Robert Scott Horning wrote:

Rob Church wrote:

...

On 23/08/06, Michael R. Irwin michael_irwin@verizon.net wrote:

...

I have a query into WikiCommons but I could not find any definitive procedure or policy regarding which file formats are accepted there and which are not.

snip

I wrote a reply to a similar thread on Foundation-l a little over a year ago, and I'd like to point out that video file formats are simply a licensing nightmare. While Ogg Vorbis is the "most free" of common video formats, it is hardly the most widely used. Another "animation" format is MNG, which is a variation of the PNG format, however that format is more for things like a free replacement of animated GIFs.... not what I think you have in mind.

Correct. I can get with a couple of stills in the lesson I am currently working on but a smooth nice animation would really grab the student's attention and show how engineering approximations move analysis from the intractable real world to properly simplied solvable analysis scenario.

I did have a long and hard discussion with the Ogg Vorbis creators back when they got started, where I complained about their licensing issues. They originally had a very pecular propritary license that was mostly free, but something more like the Microsoft "open source" license, that allowed you to look but any changes to the format specs had to be given back to the developers. They have since changed the spec licensing arrangements, and have dropped any royalty requirments.

I would urge a little bit of caution though with the Ogg Vorbis concept as there are apparently some patents that it might violate, and it is within the realm of possibility that this file format might have the same situation happen to it that also happened to the CompuServ GIF file format where a 3rd party can "claim" ownership of all content using this file format. This is one of the huge issues that is facing software developers with the proliferation of stupid patents jaming up the U.S. court system (and soon to be the EU court system as well).

Aha! Good to know. Thanks for valuable information.

Stay away from MPEG file formats if you want them to be free to use for students. The act of even hosting them on a web server can cause huge liabilities, even though there is a "non-commercial use" clause in some MPEG licensing agreements. These are completely incompatable with the GFDL, much as CC-by-SA-NC is also incompatable, and for the exact same reasons. MPEG also sucks the karma out of you if you have to ever deal with it, which is exactly why the motion picture industry is so supportive of the format. I speak from experience here.

Also valuable.

You seriously would not believe the money that has gone into the development of the MPEG file formats and related concepts like DVD-Video.... you would think it was some kind of Manhattan-style military project and other stuff that would just turn your stomach.

Actually I would. In the U.S. big business thrives on a diverse business ecology. I spent a lot of money producing a lot of multimedia content which is now encumbered by obsolete proprietary formats and difficult to use or resuse. A lot of other multimedia developers did the same. As a result of conflicting difficult multimedia formats much of the initial surge into this area has stalled. Eventually the big guys will find it a nonprofitable area (with exception of a couple of key areas like video games on ataria and xbox) and we will be back at square zero defining commonly available standard formats.

Other video file formats are even more propritary, such as Quicktime (owned by Apple), Windows Media (Microsoft), and RealVideo. We simply can't use these.

So just about all other alternatives are really messed up. I think this is one situation where some hard work is going to be needed and that Wikimedia projects, and perhaps Wikiversity in particular can play a huge role in helping to define the multimedia file formats that are going to be widely used. Believe it or not, this is bleeding edge technology and concepts here, where we are discussing "free content" video. It shouldn't be, but it is. I have spent over a decade in researching free (as in GPL, GFDL compatable) multimedia file formats and I am as close to a leading expert on the subject as can be found.

Thank you for the information Robert. I appreciate your sharing the expertise. Makes me feel less silly when google searches come up empty. Perhaps I shall need to establish a "Art of Illusion" portal or user group sooner rather than later. It is a java based open source animation package currently lower scale than 3DSMax but providing a similar type of development environment. Last time I checked it was not yet importing 3DSMax but the lead developer was very interested in assisting anyone who wanted to tackle a plugin for that purpose.

I will do some further research. Eventually this expensive animation I have will be used effectively for some purpose!

Regards, mirwin

Rob Church wrote:

On 23/08/06, Michael R. Irwin michael_irwin@verizon.net wrote:

...

I have a query into WikiCommons but I could not find any definitive procedure or policy regarding which file formats are accepted there and which are not.

All content on the Wikimedia Commons - with the exception of some images that are used as logos for the Wikimedia Foundation, and are therefore copyrighted, trademarked, etc. - needs to be under a free

licence.

Yes. I agree that is an excellent organizing principle we need to treat as bedrock. I do not see any problem with donating or "freeing" avi files under the FDL. While I will still have the proprietrary models used in the Max files to create the animation it is certainly possible to upload avi files into animation and video editing tools such as Adobe Premiere or Art of Illustration?? and modify them in various ways at the frame level.

I believe the Commons people prefer open-licenced formats. I know that most projects don't have proprietary format uploads enabled unless they are internal wikis.

Ok. Fair enough.

...

I did find a procedure citing a free tool to convert some video file formats into ogg vorbis so perhaps that is a current answer. I have no experience with accessing ogg vorbis from the middle of html or media wiki renderer. Does anyone know if it works well with translated files?

At present, MediaWiki fulfils 50% of the name; support for a lot of media files is quite crap, to be frank about it. Static images work fine, and there's recent support for rendering stuff like DjVu files inline and making nice previews, etc. but things like audio and video aren't supported as well as we'd all like.

There is, however, an active effort to improve the situation. One of our two Summer of Code developers is working on embedding and handling media and using free plugins to "make it all work nice", for want of a better term. I'm not quite sure how far progressed it is, but with luck, it won't take forever to complete.

Good to know there is work in progress. Perhaps as Wikiversity develops its portals and projects we will start to attract an active developer segment of the population.

...

Does anyone have strong feelings one way or another regarding what kind of files we should be accepting? I know some at Wikipedia felt strongly a while back that all materials served should be free formats accessible by free tools.

I suggest that we continue to advocate the use of open file formats to allow reuse of the material we publish by as many people as possible.

Ok fair enough. The precedent of the gif fiasco supports sticking to open formats as much as possible. It is the transisition from existing archives that is tricky.

...

For example all of my (mine as in I was the producer and own all rights) animation clips are currently in avi or FLC. If clear directions exist on how to convert these files to a preferred format then I will do so. If it is still vague territory suited best for linux gurus then I will probably forego the pleasure.

To take the techie's first response; tried Googling? I believe it's quite possible to convert AVI files to Ogg Theora format, although I've never done it, having stuck with converting a few sound clips from time to time. I suspect there must be one or two freeware (perhaps not free-licenced, but no-one cares *how*) applications available which will help, and the codecs themselves are, of course, free.

I will give google a shot when I get serious about using some of these pre-existing animation files in lessons. I have not been active with multimedia production for several years so most of my experience is very dated at the moment. At the moment I am checking out preliminary environmental constraints.

Perhaps this thread is a symptom of the bigger issue that we're all (hopefully) thinking, but maybe not ballsy enough to admit; is MediaWiki alone good enough for our purposes? I'm of the mindset that we are going to want expanded features which aren't part of the core code, although I can't think of a good example at this precise second.

Since MediaWiki's primary development audience is Wikimedia projects, having fairly broad changes implemented won't be an issue, and having highly specific things implemented can be done via extensions. The question is, am I right in thinking we *will* want to be able to do X, Y and Z that we can't right now, and if so, what *are* X, Y and Z?

Determining the latter might prove difficult. We don't know quite how someone wishes to learn something until that time comes; various different learning methods suit various different learning materials, teachers and their students.

When the participation grows large enough we should be able to establish requirements defined by existing deficits in the internet environment detailed enough to inspire developers; unless we stagnate into a top down mandated culture specified by "not the Wikimedia Foundation" way or "wiki only welcome here". I anticipate Wikiversity will quickly grow large enough to help put some further severe pressure on interoperability requirements of all online software. People who write free molecular analysis software and interactive nano design editors would obviously like them used productively but file formats can be a severe hurdle. Both development groups are likely to receive feedback from diverse user groups at Wikiversity attempting to work together.

Personally I am tired of enough computing power to run the Apollo program sitting on my desk and being used for little other text editing or web browsing ...... and I am not a world class programmer ..... and while in the past I ran a business that could afford 25K-50K/year for desktop video processing tools and CAD tools ..... I am not currently so situated. No doubt others find themselves in a similar situation. When the participation at Wikiversity is diverse and large some stone soup projects should emerge. Probably often by Wikiversity Portals adopting independent projects or acting as virtual user clubs for more esoteric software tools not of widespread interest to the general public.

When we get a little "free engineering" and other "free expertise" added to free software development and free learning and publishing then I think Wikiversity or the equivalent elsewhere is going wild and wooly places headed in the direction of a "singularity" even if such a event as predicted and hoped for by "singularitarians" never actually happens.

A finite elements package or molecular design package in Java embedded in free online text books to allow widespread free verification of new scientific findings will go a long way to rocking the educational/industrial complex's worldview. Matlab, Cosmo, ProEngineer, Ideas, Autocad, Adobe, 3DSMax, etc. are too expensive and therefore unavailable for most people. What point in spending a decade and billions on proving out and patenting kapton for secret use by the U.S. military in highly overpriced tools of the oldest trade when a couple of thousand teenagers mentored in a free learning environment by proven world class scholars can design better material, receive worldwide assistance in testing or verifying it, and then use proven models to "free" it to the world so it cannot be monopolized by big industrial conglomerates all within the context of an international free learning portal? Suddenly instead of keeping everything secret, at least until it is patented, and then too expensive for widespread use; there will be accelerating trends again towards rapid public development accessible to everybody.

It will start slow but it will integrate to large influences in our society. Obviously a large restoring force exists in that once people realize they are doing valuable work many will wish to drop out and exploit it. Existing society trains us this way. Still once it becomes obvious big useful things can be accomplished by hordes of people collaborating over the net the frequency of actual golden eggs should continue to rise.

User:Roadrunner at Wikiversity has some interesting concepts revolving around nonprofits and grant funding. Basically he supports small tailored non-profits that can help sustain individuals drives towards personal development, credentials, and effective R&D. He sees multitudes of these interacting with and supporting a broader Wikiversity concept than we are initializating today.

Thanks for the information regarding current wikimedia practices concerning file formats.

Regards, mirwin