|Status of the proposal|
|Reason||interesting idea, abandoned & marked as stale by creator. Pecopteris (talk) 05:52, 20 August 2023 (UTC)|
|Details of the proposal|
|Is it a multilingual wiki?||See #One wiki or many?|
|Potential number of languages||See #One wiki or many?|
|Proposed tagline||Collaborative educational videos|
|New features to require||Extension:WikiVideos|
|Additional project settings||Will use files from Commons, but may require local files too.|
Wikivideos is the wiki for collaborative educational videos. Videos are created with the familiar <gallery> syntax, but with mode set to "video". For example, the following wikitext:
<gallery mode="video" width="300"> File:Old pond.jpg | Old pond File:Frog leap.jpg | frog leaps in File:Water sound.jpg | water's sound </gallery>
will output the following video:
(This is just a simulation, see https://wikivideos.org for the real thing.)
Essentially the algorithm shows one image after another while an automated voice reads the text aloud. However, the syntax can be extended and complicated endlessly, with support for different file types (including other videos and perhaps even other wikivideos), as well as
parameter=value pairs (again similar to the gallery syntax) specifying things like different voices for different images (to simulate dialogues), moments of silence, background music, etc. For an up-to-date wishlist, see https://wikivideos.org
The project is based on the recent (2022) Extension:WikiVideos that introduces the video mode to the <gallery> tag. It's still new and unstable, but already supports not only video generation, but also automatic captions, automatic chapters for navigating the videos, multiple voices in many languages using Google's text-to-speech service, and other features. See https://wikivideos.org and Extension:WikiVideos for an up-to-date list of features.
- VideoWiki - The algorithm for generating videos seems conceptually similar, but the implementation is different. Aims to add collaborative video creation within Wikipedia, while Wikivideos aims for an independent project.
- Wikispeech - Wikimedia text-to-speech initiative.
- Wikistories - Concept to stage Wikimedia content for posting in other social media platforms.
- Wikimedia Commons - Hosts videos, but they cannot be edited except by downloading, editing and re-uploading.
- Commons:Sequencer - Sequences of media without a text-to-speech feature. Currently broken.
- https://www.wikivideos.org - Already secured
One wiki or many?
At first glance, it may seem like this project may require one wiki per language, like Wikipedia. However, having multiple wikis is extremely inefficient in that templates, support documentation and almost everything else needs to be recreated (and maintained!) dozens or even hundreds of times, by often scarce volunteers that retrace and redo the same work again and again. This is currently the case for Wikipedia and many other Wikimedia projects, but even though the Wikimedia Foundation is still unable to deliver a solution, it's not the year 2000 anymore. Nowadays we have Extension:Translate and automatic translators like Google Translate that can greatly help us to avoid such inefficiencies while collaborating across languages.
One upcoming feature for the Extension:WikiVideos are automatic translations of wikivideo captions using Google Translate, and then automatic translation of video voices using said translated captions. This will allow us to script videos in English (or any other language!) and have them autogenerated in every other language. Automatic translations are not perfect but again, it's not the year 2000 anymore. For some language pairs they are just amazing and are always getting better. Looking 10 years ahead, automatic translations are the future and in any case, we can always fallback to some system to override them when a volunteer feels like doing a manual translation.
- Hello @Sophivorus: I would prefer one wiki per language, because Google Translate is a very bad automatic translator. If you can implement one that gives more faithful and sensible translations, such as Deepl, I will be ok for one wiki. --Hérisson grognon (talk) 11:17, 24 August 2022 (UTC)
- Sophivorus (talk) 03:21, 5 August 2022 (UTC)
- Aishik Rehman (talk) 07:11, 5 August 2022 (UTC)
- Diegodlh (talk) 13:19, 5 August 2022 (UTC)
- Bluerasberry (talk) 15:03, 15 August 2022 (UTC)
- Emolga826 (talk) 13:12, 18 August 2022 (UTC)
- –SJ talk 17:29, 23 August 2022 (UTC)
- Aliyu shaba]]Talk 18:04, 23 August 2022 (UTC)
- Doc James (talk · contribs · email) 03:49, 9 September 2022 (UTC)
- —Yahya (talk • contribs.) 19:58, 10 September 2022 (UTC)
- Zblace (talk) 10:16, 12 January 2023 (UTC)
- Pharos (talk) 18:49, 27 February 2023 (UTC)
- This is a cool concept; however, I’m concerned about the practicality here. Going by the example, it seems like the highest quality level is going to be extremely low, and the level of potential enthusiasm here is probably low. Going by that, I’d say Oppose as impractical and not filling any serious niche. Dronebogus (talk) 08:16, 12 August 2022 (UTC)
- Note, there was a previous similar idea implemented that never really went anywhere - commons:Commons:Sequencer, although perhaps the bugs were not sufficiently worked out of the software to give it a chance. Bawolff (talk) 05:55, 25 August 2022 (UTC)