words, the raters themselves
are thereby rated based on the degree to which their ratings are confirmed
by later raters! ;-)

In its simplest form, this would provide overall aggregate ratings. In
order to make these more meaningful, and to provide something approaching
an effective collab filtering system, individual users should be able to
weight the ratings of other users based on similarities to their own
ratings. In other words, we should be able to selectively disregard the
"votes" of those whose tastes are sufficiently dis-similar from our own,
creating in the process a personalized pop chart, in which one's familiar
favorites are mixed in with unknown material that has been rated highly by
others with similar tastes.

One of my hopes for future recommendation systems is to ditch the emphasis
on genre - I like Miles Davis and the Flying Burrito Brothers but I can't
stand John Coltrane or the Eagles - genre-based ratings will never account
for this.

It would be possible to manipulate an aggregate chart by "spamming" the
system, but this would be expensive, and largely ineffective, because for
this to register on personalized charts (which are after all the whole
point of this system) a spammer (="independant promoter"?) would have to
make an equal investment in a statistically broad sampling of other popular
material. That is, to corrupt your own list an attacker would also have to
"stuff the box" in favor of your current top ten, which means they would be
*paying you* to listen to their music - fair enough.

If we can make this whole process as game-like as possible, I think it has
a chance to catch on. But how to make paying seem like a game? I would
suggest that after an individual has made a convenient number of "votes" -
lets say $10 worth - s/he needs to back these up in order to make them
"count". At minimum, the 'voter' needs to make a contribution to the
*system* of $10, and it doesn't really matter who gets it, as long as this
fact is made known in some verifiable way, to the rest of the network.
that is to say, anyone on the network who has accumulated enough "votes"
(either directly by creating popular products, or indirectly by being a
'taste-maker') can expect to receive payments more-or-less at random from
other participants in the game.

This eliminates the need for the third-party payment aggregators as
described in my earlier "potlatch" scenario, or rather, it devolves this
task onto the hoi polloi membership. Lots of crypto-conundrums here, no
doubt. The problem is not so much assuring payment (as nothing happens
until a payment is made) but in assuring that one's payments are credited
properly by the system, that is, broadcast to the network, in such a way
that one's "votes" remain pseudonymous. We need an open-source rating
system, not spyware.
(http://www.smh.com.au/news/0105/17/pageone/pageone6.html)

Such a system will have several measures of member status - one is the
overall dependability of a member, ie. their track record in making and
verifying payments. Another is the global "rank" of each member, ie. the
sum of "votes" received; and the local rank of each member from the point
of view of others, ie. the weighting factor for their personalized charts.
Finally, we need another global value, which is the "account balance" of
each member, ie. the total amount of "un-redeemed votes" in each member's
favour.

Which begs the question, account on what? Ultimately what it requires is
the mythical 'eternity service' - one humongous database, from which
nothing ever gets deleted.

I've blathered on far too long - kinda thinking out loud over here, does
any of this make any sense at all??


-Jim Carrico
http://www.potlatch.net

[Johan Pouwelse]

Hello,

The proposal for Jim Carrico about the recomendation and payment system are very
intresting. But the gap between the nice structure of the proposal and the harsh
reality of music labels, paypal payments, and Gnutella is may be too large.

How do you convince > 10.000.000 music lovers to start using your system and
take over the music industry and replace it with "potlatch_V2.cpp"?
Let's hope we can do this in the coming years!

Charging money for votes presented as solution for the spamming problem.
Instead of financial solutions there are also technical or even social and legal
solutions to that. Typing your credit card numer, e-wallet, or paypal info
before you can vote/participate in the system seems to me a showstopper for the
system.

By combining the recomendation system with the promotion you depend on the
success of both. Espra.net and FreeAmp are great verhicles for a voluntary
payment and recomendation system. Relatable.com seems to have build their
business model around what Jim calls spyware.

In the scientific literature there are several working examples for group
recomendation systems. It is difficult to make these things popular because the
success of your algorithm depends on the population of users and the population
of users depends on ...

In my opinion a recomendation system should not be transparant to the user,
it should be integrated with the playlist.
The playlist should be extended with a button that is labeled "send playlist to
server. This information can be used by the server to build the recomendations.
I fully agree that a recomendation system based on a subjective genre is a bad
thing.
A recomendation system database consists of simply two trackID and the number of
playlists both tracks appear in together. It is simple to create technical
measures to avoid spamming of this database. Don't you agree?
Let's call the number tied to the two TrackIDs the: CombineFactor.

Extracting a recomendation from the database is still difficult to obtain with
high performance as the database will be quite large, but stil realistic I hope.

The recomendation algorithms takes as an argument a likeness parameter and the
number of requested recomendations. The recomendation system database is just a
weighted mathematical graph. Where the CombineFactor is the wieght of the arc's
between nodes. Each track is a nod. With a simpel tree walker you can extend a
playlist of a user with new tracks. You assume that only similair tracks appear
in a users playlist and the higher the CombineFactor the more users find this
track also nice. A playlist is extended with track that have a high
CombineFactor with the playlist of the user. The degree of required likeness is
dependant on the likeness parameter.

Is this a realistic recomendation system?
Can relatable.com people comment on this or is their algorithm closed/patented?

Just my 5 eurocents,
Johan.

Jim Carrico wrote:
> Curt Siffert <***@tangrams.com> said:
> >If I had funding I'd probably concentrate on developing something
> >exactly like what I described, but I honestly think it's going to
> >be created through a combination of open-source tools in the next
> >year. The one piece that I think would take the most prodding is
> >the open-source recommender system. I can approach some academic
> >folks on CF lists for ideas and interest levels. I've been swamped
> >for the past couple weeks but I want to continue cogitating about
> >what features it would need and how it could fit in with payment
> >mechanisms, like how we were just talking about it in regards to
> >potlatch.
>
> It seems like there's 4 distinct aspects to the "dream system" -
> production, distribution, promotion, and payment.
>
> The first will be handled the way it is now, with a tendency toward the
> lean & mean over time. The second seems like it's pretty much happening -
> gnutella has apparently picked up the slack from napster, and there's
> plenty of other systems in beta. That leaves promotion and payment which I
> propose to consider together. Ideally, this process should be aggregated
> onto a public network and implemented as an open protocol that developers
> can plug into any application.
>
> my current thinking goes like this:
>
> consider the initial goal to be a rating system, intended to function as a
> parallel "pop chart" that is tunable to the tastes of individual
> listeners/end users. (lets assume that these are pseudonymous entities to
> protect individual privacy.) To achieve this, listeners are encouraged to
> "vote" for what they like, either by conscious acts (eg. clicking buttons),
> or by setting automated systems which are built into client/player/browser
> software. To prevent spamming, these votes must be backed with a small
> payment, large enough to make it difficult to "fix" the charts - let's say
> 1 vote costs 1 cent. To create incentives for participation, let's add the
> "fairshare" concept into the mix. Half the votes=payments for a given
> track accrue to the "artist"* and the other half are split proportionally
> between the previous voters=payers. In other words, the raters themselves
> are thereby rated based on the degree to which their ratings are confirmed
> by later raters! ;-)
>
> In its simplest form, this would provide overall aggregate ratings. In
> order to make these more meaningful, and to provide something approaching
> an effective collab filtering system, individual users should be able to
> weight the ratings of other users based on similarities to their own
> ratings. In other words, we should be able to selectively disregard the
> "votes" of those whose tastes are sufficiently dis-similar from our own,
> creating in the process a personalized pop chart, in which one's familiar
> favorites are mixed in with unknown material that has been rated highly by
> others with similar tastes.
>
> One of my hopes for future recommendation systems is to ditch the emphasis
> on genre - I like Miles Davis and the Flying Burrito Brothers but I can't
> stand John Coltrane or the Eagles - genre-based ratings will never account
> for this.
>
> It would be possible to manipulate an aggregate chart by "spamming" the
> system, but this would be expensive, and largely ineffective, because for
> this to register on personalized charts (which are after all the whole
> point of this system) a spammer (="independant promoter"?) would have to
> make an equal investment in a statistically broad sampling of other popular
> material. That is, to corrupt your own list an attacker would also have to
> "stuff the box" in favor of your current top ten, which means they would be
> *paying you* to listen to their music - fair enough.
>
> If we can make this whole process as game-like as possible, I think it has
> a chance to catch on. But how to make paying seem like a game? I would
> suggest that after an individual has made a convenient number of "votes" -
> lets say $10 worth - s/he needs to back these up in order to make them
> "count". At minimum, the 'voter' needs to make a contribution to the
> *system* of $10, and it doesn't really matter who gets it, as long as this
> fact is made known in some verifiable way, to the rest of the network.
> that is to say, anyone on the network who has accumulated enough "votes"
> (either directly by creating popular products, or indirectly by being a
> 'taste-maker') can expect to receive payments more-or-less at random from
> other participants in the game.
>
> This eliminates the need for the third-party payment aggregators as
> described in my earlier "potlatch" scenario, or rather, it devolves this
> task onto the hoi polloi membership. Lots of crypto-conundrums here, no
> doubt. The problem is not so much assuring payment (as nothing happens
> until a payment is made) but in assuring that one's payments are credited
> properly by the system, that is, broadcast to the network, in such a way
> that one's "votes" remain pseudonymous. We need an open-source rating
> system, not spyware.
> (http://www.smh.com.au/news/0105/17/pageone/pageone6.html)
>
> Such a system will have several measures of member status - one is the
> overall dependability of a member, ie. their track record in making and
> verifying payments. Another is the global "rank" of each member, ie. the
> sum of "votes" received; and the local rank of each member from the point
> of view of others, ie. the weighting factor for their personalized charts.
> Finally, we need another global value, which is the "account balance" of
> each member, ie. the total amount of "un-redeemed votes" in each member's
> favour.
>
> Which begs the question, account on what? Ultimately what it requires is
> the mythical 'eternity service' - one humongous database, from which
> nothing ever gets deleted.
>
> I've blathered on far too long - kinda thinking out loud over here, does
> any of this make any sense at all??
>
> -Jim Carrico
> http://www.potlatch.net
>
> _______________________________________________
> Musicbrainz mailing list
> ***@musicbrainz.org
> http://www.musicbrainz.org/mailman/listinfo/musicbrainz

[Curt Siffert]

It's great to get your comments, Jim.

I take "promotion" to mean the promotion of new artists/songs
through the system. It's one of the key areas I would really
like to see emphasized, and I think it's also something that
would differentiate the system from other CF algorithms. Giving
new songs/artists a chance to be heard, rated, and matched up
with listeners, especially in the CF paradigm, would be great
for music in general since it would no longer penalize artists
for having extremely original music. It's one of the things I
like best about CF; a piece of art no longer has to appeal to
the lowest common denominator to be recommended. (Insert rant
about the homogenization of mass-marketed music here.)

So I'd like to think about implementation strategy. There are
a few disparate systems here that would/could be working together.

1) The recommendation system (requires finding some A.I. experts
interested in doing open-source work). It would definitely
need to incorporate the promotion of new music somehow.

2) The tipping system. Seems the latest thinking is that "tipping"
isn't really the right metaphor for online remuneration, and
won't be until there's some external browser-based system
people can be permanently, securely logged into that allows
them to just right-click chuck in pennies on a whim. In
the meantime, a "pledge" system seems like a better metaphor.
And really, this wouldn't be all that difficult to implement.
It's just an extension to fairtunes/paytheband that instead
of billing people on every submission, just keeps track of
the pledges on a user-basis and asks them to hit the "Bill Me"
button in $10 or $20 increments.

I'd be glad to pony up the $30 to register the domain for
this sort of system if we can think of a snappy word for
it. "artpledge", "musicpledge" or something. I already
have "museworld" but it might be too vague for this purpose.
Or we could fold in the fairtunes folks and figure out how
to add that into their system.

3) The "artist investment" service - whether FairShare or the
other similar one I describe, those feel like separate
efforts. I think these things would probably require a
large artist/user community already put in place before
we could ensure people adopting it. Also, I have some
big reservations about FairShare.

4) Distribution - a file-sharing program that has hooks for these
other areas

5) Music identification - Musicbrainz with relatable.

I am intrigued about the ideas of tying in playbacks/recommendations
to payment, tying in Fairshare, etc - but I also agree with Johan
that requiring financial information to participate in a system is
a huge obstacle, especially for a service that would rely on a large
volume of users to work properly.

Jim, is part of your desire for the tie-in to give users more
motivation to tip? A technical prerequisite is definitely a good way
to do it - but I think the subject of finding more ways to movitate
users to tip/pledge would be another great discussion to brainstorm
on. I am not yet sold on the idea of tying in the two areas
(payment and recommendations) into one protocol, just because of
the fact that they don't technologically depend on each other -
I can't help thinking that they'd eventually splinter into separate
development efforts that wouldn't require each other's presence, at
which point a system with the two tied together might seem a
liability... ? My opinion on this matter isn't very strong here,
though, and I think the idea is very creative. We could definitely
tie in #1 and #2 above together as time moves on.

Curt

PS - I realize we are sort of splintering off from musicbrainz
discussion here (even though it's periferally involved) and
would be glad to offload the discussion into another list if
people think it would be appropriate. Any comments?


On Mon, May 21, 2001 at 01:55:45AM -0700, Jim Carrico wrote:
> Curt Siffert <***@tangrams.com> said:
> >If I had funding I'd probably concentrate on developing something
> >exactly like what I described, but I honestly think it's going to
> >be created through a combination of open-source tools in the next
> >year. The one piece that I think would take the most prodding is
> >the open-source recommender system. I can approach some academic
> >folks on CF lists for ideas and interest levels. I've been swamped
> >for the past couple weeks but I want to continue cogitating about
> >what features it would need and how it could fit in with payment
> >mechanisms, like how we were just talking about it in regards to
> >potlatch.
>
>
>
> It seems like there's 4 distinct aspects to the "dream system" -
> production, distribution, promotion, and payment.
>
> The first will be handled the way it is now, with a tendency toward the
> lean & mean over time. The second seems like it's pretty much happening -
> gnutella has apparently picked up the slack from napster, and there's
> plenty of other systems in beta. That leaves promotion and payment which I
> propose to consider together. Ideally, this process should be aggregated
> onto a public network and implemented as an open protocol that developers
> can plug into any application.
>
> my current thinking goes like this:
>
> consider the initial goal to be a rating system, intended to function as a
> parallel "pop chart" that is tunable to the tastes of individual
> listeners/end users. (lets assume that these are pseudonymous entities to
> protect individual privacy.) To achieve this, listeners are encouraged to
> "vote" for what they like, either by conscious acts (eg. clicking buttons),
> or by setting automated systems which are built into client/player/browser
> software. To prevent spamming, these votes must be backed with a small
> payment, large enough to make it difficult to "fix" the charts - let's say
> 1 vote costs 1 cent. To create incentives for participation, let's add the
> "fairshare" concept into the mix. Half the votes=payments for a given
> track accrue to the "artist"* and the other half are split proportionally
> between the previous voters=payers. In other words, the raters themselves
> are thereby rated based on the degree to which their ratings are confirmed
> by later raters! ;-)
>
> In its simplest form, this would provide overall aggregate ratings. In
> order to make these more meaningful, and to provide something approaching
> an effective collab filtering system, individual users should be able to
> weight the ratings of other users based on similarities to their own
> ratings. In other words, we should be able to selectively disregard the
> "votes" of those whose tastes are sufficiently dis-similar from our own,
> creating in the process a personalized pop chart, in which one's familiar
> favorites are mixed in with unknown material that has been rated highly by
> others with similar tastes.
>
> One of my hopes for future recommendation systems is to ditch the emphasis
> on genre - I like Miles Davis and the Flying Burrito Brothers but I can't
> stand John Coltrane or the Eagles - genre-based ratings will never account
> for this.
>
> It would be possible to manipulate an aggregate chart by "spamming" the
> system, but this would be expensive, and largely ineffective, because for
> this to register on personalized charts (which are after all the whole
> point of this system) a spammer (="independant promoter"?) would have to
> make an equal investment in a statistically broad sampling of other popular
> material. That is, to corrupt your own list an attacker would also have to
> "stuff the box" in favor of your current top ten, which means they would be
> *paying you* to listen to their music - fair enough.
>
> If we can make this whole process as game-like as possible, I think it has
> a chance to catch on. But how to make paying seem like a game? I would
> suggest that after an individual has made a convenient number of "votes" -
> lets say $10 worth - s/he needs to back these up in order to make them
> "count". At minimum, the 'voter' needs to make a contribution to the
> *system* of $10, and it doesn't really matter who gets it, as long as this
> fact is made known in some verifiable way, to the rest of the network.
> that is to say, anyone on the network who has accumulated enough "votes"
> (either directly by creating popular products, or indirectly by being a
> 'taste-maker') can expect to receive payments more-or-less at random from
> other participants in the game.
>
> This eliminates the need for the third-party payment aggregators as
> described in my earlier "potlatch" scenario, or rather, it devolves this
> task onto the hoi polloi membership. Lots of crypto-conundrums here, no
> doubt. The problem is not so much assuring payment (as nothing happens
> until a payment is made) but in assuring that one's payments are credited
> properly by the system, that is, broadcast to the network, in such a way
> that one's "votes" remain pseudonymous. We need an open-source rating
> system, not spyware.
> (http://www.smh.com.au/news/0105/17/pageone/pageone6.html)
>
> Such a system will have several measures of member status - one is the
> overall dependability of a member, ie. their track record in making and
> verifying payments. Another is the global "rank" of each member, ie. the
> sum of "votes" received; and the local rank of each member from the point
> of view of others, ie. the weighting factor for their personalized charts.
> Finally, we need another global value, which is the "account balance" of
> each member, ie. the total amount of "un-redeemed votes" in each member's
> favour.
>
> Which begs the question, account on what? Ultimately what it requires is
> the mythical 'eternity service' - one humongous database, from which
> nothing ever gets deleted.
>
> I've blathered on far too long - kinda thinking out loud over here, does
> any of this make any sense at all??
>
>
> -Jim Carrico
> http://www.potlatch.net
>
>
> _______________________________________________
> Musicbrainz mailing list
> ***@musicbrainz.org
> http://www.musicbrainz.org/mailman/listinfo/musicbrainz

[Devon Jones]

Curt Siffert wrote:

>
> 2) The tipping system. Seems the latest thinking is that "tipping"
> isn't really the right metaphor for online remuneration, and
> won't be until there's some external browser-based system
> people can be permanently, securely logged into that allows
> them to just right-click chuck in pennies on a whim. In
> the meantime, a "pledge" system seems like a better metaphor.
> And really, this wouldn't be all that difficult to implement.
> It's just an extension to fairtunes/paytheband that instead
> of billing people on every submission, just keeps track of
> the pledges on a user-basis and asks them to hit the "Bill Me"
> button in $10 or $20 increments.
>
> I'd be glad to pony up the $30 to register the domain for
> this sort of system if we can think of a snappy word for
> it. "artpledge", "musicpledge" or something. I already
> have "museworld" but it might be too vague for this purpose.
> Or we could fold in the fairtunes folks and figure out how
> to add that into their system.

I would be willin to offer up a domain that I purchased that is no longer of use
to me (I intended it to be a site for *ahem* commentary of metallica after the
high profile napster stuff)

I don't know if it will be of any use, but if you want it for this, it's yours:
musicfreedom.net

Devon Jones

--
-----

DVD Decoder:
usage:
cat /mnt/dvd/VOB_FILE_NAME | qrpff 153 2 8 105 225 | extract_mpeg2 | mpeg2dec -

Save following as qrpff.pl:
#!/usr/bin/perl
$_='while(read+STDIN,$_,2048){$a=29;$b=73;$c=142;$t=255;@t=map{$_%16or$t^=$c^=(
$m=(11,10,116,100,11,122,20,100)[$_/16%8])&110;$t^=(72,@z=(64,72,$a^=12*($_%16
-2?0:$m&17)),$b^=$_%64?12:0,@z)[$_%8]}(16..271);if((@a=unx"C*",$_)[20]&48){$h
=5;$_=unxb24,join"",@b=map{xB8,unxb8,chr($_^$a[--$h+84])}@ARGV;s/...$/1$&/;$
d=unxV,xb25,$_;$e=256|(ord$b[4])<<9|ord$b[3];$d=$d>>8^($f=$t&($d>>12^$d>>4^
$d^$d/8))<<17,$e=$e>>8^($t&($g=($q=$e>>14&7^$e)^$q*8^$q<<6))<<9,$_=$t[$_]^
(($h>>=8)+=$f+(~$g&$t))***@a[128..$#a]}print+x"C*",@a}';s/x/pack+/g;eval

-------------- next part --------------
A non-text attachment was scrubbed...
Name: devon.vcf
Type: text/x-vcard
Size: 245 bytes
Desc: Card for Devon Jones
Url : http://lists.musicbrainz.org/pipermail/musicbrainz-users/attachments/20010521/cc98b9e9/devon.vcf

[Jim Carrico]

Curt Siffert <***@tangrams.com> said:
>I am intrigued about the ideas of tying in playbacks/recommendations
>to payment, tying in Fairshare, etc - but I also agree with Johan
>that requiring financial information to participate in a system is
>a huge obstacle, especially for a service that would rely on a large
>volume of users to work properly.

this is very true, unfortunately. asking for money up front would be a big
mistake. i propose we let people start off with 1000 "votes", and then
after they're used up ask them for the $10 in order to make them "count" in
some significant way. Either for the CF system to kick in, or for the
chance to actually make a profit in a 'fairshare' arrangement, or whatever.



>Jim, is part of your desire for the tie-in to give users more
>motivation to tip? A technical prerequisite is definitely a good way
>to do it - but I think the subject of finding more ways to movitate
>users to tip/pledge would be another great discussion to brainstorm
>on. I am not yet sold on the idea of tying in the two areas
>(payment and recommendations) into one protocol, just because of
>the fact that they don't technologically depend on each other -
>I can't help thinking that they'd eventually splinter into separate
>development efforts that wouldn't require each other's presence, at
>which point a system with the two tied together might seem a
>liability... ? My opinion on this matter isn't very strong here,
>though, and I think the idea is very creative. We could definitely
>tie in #1 and #2 above together as time moves on.

see my comments in previous post.

by placing a payment system inside a "voting" system, we create a means to
collect money for material that isn't "licenced" for that purpose, because
for legal purposes it is simply a gift, perhaps to be distributed according
to the terms of the 'fairshare' formula, perhaps not; and for technical
purposes it is a "resistance" aimed at discouraging abuse of the system.

>Curt
>
>PS - I realize we are sort of splintering off from musicbrainz
> discussion here (even though it's periferally involved) and
> would be glad to offload the discussion into another list if
> people think it would be appropriate. Any comments?


maybe we should start a thread on infoanarchy.org and take it over there...


- Jim C.
potlatch.net

[Adrian Wajsbrem]

Curt- The last time I used freeamp, i.e. a couple of months ago, MB had
matched songs to incorrect metadata. I figure to resolve this situation
artists should insert data in MusicBrainz as soon as a track is written and
then release the track freely without copyright into an artist approved
'file-sharer'. But how is it possible to ensure artist remuneration without
depending on public generosity? I'm thinking along the lines of a) Selling
Freeamp playback data b) inserting adds into icecast streams from an artist
approved 'file-sharer'. What do you think?
~Adrian

[Adrian Wajsbrem]

Hi Curt, what open-source tools did you have in mind?

[Adrian Wajsbrem]

Hi Jim C:
production- I assume you mean open-source
distribution- I assume you mean content distribution
promotion-I assume you mean marketing the 'dream system'
payment.- I assume you mean transferring money from users to artist.

I would like to say your idea for collecting votes is good. How about having
freeamp simply track what is being played and upload the encrypted data to a
database? I feel when a song is heard in its entirety that's good enough to
be a genuine vote. The database is mined city per city and the information
is sold to local radio stations, money goes to the artists that are played
most. Seems like an accurate and marketable solution.
The Database is also mined to help with the collaborative filtering. Artists
upload content to the distribution system. But how should the distribution
system operate if it is to be built from the ground-up around artists?
Fairtunes is failing as a business model. The dream system needs a business
backing it. A bit like ASCAP who are not-for-profit.

Thanks, Adrian.

[Curt Siffert]

Tracking recommendations by listens -

I think it should be a user-configurable option. I personally
would not want to have it just work by tracking what I play.
There's a small subset of pieces in my playlists that really
define what my music tastes are. Like, I absolutely love
certain Bjork and Spearhead tracks but I'll also listen to
some Duncan Sheik and Paula Cole because, you know, they're
okay and have some nice moments. I think without having the
option to rate our tracks, the recommendations we receive won't
be all that personalized - all of us are already partially
"homogenized" by the mass marketing of music. But there's
nothing wrong with including that rating system as well as
something more specific though.

I think that gauging interest in a track by how often it is listened
to has its merits, but it's also flawed in that it it puts more
recently acquired music tracks at a disadvantage. I am just
bonkers over a recent Costello/Nieve track I got, and I like it
more than "In Your Eyes", but it's going to take a long time for it
to approach the number of listens that "In Your Eyes" has for me.

I'm hoping that the relatable/freeamp system will incorporate
actual user ratings at some point in the future.

Curt


On Mon, May 21, 2001 at 10:01:35PM +1000, Adrian Wajsbrem wrote:
> Hi Jim C:
> production- I assume you mean open-source
> distribution- I assume you mean content distribution
> promotion-I assume you mean marketing the 'dream system'
> payment.- I assume you mean transferring money from users to artist.
>
> I would like to say your idea for collecting votes is good. How about having
> freeamp simply track what is being played and upload the encrypted data to a
> database? I feel when a song is heard in its entirety that's good enough to
> be a genuine vote. The database is mined city per city and the information
> is sold to local radio stations, money goes to the artists that are played
> most. Seems like an accurate and marketable solution.
> The Database is also mined to help with the collaborative filtering. Artists
> upload content to the distribution system. But how should the distribution
> system operate if it is to be built from the ground-up around artists?
> Fairtunes is failing as a business model. The dream system needs a business
> backing it. A bit like ASCAP who are not-for-profit.
>
> Thanks, Adrian.
>
> _______________________________________________
> Musicbrainz mailing list
> ***@musicbrainz.org
> http://www.musicbrainz.org/mailman/listinfo/musicbrainz

[Jim Carrico]

This post might be inappropriate. Click to display it.

[Jim Carrico]

>Hi Jim C:
>production- I assume you mean open-source
>distribution- I assume you mean content distribution
>promotion-I assume you mean marketing the 'dream system'
>payment.- I assume you mean transferring money from users to artist.

I'm talking about the architecture of the music industry, and how it's
going to change. We're about halfway along I think. by production I mean
DIY, by promotion I mean promoting artists, by a sort of aggregated
word-of-mouth, facillitated by the 'dream system'.

otherwise you assume correctly.


>I would like to say your idea for collecting votes is good. How about having
>freeamp simply track what is being played and upload the encrypted data to a
>database? I feel when a song is heard in its entirety that's good enough to
>be a genuine vote. The database is mined city per city and the information
>is sold to local radio stations, money goes to the artists that are played
>most. Seems like an accurate and marketable solution.

hell yeah - What do you think, Rob - does that sound very hard from
freeamps pov? Is this stepping on relatable's toes?


>The Database is also mined to help with the collaborative filtering. Artists
>upload content to the distribution system. But how should the distribution
>system operate if it is to be built from the ground-up around artists?
>Fairtunes is failing as a business model. The dream system needs a business
>backing it. A bit like ASCAP who are not-for-profit.


or a lot of small businesses backing it, which would be the decentralized
approach. The big question is is how to kick start such a system, without
requiring a massive investment and all the centralization and
revenue-farming that would require.


- Jim C.
potlatch.net

[Nicolas Fischer]

Jim Carrico:
>I've blathered on far too long - kinda thinking out loud over here, does
>any of this make any sense at all??

>If we can make this whole process as game-like as possible, I think it has
>a chance to catch on. But how to make paying seem like a game? I would

We saw this couple of years ago with the stock market boom. These firms somehow made it seems like a game to invest in them and everybody started giving their money to them. So the user can place an "investment" or "bet" on an artist or album (vote/pay = one act).
If previous voters are credited, as Jim suggested, this should be enough encouragement for investing.

Maybe it isn't even necessary to split the money between artist/previous raters, because artists can invest in their own music.

Consider this scenario:
an artist creates a new album and votes 100 credits. The share price is now at 100 credits.
some trendsetter thinks it's promising sound and adds 10 votes by buying 10 shares paying 1000 credits, the value goes up to 110, the 1000 credits are distributed among the previous voters (only the artist in this case, who gets 1000 credits).
Now a user decides the sound is great and buys one share paying 110 credits. The share-value goes up to 111, the artist gets 100 credits and the trendsetter gets 10.
A next user decides the sound is great, buys a share for 111 credits which are distributed among the previos voters (100 for the artist, 10 for the trendsetter, 1 for the 1st user), share-price goes up to 112.
...

In this scenario the share price can only go up, shares cant be sold back.
There might also be a scenario modeled more closely to how real stock exchange markets work...

Is anything like this workable?
I think this approach (with share-value graphs on mb etc.) could trigger a lot of people to participate. It still suffers from the bootstrapping problem and legal issues, though.

[Adrian Wajsbrem]

>>The dream system needs a business
>>backing it. A bit like ASCAP who are not-for-profit.

>or a lot of small businesses backing it, which would be the decentralized
>approach. The big question is is how to kick start such a system, without
>requiring a massive investment and all the centralization and
>revenue-farming that would require

-There must be an umbrella association guiding each business into alignment.
- Association members are representatives of each business.
- The association helps to allocate the profits of each business back to its
other members in order to serve the artists and community.
- To get kick-started, the association needs to be formed, begin seeking
revenue from its members and budgets drawn up to help the slowest movers in
the pack.