I love CC-BY-SA and I love it when people contribute their creative work to the public in this way.
However, this is not always a great fit for photography Q&A, because once so licensed, a picture can be taken out of context and used however. I’m not worried about how my writing about, say, the way aperture works in photography, might be used — I want it to be shared and distributed and built on. But if I have a question about how to improve a portrait of a friend, I’d like to be able to share that portrait without agreeing that it’s fine for that portrait to show up in a global ad campaign by some big company (or politician!). It’s worth noting that even sharing a cropped and low-resolution version under CC-BY-SA may put the full image under that license.
The “no commercial use” Creative Commons clause is problematic in its own way and doesn’t really solve the problem.
I think we need a special permission granting the same things as CC-BY-SA but only in the context of the associated Q&A . That way, the site can function and answers can show modified forms of the original, and the content isn’t lost if the Q&A gets moved or reused elsewhere — and one could still, for example, make and sell a book form of the Q&A and use the images.
Fair Use
It is also often the case that people want to learn about images which are not their own. In general, this squarely meets three of the four standard factors in deciding on whether a use falls under the fair use doctrine:
Purpose and character: criticism, comment, teaching, scholarship, or research
Nature of the work: (this probably doesn’t apply)
Substantiality of the portion used: usually a reduced-resolution preview is fine
Effect on the potential market: usually none
I don’t think we want to quash such use preemptively. But, we also shouldn’t make people falsely declare that they have the right to relicense other people’s work they want to learn from.
On TopAnswers they’re exploring the idea of a code license that’s different from the post license. Perhaps we can do something similar for any clearly-demarcated content (code block, image, etc). This would be user-specified and we’d have to indicate it on the page somehow. Post-MVP, but we should come back to this.
Until we have something like this, I presume that a Codidact photography community could decide to allow links to externally-hosted images in its questions (or maybe only certain categories of questions). These wouldn’t be embedded (I think we’ve agreed that embedded images need to be hosted by us), and the community would have to decide how to manage such posts over time (links tend to rot), but it’d be a stopgap.
It’s an okay stopgap, but is generally not good — people tend to link to pages with multiple images, or to transient URLs, and the whole thing quickly becomes valueless. And of course clicking on an off-site link is always inviting a surprise of some sort.
I think it’d be very nice to have this worked out at the beginning of a photography community launch.
Perhaps we can do something similar for any clearly-demarcated content (code block, image, etc). This would be user-specified and we’d have to indicate it on the page somehow. Post-MVP, but we should come back to this.
For images, I suggest a required title, attribution, and stated license that gets displayed consistently when the page is presented. I also suggest that feature be part of MVP (however, I wouldn’t put much thought or code into the enforcement or verification of those fields at MVP).
Until we have something like this, I presume that a Codidact photography community could decide to allow links to externally-hosted images in its questions (or maybe only certain categories of questions).
100% with Matt here — it’s definitely a stopgap, but it will quickly degenerate. If nothing else, I suggest that for MVP, links to off-site images at least be treated differently (i.e., different markup or tags in the source) than just generic “look here…” links. That way it’s a lot easier to clean up or semantically re-present them as images post MVP.
I wouldn’t limit them to certain categories of questions. That seems like a minefield later on for the framework developers.