IMSLP Forums

I've received a request from two IMSLP admins to state my response on some of the points that were brought up on this forum, specifically those by U-E. I have so far avoided speaking out to avoid steering the direction of discussion in a biased way, but since a few people have requested that I clarify myself, I will do so here.

First, response to a few points made by U-E:

We regret and wonder why the removal of some of our works can have such an impact on the future running of this website.

I believe I have responded to such adequately in both my response to the first cease and desist (which I have posted on this forum; please read it if you haven't yet ), and my open letter.

Essentially, I do not believe in artificial reductions in the scope of the public domain. Why should the public domain be given inferior treatment compared to copyrighted works? Yes, the interests of the copyright holders must be served, but what about the interests of the public?

If I remove any file at an insubstantial request of any person or organization, I am essentially acknowledging that what is public domain is actually copyrighted, which in my opinion does much more harm than good, since it will only invite snowballing. Pretty soon we'll have to take works down according to the copyright law in Mexico (life+100; see my reply to the first C&D for more information), and possibly much more should someone get it into their head that copyright should last indefinitely. Rather than give in and let the public domain be chipped away little by little, I would rather remove the entire site, and restart with more solid support. Anyone remember the days when copyright lasted for 14 years after first publication and renewable for another 14? Perhaps a reality check is in order?

I have not even mentioned the fact that removing the files is not all that was requested in the letter. Without noting the fact that I to this date have requested but not received a list of all works that were "infringing" U-E copyright on IMSLP (including the jurisdictions under which they are copyrighted), IP filtering is requested.

And with regards to IP filtering, U-E's retainer has agreed that IP filtering is not a waterproof method of preventing infringement, since it will just simply force people to use proxies. And so he suggested, of course, to also ban all proxies. I will not comment here on the feasibility of this course of action, and leave it as an exercise to the reader, but I hope the reader will notice the trend this is taking.

[This post is a work in progress... have to prepare for classes now ]

A few more answers to points raised by UE:

1. UE did not close down the site. UE merely requested that a limited number of works be removed. There was absolutely no reason whatsoever to take the site down. As the former CEO of a website legally distributing over 15,000 sheet music titles from over 200 suppliers, I am most perplexed as to why this decision was taken.

I've replied in the above post.

2. UE most certainly made repeated polite and direct attempts to discuss in an amiable manner the copyright infringements taking place on the site. IMSLP has deliberately decided not to show you this correspondence, in the (successful) attempt to give the impression that UE was not prepared to enter into a dialogue and to resolve the issue in a friendly manner.

Actually, I thought I was doing U-E a favor by not showing the first cease and desist request, as it acknowledged that the works are legal in Canada, among a few other things (including wanting 3,000 Euro per "infringement"). But since U-E has requested the disclosure (and a few other people too), I have put up both the first Cease and Desist request, and my response in this forum.

3. The arguments presented to us by IMSLP basically amount to a rejection of existing copyright laws in a number of countries. Discussion of international law is by all means welcome and can be very interesting, but you can’t just tell a traffic cop that you don’t like the speed restrictions on the highway because another country has different rules.

I'm not too clear what the parallel is. IMSLP is based in Canada, and do not drive cars on Austrian roads. However, whether or not you choose to import Canadian cars and drive them on Austrian roads ignoring the speed limit is your choice. Or do you mean that the manufacturing of specific types of Canadian cars should be banned altogether, because they can be imported into Austria and are capable of surpassing the speed limit there?

4. UE has no problem whatsoever with Canadian users in Canada downloading music which is public domain in Canada. We have repeatedly said this – but this information has been deliberately withheld from you.

Please reveal this information, I am interested also.

Or do you mean that the manufacturing of specific types of Canadian cars should be banned altogether, because they can be imported into Austria and are capable of surpassing the speed limit there?

I guess if you want to sell a Canadian made car in Austria, it has to meet Austrian safety standards. Even if those are stricter than the Canadian standards. That's the whole point here.

PI wrote:I guess if you want to sell a Canadian made car in Austria, it has to meet Austrian safety standards. Even if those are stricter than the Canadian standards. That's the whole point here.

Then YOU ban the importing of the car. Tell UE to ask Austrian ISPs to ban the portion of IMSLP that is copyrighted in Austria. It is NOT substantially more difficult than me implementing the feature. In fact, it is probably much simpler since they don't even need to set up an IP filter. Its not like there are no such setups for other countries.

In fact, to follow up with my previous post, there are many good things about having the filter at the ISP:

1. The ISPs actually have the manpower to do this. Its not like there aren't ISPs around the world who filter content.
2. Instead of having the nth webmaster learn the copyright laws of every country in the world, the ISP can just focus on the copyright law of one country.
3. The ISPs don't need IP filters, and so their implementation is extremely simple.
4. The ISPs are actually the ones who import the copyrighted content into the country in question.

I think it is only logical to prefer an ISP-level filter against a filter that has to be implemented at every site.

imslp wrote:I think it is only logical to prefer an ISP-level filter against a filter that has to be implemented at every site.

Surely it's much harder for ISP's to accurately filter content than for webmasters to do so? A webmaster has to look, in reasonable depth, into which laws his/her content is in danger of infringing in different countries around the world - and yes, there are dozens of such systems, and if the level of 'reasonable depth' is set high, this is very difficult for an individual webmaster (or even a small group) to do.

But an ISP, to accurately filter content, would have to look at all the websites around the world that might infringe their local copyright laws - and there are millions and millions of those. An ISP asked to filter content would be most likely to ban at the level of entire sites, which would end up causing far more loss to users.

(By the way: thank you so much, imslp, for having provided such a wonderful resource! Although I disagree with this proposed solution, I do hope that a way is found to get the library running again, and to persuade/compel commercial publishers to be more tolerant.)

Pit-Trout wrote:Surely it's much harder for ISP's to accurately filter content than for webmasters to do so? A webmaster has to look, in reasonable depth, into which laws his/her content is in danger of infringing in different countries around the world - and yes, there are dozens of such systems, and if the level of 'reasonable depth' is set high, this is very difficult for an individual webmaster (or even a small group) to do.

I think that

1. It is unreasonable to make the webmasters of all user-contribution projects to use IP filters.
2. It is humanly impossible to know the copyright law of every country in the world.

But an ISP, to accurately filter content, would have to look at all the websites around the world that might infringe their local copyright laws - and there are millions and millions of those. An ISP asked to filter content would be most likely to ban at the level of entire sites, which would end up causing far more loss to users.

No no, I'm not asking the ISP to filter content on their own. After all, the burden of the enforcement of copyright is on the copyright owner. Instead, the copyright owner provides a list of URLs to the ISP to block, and the ISP blocks them. I think that is very simple

Hello IMSLP.

Your proposal about ISP filtering seems nice but, since there are so many ISP spread all over the world, it would be hard to be applied in the short period, and a day more of closing for IMSLP is a loss for worldwide culture.

A question:

Since we're talking of Bytes, so all questioning about parallels in reality (buy something and bring it somewhere else where it is not legal, etc) can be right and can be wrong, I'd like to know one detail:

What is not legal, having the score online so that everybody can read it or having the score ready for download, so that people can print it and possibly create pirate editions?

Because, if having a score online just for checking it is legal, the pdf's of the scores in UE's list could be re-uploaded with a print locking option (I know that this feature is avalable in the acrobat pdf format).

This should give everybody the chance of checking out the book (as in a library) but without the chance to reproduce it.

Obviously, the pdf protection can be cracked, but - as every crack - this is a personal crime of the person. IMSLP does something effective to contrast the printing out of the edition.

This seems to me much more easy to be realised than ISP filtering (which should be done on all ISP all over the world. Moreover, somebody from Mexico could dial a Canadian ISP or login to a US proxy without being against the law) and local IP filtering, which as well would not work in an efficient way (for the same reasons as before)

Regards.

Dietrich

Buxtehude wrote:Your proposal about ISP filtering seems nice but, since there are so many ISP spread all over the world, it would be hard to be applied in the short period, and a day more of closing for IMSLP is a loss for worldwide culture.

Actually, I think it is even more unreasonable to make a single webmaster learn the copyright laws of the entire world. Also, in this case, UE seems to be intent only on enforcing EU copyright, in which case there really aren't that many ISPs.

Since we're talking of Bytes, so all questioning about parallels in reality (buy something and bring it somewhere else where it is not legal, etc) can be right and can be wrong, I'd like to know one detail:

What is not legal, having the score online so that everybody can read it or having the score ready for download, so that people can print it and possibly create pirate editions?

Because, if having a score online just for checking it is legal, the pdf's of the scores in UE's list could be re-uploaded with a print locking option (I know that this feature is avalable in the acrobat pdf format).

This should give everybody the chance of checking out the book (as in a library) but without the chance to reproduce it.

Obviously, the pdf protection can be cracked, but - as every crack - this is a personal crime of the person. IMSLP does something effective to contrast the printing out of the edition.

I'm very much unsure that UE will be satisfied with this action. Plus, actions such as this focus mostly on getting IMSLP up in the short term, and not the long term consequences. Rather than satisfying the desire of some IMSLP users in the short term, I believe it is much more important to plan for the long term, so that IMSLP can actually stay up.

This seems to me much more easy to be realised than ISP filtering (which should be done on all ISP all over the world. Moreover, somebody from Mexico could dial a Canadian ISP or login to a US proxy without being against the law) and local IP filtering, which as well would not work in an efficient way (for the same reasons as before)

Actually, this would be another argument for ISP filtering: the ISPs can just filter all proxies, instead of having every webmaster do so. And indeed UE have requested the filtering of proxies. Don't you think it would be much easier just to let ISPs filter them, instead of every user-contribution website on the net?

imslp, I side with your arguments against the cease-and-desist nonsense leveled against this site, but there are many problems, assumptions, and general innaccuracies in your post:

imslp wrote:In fact, to follow up with my previous post, there are many good things about having the filter at the ISP:

1. The ISPs actually have the manpower to do this. Its not like there aren't ISPs around the world who filter content.

Few ISPs filter content. In fact, I cannot think of one and I am familiar with ISP operations in general. I do not think you can name one either. Now, ISPs may do packet shaping, but that's not the same as content filtering. One reason they wouldn't want to do filtering is that filtering, at least in theory, creates liability for content. This was explained to me by a well lawyered ISP owner here in the United States. Other reasons would include general technological feasability of caching and examining markup while meeting latency expectations.

I do not understand why you would think an ISP would have spare manpower to do what it's not required or in business to do.

2. Instead of having the nth webmaster learn the copyright laws of every country in the world, the ISP can just focus on the copyright law of one country.

This would be much more easily done at the server's end. Granted, there would be ways to circumvent it, but there it is. To do this on the ISPs end would require a lot of changes for this to happen and the internet may well just suck afterwords. It would simply just be easier for them to block your IP or make a DNS exception to not route traffic to you at all rather than figure out what content of yours is illegal.

3. The ISPs don't need IP filters, and so their implementation is extremely simple.

Short of completely denying access to your site in one way or another, how do you think this would be accomplished?

4. The ISPs are actually the ones who import the copyrighted content into the country in question.

Quite simply, no. the ISP is the local gateway to the backbone of the internet, linked through one or more tier I or II providers to get there. They don't import, transmit, or cause anything to happen to the material itself. The client computer and remote server are the ones responsible for the requesting, transmission, acceptance, and interpretation of data. The ISP just passes along packets -oblivious to what's in them.

The problem for the publisher is that it has no way to identify the only one's actually breaking the law, which are the end users. That's simply too bad for them.

I think it is only logical to prefer an ISP-level filter against a filter that has to be implemented at every site.

It's logical to admit that divergent international copyright policies simply do not mesh well with the internet. So either the Internet needs to go away, international copyright needs to be rethought, or some nations and parties in those nations just need to accept that their soveriegnty ends at their border. ISPs are a poor platform to solve these sorts of issues. It would be comparable to requiring the post office to spell-check and fact-check all mail.

respectfully....

Thanks for the good work!

4. The ISPs are actually the ones who import the copyrighted content into the country in question.

Quite simply, no. the ISP is the local gateway to the backbone of the internet, linked through one or more tier I or II providers to get there. They don't import, transmit, or cause anything to happen to the material itself. The client computer and remote server are the ones responsible for the requesting, transmission, acceptance, and interpretation of data. The ISP just passes along packets -oblivious to what's in them.

I have to disagree with this. Local ISPs HAVE to import and transmit the material into the country in question. That is what they do, as ISPs.

If some data enters a country on leased lines of an international company, then that company is responsible for importing the data into the country in question. I'm not a lawyer, but that seems obvious enough to me.

The client computer may make a request to a foreign IP address, but it only sends data between itself and local routers/servers. Transmission is not facilitated in any way solely by the requesting or receiving server, but by intermediate routers along the way. I have some traceroute paths I can send you if you believe otherwise. Acceptance of a request is likewise negotiated and/or transmitted through intermediate routers/servers.

Interpretation is meaningless, since that is a purely subjective matter. If I am a research project spidering the internet for available ports, should I be responsible for copyrighted material that I never read, understand, or interpret in any manner other than to receive it?

Also, the ISP is never 'oblivious' to the data that passes through its lines. The fact that content shaping even occurs should make this obvious.

The Internet is not an international phenomenon devoid of national law. It is a collection of interconnecting telecommunications points in different countries, each of which is subject to the national laws where the data between such points can be transferred.

The problem for the publisher is that it has no way to identify the only one's actually breaking the law, which are the end users. That's simply too bad for them.

Indeed. Which is why in the U.S. the RIAA stopped going after ISPs, and for the most part started going after music sharers - i.e. started litigating against end users who downloaded copyrighted content.

I dont'think that imslp talks about filtering of the content. That is obviously impossible.

I mean:
if the composer John Brown is copyrighted by the Whoknows Editions in the Neverneverland country, if the Jown Brown compositions will be linked and hosted on a path like

imslp.org/composers/John_Brown/

all they have to do is to deny the access to this path to any user connected through the specific ISP (which is supposed to be a citizen from Neverneverland, being a foreigner present in the country or a foreigner dialing the isp from abroad will be obliged to compy to neverneverland laws).

UE then prepares the lists of the requests, specifying reason, country and expiring date.

IMSLP communicates to the ISPs the pages to be locked, country by country. Or better, prepares a list of its pages, specifying "locked in x and y and z") and adds on the composer's page a warning that "if you live in neverneverland you get a copyright violation by downloading this file".

This solution is, as imslp says, the best one to have long life for imslp, I agree with him and am available to help for anything I can do.

Regards.

Dietrich

I have to disagree with this. Local ISPs HAVE to import and transmit the material into the country in question. That is what they do, as ISPs.

But they don't cause it to be transmitted, that's the client and remote server doing that.

If some data enters a country on leased lines of an international company, then that company is responsible for importing the data into the country in question. I'm not a lawyer, but that seems obvious enough to me.

It depends on local law, but any local law saying thus is obviously aloof of how the internet works. To the ISP, they're just packets.

The client computer may make a request to a foreign IP address, but it only sends data between itself and local routers/servers. Transmission is not facilitated in any way solely by the requesting or receiving server, but by intermediate routers along the way. I have some traceroute paths I can send you if you believe otherwise. Acceptance of a request is likewise negotiated and/or transmitted through intermediate routers/servers.

I think it's important that the routers are doing nothing but following directions between the client and server in this. They are the equipment it charge. Yes, it talks to the ISP's router, and that passes on the message, but it's a little like prosecuting AT&T itself for a customer calling in a bomb threat, since afterall, it's not the culprit's voice they're hearing, but a sonic reproduction transmitted by analog carrier via the phone company.

Interpretation is meaningless, since that is a purely subjective matter. If I am a research project spidering the internet for available ports, should I be responsible for copyrighted material that I never read, understand, or interpret in any manner other than to receive it?

I don't think it's meaningless or meaningfully subjective. The ISP simply has no idea what's being transmitted. Software is necessary to interepret the result into anything resembling a copyrighted work. On the network it's just packets, bits, and routing information.

Also, the ISP is never 'oblivious' to the data that passes through its lines. The fact that content shaping even occurs should make this obvious.

You have a fundemental missunderstanding on what packet shaping versus filtering is. Content filtering requires making a judgement based on content. Packet shaping simply looks at port and protocol information and delays or throttles bandwidth based on that.

The Internet is not an international phenomenon devoid of national law. It is a collection of interconnecting telecommunications points in different countries, each of which is subject to the national laws where the data between such points can be transferred.

Well, the problem is that you can only expect one nation to enforce another's laws by consent. In that way it's very difficult to bring the whole internet under international law as it's purely cooperative. The notion that web information is subject to law at every point at which it's transfered is completely ridiculous and unenforcable. Imagine the possiblity of someone in Detroit being arrested for making a critisism of the Chinese government after his email was routed through the Asian Mainland to get to Australia!

Indeed. Which is why in the U.S. the RIAA stopped going after ISPs, and for the most part started going after music sharers - i.e. started litigating against end users who downloaded copyrighted content.

They stopped going after the ISPs because Congress correctly saw the role of the ISP as nothing but a nuetral carrier. Consider that nations that see differently are just plain wrong.

all they have to do is to deny the access to this path to any user connected through the specific ISP (which is supposed to be a citizen from Neverneverland, being a foreigner present in the country or a foreigner dialing the isp from abroad will be obliged to compy to neverneverland laws).

Technically, may be feasable. You could rely on certain methods DNS caching to do this by requiring special software on the client computers, or by using proxies for dns and such, but besides that I'm not certain how feasable it would be to read and interpret path in real time as that info is contained in packets and not involved in actual routing. In any event, it would be an administrative nightmare for the ISP to track page additions and changes. Also, why consider that they'd have to do this just for IMSLP? There could be thousands of websites required to do this for! This would be horrible.

Anyway, none of this is going to happen. The only choices that are on the table are to fight/ignore these letters or leave IMSLP's score database offline.

All in all, I see this as a momentary blip interrupting the public domain. The info will resurface shortly in one form or several.

Local ISPs HAVE to import and transmit the material into the country in question. That is what they do, as ISPs.

But they don't cause it to be transmitted, that's the client and remote server doing that.

National carriers have no obligation to data packets, regardless of origin or destination. Go ask China. I'm not saying discrimination is the right thing to do, but sovereignty of national laws means that localities have a right to mangle imported or exported data however they wish.

If some data enters a country on leased lines of an international company, then that company is responsible for importing the data into the country in question. I'm not a lawyer, but that seems obvious enough to me.

It depends on local law, but any local law saying thus is obviously aloof of how the internet works. To the ISP, they're just packets.

"How the internet works" is purely a function of how national authorities wish it to work for those in their nation.

I think it's important that the routers are doing nothing but following directions between the client and server in this. They are the equipment it charge.

Routing behavior is not guaranteed to comply only to client or server directions anywhere, even in the U.S.. See, e.g., AT&T rerouting data to secret data-collecting servers belonging to the NSA inside the U.S.

Yes, it talks to the ISP's router, and that passes on the message, but it's a little like prosecuting AT&T itself for a customer calling in a bomb threat, since afterall, it's not the culprit's voice they're hearing, but a sonic reproduction transmitted by analog carrier via the phone company.

Since you're in the U.S., call in such a threat to someplace where the U.S. doesn't have specific telecommunications treaties in place, say, Mauritius, and let me know when you've been prosecuted on U.S. soil.

Interpretation is meaningless, since that is a purely subjective matter. If I am a research project spidering the internet for available ports, should I be responsible for copyrighted material that I never read, understand, or interpret in any manner other than to receive it?

I don't think it's meaningless or meaningfully subjective. The ISP simply has no idea what's being transmitted. Software is necessary to interepret the result into anything resembling a copyrighted work. On the network it's just packets, bits, and routing information.

My point here originally was that for something to be truly meaningful, more than software is required. Wetware is needed, a conscious mind capable of discerning meaning from the multitude of data. And, again, packets, bits, and routing information are manipulable as they pass across a network.

Also, the ISP is never 'oblivious' to the data that passes through its lines. The fact that content shaping even occurs should make this obvious.

You have a fundemental missunderstanding on what packet shaping versus filtering is. Content filtering requires making a judgement based on content. Packet shaping simply looks at port and protocol information and delays or throttles bandwidth based on that.

By your own definition, packet shaping looks at port and protocol information in a packet. Thus, the ISP is not 'oblivious' to the data in the packets that passes over its lines as I asserted, no?

The Internet is not an international phenomenon devoid of national law. It is a collection of interconnecting telecommunications points in different countries, each of which is subject to the national laws where the data between such points can be transferred.

Well, the problem is that you can only expect one nation to enforce another's laws by consent. In that way it's very difficult to bring the whole internet under international law as it's purely cooperative. The notion that web information is subject to law at every point at which it's transfered is completely ridiculous and unenforcable. Imagine the possiblity of someone in Detroit being arrested for making a critisism of the Chinese government after his email was routed through the Asian Mainland to get to Australia!

And this, my friend, is entirely the point. Why should IMSLP, operating in Canada, be beholden to copyright restrictions in place in Europe? Imagine the possibility of someone in Toronto being levied a $180,000 Euro fine for delivering another Canadian citizen a music score that is clearly in the public domain in Canada, because some ISP in between happens to route part of their traffic through Ireland because it's cheaper. Is that fine justified in any way?

Indeed. Which is why in the U.S. the RIAA stopped going after ISPs, and for the most part started going after music sharers - i.e. started litigating against end users who downloaded copyrighted content.

They stopped going after the ISPs because Congress correctly saw the role of the ISP as nothing but a nuetral carrier. Consider that nations that see differently are just plain wrong.

Great. Consider that many of us here on the IMSLP forums consider UE's claims equally wrong. I assume, then, that you're on IMSLP's side, and are perhaps suggesting that Universal Edition should be going after copyright infringers in the territories covered by their copyright instead of after IMSLP?