They used to be peer-to-peer, and they'd give "free" service to employees of certain big tech companies because they could take advantage of the upload bandwidth of the big tech offices.
They culled the P2P tech in 2014 and went to a standard server-client model.
I get the feeling that a lot of p2p efforts fail because of patents.
For example I understand the reason Facetime calls don't go phone-to-phone and instead go through a server is because of a patent [troll]. I'm curious to see how WebRTC fits into that.
I think that was more a lesson from observing the problems with Skype and other video conferencing software. P2P video is great when everything is one on one, start adding in additional people and the whole experience grinds to a stuttering pixelated mess.
As someone whose PhD touched this: it is actually ok because it stabilises towards using the most available paths.
As someone who set this up at an ISP: you can QoS the hell out of P2P and keep everyone happy. That's unless you tried blocking it and now find yourself in a stupid arms war, against folks with more free time than yourself.
There only issue I can think of is upload speed, which is improving for many, and would improve more prevalently if more average users had a clear need.
The original BitTorrent wasn't aware of network topology at all, so it could randomly use a relatively expensive or congested link just as easily as a LAN connection. You could get better use of network resources overall if you usually preferred to get pieces from peers that are closer to you on the network.
I think the BitTorrent company and/or community were studying this issue at some point and so they may have come up with alternative algorithms that prefer at least AS-locality or something.
As an extreme case, imagine that there are two seeds, one being another computer on your LAN and another being a host in Timor-Leste (which at one point in the 2000s reportedly had a single T1 connection for the entire country). It will be nicer for you and for the Timorese if you can somehow arrange to simply download the file from your LAN seed.
I like the overprovisioning concept fine, but it's going to be pretty expensive to overprovision for the possibility of high-speed connections between all random pairs of hosts everywhere in the world regardless of geography. We'll probably have to persuade a lot more people to value this degree of connectivity ideologically, because demand for it isn't really getting expressed much right now.
I don't want to sound like a Bellhead here, but the packet switching infrastructure that's been built out has embodied some assumptions about its likely patterns of use. To change those patterns of use a lot (which I'd like to see too) probably needs more infrastructure. Even things like ADSL are a problem here. I agree that path asymmetry doesn't feel at all healthy for the Internet -- it feels like it's baking in producer and consumer roles -- but there are a few counterarguments that may give us some pause:
* It's never been cheaper for people who specifically want to provide services to others to rent servers in colocation facilities. It's getting cheaper every year. So, maybe people who care to do so could just rent some servers. (Of course, this might give them less autonomy because there are more people who can decide to kick them off or shut them down, and they might still not have the infrastructure to distribute media that they produce at home to the servers they rent. On the other hand, running things at home may make anonymous publishing harder because the associated broadband service is usually tied to a physical address by the ISP.)
* There's never been a huge amount of expressed demand for symmetric network connections, even when they've been available as an option. (Maybe this is different in different countries?) The economic division of network operators into "eyeball networks" (major net traffic sinks) and "provider networks" (major net traffic sources) seems to be getting sharper over time. The network operators themselves might have an economic incentive to try to get their users to equalize the traffic in both directions -- especially in terms of arrangements around settlement-free peering -- but they haven't managed to change most users' habits very much this way.
* I think of myself as a very pro-decentralization Internet user, yet whenever I run "ifconfig wlan0 | grep bytes", my download volumes consistently dramatically outpace my upload volumes, even though I have a symmetric broadband connection both at home and at work! So apparently I'm not being the change I want to see in the world.
Your first point is actually a boon for P2P. If it is cheap enough for users to rent bandwidth closer to each other, that will solve the same problem CDNs solve.
As for points two and three, decentralization does not mean 100% symmetry, just relatively more.
I have a symmetric connection, and can forward ports without any fuss, and rather enjoy those advantages. Even so, I download quite a lot more than upload. I don't expect that to change much, even with significantly more P2P traffic.
There are a few things where I am really frustrated with other's inability to forward ports, especially video games and VOIP.
Well, ISPs did not build their systems with equal upload and download in mind. There is no pure technical reason they couldn't have, but it was what they did. To replace all of that would be difficult, and the large players in the existing system will have economic motivations for fighting any change.
Not particularly. Symmetry is totally arbitrary. It's having the same amount down as before, while having more up that is the problem.
> and the large players in the existing system -will- have economic motivations for fighting any change.
That's the real problem. The solution is to give them an economic incentive to change. Average users wanting more bandwidth/symmetry is an effective way to do that.
Symmetry isn't totally arbitrary. In many deployed access technologies, scaling head end send speed is easier than scaling user end send speed; sometimes there are settings to increase upload speed at the expense of download speed (ex annex M for adsl2), but sometimes it's an inherently asymmetric system.
Well, doesn't your answer of "it's improving for many" admit to the reality that it sucks right now, answering your own question? Besides, even if we have more bandwidth, it doesn't change the fact that p2p would require more of it.
Also, if you asked the average person if they want Spotify to be peer-to-peer + use more bandwidth vs centralized + use less bandwidth, do you think they are in a position to care much beyond the bandwidth difference? Especially on mobile devices?
I certainly would choose the latter. What do I care about making Spotify's operations cheaper at my own expense?
> it doesn't change the fact that p2p would require more of it.
But P2P doesn't require more bandwidth. It requires more upload bandwidth, which is arbitrarily constrained.
> if you asked the average person if they want Spotify to be peer-to-peer + use more bandwidth vs centralized + use less bandwidth, do you think they are in a position to care much beyond the bandwidth difference? Especially on mobile devices?
Of course not. Spotify is already tailored to the current state of internet infrastructure, which heavily favors centralized content delivery.
Spotify can afford to foot the bill for their content delivery network. That is why it is no longer "better" for them to use peer to peer networking.
> What do I care about making Spotify's operations cheaper at my own expense?
For Spotify or Netflix, or any other site that already has a centralized CDN, there is no advantage.
For everything else, however, P2P is a great way to scale distribution, and the more we do to incentivize ISPs to increase upload speed, open ports, etc. the better.
It doesn't scale well. It saturates the upload bandwidth and grinds the connection to a halt. It ruins the internet for the entire family or office. Office sysadmins will block the software and start a witch hunt once they realize the problem.
Also, connecting doesn't work with NAT, carrier grade NAT and firewalls. They need a client-server model with a server that is accessible.
Just because ISPs love to skimp on upload speed does not mean that it is reasonable to do so.
Infrastructure is reactionary. If users need more upload speed, they will pay for it, and ISPs will update the infrastructure accordingly.
> Also, connecting doesn't work with NAT, carrier grade NAT and firewalls. They need a client-server model with a server that is accessible.
That's another arbitrary issue caused by overbearing ISPs. There is an awful lot that people would be able to do if they had reasonable control over their ports.
ISPs - especially the 6 massive incumbents - build the absolute minimum infrastructure they can get away with. That is simply their fiscal incentive. We shouldn't be afraid to give them new incentives.
This is a controversial stance but I too don't understand why patents are generally limited to 20 years but creative copyright is granted until 70 or more years after the death of all participating authors. If you cannot make money off of your work within 20 years, it's highly unlikely that this will change.
I don't think companies like Disney lobby to extend copyright endlessly because they can't make money in the first 20 years, but because they can continue making money after 20 years and would prefer to do that than not. Especially with things like Disney cycling availability of older films and stuff, I'm sure they make loads of money on their older material and just want to keep doing that even if it was successful when first released.
This is exactly my point. It only benefits the already big successful cooperations and people. It does nearly nothing for small artists while actively hindering people from enjoying cultural goods. Every song that is being created today will still be under copyright when you die. Meanwhile if someone creates a new way to manufacture a certain product, it's "only" protected for 20 years (which is fine).
If Taylor Swift lives until she is 80, her songs will be free in the year 2140 - maybe even later if someone like a composer lives longer. Isn't this just absurd?
When an artistic pursuit is personal, it seems like it could maybe be passed on in a will or something similar. This might be a ridiculous idea, but I guess I find complaining about someone’s artistic endeavors not being free to be absurd in it’s own way.
Copied? Snow White was a groundbreaking feat of animation brilliance, a result of tireless effort, craftsmanship and detail that experts widely agree rivals any animation project produced to date. It was literally the first feature-length color animated film with sync audio ever produced, and you’re cutting it down for it’s screenplay being inspired by a fairy tale? This is no weakness. This is consistent with filmmaking tradition and the nature of humanities efforts in general. The line is not a sharp one, but in terms of Snow White, if I understand your accusation correctly, your point is lost.
I am open minded but, with all due respect, it appears you might not be aware of some critical truths of how art gets made.
Either way, what do you suggest?
We probably generally agree on the topic at hand. I am saddened seeing many early films eroding which have yet to be digitized because they are not accessible.
>you’re cutting it down for it’s screenplay being inspired by a fairy tale
That wasn't my reading of the parent at all.
Snow White the animated film wouldn't have been made if copyright was perpetually inherited and some distant Grimm relative refused to give Disney permission to make the film. This is relevant to your original point - that "I find complaining about someone’s artistic endeavors not being free to be absurd in it’s own way." You may find it absurd, but if it were not so, a work you seem to admire might never have been produced.
There is a cultural component to art. It draws from the culture and gives back to the culture. Long copyright terms - and I find current copyright terms very long - prevent future artists from drawing on the culture to produce new, transformative works.
Disney was able to draw on cultural material from the prior 100 years to make Snow White. But if the current copyright-extension-cycle continues, it seems plausible that Disney films - and all works created in the same period - may never fall into the public domain in the USA, and future artists will not be able to enjoy the same free use of cultural source material as Disney itself has done.
Thanks for posting your response. This makes a little more sense.
Of course, my point was about distribution of original works, not inspiration or even creative appropriation. Maybe I needed to specify that? Or maybe this was well implied by the wider conversation around Spotify’s distribution of original works. Either way, it’s a separate discussion.
To be clear, Disney did not publish and distribute copies of Grimm’s Fairy Tales.
And let’s consider what Disney would have done if prevented from using the Snow White tale for his animation project? Would he have just given up?? Let’s not kid ourselves. He would have done what great artists tend to do: be creative and find another solution. There were mountains of hurdles in making this film and that would have been a mole hill. It would have been a different film, sure. It may have been better. We do not know.
The modern Disney corporation is a separate entity, to be sure.
You open up a new basket of issues here. For example, films are commonly produced from narrative works in other domains, personal accounts, etc. The royalty system allows for more than simply profit. In films, it also protects from the original authors, or storytellers and their families from what can be something like libel. It allows them to deny permission when accounts are exaggerated for audience appeal, when it may unjustly disgrace their reputation. This is only a basic example of why these things exist. It does sort of sometimes suck for artists but artists but they tend to better understand the issues at hand.
Kinda off-topic here, but as someone who co-authored a thesis on the subject, snow white was indeed ahead of it's time in terms of fluid animations and mood making, but it still featured the passive damsel-in-distress archetype which put it way behind in terms of sociological evolution.
Compare it to female counterparts in, say, Hayao Miyazaki cartoons where women are proactive and actually have character development.
> This is consistent with filmmaking tradition and the nature of humanities efforts in general.
It was consistent with filmmaking tradition when Snow White was made. It's not anymore because the previous century's worth of work is locked up behind essentially perpetual copyright.
Disney used public domain works as inspiration but then pulled up the ladder behind them so that future artists can't do the same.
> If you cannot make money off of your work within 20 years, it's highly unlikely that this will change.
Is your argument that if I can't make money off my own creative work in 20 years, anyone should have the right to give it a shot after that, even if I'm still around?
That sounds pretty shitty to me. I'm all for copyright expiring, but I'd hate to lose the rights to my own work while I'm still living.
One reason is that patents are much broader than copyrights. A copyright only prevents someone from copying your specific work. You can copyright Batman, but someone else can make Black Panther.
Patents cover independently developmed products and can cover fundamental methods or products. Inventions are often nearly developed by diffrent people. If the Wright Brothers didn't invent the airplane, someone else would have within ten years.
Someone else wouldn't have written To Kill a Mockingbird. Some other novel may have captured the same fame and critical praise, but that copyright doesn't stop that novel from exisiting.
It's a double edged sword. Less capital would flow into creating new technologies if there was no way to protect that investment. On the other hand onerous copyright protections slow the pace of innovation.
See, the "an unjust law should not be obeyed" rhetoric works better when the principle being upheld isn't "I want to be entertained for free".
I can get behind the "the information wants to be free, man" line when we're talking about overpriced textbooks and academic journals and bytecode for tractors, but when it comes to entertainment, I have a hard time caring. I don't think producers have a fundamental right to control recording, performance and distribution, as granted by the gub'mint, but I also don't think the consumers have a fundamental right to consume any cultural artifact for free.
But I do think that people should obey most laws, even ones they disagree with, until some ethereal line has been crossed, simply because it is the law, and I think "I want to be entertained for free" is way too low a bar, so I get kind of disgusted when pirates try to take a high-and-mighty stance.
There are a lot of definitions you could use: a law which denies fundamental rights; a law created by an illegitimate government; a law which makes the world worse.
But if your definition is basically "any law I personally disagree with", you're just advocating for simple anarchy.
All three of the definitions you provided will work fine.
I am denied the fundamental right of expression by copyright.
I am denied the fundamental right of security by DRM.
I am denied the fundamental right of privacy by DRM and copyright enforcement.
The United States bullies other countries like New Zealand and Sweden to enforce its copyright.
Now that Spotify exists, and can afford royalties, all innovation must be done by Spotify, and a select few other incumbents; who are all more likely to favor stability. To do as Spotify did originally (P2P file streaming) is illegal.
Section 1201 of the DMCA prevents lawful security research on anything that includes any form of DRM, making the world less secure, making actors less trustworthy, and seriously degrading privacy.
It is law that I adamantly disagree with for very real reasons. I am sharing my disagreement with anyone who will listen, and hoping for change. That is not a slippery slope to anarchy: it is democracy.
If it's "just entertainment," then why should the government (and by extension, we the people) be engaged in propping up the profitability of their industry?
If you can't find a way to make your entertainment profitable without legal protectionism, maybe you don't deserve too.
I don't really care if the government is in the business of "propping up the profitability" of the entertainment industry. If the laws were changed to reduce or eliminate copyright protections, I'd be fine with that.
I do care about rule of law. If copyright laws are the law of your land, I think that is sufficient reason to follow them, even if you are actively working to get them changed, either because you don't like them specifically or because you don't believe the government should be engaged in unnecessary activities. I don't believe that the existence of copyright law abrogates a fundamental right or denies a fundamental need, so I don't believe there is a legitimate basis for "protest" non-compliance.
It's simply nonsense to claim that sampling and mashups have been stifled by copyright law.
Ignoring a few ancient parody/satire acts like Negativland, samples are made, used, cleared, and paid for with no serious negative consequences. Likewise remixes.
There's some extra paperwork, money changes hands, and on the whole everyone is happy.
What's been far more damaging is Spotify's insanely exploitative model, which sees its officers making millions while the artists who provide the content get pocket change - if that.
> Spotify's insanely exploitative model, which sees its officers making millions while the artists who provide the content get pocket change - if that.
That's not just Spotify's model; it's music labels' model.
Spotify would gladly pay the artists all the money and it could. And it can’t. Precisely because of copyright laws: you end up paying the labels, not the artists.
Independent artists with whole or majority ownership exist and do well on Spotify.
For example, a signed artist would have to stream 1,117,021 songs to make $1,260 in a month (what the attached infographic says is minimum wage for a month). However, an independent artist who owns their work only needs to stream 180K to make the same amount [1]. This data is in 2015, but is the best data I've seen in quite a while. FWIW, they would have had to sell ~100 (~450 for signed artists) albums in a month to make this same money vs. streaming.
At this point, affiliating with a label is only worth it when you know you're going to break out.
Most musicians, even niche ones, have learned how to use the system their confined in, and do well off of a royalty-based model vs the traditional selling-CDs-at-a-show [2].
FWIW selling 100 albums/month, if you’re touring, was far far incredibly far more feasible than getting 180,000 streams is now.
Also it seems a bit unclear whether Spotify’s payouts are as egalitarian as they claim. It may be the case that a bunch of cash went towards getting existing catalogues signed up.
I’m all for Spotify btw, I just think it may not be as idealized as we like, and has some structural problems that make it infeasible for certain types of artists to make any sort of living.
I’m not advocating for touring being better. I hate touring. But the streaming amounts we’re discussing are really tiny in the end. The guitar player you linked to is making less than a tenth of what someone should hope to make to even think about calling their music income a success.
> The guitar player you linked to is making less than a tenth of what someone should hope to make to even think about calling their music income a success.
I'd surely hope $7000/month is considered a success. A passive $700/month from streams would be considered a success around here.
This is the assumption the stream counts continue, of course.
Copyright is exactly how a a musical artist is able to turn their creative output into an asset on which they can make a living from. The copyright is the very thing that allows an artist to sell and license their recorded work. An artists record company buys tor licenses he copyright to the master recording from them.
There will always be weasels and bottom feeders that seek to exploit but do you really believe copyright has somehow held back musical innovation? It was this copyright that enabled a business model that brought recorded music to the masses.
Copyright has been abused to the point where it is more of a detriment than it is worth.
There will always be weasels and bottom feeders that seek to constrain every other individual in the world from extending an artistic work or idea, even generations after his or her death; but do you really believe copyright/patent law has somehow pushed forward innovation?
>"Copyright has been abused to the point where it is more of a detriment than it is worth."
Do you really believe that an artist who makes good money licensing their music for use in TV or movies views copyright as "more of a detriment than its worth"? The only reason that revenue stream is possible is because of copyright and the sync royalties paid out on them.
>"...but do you really believe copyright/patent law has somehow pushed forward innovation?"
That's not the point I was making, my point was that it hasn't prevented musical innovation creatively. Not every instance of copyright usage is automatically detrimental to society.
Quite a few people make money with copyleft works. Even so, it is difficult, because strongly enforced copyright, and propaganda have created a culture that tries desperately to require copyright.
> Not every instance of copyright usage is automatically detrimental to society.
Of course not! But many are:
I can't watch Netflix in 1080p using Linux because of DRM.
Security research is illegal because of DRM.
I can't start a P2P music streaming service like Spotify did, because I don't have the capitol to pay for licenses. Any innovation in this sector must be done by incumbents; who have already found success, and are more content with stability.
I can't upload a video with fair use copyrighted content to Youtube, because Google has been bullied to/shows off their "capability" to weed out copyrighted content. That also sets an unfeasible precedent for competitors; who will never have the resources to do anything similar.
s/I can't/Practically no one in the world is allowed to/
Something need not be 100% detrimental to be more detrimental than it is worth. Copyright is a clear example of that. It began somewhere near middle ground, but has since drifted extremely to the benefit of media corporations, and the detriment of everyone else.
>"Enforcement is not a requirement to possibility."
That statement is gibberish. Today an artists are able to get paid because their creative output becomes an asset by virtue of a copyright. An artists can and many do start their own publishing and record companies are fully in control of their copyright regime.
>"I can't watch Netflix in 1080p using Linux because of DRM."
DRM and copyright are orthogonal.
>"I can't start a P2P music streaming service like Spotify did, because I don't have the capitol to pay for licenses."
No you can go to Media Net and get your streaming service up and running pretty quickly for as little as $1K a month only increase royalties as your service grows. See:
It's not clear to me that streaming music has actually resulted in better music or a better listening experience. It's just resulted in more plays. This benefits Spotify and advertisers the most.
Music streaming has resulted in a better discovery experience, which is valuable to users and creators.
There is an awful lot of innovation being held back. You are likely unaware, since it is not you who wants to innovate, and you aren't experiencing those innovations.
Just not sure that’s really the case. Discovery hasn’t changed much since Pandora or Last.fm.
Vinyl has made a massive comeback. Probably growing at a rate faster than Spotify at this point.
I’m really not sure what amazing “innovation” is being held back. The vast majority of social music applications just create “value” by taking revenue from someone else, not by creating new value of any kind. Not really what I think of when I think innovation.
The CD was innovative. The iPod was innovative. Spotify was just the last man standing in a long line of music streaming applications that came after music piracy hit its peak. I never found it unique or innovative. They all offered the same damn thing. Most of Spotify’s growth came from spam on the Facebook newsfeed when that was still allowed.
The whole social <whatever> app is a bit of a farce. It’s mostly based on quasi-legal theft. If that’s what counts for innovation these days then I don’t want to innovate anymore.
Though that was from a beta and in an unsure time of how the service would turn out.
Just because a service starts by doing dubious things, does not mean it will continue doing so.
I got in post-beta, but invite-only free version in 2008, and I remember it having a lot of albums where names of which were still scene releases because (I presume) the parsing was off. I also remember specific releases having audio compression artefacts matching what I streamed from Spotify.
They replaced everything in phases, because I remember all the URI:s to songs kept breaking in my playlists before they implemented auto-matching of your songs to new location once a specific album was rendered unavailable.
It is my belief that their library was in some portion built by exflitrating data from user's computers.
I have an enormous collection of lossless audio and noticed it using lots of bandwidth. The desktop versions insist on loading when the machine boots, and use subsantial amounts of bandwidth/memory at idle.
I'm pretty sure that's just the absurd inefficiency of today's trendy software.
I've also long shunned the spotify "deskop apps" and stuck to the website because firefox can contain the damage a bit, and ublock origin can make a surprising difference in the cpu usage of spotify! The ads are crazy javascript sometimes, and I'm guessing the same crap is loaded into the desktop electron apps.
