Monday, August 24, 2009

It's Alive!!!

Update: With version 0.1.2, opening up a page is much easier, so I have edited the instructions below.

After nearly a month of (very sporadic) work on it, I have finally done the impossible: embedded a fully-featured web browser in Emacs. Behold:



I am calling my creation "Ezbl," pronounced "ease-able." It is a combination of Uzbl and Emacs. Right now, Ezbl is still in the early stages, but it can display a Uzbl window and receive commands from Emacs (over Uzbl's STDIN).

Currently, it requires a patched version of Emacs (available here) which supports embedding widgets and windows, such as Uzbl. It is still in an early state, and any of the drawing-related problems will be caused by the xwidget code. Joakim Verona is the genius behind the xwidget code, so talk to him if you have suggestions about that aspect.

Right now, I'm using the following code to start an instance of Uzbl (which can be called interactively):


(ezbl-open "www.google.com")


You will notice that it creates a buffer called "*ezbl-display-xxxx*", where "xxxx" is the pid of the Uzbl process. To browse to a different page, execute:


(ezbl-command-uri 1234 "www.yahoo.com")


Where "1234" is the pid of the Uzbl process. This will be cleaned up in the future with an "Ezbl" major mode and interactive commands.

Just recently, I've added a live(ish) updating display of the current page title and URL to the mode-line, shown here:



The interface still needs quite a bit of work, and it is not yet possible to type into text areas of the browser window. If you are interested in helping out, I have a GitHub project here to which I regularly push.

Eventually, I plan to have Emacs manage history and cookies, as well as multiple "tabs," which will be done by having one Uzbl window per buffer. It has a long way to go, and is not at all useful for real browsing at this point, so help is greatly appreciated.

Enjoy!

Sunday, August 2, 2009

The music industry (again)

My dad wrote me another interesting email on the music industry. I couldn't help but to respond.


Interesting, but he equates "large recording companies" with "the music industry". Ignores live performances, and, apparently, any direct artist to consumer sales.


Which, importantly, are the ways that the vast majority of (especially big-recording-label) artists make most of their money. Most artists would be (and, in many cases, are) happy to give their recorded music away for free in order to attract more people to their concerts.

The last (only?) large music concert I went to was a Maroon 5 (a fairly popular main-stream band signed to Universal) concert Sophomore year. The tickets were either $40 or $60, and it was held in the TD Garten (formerly the Fleet Center) in Boston, which has a capacity of just shy of 20,000 people. I don't know exactly how many people were there, but it looked mostly full, so let's say 15,000. For a 4-hour performance, a gross of $900,000 was made. I don't know exactly how much they owe the venue, the lighting crew, their agents, etc, but apparently, walking away with 50-60% of revenues is the average for large concerts.

Apparently, this is a full flop from the 70's, when people did tours to sell their albums, according to this interview, artists are much more likely to give away their recorded music to sell their tickets. He says that the difference (for the large artists, at least) between the money they make from album sales and what they make from concerts is massive, not even comparable. Compare this with an artist's share of record revenues, to see how bad of a deal the bands get when they sign up for major labels.


On the other hand, it would be terrible if it ceased to be practical to be a professional musician.


I agree, to some extent, but it might be inevitable, and wasting taxpayer money to delay the inevitable seems like a bad idea. If it happens that the Internet makes professional music economically impossible, then we have to figure out what to do next; but trying to sue an industry into profitability doesn't seem like a worthwhile use of anyone's time or money.

Now, I happen to think that professional music is not in any serious danger, and may in fact be much healthier today than it was a decade ago, but it is distributed much more evenly. It is no secret that only 1 in 10 albums produced by the (major) record companies turn a profit, so the major labels are very much dependent on the album sales of a small number of large stars. They have a lot to lose from the "long tail" industry model, since their whole business is set up to maximize their profit from the "head" of the market.

In fact, they are so adverse to competition that they resort to illegal means, such as "payola" to promote their own music at the expense of independent artists who can't afford to pay. It is a scam worthy of Wall Street and goes like this: the music producer pays music radio stations to play its songs. This wouldn't be a problem, except that radios are required to disclose when they are playing sponsored music, and music selected by the DJ for "regular airplay," is not allowed to be purchased. The loophole the industry now exploits is that they can pay a third party "independent music promotion" company to pay the radio
broadcaster to play the music company's songs.


There is a level of accomplishment that can only be achieved with the total dedication to the work that only professionals can afford. But one has to be able to make a living doing it. If everyone feels entitled to steal the professionals' output, because it is easy, then soon the entire quality of music available will collapse. That would be a tragedy.


But if an artists work can be advertised with a minimum of friction (either in the form of cost or legal blocks), then they can reach a maximum audience, then they can attract large crowds to their concerts.

Also, there is the question of holding back progress that benefits everyone for the sake of a minority to whom we have an emotional connection. The printing press effectively destroyed the notion of a scribe as a professional and socially elite role, but it clearly did a lot to increase the viability of democracy. Yet, we are left with a world with no (or only very few) professional scribes.

Also, there has been music since the beginning of time, and nothing, not even government death squads, could stop people from creating music. Nowadays, there is a huge amount of money being spent towards "music" which doesn't directly support the ability of artists to live off of playing music. A huge portion of the spending on recorded music does not go to the artist, so all of that could be eliminated while still supporting the artist's cut of the profits.


What should the penalty be for stealing music. Let's make a few assumptions that are unlikely to apply in real life, but illustrate the argument.


From here:

In a 2003 case involving State Farm Mutual Automobile Insurance Co., the Supreme Court invoked the Constitution's due process clause in ruling that punitive damages in most cases must be capped at 10 times compensatory damages.


The damages in this case were "statutory," but there are a number of people who argue the same rules should apply to statutory fines as well.

From the Ars Technica article discussing the verdict:


Tenenbaum filed a motion to dismiss the plaintiffs’ statutory damages claim on constitutional grounds, but Judge Gertner deferred ruling on the issue unless and until there was actually a damages award handed down by the jury.



We know exactly how many people downloaded each copy of each song shared by the primary thief.


Maybe. The software doesn't necessarily keep a record of this, and it could easily have been deleted sometime long before the trial started (so it wouldn't be obstruction of justice or anything). The only way (barring a wiretap or spyware) to get that information would be to look at the log from Tenenbaum's P2P program. The whole point of P2P is that no central server keeps track of (or even knows about) the transfers going on, so if Tenenbaum deleted the record, there would be nothing to go off of.


Say the criminal posts 100 songs online, a small time crook, but a nice round number. Let's say each song is stolen by 10 people. So 1000 songs stolen.


I obviously take issue with the use of the word "stolen," as nobody lost any property in transaction, but the numbers seem fine.


Now, lets say that each stolen song represents an album that was not purchased. I know, criminals will not necessarily buy things that they would be happy to steal, but again, simplifying assumptions.


Hold on. First of all, an illegally downloaded song would correspond to at most one legally downloaded song, at $1 on iTunes or Amazon. To say that a pirated song represents one album not purchased is like saying (modulo the lack of equivalence between digital copying and physical theft) "stealing one car represents the loss of 20 cars to the dealership." It makes no sense.

As far as the impact of piracy on music sales, here are two reports on the matter.

One report says:


This study reveals that for every five albums downloaded results in one less sale


Assuming the same ratio applies to songs (which may not be the case, because of the tendency of albums to contain filler songs), 1000 songs downloaded would mean 200 songs would not get bought.

According to another report:


The results suggest that, for the group of users of peer-to-
peer systems, piracy reduces the probability of buying music by 35% to 65%


So this means between 350 and 650 songs bought. Let's be generous and say 500 songs would have been purchased would it not have been for the piracy of those 10 original songs. Not an order of magnitude difference, by any means, but a change from 1,000 albums.


So 1000 album purchases forgone. Say the albums average $10 retail. So $10,000 lost to the legal music industry due to criminal activity.


No, it would be closer to $500, because the substitute for an illegally downloaded song is a legally downloaded one, which costs anywhere from $0.50 (on eMusic) to $1.29 (on iTunes). Most big-label songs can be had on Amazon for between $0.80 and $1, so let's make it nice and even and say $1 per song.


Does this mean that stealing these 100 songs should cost the thief $10,000? Of course not. That is just the lost income restitution.


So, according to the Supreme court, the maximum penalty should be $5,500; $500 for compensatory damages plus $5,000 for punitive damages.


There is the cost of investigating and prosecuting the criminal.


Yes, but that has nothing to do with the punishment awarded for each song. I don't know exactly how lawsuits work, so it may be the case that the losing party has to pay the winning party's legal fees, but again, that is a separate issue from how much is owed per song. In this case, the jury decided on $22,500 in damages per song. That is what is insane.

If the damages were determined by the cost of investigation, I could just buy a bunch of private jets, crash them into a mountain, and claim that as part of the investigation cost. If the record company is only going to get $5,500 out of a trial, then it may well not be worth it for them to go after someone so small-time.


Now who should pay for the fact that the industry has to do many investigations to get one successful case?


This is false. Of the 18,000 people targeted by the RIAA, this is only the second which has gone to trial (the first was Jammie Thomas-Rasset). All of the other ones have settled out of court due to the embarrassing brokenness of our justice system for things like this (that's another story, though). They have a history of using illegal means to attempt to find individuals they suspect of file sharing, some of whom had never used a computer.

So who should pay for the illegal investigations? The idiots who are doing these investigations.


I would double the costs, at least. That way a successful prosecution pays for a set of lesser investigations that do not go all the way (losing in court does not appear to be a possibility).


Again, this has absolutely nothing to do with how much infringement of a single song should be punished.


Unless it was a short and easy investigation and prosecution, sounds like Tenenbaum got off light.


Well, part of the problem was that the judge denied a Tenenbaum any chance to argue his case. From the Ars Technica article:


Tenenbaum's case was dismantled piece-by-piece by a series of adverse rulings over the past several months. Judge Gertner dismissed his abuse-of-process claims against the plaintiffs and the Recording Industry Association of America; excluded four of his proposed expert witnesses and limited the scope of a fifth; and, in a coup de grace delivered less than eight hours before the start of trial, barred him from arguing fair use to the jury.


It's definitely not over yet.


Here is the future of the recorind business, as discussed on the NYtimes op ed page.


This was an interesting article, though it got some facts wrong.

Swan Songs?


The speed at which this industry is coming undone is utterly breathtaking.


Only to those who haven't been paying much attention. Napster was the sign that they were on their way out. A big, centralized distribution industry is unnecessary in an age when a college kid can, in his spare time, make a global music distribution platform that millions of people find more convenient than going to a record store. Distribution used to be hard; it isn't anymore, so it's not worth money.


First, piracy punched a big hole in it.


This is not nearly as true as the music industry likes people (Congress, especially) to think. They are not above blatantly fabricating numbers to make the problem seem much bigger than it is. This includes claiming (without so much as a shred of evidence) that the loss to the US economy due to piracy is greater than the "combined 2005 gross domestic revenues of the movie, music, software, and video game industries."

These companies would love nothing more than a law to prop up their failing and obsolete business models.


Now music streaming — music available on demand over the Internet, free and legal — is poised to seal the deal.


He might be right on this. I personally buy much less music as a result of listening to Pandora (which is now ad-supported). Well, actually, I think Pandora has caused me to discover and purchase a lot of new music, but I am kind of weird as far as music listening habits go.


This is part of a much broader shift in media consumption by young people. They’re moving from an acquisition model to an access model.


Now this is something with which I can agree. A lot of the music I listen to is used as background, filler music. I want some music to play, but I'm not as picky about exactly what it is. Pandora fits this situation perfectly. I don't necessarily need to own all of the music, but I just want to listen to it once in a while. There is a different class of songs which I do want to own, and am willing to buy, but it is smaller than the set of songs which I want to hear a few times and then am done with.

I think the problem is that they are used to having a constant stream of throwaway hits; songs which everyone has to hear a few times and then grows tired of. With physical distribution, the only way (aside from the radio) to hear these songs repeatedly was to buy the full album, which may only contain 2 or 3 worthwhile songs. To the RIAA, there wasn't a difference between someone who bought an album, listened to two songs a few times, and then let it sit on a shelf and someone who bought an album and listened to every song on it for years.

With streaming entering the mix, songs that fall into the "disposable" category can be streamed, rather than purchased, and since people aren't going to revisit them anyway, the lack of "ownership" of the songs isn't that big of a concern to the people listening to them. I imagine that this effect is more pronounced among the more technically-savvy and disposable-music-inclined 14-18 demographic, which accounts for the drastic drop in their purchasing habits.

In order to get album sales, the recording industry will have to produce music which is worthy of being listened to multiple times over an extended period, not just "summer hits" that are forgotten as quickly as they rise up the charts.


Even if they choose to buy the music, the industry has handicapped its ability to capitalize on that purchase by allowing all songs to be bought individually, apart from their albums. This once seemed like a blessing. Now it looks more like a curse.

In previous forms, you had to take the bad with the good. You may have only wanted two or three songs, but you had to buy the whole 8-track, cassette or CD to get them. So in a sense, these bad songs help finance the good ones. The resulting revenue provided a cushion for the artists and record companies to take chances and make mistakes. Single song downloads helped to kill that.


I have no sympathy for this problem. The only reason they were able to get away with this is that the distribution medium forced people to take the bad with the good. In essence what they (and the columnist) is saying is that "we got used to being able to charge $20 for 2 songs, and as soon as we give people a say in the matter, they are only buying what they actually like!" Cry me a river. If you want me to buy a whole album, make a whole album worth of music I am interested in buying. Competitive markets are tough, aren't they?


A study last year conducted by members of PRS for Music, a nonprofit royalty collection agency, found that of the 13 million songs for sale online last year, 10 million never got a single buyer and 80 percent of all revenue came from about 52,000 songs. That’s less than one percent of the songs.


this is a flat-out, baldfaced lie. Take, for example, eMusic, which only carries independent music. In 2008, it sold 75% of its 5 million songs, or 3,750,000 tracks. This excludes all major label music, so the "10 million never got bought" claim is patently false.

Additionally, PRS is hardly a neutral party. It, and similar companies, have been found by the European Commission to be in violation of antitrust laws. They have a habit of suing people for listening to the radio while at work.

All in all, the music industry is facing a similar problem to the newspaper industry: their main role was a distributor, and now distribution is essentially free. As with newspapers, there are a lot of complex issues at play, an just as "news" gets incorrectly conflated with "newspapers," so does "music" with "the major record labels." Music existed before the record labels and it will exist after them, and we will all be a lot better off once these gatekeepers are gone.