Rosetta Code talk:Copyrights

From Rosetta Code
Jump to: navigation, search

Perhaps you should consider a license for this project such as Creative Commons or GPL3.

Just a thought. SiliconJesus

Yeah. That'll be one of the things to do before we get out of beta. --Short Circuit 00:22, 9 January 2007 (PST)

Don't you think this page should be protected too? --Mwn3d 10:48, 19 December 2007 (MST)

It is...This talk page isn't, though.

Should there be a "not suitable for any purpose; use at own risk" clause, considering the wide range of programming abilities represented here? There is not even any guarantee that examples here been compiled, run or tested. --IanOsgood 14:59, 8 May 2009 (UTC)

Contents

[edit] Lincenses more permissive...?

In the section Contributors it is stated You own license to your changes; then it is said If you wish to license your contributions under terms more permissive than the GNU FDL... at FSF they suggest not to use the GNU FDL for code (as noted also here, GNU FDL and GPL are not really compatible); then in the Citations it is written to cite the source in case the original is licensed even under any copyleft license, e.g. GPLv2/3. I interpret this with the fact that RosettaCode can contain e.g. GNU GPLed code, taken from "outside", if cited... But... can RC contains GNU GPLed code "from inside" without citation?!

In other words, can I state in my user page that all my contributions can be considered (for the code part) released under e.g. GPLv3, or to be able to do so, I must before upload the code on my site, and then cite myself from the page where I used that code? --ShinTakezou 17:25, 8 May 2009 (UTC)

I don't think that GPL v3 qualifies as "more permissive" than the GFDL, in part due to the patent-related clauses, but I suppose I can re-word it as "dual-license", and strike the "less permissive" mention.
And, yes, GFDL is inappropriate for source code. I very much wish I'd used a Creative Commons license for Rosetta Code when I created it, rather than GFDL, but I wasn't familiar with the incompatibilities of the GFDL and most uses of source code. I don't think I can switch the license now, unless I take advantage of the convert-to-CC window GNU added recently. --Short Circuit 22:11, 8 May 2009 (UTC)
You can't, because the licensing condition is explicitly GFDL 1.2 (without the "or, at your option, a later version" clause). Since the relicensing window only applies to 1.3, AFAIU relicensing the site is not possible (unless each and every author explicitly agrees to the relicensing, of course).
Also note that the GFDL link at the end of the page is now wrong (it points to the current 1.3 license instead of the older 1.2 one). GFDL 1.2 is now at http://www.gnu.org/licenses/old-licenses/fdl-1.2.html --Ce 14:55, 10 May 2009 (UTC)

[edit] How to switch

Why not switch over to a better license incrementally? Create categories for 'GFDL 1.2' and 'whatever new license', dump everything current into the old-license category, encourage new pages to go into the new license, then review old pages to contact the authors. --Kevin Reid 18:15, 10 May 2009 (UTC)

The question then becomes, to which license? I dislike GPL 3, as I find its stance on patents to be counterproductive. GPL2 is fine, but I tend to prefer attribution licenses (i.e. CC-Attrib, without the SA or NC sections) or even public-domain. (Though I believe not all countries have a concept of PD.) And, ultimately, I don't have the resources to enforce the licenses, and I don't like the idea of handing over that responsibility to the FSF or similar, as then it becomes limited by what suits their aims.
But then, while this is my site, and I have the ultimate say, if I screw up here, I'll drive off many of my most valuable users--the people who actually edit, review and contribute tasks and code. So if we do any sort of sitewide license switch, I'll need the involvement of as many of Rosetta Code's active contributers as I can get in the discussion. And judging by the history of source code and license philosophy, it's likely to be a messy, convoluted one. I don't think wikitalk will be a good format for it. (I wonder if a NNTP hierarchy could be built out of the MediaWiki categories, and the Talk namespace pages be replaced with an interface to it. I'd be most comfortable in an environment like that. But that risks getting expensive in bandwidth and disk space.) --Short Circuit 06:01, 11 May 2009 (UTC)
The "incrementally" idea is good to me. I believe a not-so-far-from-GPL(v2) license or "compatible" can be ok. Anyway still some shadows (from the "legal stuff" point of view). As already said: RC allowed GPLed code to be here, if cited properly; this means that RC even though mainly licensed with GNU FDL for "direct" contributes, contains (and can contain) also GPLed codes. As said (again), if this was possible (and still possible it seems), then it seems that if I dislike the licensing of RC, I can write my own code on "my site", and then cite it, so I keep my GPL... If it will be no possible in future, it means that "cited GPLed code" must be removed...? (Or special exception should be created?)... --ShinTakezou 18:25, 11 May 2009 (UTC)
FWIW I don't have a whole lot of time for programming hobbies these days (hopefully again soon) but personally I'd be happy to declare anything I, personally, posted here as CC-Attr and leave it at that. And, really, PD would be fine with me - if I didn't want people to use my code, why would I put it on a website that is all about "how to do certain things in a certain language"?. Quite frankly I doubt the sincerity (or sanity) of anybody who puts code world-visible on the internet but then declares they don't want it to spread... Maybe a quick email dashed off to everybody who has contributed and in the best case they all simply agree on a CC-Attr type license and it'll all be really painless...Sgeier 21:57, 11 May 2009 (UTC)
To me (and as long as I can understand legal matters) PD is not acceptable (for "useful" code which is a little beyond the "learning the basic" of a language). I am happy with the idea something I've written can be used by someone else, even to earn moneys; I am not happy with the idea that they can "take" the work as if it were all their own work (maybe PD does not allow this, since the material must remain in the Public Domain... but PD is a rather foggy definition, and the fog density changes from country to country). So a CC or similar with attribution is, to me, the minimal acceptable license. It could be relaxed for "teaching code" (where maybe PD is the best indeed), but I think raising too many exceptions can confuse, and exceptions are needed for some code that "teachs" (everything can) in the RC spirit, but it is also original and useful outside the "teaching scope". I think attribution does not harm, and it is reasonable for "publically available" material. --ShinTakezou 10:26, 12 May 2009 (UTC)
It should be noted that in some parts of the world (e.g., the EU) PD is pretty much equivalent to CC-by; it's not possible to totally disclaim moral rights like the right to be identified as the author of the work. On the other hand, it's most certainly possible to disclaim all economic rights, and in other parts of the world (e.g., the US) only economic rights exist. Licensing of “free” works is really complex at the global level! (No, I know I'm no expert, but I've talked to people who have talked to experts...) –Donal Fellows 18:22, 26 January 2010 (UTC)

[edit] GPL

The copyright page tells one can not provide GPL contents on RC, because the FDL is not compatible with the GPL. This is partially true, and also partially false, because in fact it is more complex than this. In fact it is true if one wishes that both of these licenses apply at the same time on the same content (an and basis), but it is right possible for an RC contributor to write on his page that the content that he provides can be used under GNU/FDL or GNU/GPL. On an or basis, the statements of the 2 licenses don't apply at the same time and don't interact with opposite statements. As an example of this there are several libraries available on the web which can be used under the GPL or a proprietary license (which you buy). In these cases the GPL and the proprietary license are not compatible and are not applied both at the same time on an and basis. --Blue Prawn 20:16, 25 January 2010 (UTC)

This said, IMO only or licensing should be allowed on RC, and any and licensing should be prohibited on RC because this makes it very difficult to respect for people willing to reuse contents. --Blue Prawn 20:16, 25 January 2010 (UTC)

The way the page is currently worded, contributors are able to grant additional privileges, but not remove them. That that only applies to their individual changes is certainly problematic for someone wanting to reuse the code, and that's something I wish were easily remedied, but the only sure remedy I can think of that allows for some degree of consistency while maintaining technical simplicity is to gradually migrate the entire site over to a more permissive license like CC-By or CC-SA. (I'm not personally interested in promoting the NC component.) A migration would involve deprecating all current tasks and pages, and starting from scratch under a new license. That's not impossible; we may very well be at the perfect relationship of content volume and community size to do it. I could even create a namespace to move the current GFDL content to. (There's still a question of category and template content, which isn't so easily shuffled around.) It would be a lot of work, though, and not something I have time to effectively oversee. --Michael Mol 04:10, 26 January 2010 (UTC)
Well you perhaps have noticed that the licensing on Wikipedia have migrated from GNU/FDL to CC-by-sa. Something I am sure of is that noone contacted me to ask me any authorisation to change the license. Someone from my LUG explained me that the license migration process have proceeded as follow: the license was FDL version X or greater, they wrote a new FDL version so undated WP to this new one, and this new one was re-written in a cc-by-sa compatible way, so they have switched WP contents from FDL to CC-by-sa. You should verify if the process was exactly this one. Also we should update the templates like the one on this page Mandelbrot_set because it tells the WP content is FDL while it is not anymore. Now should RC also switch from FDL to CC-by-sa, I have no opinion on the question. --Blue Prawn 17:39, 26 January 2010 (UTC)
Wikipedia was using GFDL 1.3 at the time, which explicitly allowed that kind of a transition, for a limited period. I missed the window. (Largely because of a series of 60-80 hours work weeks piled on top of other responsibilities. Been having a similar schedule for the past three months, actually.) I've looked at that transition quite a number of times, and have had to give a lot of thought about how to get around not having that window available. :| --Michael Mol 20:52, 26 January 2010 (UTC)

[edit] Book: Picolisp by example derived from the site

I've found that the book picolisp by example is out and am concerned about the statement of authorship given on the second page as it doesn't mention this sites contributors even partially.

I am also concerned about how difficult they make it to find out such authorship. By selecting just what links they left in the book they make it difficult to track authorship of tasks, but by leaving in external links from this RC site they show that it is possible to leave them in.

What I don't want to do is attribute any malice from the authors in doing this, but I don't think enough attribution has been given to authors and an easy method for that attribution to be found out, i.e. linking to tasks on this site, has been elided. --Paddy3118 06:02, 2 September 2012 (UTC)

P.S. Have I missed some announcement of this book on RC? I just googled and found nothing.

Wouldn't that just be Abu anyway? --Ledrug 06:23, 2 September 2012 (UTC)
I don't know any of the "authors", so I am giving them the chance to comment. If you have knowledge of other cases that might be good reading then by all means, add some links. --Paddy3118 06:52, 2 September 2012 (UTC)
Hmm? No, I mean the copyright concern. Assuming Abu is the author of said book, how many picolisp examples on RC are not written by Abu? Even though code he released on RC is under CC (or is it GFDL?), he himself can still use his own code however he pleases, such as publishing somewhere else without mentioning RC. Unless you do find in his book substantial code written by other RC users, there's no reason be be concerned about. --Ledrug 07:50, 2 September 2012 (UTC)
I was contacted beforehand for advice. It's my understanding that all of the code examples were written by Abu, so I believe it's fine. I've also been asked to write a forward for "PicoLisp by Example", which I intend to get around to today. --Michael Mol 10:08, 2 September 2012 (UTC)

The book has around 600 task descriptions copied just about verbatim from RC. Each section of the book begins with the task description and yet only Alexander Burger and Thorsten Jolitz are cited as authors. No attempt seems to have been made to even inform RC contributors of the existence of the book prior to publication.

Personally, If they had acknowledged that the task descriptions take time and expertise to create and are NOT authored by them, added an explanation of how to look at authorship through RC page histories and added links to each task page from every task they had used (in the style of some of their RC external links that they have left in), then that may help; but they seem to have concentrated merely on the examples they may have written in PicoLisp and, maybe by ommision, they claim authorship of the task descriptions too. --Paddy3118 10:53, 2 September 2012 (UTC)

First, chill! Second, I was informed a few weeks ago in IRC, when they asked me for advice wrt copyrights and GFDL. I gave them my understanding of how things work. Second, I was asked to write a forward nine days ago (whereupon I asked for a reminder in nine days, as I was about to have an extremely busy week, fitting 40 hours of work into three days, a wedding and a family emergency) Third, they're not required to give anyone notification, so long as as they conform to the license requirements. Now, they glitched, but it's fixable. Now here's the beauty of it: It's on Github, which means it's eminently reviewable and fixable. Someone reported the copyright glitch to Github, landing them in hot water. So now tempers are high all around. And that makes things difficult. All that's required to get them in conformance is a list of authors of the task descriptions, because it's a given that they already have the copyright over the code examples they're using. I devised a reasonably simple algorithm to come up with the list for a given task. I don't have time to flesh it into code, but I'll drop it into Task Description Authors as a draft task momentarily.--Michael Mol 23:04, 2 September 2012 (UTC).
Hi Michael, thanks for your explanation. I guess the picolisp authors aren't required to join in this discussion either, which is what I would have preferred.
If they had given the site a heads-up before publication then I wouldn't be surprised if RC contributors could have helped them in their endeavours. An announcement by them, with a meaningful sub-heading or a user sub-page off their User page, would have got some attention here, but then, they aren't required to do that either. --Paddy3118 03:45, 3 September 2012 (UTC)
I explicitly told them to simply hunker down and code, and that I'd see what I could do on this end. Tempers being as high as they were, it was obvious to me I needed to step in as a moderator/mediator before mixing things up again.
They did give me a heads-up, but I didn't relay the announcement at that time, owing to the extraordinary time crunch I described above.
As for whether it was appropriate for them to contact me directly, rather than leaving a note on their user page, the etiquette on that kind of thing is really very unclear. I'm sometimes amazed at how unfamiliar people are with wiki etiquette when I talk with them face-to-face on such things. At the same time, most signs looking for labeled authority on RC lead back to me, so I give the appearance of a single (or, at least, first) point of contact. So perhaps there's a better way for this kind of thing to be handled in the future.
I presumed the RC community would still be willing to to lend a hand and get this thing rolling. The infrastructure and effort required to set up a book like this (and the experience gained therefrom) can be turned around and applied to every language on Rosetta Code, which I saw as a great opportunity to raise awareness of more programming languages, provide more consolidated documentation for some, show off the great work that the RC community as put into creating this corpus, and to highlight editorial need. (Seriously, read through the book; presented in this way, you can see the variance in quality and organization of RC tasks, and it strikes me as a great view to help come up with a better task layout and template.)
I really think the RC community should step up and lend a hand, even if only by looking at Task Description Authors and applying it to each task on RC. The PicoLisp by Example book looks (to me) like a great opportunity to drive improvement on RC, and to bring what we've done to more people (while setting up a framework for doing similar things going forward). --Michael Mol 13:51, 3 September 2012 (UTC)
Hi Michael, so the two mentioned as authors know about this discussion and just choose to ignore it? It would be good if they joined the discussion and gave their own view. --Paddy3118 03:51, 4 September 2012 (UTC)

P.S. Doesn't the license state that some attempt at stating authorship needs to be given? --Paddy3118 10:53, 2 September 2012 (UTC)

Here is how contributors to the Python documentation are listed. It gives a list of around 230 names and allows for consideration of additions. Nothing like that is done in this PicoLisp book. It seems to belittle the contributors of around a quarter of the text of the book. --Paddy3118 11:09, 2 September 2012 (UTC)

What a great idea. The task that I created when I first came to this wiki, best shuffle, is also in there. Cool! -- To have a long list of contributors and task authors in a book like this, I don't see what purpose that would serve. Nobody cares. Fwend 17:53, 2 September 2012 (UTC)

The problem is that with who-knows-how-many authors in various bits throughout RC, someone might care. So it's generally best to act on a pessimistic reading. CYA sucks, but it is what it is.--Michael Mol 23:04, 2 September 2012 (UTC)
I find this situation ridiculous.
You may be technically correct that the book authors should cite original contributors for each task (I'm not even sure about that, GFDL 1.2 doen't seem to say so for verbatim copy), but what a lot of good that will do. Given a task name, it's straightforward to look it up on RC and check its history, and the book makes very clear that RC was the source. I don't think they are obliged to provide a link for each task, which wouldn't make sense in a printed book any way; to satisfy the technicality, they however would have to look through each task's history and figure out who should be classified as its significant contributors, probably by hand (MikeMol's proposed automation isn't really going to guarantee complete correctness in this department). What would all this undue burden give us? Distracting and goofy attributions like "task originally by Paddy1234" or "by Ledrug" or some such, which neither the book authors nor readers really care -- and those aren't even real names. All this wasted effort, for something one could easily look up in the first place.
GFDL, CC and various other licenses were designed to promote share and reuse of ideas; getting all worked up as if Abu & Co. were trying to steal your life's work (I'm pretty sure they weren't) is quite contrary to that spirit. Technically it's within your right to demand attribution, but, to me, that doesn't make your desire of being credited wherever possible any less petty.
(The above are my personal, original opinions; they would normally have been available under the terms of GFDL 1.2 per RC copyright policy, but I hereby release these four paragraphs, including this one, into public domain.) --Ledrug 02:08, 3 September 2012 (UTC)
I have seen it written that book authorship and where your work has been cited is currency in academia, reviewed in job applications and on promotion boards. In computing, job applicants are told that contributions to open source projects help their case. It would have been polite, when you can look at a page of the book and see a task description longer than the picolisp code and which could have taken more effort to put together than the code, it would have been polite if one had known it was coming.
I have learned and derived pleasure from writing and modifying task descriptions as well as code. I think its good to see their improvement over time, and think they are of intrinsic value. I cannot say what individual task authors would think, but with so much of the book being RC stuff not written by the picolisp authors then I would think it was blindingly obvious to someone that has an account on the site enough to write all that picolisp, that discussing the use of such text on the site would be a good idea. But they don't have to, and have (still) chosen not to do that. --Paddy3118 04:27, 3 September 2012 (UTC)
Personal tools

Variants
Actions
Community
Explore
Misc
Toolbox