I'm a bit uncomfortable with some of the approaches suggested - they seem to be oriented to the philosophy of OA rather than the functional results. "can I read this, right now, without paying or signing anything". Any other information is wasted on almost everyone, and the more information we try and provide in our signalling system, the less useful it becomes. We can convey one or two things clearly ("YES, NO"), but if we try and convey subtle details, everyone gets confused. I agree it's sometimes useful to know about licenses etc... but most people, most of the time, don't care, and those of us who do care can follow the link and find out. Is it really important for us to maintain this information in Wikipedia citations? Has anyone ever said "maybe I won't follow that link, it's CC-BY-NC"? I'm not seeing the real benefit here. I would suggest we need to identify two things: a) This article (or this copy of this article) is "open access", gold or green - you, yes you, can follow this link right now and read it. It might be gold in PLoS One, it might be a repository copy with an expired embargo, it might be a postprint on arXiv, but you can read it, and maybe you thought you couldn't. b) This article is locked\paywalled and you cannot read it without special access. This symbol works for both the Elsevier Journal of Expensive Research and for an article in the New York Times. Open padlock, closed padlock. Maybe differently-styled padlocks (the curvy OA one versus a squared-off closed one?). Nice and simple and widely understood. Anything else is more useful to us, as people who care about open access and debate definitions, than it is to the general public. Andrew. [disclaimer: I am on my lunch break. this does not necessarily represent the OA position of my employer] On 9 September 2013 13:08, Stuart Lawson <stuart.a.lawson(a)gmail.com> wrote: -- - Andrew Gray andrew.gray(a)dunelm.org.uk

Hi all. Andrew raises a very goog point (I remember Ocaasi saying something very similar). It's a very important and pragmatical position (a very "librarian-oriented" one, I would add :-) and I fully respect that. But, at the same time, I feel that we do not want to associate the orange padlock, which is a sort of default symbol of Open Access, with just free to read. I don't really want to do "The Stallman" here, "free to read" is not Open Access, and, in the long run, this matters. This is why I proposed a 3 icons system, intead of binary one. 3 icons, I argue, convey more meaning than 2, and we should want that nuance signalled. We can choose different icons, if the ones I proposed are not OK. But I have the (weak) opinion that we can make this system more useful and "sustainable" if we associate the orange lock to real Open Access. I would think it's important for the OA movement and the Wikimedia one, at the same time. We push CC-BY-ish licence with GLAMs and in every kind of outreach. It's important, because make us *interoperable* with the world outside. Having said that, I won't mind if the collective decision is just for 2 icons, really. I just wish we could be easy and simple and clear even conveying information we care about. Aubrey

... 1. *Closed access* - not free to read (subscription required, paywalled) - *closed gray lock* 2. *Free to read but not to reuse* - i'd like to see some lock variation proposed or a *book icon* 3. *Free to read and to reuse* - at least as free as CC-BY-SA (including CC-BY) - *open orange padlock* * * Jake On Tue, Sep 10, 2013 at 9:20 AM, Jake Orlowitz &lt;jorlowitz(a)gmail.com&gt; wrote: -- Jake Orlowitz Wikipedia: Ocaasi <http://enwp.org/User:Ocaasi> Facebook: Jake Orlowitz <http://www.facebook.com/jorlowitz> Twitter: JakeOrlowitz <https://twitter.com/JakeOrlowitz> LinkedIn: Jake Orlowitz<http://www.linkedin.com/profile/view?id=197604531> Email: jorlowitz(a)yahoo.com Skype: jorlowitz Cell: (484) 684-2104 Home: (484) 380-3940

Provided we can get a consistent meaning, I don't see why this has to be only for journal articles. I was originally thinking we should aim for one icon per reference though, so a newspaper article that's free to read would only have the single book icon, because it's not 'closed' in the sense of 'you must pay to read it', and the closed lock suggests you can't actually access it. In that sense the 3 icons are in order of accessibility/oa-ness and you'd only have one per reference. Jake On Tue, Sep 10, 2013 at 12:15 PM, Stuart Lawson &lt;stuart.a.lawson(a)gmail.com&gt;wrote;wrote: -- Jake Orlowitz Wikipedia: Ocaasi <http://enwp.org/User:Ocaasi> Facebook: Jake Orlowitz <http://www.facebook.com/jorlowitz> Twitter: JakeOrlowitz <https://twitter.com/JakeOrlowitz> LinkedIn: Jake Orlowitz<http://www.linkedin.com/profile/view?id=197604531> Email: jorlowitz(a)yahoo.com Skype: jorlowitz Cell: (484) 684-2104 Home: (484) 380-3940

My concern is that we are still providing a service to our readers, who want to know what they can do with a source? Do they have to pay to see it, or can they just click through and read it? That's their primary concern: can they read it. The second issue is whether the source is free for reuse in the libre sense. We want to signal that because we do want to highlight those sources, I think. I'm not sure I see how green and gold fit into this, as they don't necessarily impact the pay-read-reuse structure. Green articles may not be free to reuse, and same with Gold while a Gold article may be free to reuse while a green may not. So I think that's a side issue that we're actually not wading into with the pay-read-reuse structure. And I agree we shouldn't weigh into that broader debate as it's very much still up in the air. On 9/10/13, Andrew Gray &lt;andrew.gray(a)dunelm.org.uk&gt; wrote: -- Jake Orlowitz Wikipedia: Ocaasi <http://enwp.org/User:Ocaasi> Facebook: Jake Orlowitz <http://www.facebook.com/jorlowitz> Twitter: JakeOrlowitz <https://twitter.com/JakeOrlowitz> LinkedIn: Jake Orlowitz<http://www.linkedin.com/profile/view?id=197604531> Email: jorlowitz(a)yahoo.com Skype: jorlowitz Cell: (484) 684-2104 Home: (484) 380-3940

I'm quietly watching the discussion and have been wondering whether anyone has thought about people with disabilities. Green, gold, orange, etc....can a colorblind (or fully blind) person see these distinctions? I'm not an expert in web design, but I believe that web designers are told to stay away from depending on color alone to guide users. Isn't it important to be able to provide open access to those for whom much of the world is particularly closed? Bob Kosovsky, Ph.D. -- Curator, Rare Books and Manuscripts, Music Division, The New York Public Library for the Performing Arts blog: http://www.nypl.org/blog/author/44 Twitter: @kos2 Listowner: OPERA-L ; SMT-TALK ; SMT-ANNOUNCE ; SoundForge-users - My opinions do not necessarily represent those of my institutions - On Tue, Sep 10, 2013 at 5:08 PM, Stuart Lawson &lt;stuart.a.lawson(a)gmail.com&gt;wrote;wrote:

Interesting discussion - thanks to all who have joined in so far! So far, we have covered the perspectives of readers, Wikipedians, OA advocates (in whatever form) and librarians. Now let's think about implementation: what sources are available to gather the information whether some reference falls into one of our classes that we want to signal? Of course, we could have a bot follow all the links provided with references, and whenever it hits a paywall, mark things as paywalled. But what if there is a non-paywalled version out there, how are we going to find it? Yes, Google Scholar has much of that information, but they do not allow systematic querying. And even if the bot (or someone) finds a copy somewhere, how do we know that it is legally up there? And for how long? Or, conversely, if we don't find anything now, perhaps an embargo is over somewhen down the line, and maybe someone will post the thing somewhere then - should we chase for these? And what about different file formats? JMIR, for instance, offers HTML for free (CC BY) but charges for PDF. Nature offers XML for "free to read by machines" but charges for PDF and HTML. How do we go about that? Those kinds of problems do not occur if we just stick to labelling stuff that has a CC license associated with it in trusted sources of metadata (like DOAJ or CrossRef). The rest is (to be) covered by the OA button approach, a representation of which should ultimately be part of our signalling scheme. The next question, then, would be whether we go for "any CC license" or some subset thereof for the part that we implement ourselves. I agree that most readers don't care, while many Wikimedians do, and I think we should cater to both groups. This could be achieved in multiple ways, e.g. by having a free to read icon on Wikipedia and some representation of the licensing on Wikidata (once we have individual references represented there). On the other hand, a signalling of OA-ness on Wikipedia should not be seen in isolation of the same thing being signalled elsewhere - be it in repositories or on the pages of publishers, on blogs or on the homepages of researchers. If all of these places were to use a certain signal to convey the meaning of "free to read" (in fact, many use the orange open lock icon for that), we are a small step ahead of a system in which readers always have to find out by themselves whether they can read something or not. If, on the other hand, we can agree on what precisely we mean by OA, then a layer of tools and services (or "apps", if you prefer) could be build on top of that - just imagine an "import this image to X" button associated with images in places like PubMed Central or even at publishers (X, btw, could be Wikimedia Commons, or a blog, or an OER), or authors proudly displaying stats of the GLAMorous kind (cf. https://commons.wikimedia.org/wiki/Category:Media_from_highly_reused_open-a… ) in their lists of publications. If we go for anything here that includes -NC clauses, nobody other than the copyright holders could make money off those apps, while -ND stuff could not be edited, so it would not fit a wiki environment. This leaves us with Commons-compatible licenses, and considering that there are almost no scholarly papers licensed CC BY-SA, going for CC BY compatibility would not be a problem (a quick Google search on PubMed Central currently lists 13 articles under CC BY-SA, at least some of which - including one of mine - are false positives that talk about that license, rather than being licensed that way: https://www.google.com/search?q=site%3Awww.ncbi.nlm.nih.gov%2Fpmc+%22http%3… ). So I imagine that we signal the following: - initially, just" CC BY compatibility" (either through the CC BY icon or some non-orange version of the open lock - my favourite would be blue as in http://dx.doi.org/10.5334/bha.2114 or https://twitter.com/OpenAccessEC ). Not sure whether CC0/PD stuff should be signalled separately - probably yes if we go for a CC BY icon before, perhaps not if we go for something else, e.g. https://commons.wikimedia.org/wiki/File:Sz%C3%A9chenyi_l%C3%A1nch%C3%ADd_in… ). - once this works, we could also signal "free to read" for stuff that is under other CC licenses (e.g. via a book icon; this may cause confusion, though, in the sense that some may believe that CC BY means not free to read, at least if we do not also use the free to read icon alongside the CC BY icon) - in the long run, references not being labelled by any of the two above would get a representation of the OA button's score (or whatever they have then) for that article (which may include a signal that some "free to read" version is available). so long, d. On Tue, Sep 10, 2013 at 11:36 PM, Stuart Lawson &lt;stuart.a.lawson(a)gmail.com&gt; wrote: