Talk:Gracenote/GPL

GPL[edit]

Thank you for stepping in. Note that I just removed a link put in by one of the more controversial editors here. It was a selective link from the wayback machine that contained incorrect text that was corrected in a later version of the page. That link is at this wayback machine link, or any later page in the wayback machine. Ti Kan never released the data under the GPL, and he asked for the page to be corrected when he noticed the error. The page was corrected in 1998 about the time of the archive link above. For proof that the database was not released under the GPL, see the oldest extant version of the archive (circa 1994): The accompanying README file and The archive itself. Note that every entry has a copyright notice at the top that says (with varying year, of course):

xmcd 1.0 CD database file

Download the archive and look for yourself if you like. Also download later ones if you like, and you'll see the same thing. This is not the GNU Public License. This is a standard copyright notice. Note that the archive contains no other copyright information. So it's pretty clearly not released under the GPL, which has a very long and distinct document that must be included with the product. Lastly, as anyone with familiarity with the GPL can tell you, you cannot license data under the GPL. It only applies to source code, so the data could not have been legally GPLed, regardless. The GNU Free Documentation License might have been applicable, but it did not exist yet. One last note. Until the day Escient acquired CDDB, the database was wholly owned by Ti Kan, as can be seen in the copyright notice. There was no official CDDB organization until minutes before CDDB was legally acquired. Also, Ti Kan did not run the website, as you can see from the original wayback machine link. It was run by and copyright Steve Scherf, and any error in the text on the page is irrelevant because only Ti Kan would have had the right to release the data under some other license than the one in the database archive package. Steve Scherf 21:00, 30 November 2006 (UTC)[reply]

I have replaced the archive.org links, deleted by Scherf, to the original cddb website, and added text to explain why the gpl notice that was on the cddb website for more than a year caused confusion for many users who assumed that the whole service was free. Scherf or other Gracenote employees, please do not selectively delete inconvenient information and make excuses about this issue regarding the gpl, since freedb has been licensed under the gpl for years as well, with no issues that you discuss. Also, please note that the link to the readme file that is posted is undated and could have been changed at any time. If you would like this included, please find one that is dated. Fatandhappy 22:34, 30 November 2006 (UTC)[reply]

As the GPL was up for quite a while, it should be included in the article. Its not like it was a week or something it was a year and previous versions of that page also include that information. Although I have deleted the sentance about people being mad about it becoming a private company due to no sources. --Simonkoldyk 22:53, 30 November 2006 (UTC)[reply]

Okay so I found this [1] dated Nov 10th 1997. I'm currently looking for other sources that can verify this. Although I think the confusion should still be written in the article; but, not quite in a way to say that is was ever licensed under the GPL. Steve and Fatandhappy, I would ask that you do not edit this part unless you discuss your changes here first.--Simonkoldyk 23:06, 30 November 2006 (UTC)[reply]

Okay I also found this [2] for year 1996 and 1997; while, it does say copyright you Steve, it includes a GPL license and discusses that in the help file. --Simonkoldyk 23:20, 30 November 2006 (UTC)[reply]

Simonkoldyk, please look at my text above. It has a link to the database itself. The "cddbd" link you refer to above is not the database. That is a link to the CDDB software, which is completely different. There is no dispute that it was GPLed, that's clear. Please look at the database link and you'll see that it's something entirely different. As for fatandhappy's claim that freedb has licensed the data under the GPL: you can attach whatever license you want to anything. You can tape a copy of the GPL to your pet fish and nobody's going to stop you. However, that does not mean you can actually GPL your fish by doing this. This is part of the reason why the GNU organization came up with their documentation license. So your argument is fallacious. In addition, if you simply go to the ftp.x.org page and scroll down you will see the timestamps next to the CDDB database and readme files. You cannot fake the timestamp. Are you accusing the people at ftp.x.org of changing things out and resetting the timestamps, just for me? Or are you accusing someone of somehow breaking in and changing dates on files? I find your attitude that we are somehow capable of things like this very very tiresome. And I'm getting tired of your apparent need to hurt Gracenote any way you can. But if you are determined to continue digging and digging for anything you can, at least cut the inuendo WRT dishonesty. I think it was you (forgive me if I'm mistaken) who questioned the validity of the legal documents on our website. That kind of appalling attitude is oh so tiresome and inappropriate. —Preceding unsigned comment added by Scherf (talk • contribs)

Thank you for that clarification, makes sense. --Simonkoldyk 00:27, 1 December 2006 (UTC)[reply]

FYI, I was just revisiting my previous paragraph, and I realize it's not necessarily clear who I was addressing. I was addressing Simonkoldyk until the "as for fatandhappy...", after which I was addressing fatandhappy. Sorry, I don't want to seem like I was addressing the wrong person's statement. That was kind of a weird and sudden transition. Steve Scherf 06:29, 1 December 2006 (UTC)[reply]

Simonkoldyk, thank you very much for helping out here; your fair-mindedness is highly appreciated. Mr. Scherf, you seem to be saying that data cannot be GPL'd, however, according to the GNU (the people who maintain the GPL), "The GNU GPL can be used for general data which is not software, as long as one can determine what the definition of "source code" refers to in the particular case." [3] Do you somehow have better information about the GPL than the GNU? No license is perfect, but the GPL can in fact be used for data. Freedb claims that the entire database is GPL'd and has done so for years without challenge. Furthermore, you seem to be implying that the GNU Free Documentation License was created as a kind of solution to a fictional deficiency in the GPL regarding data. Referring to the same GNU site, however, reveals that the FDL is a "license intended for use on copylefted free documentation. We plan to adopt it for all GNU manuals. It is also suitable for other kinds of useful works (such as textbooks and dictionaries, for instance). Its applicability is not limited to textual works ("books")."[4] Are you trying to insinuate that the database is some kind of "book" or "manual"? Finally, I find it interesting that you, on the one hand, claim that Ti Kan owned the database by virtue of his copyright notice which was hidden away in each database file, not clearly visible to anyone except a person who digs into them. Then, on the other hand you deny the applicability of the GPL for the database even though the GPL is in fact applicable to a database and was clearly posted for a long period of time on the website you created with Mr. Kan's blessing. The fact that the GPL was mentioned on original CDDB website for more than a year did in fact lead many contributors and software developers to believe that the database was intended to be open and free forever - a major reason for the controversy once the data was pulled down and locked away by your sale of the cddb. Fatandhappy 15:38, 1 December 2006 (UTC)[reply]

The GPL can be used for "genera data" if its defined as the source code the problem I'm having with that is this readme [5] says nothing about the GPL. Do you have a copy of a readme for the database that has info about the GPL? Even though xmcd is under the GPL it does not automatically make the database it uses under the GPL, it must be specifically defined by its developer. --Simonkoldyk 18:29, 1 December 2006 (UTC)[reply]

Simonkoldyk is correct, the issue is how you define the term "source code". In software development, it is not uncommon for software source files to contain data tables of all sorts. In such cases, the data in those tables can be considered part of the source code. Raw data, without any software, as with the database distribution here is quite a different thing. I am not a lawyer, and I am only repeating what at least three different lawyers have told me about how the GPL applies to data. But that's beside the point here. Ti Kan had little or no involvement in the website. That was my idea and my venture. You can see that plainly, because the home page credits the contributors and copyright holders of the website. Ti Kan is not included in that list. In fact, at first Ti resisted even having a website, but I went ahead and did it anyway. And the database was Ti's baby. He collected all of the data and made releases of it, and all entries were copyrighted by him - it was his. Until late 1998, we were simply friends working on our own parts of a project together. There was no company, no official organization, nothing, no legal entity whatsoever. He didn't even visit the website much, obviously, because he didn't notice the error in the text for a long time. So you cannot say that the data was released under the GPL, because there was never a GPL notice released with the data package or the readme file, and because it was TI's data and not mine. Nobody can just take someone else's work and change the copyright terms at will, sorry. So my website saying the wrong thing about Ti's data is totally irrelevant WRT the actual copyright status. Sure, some people got confused, no question. But so what? Is that tantamount to some sort of deception? No. So the text here is inappropriate, because it seems to imply that we somehow tried to backtrack and change history. If the error is going to be mentioned, it should be described as an error. (By the way, having a copyright notice at the top of each document in a bundle of distinct documents is a completely valid method of copyrighting something. In fact, at least in the US, there is an implied copyright for every document even if that document is not explicitly copyrighted. So including over 100,000 copyright notices in the database package wasn't even required in order for Ti to claim ownership of the data.)Steve Scherf 22:26, 1 December 2006 (UTC)[reply]

Someone correct me here, but if all creative works (which is apparently what the manually created CD tags are) are implicitly copyrighted by their authors, doesn't that mean that the copyright STILL belongs to the original authors? As far as I can tell, no one signed over a document stating that their submissions were subjected to a specific copyright license, which means that all entries are subject to standard copyright. In that case, Ti didn't own the data, nor did Scherf - the original authors did. Anyb ody willing to comment on this? RFT, Dec 1st, 15:50 PST

Hello, when putting them into the database you sign over your copyright rights. --Simonkoldyk 01:17, 3 December 2006 (UTC)[reply]

Simonkoldyk, I believe that Mr. Scherf's comments are misleading at a minimum. At the archive you linked to earlier, there is an included file called "cddb.howto.z" which contains the following text describing the database, and associated server and other related code. I made applicable portions of the text bold. Apologies in advance for the lengthy reply:

WHAT IS THE XMCD CDDB

Xmcd is a CD-audio player software package for many computer platforms (primarily UNIX) running the X11 window system. It uses the OSF/Motif toolkit for its graphical user interface.
Users of xmcd can enter the CD's artist, disc title, track titles, and extended information such as band information, song lyrics, etc., and save the information on the computer's local directories. The next time a user loads the same CD in the drive, the saved information is automatically retrieved and displayed. Further more, xmcd has a "Send" feature which allows its users to submit CD information entries to a central archive via Internet electronic mail. Over time, this archive has grown to contain a substantial collection of CD information and is continuing to grow at a rapid rate. This collection of CD entries is known as the CD database (CDDB).
The CDDB were made available via FTP on the Internet, so that users can download this archive onto their computers and use it with xmcd. We also periodically update the FTP-able CDDB such that user submissions are incorporated.
Recent versions of xmcd began supporting the concept of CDDB servers, such that users no longer need to download the entire CDDB to make full use of it. Rather, a number of CDDB server hosts have been set up on the Internet around the world, and xmcd running on a system which is connected to the Internet can connect to one or more a CDDB servers and query the CD database information.
The CDDB and servers are designed to be open, and can be used by other CD players applications or other software requiring CD information. Indeed, the concept is popular and several other CD players have already adopted it and more are modified or being developed to use it.

CDDB USE RESTRICTIONS

A lot of time and effort went into the development of xmcd and the CDDB server software. Since the xmcd package, the CDDB server software, and the CDDB information are all distributed for free, we would like to foster the concept of free software. Moreover, the public CDDB servers all run on sites that have graciously volunteered their disk space, computing and network resources, not to mention occasional maintenance and support chores, all for free.
Given this, we provide full assistance to freeware authors who desire to incoporate the use of CDDB into their software. In addition, shareware programs are also supported, because it is our understanding that shareware authors, like freeware authors, usually develop their software because they enjoy doing so, and they rarely make significant enough money from their shareware programs.
While we don't have anything against commercial software per se. we take a dim view toward having someone profitting from the use of the CDDB and servers. The reason should be obvious given the above circumstances. However, the CDDB archive itself does not carry any licensing restrictions, nor is there any connection restrictions imposed on the CDDB servers. Thus, commercial software are not expressly prohibited from using these facilities. We will provide support and assistance if the commercial software will be submitting entries to the CDDB archives. This is in the interest of ensuring that the submissions will not compromise the integrity of the CDDB, and because such software would contribute to the growth of the CDDB, With this noted exception, we do not otherwise support nor endorse the use of the CDDB archives and servers by commercial software.
The xmcd and CDDB server software are both released to the public with full source code. You may inspect the source to see how it works. but please be aware that the source code to both of these packages are released under the terms of the GNU General Public License. The full text of the GNU GPL is in the COPYING file in each of these packages. If you plan to use the CDDB and/or the CDDB servers in your software, please notify xmcd@amb.org of your intention. Also, we appreciate that you keep us posted as to your development/test progress and release schedules.
You must explicitly give credit to the authors (Ti Kan, author of xmcd and creator of the CDDB concept and maintainer of the CDDB sites; and Steve Scherf, author of the CDDB server software and co-designer of the remote CDDB interfaces) both in all documentation and when the product
is operating. The latter need not appear conspicuously, but should be viewable by the user through some obvious means.

Mr. Scherf's claim that the database is somehow copyrighted via the notices appended to the data entries and therefore somehow proprietary certainly seems confusing. This also seems somewhat disingenuous in light of the fact that the documentation above clearly states that the CDDB archive itself does not carry any licensing restriction, and that Since the xmcd package, the CDDB server software, and the CDDB information are all distributed for free, we would like to foster the concept of free software.

Further, Mr. Scherf states that the copyright notices are somehow a valid way to copyright each entry in the CDDB data itself. Each entry in the database is in fact a small text file that contains title information about CDs. According to the US Copyright Office website, Titles, names, short phrases, and slogans; familiar symbols or designs; mere variations of typographic ornamentation, lettering, or coloring; mere listings of ingredients or contents are not eligible for copyright protection. To use Mr. Scherf's earlier style, you can place a copyright on your chocolate chip cookie recipe incredients, but that does not mean it is copyrightable. So the database itself would appear to free and clear, though the GPL notices in multiple locations (documentation, website pages, etc.) would certainly seem to be a good cause for confusion in the CDDB user community when the database was removed from public access, and commercialized. Fatandhappy 03:30, 2 December 2006 (UTC)[reply]

I'm really sorry to tell you Fatandhappy; but, the above readme you presented me says that the CDDB database software is under the GPL, not the actual data inside it. --Simonkoldyk 04:08, 2 December 2006 (UTC)[reply]

Thanks for keeping up on this. I agree that the data inside is not clearly GPL'd, in fact the readme says it does not carry any license restrictions. The link to the copyright office site shows that a collection of information like the cddb data is not copyrightable, as well, so the point about whether or not the database is GPL'd or not is probably mute from a legal standpoint, anyway. The issue I was trying to focus on here was that there is lots of reasons for users and developers at the time to have been confused (and angered) by the change from a free and open model to the commercial and closed model path that Gracenote took. Fatandhappy 04:25, 2 December 2006 (UTC)[reply]

Seems Simonkoldyk gets it, so I really have no need to comment further. But I thought I would add, just for the record (as is also described in the Wired interview), that we took a dim view of commercial software using the service because most of the servers weren't ours. They were donated resources, and it would not have been ethical for anyone to use them for profit. That's part of the reason we went commercial, so we could get funding for commercial servers. Though we did let small time shareware guys use the servers. But our database use term evolved over about 6 months before going commercial, so the change was not exactly sudden.

So Scherf, there was a change in terms in the last six months leading up to the commercialization?? Oops. Glad we could clear that up. These license changes from being a freely available, no strings attached service for a large number of users and developers is one of the things that caused all the animosity towards Gracenote which has been widely documented. It is hardly an "anti-Gracenote", religious sort of thing, which you imply through out your comments. Many people were confused and upset that you guys changed your story after getting all those users to submit data to a service they thought would continue to be free. Developers were upset that they implemented a service in all kinds of software thinking it would continue to be free based on your own documentation. Then you went and changed the terms. I find it appalling that you on the one hand argue that you did not want the volunteer servers to be commercialized, but on the other hand went out to charge (what eventually turned into extremely high) license fees to all the commercial developers you and your group lured into the free CDDB service. Talk about bait-and-switch style tactics. I do not see how you can debate this further. Fatandhappy 06:59, 2 December 2006 (UTC)[reply]

There you go again... Whatever, dude. Nothing new in my previous post. Guess you didn't actually read the full Wired interview. Just for laughs, since you didn't: CDDB is still free. No user has ever paid a cent to use it, and they never will. That's my definition of free. And freeware developers can use it for... free. Hmm. Commercial developers pay based on our actual cost to serve their users, not some crazy fee. Not sure why it's okay for them to make money off CDDB while we're somehow the only ones not allowed to, even though we go through the work and expense of running it for them? (This is mainly for the edification of people browsing the comments here. I don't expect frothing-at-the-mouth types to listen.) <EOT> Steve Scherf 07:48, 2 December 2006 (UTC)[reply]

Summary[edit]

From both sides some good arguments, now may I say the changes I would like to make. I would like to change all comments about the database ever from being under the GPL removed, although I would like to add a little sentance or two talking about how there was some confusion by people that thought the database under the GPL due to an error on the website that was not corrected for about a year. --Simonkoldyk 01:17, 3 December 2006 (UTC)[reply]

Thanks Simonkoldyk. This is fine. You may or may not also want to refer to Feist Publications v. Rural Telephone Service which is an important legal precedent and related to the issue of Ti Kan's claim of copyright to the cddb (the database itself). The case established that mere collections of facts, such as the cddb are not copyrightable since they lack originality. This was one of the reasons that many developers at the time thought that the system would remain open and free: the database is not copyrightable, and the software was released under the GPL. Hence, though many users may have been confused about the data being GPL'd, more knowledgable users who happened upon Ti Kan's copyright notices would have assumed that the information was not copyrightable. Since Scherf did mention that there were many mistakes made with regards to the placement of GPL notices and such, third parties could just as easily thought that the inappropriate copyright notices were placed in the database entries by mistake as well. Fatandhappy 10:04, 3 December 2006 (UTC)[reply]

Indeed, this is a critical point; either the database entries are not copyrightable, or baring a copyright assignment by the authors of the entries the copyright resides with the original author of the entries. Only the latter appears to have any effect on using the data in a non-publicly available project, and one assumes that Gracenote has retained competent legal counsel in this regard. Dondelelcaro 11:06, 3 December 2006 (UTC)[reply]

I just read that Fatandhappy, and yes you are correct, technically the information is not copyrighted. The collection of it, the use of how it is used in the database, and how you can use in in another music player is though, and you don't have any copyright over the information due to you are not organizing it, they are. I've changed the wording on the original article. --Simonkoldyk 18:52, 3 December 2006 (UTC)[reply]

I never said there were "many mistakes", just one error made on MY website, run by ME and not Ti Kan. Simonkoldyk and Dendelelcaro seem to understand the facts (though I have not looked at the edits made yet, I assume they reflect this). Steve Scherf 19:23, 3 December 2006 (UTC)[reply]