Permissible Copyright for Re-typesets

[imslp]

This is a discussion moved from my talk page. Basically, the question is:

What license(s) should we allow recent "re-typesets" to be licensed under?

The question arises from the fact that non-commercial licenses prohibit commercial performances of the work, and so theoretically performances of "re-typesets" licensed under a N-C license cannot be commercial in nature. And thus arises the question of whether we should encourage such actions or not.

The assumption here is that we treat the "re-typeset" essentially as a new edition, because we cannot be sure of whether the typesetter added anything to it or not. Otherwise it would not have been copyrightable in the first place.

I can understand why someone would want to put their re-typeset under a N-C license, for example, to prevent commercial publication of the typeset. I'm pretty sure that the side effect of no commercial performance was not intentional, but would it be a legal issue that should be dealt with? Or is that part practically unenforcible as to be negligible?

[emeraldimp]

I favor more free when possible. I prefer CC-BY-SA myself (though it's arguable whether most of the typesets I've done would qualify), because it seems to be the right balance: you can do whatever you want, but if someone distributes the typeset, anyone who receives it knows where they can find the original, and what they can do with it, even if the distributer tries to restrict it (I've known some that do).

I doubt you could enforce public performance, but I'm not sure I would risk it. Maybe we need to contact CC and work for better sheet music-based licenses?

[imslp]

I favor more free when possible. I prefer CC-BY-SA myself (though it's arguable whether most of the typesets I've done would qualify), because it seems to be the right balance: you can do whatever you want, but if someone distributes the typeset, anyone who receives it knows where they can find the original, and what they can do with it, even if the distributer tries to restrict it (I've known some that do).

I have not checked the license on this one, so I may be wrong, but won't any recording of a performance of a work licensed under CC-BY-SA also be required to be licensed under CC-BY-SA or similar? How much do we want the performance to be restricted by the license of the "edition"?

I think the entire issue stems from the fact that the CC licenses were designed with literary works in mind, and not music works. I don't blame them; music is not something that is intuitive to design licenses for... especially with the composition/performance divide in western classical music.

If we want, we can write off the whole issue about performance as minor because I highly doubt that such nitpicking would actually evolve into court cases. I mean, there is neither a moral nor a monetary incentive to sue (I'm assuming here that the retypesetter means NC as in non-commercial use of the score, and not the performance), and it would be practically impossible to enforce even if the typesetter wanted to do so. However, if we can get it sorted out we may spare ourselves much trouble and misunderstanding later on...

Just suggesting alternative ideas

Edit: Second to last paragraph edited to remove misunderstanding

[Leonard Vertighel]

If we want, we can write off the whole issue as minor because I highly doubt that such nitpicking would actually evolve into court cases (I mean, the files *are* offered for free in the first place, so its not like there's a monetary incentive to sue).

If one doesn't want money for it, why restrict it to NC in the first place? And conversely, if someone does expect remuneration for commercial usage, why wouldn't that be an incentive to sue? To me, this sounds like sweeping the issue under the rug.

Could it be of any use to ask the people at Mutopia about it? They currently admit three licenses, cc-by, cc-by-sa, and pd. Maybe they have already thought about the consequences...

[imslp]

If one doesn't want money for it, why restrict it to NC in the first place? And conversely, if someone does expect remuneration for commercial usage, why wouldn't that be an incentive to sue? To me, this sounds like sweeping the issue under the rug.

Well... I was thinking more along the lines of "philosophy", as in GPL vs. BSD... and I meant "write off" as in the performance issue, and not the publication issue, which I assume is what the typesetter actually wants to prevent the commercial usage of. But me suggesting "write off" is just as a last resort, and not because I would like to write the issue off I mean, the phrase "write off" itself has negative connotations (just clearing up misunderstandings here)

Could it be of any use to ask the people at Mutopia about it? They currently admit three licenses, cc-by, cc-by-sa, and pd. Maybe they have already thought about the consequences...

Hmm, I'm not familiar with them... but it'd be a good idea if someone who knows them (or is part of the project) would ask about this

[Leonard Vertighel]

Let me add some thoughts from a more or less pragmatic point of view.

First of all, what is the reason for choosing an NC license? I suppose it's not "nobody must ever make money from it" (which would be quite stupid - paper for example has a price, so any distribution of printed copies will by necessity be commercial in nature - by the way, I quite like the new "Purchase Printed Copies" feature at Mutopia, even though it is limited to a few selected titles), but rather "if someone makes money from it, I want my share." Which is somewhat understandable, but there are also some major problems with it.

Firstly, it may be impossible to trace down the typesetter/editor of a score, if all you have is a nickname in a wiki. Thus with the passing of time, it seems likely that a significant amount of material will be effectively locked into NC-only usage. This is different from the case of contemporary composers, who frequently will be members of a copyright collecting society.

Think also about the possibility of distributing IMSLP content e.g. on DVD, either by ourselves or a third party (as has been done with the German Wikipedia). IMHO this would be a very good thing, and not only for those who happen to live in an area of this world where flat fee broadband internet access is not available. (You can rest assured that this would not detract users from IMSLP - if anything, it would make IMSLP known to a wider audience.) Tracing down and dealing with a significant number of editors individually would be impractical at the very least, making it likely that NC-licensed material would be excluded a priori from projects of this kind.

Sorry for this rather convoluted rant...

[emeraldimp]

I have not checked the license on this one, so I may be wrong, but won't any recording of a performance of a work licensed under CC-BY-SA also be required to be licensed under CC-BY-SA or similar? How much do we want the performance to be restricted by the license of the "edition"?

Yes, I think it would have to be licensed CC-BY-SA, although I suppose an 'editor' could also include a special exception for performances.

I think the entire issue stems from the fact that the CC licenses were designed with literary works in mind, and not music works. I don't blame them; music is not something that is intuitive to design licenses for... especially with the composition/performance divide in western classical music.

Yeah, we definitely need to work with CC on this...

Also, I'll contact Mutopia if you like; I'm a submitter there, as well (though far from regular).

Also, I concur with Leonard.

[imslp]

You know Leonard, I find myself agreeing pretty much completely with you. On top of that, I've also found a way (perhaps obvious, but for some reason I didn't think about it before) to make the selection of licenses clear to the submitter. I will be implementing this in the add file forms soon (hopefully today, though I have a few other things planned)

[Carolus]

I can understand why someone would want to put their re-typeset under a N-C license, for example, to prevent commercial publication of the typeset. I'm pretty sure that the side effect of no commercial performance was not intentional, but would it be a legal issue that should be dealt with? Or is that part practically unenforcible as to be negligible?

From what I've seen of the considerable number of retypesets out there at Mutopia and WIMI, most commercial publishers would not be very interested in using them. There are exceptions, of course. While a few of the retypesets are of professional level engraving quality, most are immediately recognizable as the work of amateurs and hobbyists. There are also a few publishers who don't care about such niceties and would gladly issue anything if they thought it would be profitable. I see the point, but it might not be nearly the significant issue that one might think it is at first glance.

As for performance rights in editions and re-engravings, performing-rights societies basically do not recognize them. An edition is recognized for performing-rights purposes only when it contains enough original material - like realizations - that it would qualify the editor a full life+70 term instead of the considerably more limited term reserved for urtext works. Neither ASCAP, BMI, PRS, GEMA or any of the other rights societies I am aware of collect performance royalties for urtext editions. A few of the ones in the EU might collect something for an edition which includes very extensive realizations and reconstructions on the editor's part.

[imslp]

I believe that this issue is both what Leonard and Carolus said; on one hand it is in many cases not terribly significant, however it may also be a good idea to just get this sorted out, to prevent future misunderstanding, and to clarify the permissible submissions on IMSLP.

My goal with part of this is also educational: to make people actually think about the copyright that their work is licensed under. Even though the non-commercial part may be in 99 out of 100 cases completely negligible, I think it is still a good idea to be clear not only for the 1 out of 100, but also to make people conscious of possible licensing issues of future submissions (especially the reason Leonard gave).

Another reason is to homogenize the IMSLP spectra range of permissible licenses as much as possible, and to include the possibility of commercial usage in as many of the submissions as possible. In other words, to make all the submissions as close to the public domain as possible (besides the case of new compositions, in which case it may be impossible due to external restrictions). This would have benefits in the example that Leonard gave of distributing DVDs of IMSLP archives (probably Blu-ray or HDDVD though, the archive is already more than 10GB in size ), and similar endevours.

Thirdly, I hope this restriction of licensing of re-typesets will ease the burden on copyright reviewers. I know, Carolus, that you've spent many hours on some late nights doing copyright reviews (thank you!), and I would like to make the licensing as clear and as easy to maintain as possible (by making everything as close to the public domain as possible), which I hope will ease the load on copyright review, especially should we choose to do something extra (like releasing DVDs as Leonard suggested, an idea which I'm rather partial towards).

By the way, I'm going to restrict the use of the "public domain" (and other non-permitted copyright tags) for the corresponding submissions ("typeset", "scan", etc) on the file submission page itself, so that you no longer have to go around asking everyone which license they put the file under, Carolus

[Carolus]

By the way, I'm going to restrict the use of the "public domain" (and other non-permitted copyright tags) for the corresponding submissions ("typeset", "scan", etc) on the file submission page itself, so that you no longer have to go around asking everyone which license they put the file under, Carolus

As you can imagine, I'd be delighted if you'd implement such a restriction - especially for the "typeset" submissions. The law varies considerably from place to place on the degree of protectability for retypesets, and the difference between a retypeset and a new edition is not always easily determined. In contrast, scans of both manuscripts and printed scores are likely to be public domain if those submitting files follow the simple guidelines that have been set up on IMSLP.

[imslp]

Done All the categories have the appropriate restrictions (as according to this page).

By the way, I added a new submission type, if you are wondering what "new composition" is

[Conrad]

I am not sure I follow everything said in this thread, but it seems to me that if someone, for example on mutopiaproject, re-typesets a piece of music and releases that typeset into the public domain, it should be uploaded to IMSLP as a Public Domain Typeset.

When I tried to add such a public domain typeset, the IMSLP wizard rejected it and told me to choose a different license. May I release unaltered Public Domain work under a Creative Commons BY that gives credit to someone else, when they have asked for it to be released into the public domain?

[pml]

Hi guys.

Here's something for you to consider.

I found one of my full scores had been uploaded here and copyrighted under CC by (3.0), which was the contributor's best guess at interpreting the copyright terms I'd put on the file over at the other website. The piece (Fauré Requiem) is a retypeset with (what I judge to be) a significant amount of editorial additions and other work done to it.

This was published under a personal copyright (deprecated here at IMSLP since May this year, I note) and a fairly open one at that, except on the question of performance rights, as I have purposely withheld the orchestral parts to go with the full score (there are no such limitations on the organ/vocal score).

So although this suggests that this I might have preferred a performance restriction licence (e.g. pr-by-nc-nd), in fact I have no issue whatsoever regarding performance (as without parts, performers are obliged to contact me to obtain these). What I do have an objection to, in this case, is that the default CC-by or CC-by-sa (3.0) licences allow anyone to make derivative versions (and you can probably guess where this is heading).

As IMSLP currently demands that typesets have CC by/by-sa licences, I quickly discovered CC by-nd (i.e. attribution, no derivatives) would be more appropriate, which is contrarily regarded as only applicable to new compositions.

I think strictly speaking, that since copyright arguably can be claimed in new typesets, and as it may be non-trivial to judge whether non-commercial or no-derivative licences would or would not apply, hence the page on permissible licenses should not proscribe the possibility that a submitter may wish to submit re-typesets under a more restrictive licence.

Regards, Philip

[emeraldimp]

I think the issue is more of a philosophical one than a practical one, favoring freedom (to create derivatives, in this case) over convenience (greater selection). If a piece has substantial editing, then, if the editor doesn't agree to an approved license, it shouldn't be put on IMSLP -- or removed if it is.

As to the issue of how much editing needs to be done for a piece to be copyrightable, well, that's why we require licenses on retypesets. (Also, from what I understand, the EU and other countries don't allow [don't allow? what the heck?] a copyright holder to place their work into the public domain; therefore, to be safe, we use CC-BY).