That sounds very much like Google street view (you may get the same problems they had too, i.e. mandatory blurry faces).

Also, I believe Wiki Love Monuments folks would have a field day doing this kind of project.

In T133526#2234765, @Darkdadaah wrote:

That sounds very much like Google street view (you may get the same problems they had too, i.e. mandatory blurry faces).

I suspect the auto-photography of places and people without consent would be out of our budget once we factor in all the cars. ;) but yes -- there's definitely some overlap in a viewing model with linked 360 panoramas.

Also, I believe Wiki Love Monuments folks would have a field day doing this kind of project.

Yes yes yes. :) There's an aerial photography contest going on too, these sorts of things would go lovely with them.

Yes, yes and yes. One of the challenges is that Commons seems to be the default crossroads for these types of projects, and they are (understandably) very conservative, not just with copyright but about adopting anything that will break functionality for all the projects that rely on it. For example, even getting 3D object upload working has been a long slog.

This convinces me more each day we need an experimental "Tools Server" for just playing with media. Or we need a separate project/foundation just for the multimedia. But WikimediaMedia just sounds bad. :)

Related: T121044: Vision discussion: The Future of Wikipedia (@Yurik's I Dream of Content proposal)

Removed sub-tasks of T3790.

I just discovered an experiment how 3-D models can be combined with Wikipedia articles : http://en.volupedia.org/wiki/Stegosaurus

Awesome to see this finally in motion

Seven years later, I'm tempted to surface this for the Wikimedia Hackathon 2023, or at least discuss the best possibilities using open-source tools.

One of the best non-open tools is Kuula.co, which provides a full "Myst" type authoring environment similar to Hypercard of the 1980s. It uses 360 photos as its basis, and creates virtual tours linking them together.

The best open-source analog to that right now seems to be Marzipano, but it is only a set of node.js modules that provides to tooling, but not the authoring environment.
https://github.com/google/marzipano

I could imagine providing a way to specify a virtual tour info on a wiki page, and then pointing to a toolforge implementation of Marzipano to parse it, and provide the immersive experience. More thoughts and discussion welcome.

For the 2024 Hackathon in Tallinn, I prototyped an example virtual tour using the underlying Panellum software, which takes a JSON definition file for nodes, images, and hotspots. An example can be seen here:

https://panoviewer.toolforge.org/tourbeta/

The relevant definition file, converted to TOML can be found at:
https://commons.wikimedia.org/wiki/User:Fuzheado/Panellum_Tour