Commons:Requests for comment/MP4 Video/Analysis of responses

From Wikimedia Commons, the free media repository
Jump to: navigation, search

This page is for editors to work collaboratively on understanding the nuances of what the community has said, and learning for the future. Everybody should be free to help out with that analysis, including editors who have expressed strong opinions in the discussion. Please keep the analysis neutral, and bear in mind that this page is not a suitable place to re-discuss or to attempt to overturn the clear consensus that has just been reached, namely that the community prefers no MP4 support. --MichaelMaggs (talk) 15:49, 14 February 2014 (UTC)


  • Doesn't look like this page is going to garner a lot of attention any time soon. Is this for meta-discussion about the RFC, like what points were raised in it and so forth? Well, there was an interesting point raised in the other sections of the RFC, under Comments, which was about the distinction between the definitions of free; free-as-in-free-speech (gratis) versus free-as-in-free-beer (libre) described on Wikipedia as gratis versus libre. TeleComNasSprVen (talk) 00:42, 21 February 2014 (UTC)
    Yes. Knowledge must be free for reuse. But the desire for the rest of our toolchain to be free, for forking or reproduction, is a different level of priority... parallel to free access to knowledge. Making changes that help people transform non-free formats into free ones should be high on our priority list: that clearly contribute to knowledge freedom. --SJ+
  • Another Point was, if support of video formats is a task of the Wikimedia Foundation or the browser/OS vendors. MP4 supporters suggested that this is the task of the Foundation, while those opposed see the vendors responsible. 87.78.28.226 16:26, 27 February 2014 (UTC)
  • A new RfC should be held focusing entirely on allowing conversion-on-upload, and never showing the format to readers. This RfC wasn't responsive to the concerns expressed, so most of the discussion was about why MP4-everywhere is a bad idea.
    • Most of the comments made, including many of the opposes, were compatible with transparent processing and conversion of MP4 formats on upload. ("a method to help uploaders deal with OGG and WebM could be developed, that would be great", "I think the easiest et logicalest solution is having a online intregated convert software", "A one button off-Commons Wikimedia-hosted tool that let you... Global Login and upload an .mp4 file, convert it to a free format")
    Some show opposition to spreading non-free formats, but again are compatible with conversion on upload. ("knowledge isn't free if it's not in a free format", "we should be creating free formats", "we should not be spreading patented materials", "Patented formats would encumber reuse", "Free formats can be used by anybody, the non-frees cannot.", "[ensure] people are not forced to use proprietary software in order to read, modify, and redistribute"). --SJ+ 20:18, 2 March 2014 (UTC)
  • A few comments made were that "people would upload lots of trash", or "allowing a popular format means a lot of crap would be contributed". That idea is not in line with our mission.
    Creating a draft space on Commons could help deal with this and with issues around bulk uploads (cf. the 1M images recently released by the British Library under a free license). All of these media, when the authors wish to release them under a free license, should be gree to cache and curate and revise on a Wikimedia project. --SJ+
    I think the argument was not so much about bulk uploads and more about expected trash especially from mobile users. Bulk uploads weren't really part of that discussion, I think - no wonder, as file formats are less of an issue for planned bulk uploads (where bulk conversion into a desired format can be planned as the first step). Gestumblindi (talk) 11:45, 3 March 2014 (UTC)
    The "lots of trash" phenomenon is something the Commons community is already dealing with, in my opinion very much exacerbated by the mobile uploads, which (speaking as somebody who monitors and processes Deletion Requests fairly often) would seem to have an incredibly high garbage-to-useful ratio. I agree that our mission guides us to reduce barriers to uploading, however I believe it is essential that the organization that created the mobile upload app (WMF) be paying attention to its impacts in this regard, and that it consider improving the mobile upload app in ways that would tend to encourage more high quality uploads. If the user were provided with any information at all about the Commons mission, in the software interface, I believe it would have a positive effect. (I think the regular Upload Wizard is a good model for this: there is a cartoon and also numerous points in the wizard that help the user contemplate what is good to upload.) I am hopeful that work is being done on this behind the scenes, since there have been vigorous efforts to keep the kind of category that tracks such things, which I think would also apply to something like Category:Mobile upload. @Gestumblindi: -Pete F (talk) 17:38, 19 March 2014 (UTC)
  • Superm401 offered a list of great things we should do to support open formats. Prompt people to install codecs, offer alternative playback and encoding mechanisms, support Google's efforts to make WebM hardware encoding fast. We do some of this now but could do more, and more openly. Mike Linksvayer added that we need to be working actively to make open formats become more prevalent. Otherwise this feels like a reduction in our commitment to open formats. --SJ+ 20:18, 2 March 2014 (UTC)

MichaelMaggs, TeleComNasSprVen - what do you recommend doing with this page? I'd like to find a way to implement conversion for uploaders who don't have ready access to a converter. Both individual and bulk uploaders could need such tools, though as Gestumblindi notes the bulk uploaders are spending more time in preparation and may make up their own solution. Still, this seems like a straightforward way to expand the presence of WebM files on Commons. That seems like the most useful constructive outcome from the discussion here: there are lots of contributors who currently find this to be an obstacle. A dozen ways have been proposed to handle it, from the extremely circuitous to the direct. --SJ+ 10:56, 14 March 2014 (UTC)

It's really a forum for thoughts and ideas, both for editors and for WMF staff who put the proposal up in the first place. I had no preconceptions about what might come out of it, if anything. --MichaelMaggs (talk) 08:01, 19 March 2014 (UTC)
Well, this certainly seems like an interesting idea to pursue. There exists a couple of choices of proprietary software and perhaps even freeware floating on the Internet able to convert to the Ogg file format, but their number is relatively low. If an internal MediaWiki extension were created specifically to convert everything to Ogg, we would certainly help the community by adding our name to the list. TeleComNasSprVen (talk) 10:12, 19 March 2014 (UTC)
See my comment immediately below -- ffmpeg is a widely used FOSS process for such conversions. -Pete F (talk) 17:38, 19 March 2014 (UTC)
@SJ:
  1. The "legal" page on the ffmpeg web site is encouraging: this is free software as far as copyright goes, and the creators seem to express confidence that it infringes no patents.
  2. It seems to me that it would be relatively simple and uncontroversial to add a piece of free software to MediaWiki, make it accessible at least as an "opt-in" feature, and then invite further discussion once people can try it out. (I am not really a programmer, so maybe it's not as straightforward as it seems to me; but it does seem likely that code as simple as "ffmpeg -i <inputfile.MP4> <outputfile.webm>"... should be pretty easy to include in a script somewhere, and then expose in the upload wizard via a user preference.) Have you considered simply opening a bugzilla request and seeing who jumps on it?
  3. Before it goes much further than a pilot/opt-in feature: Two things we value highly in this movement are responsible archiving, and remix culture. These two values in combination make it especially important that we create an environment in which any data loss due to conversion among lossy digital compression formats is (a) minimized, and (b) clearly communicated to the uploader, remixer, and viewer in relevant parts of the user interface. For instance, when I recently converted this screencast from AVI to Webm using ffmpeg, some small but noticeable artifacts of data loss were evident. (Also the file size was reduced by about half.) I would have valued the software communicating to me at that time, in the most minimally technical way possible, what would happen, whether I had other alternatives, etc.
  4. FYI, somewhat related: in our course Writing Wikipedia Articles, we have produced a lot of these instructional and class videos. As the project draws to a close in the next month or so, I have asked our resident expert on doing video conversions to write up a "how-to", documenting the process he has used (all free software). Any suggestions where we should publicize that when it is posted, to best inform discussions like this one? I had initially only intended to spread the word among education enthusiasts, but I can see it might have broader relevance.
-Pete F (talk) 19:03, 14 March 2014 (UTC)
Another thought: it might be worthwhile to create a feature for the Upload Wizard, similar to the Flickr ingesting feature, that permits automatically ingesting videos from YouTube (or other video sites that invite selection of a free licenses). For instance, this video is currently linked prominently on the English Wikipedia article about MOOCs; if it were easy, I'd just transfer it to Commons so it would be displayed inline. (It's possible, of course, that others would disagree with displaying it inline, for notability reasons or something; but it's not desirable to having a technical barrier to importing such videos.) -Pete F (talk) 22:46, 18 March 2014 (UTC)
@Sj: and others, are you aware of wmflabs/videoconvert? (NOTE: Tool is intermittently offline as wmflabs environment changes) I had not known about this. So, it looks like the software has already been installed on Wikimedia servers, all that remains is to find a sensible way to permit users to enable it. My guess is:
  • Whatever team produced this for WMF plans to do so at some point
  • That team is not aware of this discussion
  • That team has fleshed out plans somewhere of how to move forward
I hate playing the role of the pessimist, but I think we've seen this before. It's going to end up being complicated, and a drain on community resources, when it doesn't need to be. -Pete F (talk) 20:48, 25 March 2014 (UTC)

Dear Pete, that tool looks gorgeous. It seems to be maintained by Prolineserver - like most tools, I would venture that it was not produced by or for the WMF. I see no problem with it as presented, and would be delighted to see it gain wider use (simple and welcome, not complex). --SJ+ 03:40, 26 March 2014 (UTC)

@Sj: My comment was clearly out of line. It would have been out of line even if I were right about it being an official WMF initiative. I'm sorry.
I agree, it looks like an excellent tool, and I'm looking forward to giving it a try. -Pete F (talk) 15:08, 26 March 2014 (UTC)
Count me in. I have initiated m:Grants:IdeaLab/Making a Wikimedia-produced feature-length documentary film for theatrical release. Of course, it will also be uploaded to Wikimedia's servers. Cheers! BD2412 T 17:38, 19 August 2014 (UTC)

Just for reference, there was some discussion about how this applies to labs at Commons:Village_pump#Toollabs_facilitating_MP4 (perma-link). Bawolff (talk) 00:30, 11 September 2014 (UTC)