Commercial streaming platforms such as Amazon Prime Video, Netflix or Disney+ pay a lot of money to produce or license films and series – and are happy to pass these costs on to users. With the new platform WikiFlix there is now a streaming offer that is designed for public domain films. There are currently more than 3,100 freely available films to be seen there, from silent film classics to current animated cartoon (short) films. The platform is based on linking publicly accessible data sources. The video files themselves are located on the servers of Wikimedia Commons, the Internet Archive and YouTube. Following the example of large streaming services, you can browse WikiFlix by different genres – for example B-Westerns, films by women or films in Tamil.
Advertisement
The platform was programmed by Magnus Manske, a Wikipedian from the very beginning. Among other things, he wrote the fifth Wikipedia entry and the MediaWiki software, as well as numerous tools. In the interview he talks about the creation and background of WikiFlix.
How did you come up with the idea for WikiFlix?
I had a chance contact with a user who had already created a few manually maintained wiki pages with public domain films that were available on Wikimedia Commons. Then I thought, hey, this can be made more visually appealing and automated so that you can, for example, group films by country, genre, director or actors.
Magnus Manske at the third birthday of “Wikidata”.
(Image: Wikimedia / Jason Krüger / cc by-sa 4.0)
What is the added value compared to YouTube or the Internet Movie Database?
If you really just want to look for data snippets, you’re probably better off using Wikidata or IMDB. And if you know exactly what you want, you can also search on YouTube or Google. I just wanted to show how many freely available films there are and that there are some gems among them. And I wanted to create the experience of streaming services like Netflix and Amazon Prime that you can browse without knowing exactly what you’re looking for. It was initially intended as a demo, but then it quickly became very popular.
Only films whose copyright has expired are included – i.e. those that were released more than 70 years ago?
No, it’s more complicated. The files on Wikimedia Commons, YouTube and the Internet Archive are all hosted in the USA, so US law applies with a protection of 95 years. However, there are various exceptions. In India, for example, copyright lasts for a shorter period of time. And in the US there is a certain period of time in which you have to renew the copyright. And that often didn’t happen. The question is sometimes a bit difficult. But my experience from more than 20 years shows that the community is very good at quickly correcting any errors.
What do you do if files disappear from servers? Don’t dead links pile up at some point?
There are various bots on Wikidata that are run by volunteers and check the integrity of links. If that does become a problem at some point, I can relatively easily write something that throws out everything that no longer exists.
The basis for WikiFlix is the open database Wikidata, which contains information about around 31,000 films. Many of these films are also freely available, especially on the platforms Wikimedia Commons, Internet Archive and of course YouTube. In order for such films to appear on WikiFlix, the Wikidata entries must be linked to the corresponding video file. Everyone can help with this. It works like this:
This page lists which films are marked as “public domain” in Wikidata but are not yet linked to a video file. If you click on the film, you end up on the associated Wikidata page. There are also buttons that initiate a search by film title and year on the three platforms. A green arrow behind a button means that the search produced at least some results. You don’t even need to click on buttons with a red check mark – there obviously isn’t a corresponding video file.
However, if there are one or more hits, human helpers are required. You have to decide whether a video file you find really corresponds to the film you are looking for by comparing key data – such as title, year, director, running time.
There are often suitable video files on all three platforms. “You can enter all three of them,” says Magnus Manske. If that’s too complicated for you, you should concentrate on Wikimedia Commons. This archive is the most “stable”, so the files do not disappear from the internet so quickly.
Especially on YouTube you often come across a confusing abundance of the same or similar films. Which one to choose? Manske recommends choosing the longest and “cleanest” version, i.e. without subtitles, watermarks or references to remastered DVDs.
Next, the “Identifier” must be copied out, i.e. the unique identifier of the video file. For YouTube, this is not the entire URL, but just the string in the URL after “v=”. At Wikimedia Commons, the identifier is placed between rectangular double brackets [[]]. It can be copied by clicking on “Copy”. At Archive.org, the identifier is explicitly listed as such in the meta tags.
How do these identifiers get into the database? This is a bit counterintuitive at the moment: you click on “add statement” at the bottom of the Wikidata page and manually enter “video” in the field that opens. The copied identifier is then inserted into the free field that opens next to it. Then all you have to do is click on “publish” and the corresponding film should appear in the film portal within an hour.
With all the volunteer work for Wikimedia, do you still have time to watch films yourself?
Yes, occasionally I have to switch off and look at something.
What are your favorite films? Have you discovered any special treasures?
Anything by Charlie Chaplin is very entertaining. Metropolis is of course a classic. And Plan 9 from Outer Space is also public domain, even though it’s only from 1959. It is considered the worst film of all time – so bad that it has become a cult again. And if you like horror: We have Nosferatu and Night of the Living Dead. I also really like Nosferatu because the film shows how copyright was already a problem back then. The director originally wanted to film Dracula, but didn’t get permission from the Bram Stoker Estate. He then called the whole thing “Nosferatu.”
(grh)
To home page