Wikipedia talk:WikiProject Open/Open access task force/Signalling OA-ness

I support this - a brief private discussion where I suggested:

I think the divisions are right. The average reader wants to know "can I read this?" The average editor may also wish to know "Can I re-use material from this in other places?".

I think there is a practical distinction between electronic articles and hardcopy journals, magazines, books. A hardcopy book isn't "paywalled" as such and doesn't normally have an OA-able content. The Crick/Watson example was originally hardcopy and is electronic because the publisher added value by scanning it. Personally I think a publisher should have at best a limited embargo on this if they do it voluntarily. We may also need to highlight scanners such as GoogleBooks, JStor, etc which have complex restrictions and availability.

I think the initiative has huge potential for Open Access beyond the closed world of academia. It could bring in thousands of readers and editors and create a metadata resource which is minable. Petermr (talk) 12:46, 22 July 2013 (UTC)[reply]

OA Button and signalling OA-ness in Wikipedia[edit]

"A system aimed at signaling the non-OA-ness of articles is currently being developed - the OA button. How could the two be combined?" - I've been talking to the creators of the OA Button about this, about finding ways they can integrate their ideas in Wikipedia. They'd love to hear from any Wikipedians who have ideas about this and be contacted here. Lawsonstu (talk) 20:04, 25 August 2013 (UTC)[reply]

I am in contact with them since the BMJ hackathon and think the OA button could well fit into this OA signalling scheme, though details remain to be worked out. -- Daniel Mietchen (talk) 08:58, 4 September 2013 (UTC)[reply]

The OA Button is still in an early stage of development and a first working prototype should be launched during Open Access Week (21-27 October). They'd like to work on Wikipedia integration after they've got the basic system working stably. It could prove very useful for automatically flagging up the OA-ness of article, once we have a signally system in place. Lawsonstu (talk) 12:11, 9 September 2013 (UTC)[reply]

The Open Access Button has now launched! You can get it from their website. It will be interesting to see how it develops and how it might be integrated with Wikipedia. - Lawsonstu (talk) 21:05, 18 November 2013 (UTC)[reply]

Nice page![edit]

Just discovered this through Daniel. Kudos, and thanks for putting this together. I will definitely link to it from the molecular biology projects we're running. Klortho (talk) 19:41, 26 September 2013 (UTC)[reply]

Existing approaches to conveying info about links[edit]

I like this idea as well. It would be wise consider what Wikipedia already does to convey "something-ness" about a link to the reader. (I'd like to add this idea to the main signalling page, but want to think/talk it through first.)

Internal links are either red or blue, indicating whether the page exists.
External links are a slightly different shade of blue, and have a small arrow link indicating they are external.
External https links and have a small padlock link instead of the arrow.
PDF files carry the PDF icon.
There are several icons described on the main page here (see here) relating to open access, closed access, etc.
…others…?

In my opinion, the existence and presentations all these icons should ideally follow from a discussion of "what kinds of things should Wikipedia be signalling to its readers?" In my opinion, it is much clearer to me that the reader benefits from knowing about a link's OA status than, for example, whether it is in the PDF format.

But, removing other icons is probably not something worth taking on. Given that they exist, I would love to see any changes under discussion here implemented in such a way that their icons match the arrow, padlock, and PDF icons in their visual style, and in their placement next to the link.

Of course, it may require some fancy technical work to adjust the placement properly. -Pete (talk) 00:38, 10 October 2013 (UTC)[reply]

I agree that keeping in mind these existing signalling channels is useful. However, OA-ness is the property of the content of a document (and potential copies thereof), not of a link. There may actually be several links pointing to the same document or content-identical copies. They should all (or in the majority of cases at least - see this text file and this external link for examples of possible exceptions to your schemes above) display OA-ness in the same way if we get the system right.

Getting it right might involve turning the document property into a link property. For instance, it could make sense to have links starting with http://creativecommons.org/publicdomain/zero (e.g. http://creativecommons.org/publicdomain/zero/1.0/ ) iconized as something like

, which could indeed be automated once we have the information about the licensing available in a consistent machine-readable manner. -- Daniel Mietchen (talk) 08:49, 18 October 2013 (UTC)[reply]

License of "Biodiversity Assessment …"[edit]

Thanks a lot for this project! I only found a small inconsistence I did not know how to correct: The license of the "Biodiversity Assessment …" article is said to be CC-BY 4.0, Wikisource says it is CC-BY 2.5 and at ncbi.nlm.nih.gov it is said to be CC0 if I understand it correctly. You might match it. Thanks, --Marsupium (talk) 14:49, 5 February 2014 (UTC)[reply]

Thanks! At PLoS One it is CC0. I changed the link but I didn't find the right graphic. --Hroðulf (or Hrothulf) (Talk) 22:38, 6 February 2014 (UTC)[reply]

Simplify[edit]

Hi! I really like the idea of these indicators, but it becomes fairly complex. I think starting with a single marker that shows 'you'll be able to read this content' would be a great start. Later on details could be added if necessary but probably aren't the most important thing for most readers, and researchers can do their own drilling down. heather walls (talk) 21:06, 11 March 2014 (UTC)[reply]

Thanks for looking into this. Efforts to simplify are underway, e.g. here. -- Daniel Mietchen (talk) 00:22, 27 March 2014 (UTC)[reply]

Keep it broad[edit]

I've been meaning to comment on this again for some time - apologies for the delay, and if you've seen these comments before. I know I've definitely discussed it with Daniel a few times :-)

We should try and keep as few extra icons as possible; every additional nuance in the signalling means more complexity which means fewer readers will understand it or remember it. So one/two symbols, no colour coding, etc.
We should not presuppose much awareness of OA issues - so symbols like padlocks or dollar signs for "closed/paywalled" are good, ones like open padlocks for "open" good, but things like the CC-BY "person" or CC-0 circled-zero are somewhat more confusing.
We should remember that what 95% of readers want is to know "can I read the document at the other end of this link". "Can I reuse this" is secondary, and (to my mind, for humans) best conveyed by the resource itself; however, embedding this as metadata in a future Wikidata-based citation system, for machine-reading or for opt-in highlighting, seems worthwhile. (It seems notable that both the NISO and JISC systems have come down in favour of this approach - "readable" is the first question and licenses are an additional detail)

Taken together, these would suggest to me that a) we should have one symbol, and b) it should be used to convey "free to read". The OA padlock seems a pretty good fit for this. (There is still an open question when you stop using it, though - what's no longer OA and just business as usual?)

More generally, I think that using a hard definition of OA (like the BOAI one) to apply the labelling would be counterproductive - yes, it's a valuable position to take, but if we only highlight BOAI-compatible material we're losing a valuable opportunity to highlight material held in repositories to the reader. This is especially true as a lot of use-cases for green-OA repository copies would involve a link to the repository version alongside the canonical (locked) DOI, and we need to signal that the two links are different in some way (see, eg, my attempt to handle this in ref #4 of Lowest temperature recorded on Earth). Andrew Gray (talk) 20:52, 10 April 2014 (UTC)[reply]

How does this look?[edit]

Hey all, I've used the open and closed access pictures in South American dreadnought race. Am I using them correctly? Ed ^{[talk] [majestic titan]} 03:21, 29 July 2014 (UTC)[reply]

Hi Ed, that's quite a complete coverage in terms of signalling accessibility. The open orange lock icon was designed for scholarly stuff that is compliant with the Budapest Open Access Initiative (i.e., in essence, CC BY compatibility) but it has been used on many occasions to simply mean "free to read", just the way you used it. I have not seen it used outside of scholarly contexts. The closed grey lock icon was designed to mean the opposite of the orange one, so I think your use is fine here too. The signalling OA-ness project, however, is about signalling openly licensed scholarly materials. -- Daniel Mietchen (talk) 09:00, 29 July 2014 (UTC)[reply]

Note that the links to biblioteca.dphdm.mar.mil.br (example) under "Official sources" did not work for me. -- Daniel Mietchen (talk) 09:03, 29 July 2014 (UTC)[reply]

Yeah, I need to add another footnote linking to this. That website has been down before, but this is probably the longest it has been out. Ed ^{[talk] [majestic titan]} 13:27, 30 July 2014 (UTC)[reply]

Google Books Preview[edit]

Hi, at Talk:Phineas Gage#An Odd Kind of Fame - open access? we are discussing a fairly common situation where a book on Google Books is almost completely available for reading, but isnt 'open access' in the strictest sense. It would be nice to show readers that the link does contain a very usable edition. The reality of OA-ness is it is a murky world, and Google Books Preview is a high profile negotiated level of open access which many academic authors (and readers) consider to be an appropriate balance being struck. John Vandenberg ^(chat) 21:58, 3 October 2015 (UTC)[reply]

Template:Subscription required[edit]

Is usage of Template:Subscription required encouraged? Could it be added by bot or some other automatic way (like a DOI lookup)? Nemo 08:55, 15 May 2016 (UTC)[reply]

Within the OA Signalling project, we focus on indicating that a given reference is unambiguously available under an open license, which would allow for reuse on Wikimedia projects. Templates like {{Subscription required}} or {{Closed access}} are thus not currently part of the project, so I think we neither en- nor discourage their use. Yes, they could be added by bots in principle, but I'm not aware of anyone working on that. Besides, while the version of record may be paywalled, another version may often be available for free (possibly after some embargo), and keeping track of such cases is not trivial. -- Daniel Mietchen (talk) 00:09, 17 May 2016 (UTC)[reply]

RFCs on citations templates and the flagging free-to-read sources[edit]

See

Headbomb {talk / contribs / physics / books} 16:21, 29 October 2016 (UTC)[reply]

Help talk:Citation Style 1#Auto-linking titles with free DOIs[edit]

Please comment above, since this is directly related to signalling OA-ness. Headbomb {t · c · p · b} 13:53, 23 April 2020 (UTC)[reply]

Thanks for pushing on that change. – SJ +

Other levels of signalling[edit]

Citation-level signals[edit]

Are there any scripts or tools for estimating, for a given closed-access source, how substitutable it is for an open-access one? Perhaps as plugins that suggest improvements to editors after saving? – SJ + 17:22, 15 March 2024 (UTC)[reply]

Article-level signals[edit]

It would be nice to cache and present some article-level overview of the OA-ness of the sources used in the article. Are there any past proposals on ways to do this or to visualize it? – SJ + 17:22, 15 March 2024 (UTC)[reply]

Project-wide signals[edit]

A list of the most widely used closed-access cites, combined with an estimate of how substitutable they are for an open one, would be helpful for increasing the openness of an entire project. This requires more of a research effort up front, but could output a useful set of open tasks. – SJ + 17:22, 15 March 2024 (UTC)[reply]