Entries from October 2009 ↓

One of the very last sessions for the GSoC Mentor Summit was about Media players. There were lead devs from Amarok and XMMS2, and it was cool to speak with them in person. One frequent issue that Amarok (I can’t remember if it was also an issue for XMMS2) was that lyric sites keep going down and changing their format, sometimes adding ads in the middle of the lyrics. Another was that Amazon no longer let’s them use the album cover art, and the substitute of last.fm has very small cover art images.

My suggestion for both, but which would need to be implemented in somewhat different ways, would be to use a variety of lyrics sites, then use text similarity matching to work out what the actual lyrics part of the page was. For images, you could use google image search, and then return the image that was most frequent, as well as having some heuristic for preference of square images. I think that, although not perfect, this would make the the system a lot more robust against further changes.

Text similarity and overlaps is well understood as a computer science problem. It’s used by the shotgun sequencing approach for DNA sequencing… as well as variety of search and indexing problems. Hopefully I’ll release a usable library for it over the summer – I’ll call it libshotgun-lyrics

One of the very last sessions for the GSoC Mentor Summit was about Media players. There were lead devs from Amarok and XMMS2, and it was cool to speak with them in person. One frequent issue that Amarok (I can't remember if it was also an issue for XMMS2) was that lyric sites keep going down and changing their format, sometimes adding ads in the middle of the lyrics. Another was that Amazon no longer let's them use the album cover art, and the substitute of last.fm has very small cover art images. My suggestion for both, but which would need to be implemented in somewhat different ways, would be to use a variety of lyrics sites, then use text similarity matching to work out what the actual lyrics part of the page was. For images, you could use google image search, and then return the image that was most frequent, as well as having some heuristic for preference of square images. I think that, although not perfect, this would make the the system a lot more robust against further changes. Text similarity and overlaps is well understood as a computer science problem. It's used by the shotgun sequencing approach for DNA sequencing... as well as variety of search and indexing problems. Hopefully I'll release a usable library for it over the summer - I'll call it libshotgun-lyrics ;-)

Someone I know is quite vehement about the obsolescence of copyright, or that it at least needs to be radically reworked to be tenable in today’s environment. The environment of (almost) zero cost duplication for many copyrighted products. When it comes down to it, writing is data, music is data, and potentially, even physical objects will easily be duplicated. I’m close to that camp, but I don’t believe all data should automatically be free.

On creating something, I think you should be able to profit from your labour, but attempting to control unofficial spread of something is usually futile [1] – the big music industry would be well advised to learn something from that, except I’m sure they’ll opt to go down kicking and screaming.

Continue reading →

Someone I know is quite vehement about the obsolescence of copyright, or that it at least needs to be radically reworked to be tenable in today's environment. The environment of (almost) zero cost duplication for many copyrighted products. When it comes down to it, writing is data, music is data, and potentially, even physical objects will easily be duplicated. I'm close to that camp, but I don't believe all data should automatically be free. On creating something, I think you should be able to profit from your labour, but attempting to control unofficial spread of something is usually futile [1] - the big music industry would be well advised to learn something from that, except I'm sure they'll opt to go down kicking and screaming. I also believe that the exclusivity to charge for content should be safeguarded. Not indefinitely, and not for the ridiculous amount of 50 years after death. More like a 10th of the that, regardless of whether the author is alive. Hang on a sec, am I suggesting only 5 years to profit from your creations before it's free for others to use? Well yes, and here's why... There's been a lot of buzz on productivity and lifestyle blogs about creating things that have long term value. The idea is, you create something and then it'll continue to bring in money long after you've finished work on it. This is what copyright is all about, you have licence to profit from your creations for the length of the copyright. Sounds fair enough right? It appealing because it suggests we can create something wonderful, and if enough people like it, then we can sit back and reap the rewards. I'm sorry, I don't buy that. If someone does that, they're no longer really contributing to society. They're acting as a sink for wealth, even if they choose to redistribute it later. Wouldn't changing this affect the creation of new media and content, especially if there's no goal of being able to sit back and wait for $'s to roll in? Well, if you look at any prolific artist... they are hardly resting after they create something, they move on to the next project. People like working on projects[2]. People motivated primarily on how to make money already and inevitably fail as artists anyway[3]. Besides, I'm not arguing that an artist can no longer profit from their creation after 5 years. I'm just saying that it's non-exclusive. Other people are free to mash-up, remix, promote and sell your content without a license. I believe in attribution though. If you use someone's creative work, it's just decency to acknowledge that. Trying to pass something off as your own is just being a dick [4]. And as a result of attribution, who's going to benefit from that? The original artist. In today's society, I feel that the main difficulty for an artist is that we're in a constant poverty of attention. Trying to grab that attention, even momentarily, is an epic challenge. If an artist allows other people to use their old work, then they are also allowing other people to promote them through attribution. When a random consumer comes across the remixed/altered version, this in turn brings the original artist in to consumer's mind. That's the real value: attention and other peoples opinion. As an example: I'm a DJ, so I worked with a lot of official remixes of original tracks. As a result of this, I actually pay quite a lot of attention to who the remixer is - because chanves are that I'd like other remixes they've done. But before I got so involved in dance music, I really didn't care who the remixer was. I'd mainly be aware of the original artist. Now, if copyright lasts for 50 years after the authors death [5], this seems just crazy. Who profits after the author's death? Well, certainly not the creator. I'm guessing it's usually the family, a trust, or a company that holds it then. And after the copyright is up and people are free to start remixing and reusing the content? Well, that's hardly beneficial for the original artist. They're dead! There's a bias towards age and experience, towards older people remaining in high power positions despite incompetence (although I guess this is true of younger people too - but they rarely get put in those positions). The problem is the expectation that people can work hard and then rest once they've made their mark. Since the people in power, and with money, want to keep that money, they can petition and shape laws to maintain their income streams. Individual artists are not the cause of this, but record companies conglomerating their artist's copyrights are. Disclaimer: IANAL and this is based on my general understand of copyright. I welcome corrections on any factually incorrect or misleading stuff. [1] Assuming you can't completely remove it, which would be analogous to making the content completely illegal and defeat the purpose of trying to sell it. If there was a stronger spatial component to the internet, and it wasn't so highly connected, then maybe you'd have chance of controlling it. However, that would also require draconian policies that wouldn't fit with the nature of the net (I'm a doctor in simulating spread dynamics, trust me!). [2] At least the people I find myself surrounded with like working on projects. For which I'm very grateful for the resulting creative energy that motivates me. [3] Unless they're con-artists. [4] I'm using 'dick' here as a highly technical and well-defined legal concept here. [5] 50 years after death is a common minimum - some countries go up to 80 years! Even though the universal agreement between countries is for 25 years.

Following on from other’s recent discussions of crime and punishment, I offer these completely unhelpful transhumanist thoughts:

• A mind from the past can become completely different from the one that committed the crime. So is it fair to punish someone in the present, when their current mind state bears as much similarity to the mind that committed the crime in the past is it does to a completely separate person?
• A body replaces most of it’s cells over the course of many years. So it’s not really someone’s body we convict, but their structure. What happens when people can upload? Supposing we can represent that structure digitally or otherwise (but in a form of easily copy-able data) what happens to the replicates of that individual? Are they convicted as well? Does it become illegal for other people to harbour that sequence of data, even if it’s in stasis and getting no processor time? (which is essentially the same as dead, but with the difference of being revivable at a moments notice)
• Continuing from the assumption that it’s the structure of a criminal we want to punish/remove from society: Since a baby is essentially derived from the fair proportion of the parent’s structure, if the parent commits a crime, then shouldn’t the child also be considered a criminal? Even though the child takes some of it’s structure from the other, hopefully non-criminal, parent, the first point seems to imply that exact similarity isn’t required.

(Note, most of these thoughts are me just musing on a theoretical level that is not at all pragmatic. I don’t actually believe children of criminals are also guilty)

Following on from other's recent discussions of crime and punishment, I offer these completely unhelpful transhumanist thoughts: A mind from the past can become completely different from the one that committed the crime. So is it fair to punish someone in the present, when their current mind state bears as much similarity to the mind that committed the crime in the past is it does to a completely separate person? A body replaces most of it's cells over the course of many years. So it's not really someone's body we convict, but their structure. What happens when people can upload? Supposing we can represent that structure digitally or otherwise (but in a form of easily copy-able data) what happens to the replicates of that individual? Are they convicted as well? Does it become illegal for other people to harbour that sequence of data, even if it's in stasis and getting no processor time? (which is essentially the same as dead, but with the difference of being revivable at a moments notice) Continuing from the assumption that it's the structure of a criminal we want to punish/remove from society: Since a baby is essentially derived from the fair proportion of the parent's structure, if the parent commits a crime, then shouldn't the child also be considered a criminal? Even though the child takes some of it's structure from the other, hopefully non-criminal, parent, the first point seems to imply that exact similarity isn't required. (Note, most of these thoughts are me just musing on a theoretical level that is not at all pragmatic. I don't actually believe children of criminals are also guilty)