I was going through the Wikibooks study help desk, and a very surprising offer showed up recently:
http://en.wikibooks.org/wiki/Wikibooks%3AStudy_help_desk#How_to_donate_a_dvd...
"I would like to donate A video taped lecture series on Calculus which have been converted to dvd format. How do I go about doing this? The lecturer is willing to send the dvds to an appropriate uploader, but does not know much about computer programming himself. The lectures were taped at The University of Missouri - Kansas City.
"Please send any information you have to: micheleforan@sbcglobal.net
"Thankyou"
I find this an incredibly generous offer, and something that supports both Wikibooks and Wikiversity in a great many ways, although most media content like this generally is sent to Commons. There are some other technical issues here involved as well, simply due to the sheer quantity of material here that is going to be offered, which is one reason why I'm trying to vet this idea here on Foundation-l.
In addition to trying to figure out just what Wikimedia sister project this ought to be sent to for further assistance, I would like to know what the status is right now for storage of large numbers of video files. I noticed that the [[Commons:Category:Video]] has several video files listed, although almost all of them are generally very short clips. This offer as it currently stands could potentially be several gigabytes of data, which would dwarf the entire video category by itself.
There is also the comparatively minor issue of trying to convert the files from MPEG format to OGG or other patent-free (or compatable with the GFDL) formats, but I think we can find some individuals to help out with that on the various Wikimedia projects. That is not the major issue at the moment.
I guess this is also a political issue in the sense of wheither the WMF wants to get into the business of hosting DVD-video length content on its servers. The costs of doing so I would guess would be rather substantial for bandwidth costs alone, assuming several individuals are all trying to download these huge files, which has been an issue raised in the past. Video files havn't been explicitly restricted from Wikimedia projects, but then again most of the video that is used is comparatively small snippets at the moment. Any thoughts or ideas on this topic are significantly appreciated.
On 1/1/07, Robert Scott Horning robert_horning@netzero.net wrote:
There is also the comparatively minor issue of trying to convert the files from MPEG format to OGG or other patent-free (or compatable with the GFDL) formats, but I think we can find some individuals to help out with that on the various Wikimedia projects. That is not the major issue at the moment.
Conversion is trivial. There are some technical burps related to handling large files (right now max file size on commons is set to ~20megs) which we will need to iron out for other content.
The costs of doing so I would guess would be rather substantial for bandwidth costs alone, assuming several individuals are all trying to download these huge files, which has been an issue raised in the past.
The popularity of the content is a big factor. A short video viewed by 10,000 people will usually use more bandwidth than a long video viewed by ten people. :) I think in the case of this videos, if we found people viewing more than the beginnings it would be evidence that they were actually being used.. and that is a good thing..
Really I think something like this is valuable for far more reasons than the video: if the content were freely licensed it could be transcribed and then converted into wikiversity courseware.
Big files can be uploaded through Eloquence. It is somewhere on the Commons-l mail list.
Bryan
On 1/1/07, Gregory Maxwell gmaxwell@gmail.com wrote:
On 1/1/07, Robert Scott Horning robert_horning@netzero.net wrote:
There is also the comparatively minor issue of trying to convert the files from MPEG format to OGG or other patent-free (or compatable with the GFDL) formats, but I think we can find some individuals to help out with that on the various Wikimedia projects. That is not the major issue at the moment.
Conversion is trivial. There are some technical burps related to handling large files (right now max file size on commons is set to ~20megs) which we will need to iron out for other content.
Bryan Tong Minh wrote:
Big files can be uploaded through Eloquence. It is somewhere on the Commons-l mail list.
Bryan
So, does this mean that uploading and storage of very large files (> 100 MB) is an accepted practice and something done routinely on Commons? That was the whole point of starting this thread was to see if this was being done and if it was something the WMF wanted to get invovled with. Basically, I didn't want to get somebody very, very new to Wikimedia projects going with a whole lot of effort doing the conversion and more, as well as trying to upload the files without checking.
I seem to recall some discussions on Meta that discouraged this sort of media content, especially given the bandwidth and storage costs (there are costs for simply having content sitting on a disk available for download). While fiscal costs related to keeping or deleting a page are trivial, in this situation these files are far from that. I'm sure you could do some specific accounting to determine what it would cost for the WMF to host just as storage on a hard drive about 1GB of data, together with serving 1TB of data as bandwidth (assuming that this 1GB file is downloaded at least 1000 times.... a rather conservative notion given crawling engines alone not to mention interested Wikimedia users). Assuming this is multiple DVDs and several hours of video, it could be substantially more than even this rather modest amount. Using a commercial ISP this would amount to several hundreds of dollars over the course of a year for these files alone.
Owing to Anthere's general epistle (it was somewhat long) to the Wikimedia community, noting especially the comments that the WMF is operating on a shoestring budget, I thought it reasonable to ask this question. This is also a preceedent setting situation that is likely to be copied in the future as well, multiplying the issue several times over.
That all this certainly fits within the "mission" of Wikimedia projects means that it is also something that should be strongly considered, but I thought it also reasonable to vet the issues first.
The technical issues over how to actually accomplish this are in comparison very trivial, and I can muddle through and figure that out. It is nice to know that Erik would be the best contact to get this accomplished, however in a technical sense. Thanks for that bit of info.
On 1/2/07, Robert Scott Horning robert_horning@netzero.net wrote:
I'm sure you could do some specific accounting to determine what it would cost for the WMF to host just as storage on a hard drive about 1GB of data, together with serving 1TB of data as bandwidth (assuming that this 1GB file is downloaded at least 1000 times.... a rather conservative notion given crawling engines alone not to mention interested Wikimedia users).
FWIW, none of the major crawlers download entire gigabyte sized files.
For something that huge though, bittorrent seems like a much smarter choice, especially if others could be convinced to act as seeds.
Anthony
On 02/01/07, Anthony wikilegal@inbox.org wrote:
For something that huge though, bittorrent seems like a much smarter choice, especially if others could be convinced to act as seeds.
We just need to frame it as a form of donation (which it is): "If you can't donate time or money to Wikimedia, please donate a little of your bandwidth by becoming a BitTorrent seed."
HTTP is preferable as BitTorrent has a couple of problems. One problem with using BitTorrent is that end-users will need to download and set up software to receive the files. If seeders drop off, the file will be unavailable for a time. How will we interface our official BitTorrent downloads with our current file description pages?
On 02/01/07, Oldak Quill oldakquill@gmail.com wrote:
HTTP is preferable as BitTorrent has a couple of problems. One problem with using BitTorrent is that end-users will need to download and set up software to receive the files. If seeders drop off, the file will be unavailable for a time. How will we interface our official BitTorrent downloads with our current file description pages?
Put a torrent file on the media page when we have a torrent set up. Keep the HTTP download link, but also offer the torrent file. Anyone who uses BitTorrent much for software will know that the file will often arrive much faster over BitTorrent than over HTTP - I was somewhat surprised when this was the case with OpenOffice.org from a Sun website.
- d.
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
On 02/01/07, Anthony wikilegal@inbox.org wrote:
For something that huge though, bittorrent seems like a much smarter choice, especially if others could be convinced to act as seeds.
We just need to frame it as a form of donation (which it is): "If you can't donate time or money to Wikimedia, please donate a little of your bandwidth by becoming a BitTorrent seed."
HTTP is preferable as BitTorrent has a couple of problems. One problem with using BitTorrent is that end-users will need to download and set up software to receive the files.
Yes, and thus bittorrent is certainly not the solution for most files. But most files aren't a gigabyte in size, either! In terms of cost vs. convenience, I think you've gotta draw a line somewhere, and I'd say at the moment a gigabyte sized file is way over that line.
Of course, personally I prefer bittorrent for any file which takes more than 5 minutes or so to download, if for no other reason than that Firefox doesn't support "resume download".
If seeders drop off, the file will be unavailable for a time.
[[Super-seeding]] is a possibility if the WMF wants to pick up that slack. Of course the best answer is to have seeders always available :).
How will we interface our official BitTorrent downloads with our current file description pages?
With a .torrent file?
Again, I really saw this as the exception for insanely huge files rather than the rule, and was comparing it to the solution of just not hosting the file at all.
Anthony
On 02/01/07, Anthony wikilegal@inbox.org wrote:
[[Super-seeding]] is a possibility if the WMF wants to pick up that slack. Of course the best answer is to have seeders always available :).
Could a Wikimedia server act as a seed for: 1) the sake of archiving: all data is gathered in a single central point 2) the sake of robustness: if all seeds non-official happen to fall off, there will always be at least one seed and an official copy of the file will be preserved?
Is there any mechanism so that a particular seed won't be used if other seeds are available?
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
On 02/01/07, Anthony wikilegal@inbox.org wrote:
[[Super-seeding]] is a possibility if the WMF wants to pick up that slack. Of course the best answer is to have seeders always available :).
Could a Wikimedia server act as a seed for:
- the sake of archiving: all data is gathered in a single central point
- the sake of robustness: if all seeds non-official happen to fall
off, there will always be at least one seed and an official copy of the file will be preserved?
Is there any mechanism so that a particular seed won't be used if other seeds are available?
Yes, it's called [[super-seeding]]. :)
Anthony
On 02/01/07, Anthony wikilegal@inbox.org wrote:
Yes, it's called [[super-seeding]]. :)
Apologies for wasting your time, though super-seeding was something else. ;)
If the files were to be converted to ogg theora (which they should; mpeg2 isn't free, neither in speech nor beer), they would take approximately 500 MB per hour. (rough estimation). So how many dvd's are we talking about?
Bryan
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
On 02/01/07, Anthony wikilegal@inbox.org wrote:
Yes, it's called [[super-seeding]]. :)
Apologies for wasting your time, though super-seeding was something else. ;)
-- Oldak Quill (oldakquill@gmail.com) _______________________________________________ foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
Bryan Tong Minh wrote:
If the files were to be converted to ogg theora (which they should; mpeg2 isn't free, neither in speech nor beer), they would take approximately 500 MB per hour. (rough estimation). So how many dvd's are we talking about?
Bryan
This appears to be the taped lectures of an enitre university-level calculus course. It could be as much as about 200 hours of instruction or more (if it is a full year's worth of lectures). I'm guessing even with some significant compression, this would still be more than a dozen DVDs.
Of course that would be multiple files in theory, if you could have each lecture as a seperate file, but that would be just for convience to get this accomplished.
On the positive side, since it is just a lecture the audio might need to be good, but the video doesn't need to be the absolutely highest quality and could be reduced in size to reduce space/bandwidth.
If there is handwriting on a blackboard or such, image quality needs to be sufficient to read it.
On 1/2/07, Robert Scott Horning robert_horning@netzero.net wrote:
Bryan Tong Minh wrote:
If the files were to be converted to ogg theora (which they should; mpeg2 isn't free, neither in speech nor beer), they would take approximately 500 MB per hour. (rough estimation). So how many dvd's are we talking about?
Bryan
This appears to be the taped lectures of an enitre university-level calculus course. It could be as much as about 200 hours of instruction or more (if it is a full year's worth of lectures). I'm guessing even with some significant compression, this would still be more than a dozen DVDs.
Of course that would be multiple files in theory, if you could have each lecture as a seperate file, but that would be just for convience to get this accomplished.
On the positive side, since it is just a lecture the audio might need to be good, but the video doesn't need to be the absolutely highest quality and could be reduced in size to reduce space/bandwidth.
-- Robert Scott Horning
foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
Anthony wrote:
On 1/2/07, Robert Scott Horning robert_horning@netzero.net wrote:
I'm sure you could do some specific accounting to determine what it would cost for the WMF to host just as storage on a hard drive about 1GB of data, together with serving 1TB of data as bandwidth (assuming that this 1GB file is downloaded at least 1000 times.... a rather conservative notion given crawling engines alone not to mention interested Wikimedia users).
FWIW, none of the major crawlers download entire gigabyte sized files.
For something that huge though, bittorrent seems like a much smarter choice, especially if others could be convinced to act as seeds.
Anthony
Now that might be a very reasonable suggestion, and certainly an alternative to be discussed. There are some problems with bittorrents, viewing from an archival point of view, but it is certainly an option if files this size become very popular. Of course there have been several attempts to try and do a P2P version of Wikipedia as well, which has vetted many of the issues that would relate to bittorrents as well for things like this. Some infrastructure issues still would have to be addressed if this is the route we would like to go for files this large.
And as far as crawlers are concerned, who said anything about "major crawlers" here? Problems from idiots who write their own crawlers without thinking, and crazies who try to mirror Wikimedia projects the hard way can cause enough problems. Besides, are you sure the Internet Archive doesn't download gigabyte sized files right now? They constantly surprise me on what they do, and I've grabbed archived downloads from them in the past that were rather large.
And the curious alone would be a source of some bandwidth.
On 1/2/07, Robert Scott Horning robert_horning@netzero.net wrote:
And as far as crawlers are concerned, who said anything about "major crawlers" here?
I didn't think crawlers other than the major ones were causing files to be downloaded thousands of times. I could very well be wrong.
Problems from idiots who write their own crawlers without thinking, and crazies who try to mirror Wikimedia projects the hard way can cause enough problems.
Is there an easy way to mirror Wikimedia projects? Last I checked downloading the images required using a crawler of some type, but this was a long time ago.
Besides, are you sure the Internet Archive doesn't download gigabyte sized files right now?
No, but that would only be one download, not 1000.
They constantly surprise me on what they do, and I've grabbed archived downloads from them in the past that were rather large.
And the curious alone would be a source of some bandwidth.
Anthony wrote:
On 1/2/07, Robert Scott Horning robert_horning@netzero.net wrote:
I'm sure you could do some specific accounting to determine what it would cost for the WMF to host just as storage on a hard drive about 1GB of data, together with serving 1TB of data as bandwidth (assuming that this 1GB file is downloaded at least 1000 times.... a rather conservative notion given crawling engines alone not to mention interested Wikimedia users).
FWIW, none of the major crawlers download entire gigabyte sized files.
For something that huge though, bittorrent seems like a much smarter choice, especially if others could be convinced to act as seeds.
Bittorrent has some very major problems with firewalls and proxies.
On 02/01/07, Alphax (Wikipedia email) alphasigmax@gmail.com wrote:
Bittorrent has some very major problems with firewalls and proxies.
I suggest BitTorrent be the primary protocal for files over, say, 200 MB but alternative protocals should be presented. Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a BitTorrent link.
Can anyone propose alternate protocols?
On 02/01/07, Oldak Quill oldakquill@gmail.com wrote:
On 02/01/07, Alphax (Wikipedia email) alphasigmax@gmail.com wrote:
Bittorrent has some very major problems with firewalls and proxies.
I suggest BitTorrent be the primary protocal for files over, say, 200 MB but alternative protocals should be presented. Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a BitTorrent link.
That's perfect :-)
Can anyone propose alternate protocols?
HTTP or FTP is fine unless someone has other ideas as well.
- d.
On 03/01/07, David Gerard dgerard@gmail.com wrote:
On 02/01/07, Oldak Quill oldakquill@gmail.com wrote:
I suggest BitTorrent be the primary protocal for files over, say, 200 MB but alternative protocals should be presented. Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a BitTorrent link.
That's perfect :-)
Haha, you discern what I meant. For the record, I inteded to write: "Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a HTTP link.".
On 03/01/07, Oldak Quill oldakquill@gmail.com wrote:
On 03/01/07, David Gerard dgerard@gmail.com wrote:
On 02/01/07, Oldak Quill oldakquill@gmail.com wrote:
I suggest BitTorrent be the primary protocal for files over, say, 200 MB but alternative protocals should be presented. Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a BitTorrent link.
That's perfect :-)
Haha, you discern what I meant. For the record, I inteded to write: "Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a HTTP link.".
Yep. Only offer HTTP if they click on that link. Until we're sure our setup can handle it.
Handling one bit of huge high-quality DVD video would attract more right away. This might be painful in the short run, but is precisely what we really want in the long run.
- d.
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
On 02/01/07, Alphax (Wikipedia email) alphasigmax@gmail.com wrote:
Bittorrent has some very major problems with firewalls and proxies.
I suggest BitTorrent be the primary protocal for files over, say, 200 MB but alternative protocals should be presented. Perhaps on the description page the user could follow a link "Having problems with BitTorrent?" and from there there would be a BitTorrent link.
We will most likely find that straightforward use of bittorrent will scale very poorly for us.
Historically it has been the case that single bittorrent seeding hosts become CPU bound well before they hit 10mbit/sec transfer. This means that BT can only achieve high output rates if you throw a lot of CPU power at it... not a problem when you can count on a lot of seeders, but if seeding our offered files requires a half TB of storage, not many will do it.
Of course, things can be done to help this.. like packaging up BT with tools that allow a central wikimedia page to control the content of the distributed BT agents.
Can anyone propose alternate protocols?
HTTP
Lets save the fancy bandwidth savings gymnastics for when there is a measurable bandwidth issue.
The files at download.wikimedia.org haven't killed us yet, nor will a few DVDs.
On 03/01/07, Gregory Maxwell gmaxwell@gmail.com wrote:
Can anyone propose alternate protocols?
HTTP
Lets save the fancy bandwidth savings gymnastics for when there is a measurable bandwidth issue.
The files at download.wikimedia.org haven't killed us yet, nor will a few DVDs.
I'm sure it won't, just thinking about the future.
Another idea: presenting users with low-quality, highly-compressed (possibly embedded) videos which would hopefully suffice for most users while also providing a link to the full quality video...
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
I'm sure it won't, just thinking about the future.
It will take us all of 30 minutes to setup a BT link when the time comes. Less future more today.
Judging by the number of completely broken audio and video files that I've found left in articles for months with no complaints.... I don't think we're in danger with being over run with requests any time really soon. Even if we are, we can always pull the files down while we figure it out.
It will be a wonderful problem to have when we find ourselves overwealmed with requests. ... It'll be a problem we'll never need to worry about if we don't get the content up.
Content must come first.
Another idea: presenting users with low-quality, highly-compressed (possibly embedded) videos which would hopefully suffice for most users while also providing a link to the full quality video...
This, of course, makes sense... It's also what we do for still images. :)
On 1/2/07, Gregory Maxwell gmaxwell@gmail.com wrote:
On 1/2/07, Oldak Quill oldakquill@gmail.com wrote:
I'm sure it won't, just thinking about the future.
It will take us all of 30 minutes to setup a BT link when the time comes. Less future more today.
I don't see a problem with that strategy, as long as the foundation is on board with it. I already expressed my skepticism that allowing a few of these types of files will make a huge difference in bandwidth costs. So if you can convince the management of the foundation to just go ahead with it, more power to you. My suggestion of using BT was mainly an alternative to telling the person to go somewhere else.
That said, the sooner the BT links get set up, the more people that can be convinced to become seeds ahead of time, before the bandwidth becomes unbearable. In that sense, I see no reason not to offer both, if no one prefers the BT links, then only 30 minutes of time is lost.
It will be a wonderful problem to have when we find ourselves overwealmed with requests. ... It'll be a problem we'll never need to worry about if we don't get the content up.
Content must come first.
Absolutely agreed. My suggestion of BT was precisely meant to get the content up, and to worry about those few people with firewall problems (BT works perfectly fine through most firewalls), what happens if a single seed starts getting hit with 10Mbit/sec in requests (get more seeds), how to encourage people to become seeds for the least seeded files first (like maybe have a list of the least seeded files), etc., first.
In another post you mention download.wikimedia.org. I think those files, at least the most popular of them, would be an excellent choice for *also* offering via bittorrent. Take the 30 minutes, set up a server, wait a year or two, and see what happens. Worst case scenario no one uses it.
Anthony
On 1/1/07, Robert Scott Horning robert_horning@netzero.net wrote:
I guess this is also a political issue in the sense of wheither the WMF wants to get into the business of hosting DVD-video length content on its servers.
None of the files I've uploaded so far have been even near this size. Now, it seems clear to me that the Wikimedia Foundation wants to be and should be a vector for useful educational content coming online. Whether we should host it directly, host it using BitTorrent, or host it in collaboration with an external partner is up to debate. Having a serious offer of useful content helps to force the issue, so I encourage you to take the lead in receiving this content and sending a first evaluation to this very list.
My take is that we should collaborate with the Internet Archive on hosting files above 100 MB, though we should still provide metadata for the files as well.
On 1/2/07, Erik Moeller erik@wikimedia.org wrote: [snip]
My take is that we should collaborate with the Internet Archive on hosting files above 100 MB, though we should still provide metadata for the files as well.
Because of their apparently non-proactive approach with copyright and use of proprietary formats, they may at some point in the future become a highly discouraged external link target on English Wikipedia much like YouTube is today.
I'd like a clear understanding of why handing off part of our hosting to someone else is being advocated. If the simple hosting some large static files is outside of out domain of expertise, what exactly is inside our domain of expertise?
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
Because of their apparently non-proactive approach with copyright and use of proprietary formats, they may at some point in the future become a highly discouraged external link target on English Wikipedia much like YouTube is today.
Of course video files hosted there for Wikimedia would have to be made available in Ogg Theora. Building this bridge would be a good starting point for opening the discussion about proprietary vs. free formats in other contexts, and collaborating on making free formats more widespread.
If the simple hosting some large static files is outside of out domain of expertise
As far as I can tell, it's not a matter of expertise, but of cost. The cost of hosting this single file would be negligible, but if, by doing so, we expand our criteria of inclusion to cover similar video lectures and educational materials, we need to carefully consider if we want to do so alone. I do believe we need to start collaborating with other organizations, especially in the fields of education and archiving, even if they do not follow a blueprint of our ideology -- after all, no good missionary preaches only to the converted.
On 1/2/07, Erik Moeller erik@wikimedia.org wrote:
Of course video files hosted there for Wikimedia would have to be made available in Ogg Theora. Building this bridge would be a good starting point for opening the discussion about proprietary vs. free formats in other contexts, and collaborating on making free formats more widespread.
So long as you're aware of the issues involved, I'm sure we'll have success in our endeavors.
If the simple hosting some large static files is outside of out domain of expertise
As far as I can tell, it's not a matter of expertise, but of cost.
If Internet archive is paying substantially less for bandwidth than we do, then we're doing something wrong.
The cost of hosting this single file would be negligible, but if, by doing so, we expand our criteria of inclusion to cover similar video lectures and educational materials, we need to carefully consider if we want to do so alone.
The same argument could be equally applied to our hosting of images. Image account for something on the order of 40% of our bandwidth usage today. I do not see why we should consider offloading moving images to another organization when we're not doing the same with still images.
I do believe we need to start collaborating with other organizations, especially in the fields of education and archiving, even if they do not follow a blueprint of our ideology -- after all, no good missionary preaches only to the converted.
Absolutely. We should collaborate. But we shouldn't shy away from something that is a core part of our mission. Are we not sure yet if converting a lecture from text to video takes it out of our realm?
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
If Internet archive is paying substantially less for bandwidth than we do, then we're doing something wrong.
It's possible that we are doing something wrong. We should find out. ;-) But aside from whatever they are paying, they have other sources of funding than we do.
The same argument could be equally applied to our hosting of images. Image account for something on the order of 40% of our bandwidth usage today. I do not see why we should consider offloading moving images to another organization when we're not doing the same with still images.
The strength of wiki is not the power to raise enough funds to host large files. In fact, we're still pretty bad at fundraising. The strength of wiki is to collaborate, to _edit_ content and derive new works. The strength of the Internet Archive is to _preserve_ and make available existing works. If we want both our non-profits to evolve towards their natural strengths, it seems logical to me to try to figure out ways for us to do less hosting and more collaborating.
In the case of pictures, hosting them makes collaboration a lot easier when they get changed and overwritten. We need to generate thumbnails, and be sure that they get served quickly when we need them. To start with, we don't necessarily need to be able to stream videos in real time (especially the really big ones), so they are a lot more suitable for being offloaded to others.
We're also not talking about offloading something we're already doing. We do not, to my knowledge, presently host any video file above 100 MB. There is an existing non-profit which hosts thousands such files, many of them falling within the scope of content we typically cover. Should we mirror all those files, just to prove that we can? I don't think so. Should we expand our scope to cover such files? Only if we cannot collaborate.
On 1/2/07, Erik Moeller erik@wikimedia.org wrote:
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
If Internet archive is paying substantially less for bandwidth than we do, then we're doing something wrong.
It's possible that we are doing something wrong. We should find out. ;-) But aside from whatever they are paying, they have other sources of funding than we do.
Still, it seems very premature to shape our decisions around bandwidth concerns which do not yet exist. How much traffic are we talking about? If we don't know, how could the internet archive fairly accept it?
The strength of wiki is not the power to raise enough funds to host large files. In fact, we're still pretty bad at fundraising.
We are bad in terms of viewership vs income. In absolute terms, I think we're not doing so bad.. How many internet/technology projects can you name that bring in $1M USD/yr with pretty much only website panhandling? There are a lot of low hanging fruit, and I have confidence for us in the future.
The strength of wiki is to collaborate, to _edit_ content and derive new works. The strength of the Internet Archive is to _preserve_ and make available existing works. If we want both our non-profits to evolve towards their natural strengths, it seems logical to me to try to figure out ways for us to do less hosting and more collaborating. In the case of pictures, hosting them makes collaboration a lot easier when they get changed and overwritten. We need to generate thumbnails, and be sure that they get served quickly when we need them. To start with, we don't necessarily need to be able to stream videos in real time (especially the really big ones), so they are a lot more suitable for being offloaded to others.
What makes you think video is different?
We have far less video, so we do see less video collaboration.. but we do see video collaboration. Of course, if we prematurely shut of collaboration by forcing video content to a non-collaborative medium (which the internet archive is) we will of course never see the volume and value of what we have missed.
Regarding the importance of downsampling for live streaming: Have you missed google video and youtube? Real time streaming *does* matter if we care about information accessibility.. And the technology to do it (using purely Free Software tools) exists today.
I firmly believe that without realtime streaming there is no bandwidth problem... the amounts of access will be just too low.
We're also not talking about offloading something we're already doing. We do not, to my knowledge, presently host any video file above 100 MB.
We do, however, host many hundreds of files much larger than that.. Some totaling hundreds of gigabytes. (download.wikimedia.org). I believe our own backups would be a better candidate for internet archive offloading than our video content.
There is an existing non-profit which hosts thousands such files, many of them falling within the scope of content we typically cover. Should we mirror all those files, just to prove that we can? I don't think so. Should we expand our scope to cover such files? Only if we cannot collaborate.
Last I checked, many of the videos that we'd consider hosting which that have would not fit our copyright guidelines. A lot of the full length movies for example, are only free in the sense that the movie itself was not renewed. The music, plot, and other elements are still copyrighted. Internet archive gets away with this because who wants to sue a library? We could too, but we've explicitly rejected material which can only be distributed by the good graces or inattentiveness of their copyright holders.
Gregory Maxwell schreef:
On 1/2/07, Erik Moeller erik@wikimedia.org wrote:
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
If Internet archive is paying substantially less for bandwidth than we do, then we're doing something wrong.
It's possible that we are doing something wrong. We should find out. ;-) But aside from whatever they are paying, they have other sources of funding than we do.
Still, it seems very premature to shape our decisions around bandwidth concerns which do not yet exist. How much traffic are we talking about? If we don't know, how could the internet archive fairly accept it?
Hoi, Bandwidth is very much a concern because they cost money. Saying that these concerns do not exist while we are doing a fund-raiser that needs to get us $1.500.000,- is a bit naff. Then again, as I said in the past, Kennisnet has offered to consider hosting video. They have even included the ogg movie formats to their application if I remember well. You may remember, Kennisnet, the organisation that does much of our European traffic for the WMF ??
The strength of wiki is not the power to raise enough funds to host large files. In fact, we're still pretty bad at fundraising.
We are bad in terms of viewership vs income. In absolute terms, I think we're not doing so bad.. How many internet/technology projects can you name that bring in $1M USD/yr with pretty much only website panhandling? There are a lot of low hanging fruit, and I have confidence for us in the future.
I am sure you base this confidence on the fact that our needs grow not quite as exponentially as the growth we are achieving in the provisioning of our services. I am really happy that you recognise a lot of "low hanging fruits" .. this will help as we need more of them as it is.
Thanks, GerardM
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
We have far less video, so we do see less video collaboration.. but we do see video collaboration. Of course, if we prematurely shut of collaboration by forcing video content to a non-collaborative medium (which the internet archive is) we will of course never see the volume and value of what we have missed.
When it comes to videos the GFDL sucks worse than it does for images.
On 03/01/07, geni geniice@gmail.com wrote:
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
We have far less video, so we do see less video collaboration.. but we do see video collaboration. Of course, if we prematurely shut of collaboration by forcing video content to a non-collaborative medium (which the internet archive is) we will of course never see the volume and value of what we have missed.
When it comes to videos the GFDL sucks worse than it does for images.
Mmm. How precisely does one tag a video with the GFDL, and what's the transparent copy?
- d.
On 1/3/07, David Gerard dgerard@gmail.com wrote:
Mmm. How precisely does one tag a video with the GFDL, and what's the transparent copy?
- d.
Taging is fairly trivial you just add the complete lisence in at the end. Ok fairly trivial as long as your video is high res otherwise good luck. The other problem kicks in when you want to start useing other people's stuff since credit becomes an issue. I'm yet to figure out a way to include GFDL and CC material in one video. Generaly I just give up and use stuff in the public domain.
On 1/3/07, geni geniice@gmail.com wrote:
On 1/3/07, David Gerard dgerard@gmail.com wrote:
Mmm. How precisely does one tag a video with the GFDL, and what's the transparent copy?
- d.
Taging is fairly trivial you just add the complete lisence in at the end. Ok fairly trivial as long as your video is high res otherwise good luck. The other problem kicks in when you want to start useing other people's stuff since credit becomes an issue. I'm yet to figure out a way to include GFDL and CC material in one video. Generaly I just give up and use stuff in the public domain.
I would think any format with an open specification would be considered transparent, though admittedly the GFDL only mentions text editors, paint programs, and drawing editors, not video editors.
As for including GFDL and CC material in one video, that's the major problem with all copyleft licenses, not just the GFDL. Of course you can always just take the Wikipedia way out and just mix them together and call the result an aggregate.
Anthony
David Gerard wrote:
When it comes to videos the GFDL sucks worse than it does for images.
Mmm. How precisely does one tag a video with the GFDL, and what's the transparent copy?
- d.
Imagine, for the sake of the discussion here, that you have a production of a TV episode like Star Trek that is available under the GFDL. Assuming that video is available under a GPL-like distribution license, you would be free to take all of the fancy special effects and replace all of the scenes with William Shatner in them with an actor who actually knows what he is doing. :)
Of you could "insert" additional scenes, or rearrange the order of the content in a variety of ways to create something very new. BTW, this was a very popular technique for Russian filmakers during the early part of the 20th Century, where they would get hours of "stock footage" and with just some simple movie editing tools would "compile" new motion pictures. This has also been a fairly common technique for some serial television programs, such as the television series [[JAG]] in some of its early seasons of production.
For a "Wikimedia" version of this as a collaborative tool, I would imagine that something akin to a documentary could be put together in relation to historical footage, making essentially a video version of a Wikipedia article. I think it would be something fun to put together, to be honest, especially if you used the Wikipedia article (FA status, of course) as the "script" for the narration and dialog in the video, with some minor editing.
I'm not exactly sure how you could put a video/audio editing suite into a web page format, but I'm sure some creative minds could come up with something if we really wanted to try. Non-web page software certainly is available free (as in beer and FOSS) to do this sort of work as well.
More reasonable for "transparent" image, I would imagine that a lossless compression technique would be required in this situation, so you could uncompress and recompress the image without generation losses. Of course the same could be said about JPEG images as well in this situation.
On 03/01/07, Robert Scott Horning robert_horning@netzero.net wrote:
Of you could "insert" additional scenes, or rearrange the order of the content in a variety of ways to create something very new. BTW, this was a very popular technique for Russian filmakers during the early part of the 20th Century, where they would get hours of "stock footage" and with just some simple movie editing tools would "compile" new motion pictures. This has also been a fairly common technique for some serial television programs, such as the television series [[JAG]] in some of its early seasons of production.
Roger Corman did the same thing - he burnt a house down with something like twelve cameras around it and used the footage in several films in a row. He complained in his autobiography that he hadn't known then that people would one day get video recorders and compare them ;-)
- d.
Robert Scott Horning wrote:
Imagine, for the sake of the discussion here, that you have a production of a TV episode like Star Trek that is available under the GFDL. Assuming that video is available under a GPL-like distribution license, you would be free to take all of the fancy special effects and replace all of the scenes with William Shatner in them with an actor who actually knows what he is doing. :)
Anybody with the initiative to sell his kidney stones on eBay must have a great sense of humour; this is supported by his incomparable rendition of "Lucy in the Sky with Diamonds.". Some others of us might be happy to get rid of the special effects and keep Shatner in a medium that requires more imagination from the user.
Of you could "insert" additional scenes, or rearrange the order of the content in a variety of ways to create something very new. BTW, this was a very popular technique for Russian filmakers during the early part of the 20th Century, where they would get hours of "stock footage" and with just some simple movie editing tools would "compile" new motion pictures. This has also been a fairly common technique for some serial television programs, such as the television series [[JAG]] in some of its early seasons of production.
We still need to respect the originator's moral rights in the sense of recognizing what he did and what came from him, but the approach you describe seems more consistent with Jefferson's vision of intellectual property than a geriatric mouse that will soon have lived longer than Jefferson himself.
Ec
Ray Saintonge wrote:
Robert Scott Horning wrote:
Imagine, for the sake of the discussion here, that you have a production of a TV episode like Star Trek that is available under the GFDL. Assuming that video is available under a GPL-like distribution license, you would be free to take all of the fancy special effects and replace all of the scenes with William Shatner in them with an actor who actually knows what he is doing. :)
Anybody with the initiative to sell his kidney stones on eBay must have a great sense of humour; this is supported by his incomparable rendition of "Lucy in the Sky with Diamonds.". Some others of us might be happy to get rid of the special effects and keep Shatner in a medium that requires more imagination from the user.
At least with free/open source content you have the option of doing both. And more. I'm really curious what will happen with "Steamboat Willy" once it's copyright expires (if it ever does).
On 1/4/07, Robert Scott Horning robert_horning@netzero.net wrote:
At least with free/open source content you have the option of doing both. And more. I'm really curious what will happen with "Steamboat Willy" once it's copyright expires (if it ever does).
Evidence suggests not much since not much has happened to the movies that have drifted into the public domain.
That's, in part, because the ones that have.drifted into the public domain really haven't: The PD movies are PD because their copyright was not renewed, however most of them contain copyrighted plots, or copyrighted music which were renewed.
Do something really interesting and you can expect rights holders to show up with their hands out.
On 1/4/07, geni geniice@gmail.com wrote:
On 1/4/07, Robert Scott Horning robert_horning@netzero.net wrote:
At least with free/open source content you have the option of doing both. And more. I'm really curious what will happen with "Steamboat Willy" once it's copyright expires (if it ever does).
Evidence suggests not much since not much has happened to the movies that have drifted into the public domain.
-- geni _______________________________________________ foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
On 1/4/07, Gregory Maxwell gmaxwell@gmail.com wrote:
That's, in part, because the ones that have.drifted into the public domain really haven't: The PD movies are PD because their copyright was not renewed, however most of them contain copyrighted plots, or copyrighted music which were renewed.
There are a few around that are PD due to shear age. For example I can't see any copyright claims on A Trip to the Moon that could last past 2008 and that assumes a couple of worst case senarios. I can't see it not being PD in the US in any case. There are other films of a simular age and stuff adapted from pre film systems such as Kinetoscope and Mutoscopes
geni wrote:
Robert Scott Horning wrote:
At least with free/open source content you have the option of doing both. And more. I'm really curious what will happen with "Steamboat Willy" once it's copyright expires (if it ever does).
Evidence suggests not much since not much has happened to the movies that have drifted into the public domain.
But then those movies weren't so interesting to the cats that want to get a hold of _that_ mouse. :-)
Ec
geni wrote:
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
We have far less video, so we do see less video collaboration.. but we do see video collaboration. Of course, if we prematurely shut of collaboration by forcing video content to a non-collaborative medium (which the internet archive is) we will of course never see the volume and value of what we have missed.
When it comes to videos the GFDL sucks worse than it does for images.
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
The general philosophy of the GPL and GFDL is that you are free to copy and redistribute the content, provided that you don't put any additional restrictions on the people that you give the content to. And you must also give the ability to freely redistribute that same content to those who are recieving the stuff. One other basic point is that you are also free to modify and change that content without fear of reprisal from the content originator. Most of the rest of the GFDL and the GPL are fine details to make sure that these basic principles are defined in a strong legal manner.
While I may admit that a license like these may have to be written explicitly for video and music archives to achieve the same philosophical goals, I'm curious about what explicit issues you are worried about in the case of video and why it doesn't apply.
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
For that standard low res we video includeing a copy of the lisence is a serious problem. the isue that the GNU logo itself appears to be under the free art lisence is interesting complication although not one most people will encounter.
I'm not really thinking about semi professional or professional films but well in my case narrated instructional slide shows. So the number of people involved isn't a problem. The problems asside from incudeing the lisence are:
Has to be GFDL only since I can't figure out how the almagimation stuff is meant to apply to videos. Free art are and CC are out (thank goodness for Wildebeests).
Ok I've included the lisence now how long do I have to include it for? 10 sec? 1 min?
Nothing demands that you convert the license to video... That's about the least sane thing to do... :) just provide a single copy of gfdl (in text) with every copy you create of an aggregate work which contains gfdl parts.
The draft of GFDL v2 waves the requirement when the amount of text copied is small, it might be useful to include a nod to video... GFDL isn't a good idea for all forms of video but it does have desirable qualities for educational content, regardless of the medium.
On 1/3/07, geni geniice@gmail.com wrote:
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
For that standard low res we video includeing a copy of the lisence is a serious problem. the isue that the GNU logo itself appears to be under the free art lisence is interesting complication although not one most people will encounter.
I'm not really thinking about semi professional or professional films but well in my case narrated instructional slide shows. So the number of people involved isn't a problem. The problems asside from incudeing the lisence are:
Has to be GFDL only since I can't figure out how the almagimation stuff is meant to apply to videos. Free art are and CC are out (thank goodness for Wildebeests).
Ok I've included the lisence now how long do I have to include it for? 10 sec? 1 min? -- geni _______________________________________________ foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
geni wrote:
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
For that standard low res we video includeing a copy of the lisence is a serious problem. the isue that the GNU logo itself appears to be under the free art lisence is interesting complication although not one most people will encounter.
I'll admit that the requirement for inclusion is a huge issue with the GFDL, and so significant that projects like Wikinews have decided to abandon the license entirely in part because of this issue. For projects like Wikibooks, it is only a minor annoyance that only takes up a couple of extra pages when distributed elsewhere.
It would be nice to have, in a future version of the GFDL, some optional clause that would allow for smaller works to simply provide a network location to a copy of the GFDL, instead of requiring the complete text, or something alone similar lines of thought. Unfortunately, this is the wrong mailing list to try and get those sort of changes implimented.
I'm not really thinking about semi professional or professional films but well in my case narrated instructional slide shows. So the number of people involved isn't a problem. The problems asside from incudeing the lisence are:
Has to be GFDL only since I can't figure out how the almagimation stuff is meant to apply to videos. Free art are and CC are out (thank goodness for Wildebeests).
Ok I've included the lisence now how long do I have to include it for? 10 sec? 1 min?
This is a very good question. In theory, one video frame might be all that is legally required, but that hardly seems to be fitting. Clearly this is one area that the GFDL does fall flat on its face as it is not very clear what is needed here.
Wikinews also accepts nd (I.e. Obnoxiously not free content) images, so I'm not inclined to give any weight to their decisions.
Most of the opposition to gfdl images I see on commons are half truths and simple misinformation.
As mentioned, the GFDL v2 draft does allow some reproduction without the text.
I just don't see why there is so much flak for the gfdls attempt to ensure recipents know their rights vs cc's effort to promote their brand and website (esp when the creative commons site promote licenses which are not free content licenses and the majority of uses of cc licenses are using the non free ones).
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
geni wrote:
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
For that standard low res we video includeing a copy of the lisence is a serious problem. the isue that the GNU logo itself appears to be under the free art lisence is interesting complication although not one most people will encounter.
I'll admit that the requirement for inclusion is a huge issue with the GFDL, and so significant that projects like Wikinews have decided to abandon the license entirely in part because of this issue. For projects like Wikibooks, it is only a minor annoyance that only takes up a couple of extra pages when distributed elsewhere.
It would be nice to have, in a future version of the GFDL, some optional clause that would allow for smaller works to simply provide a network location to a copy of the GFDL, instead of requiring the complete text, or something alone similar lines of thought. Unfortunately, this is the wrong mailing list to try and get those sort of changes implimented.
I'm not really thinking about semi professional or professional films but well in my case narrated instructional slide shows. So the number of people involved isn't a problem. The problems asside from incudeing the lisence are:
Has to be GFDL only since I can't figure out how the almagimation stuff is meant to apply to videos. Free art are and CC are out (thank goodness for Wildebeests).
Ok I've included the lisence now how long do I have to include it for? 10 sec? 1 min?
This is a very good question. In theory, one video frame might be all that is legally required, but that hardly seems to be fitting. Clearly this is one area that the GFDL does fall flat on its face as it is not very clear what is needed here.
-- Robert Scott Horning
foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
Wikinews also accepts nd (I.e. Obnoxiously not free content) images, so I'm not inclined to give any weight to their decisions.
Most of the opposition to gfdl images I see on commons are half truths and simple misinformation.
As mentioned, the GFDL v2 draft does allow some reproduction without the text.
I just don't see why there is so much flak for the gfdls attempt to ensure recipents know their rights vs cc's effort to promote their brand and website (esp when the creative commons site promote licenses which are not free content licenses and the majority of uses of cc licenses are using the non free ones).
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
geni wrote:
On 1/3/07, Robert Scott Horning robert_horning@netzero.net wrote:
How so? I will admit that the GFDL is written primarily for textual manipulation and content, but that is about like saying the GPL is lousy for copyright purposes in a book. Other than the "authorship" clauses being blurred so much for video that it is hard to determine exactly who participated with the development of the video (motion pictures can sometimes involve hundreds of people in the production of one "work"), I'm not exactly sure what the issue is here.
For that standard low res we video includeing a copy of the lisence is a serious problem. the isue that the GNU logo itself appears to be under the free art lisence is interesting complication although not one most people will encounter.
I'll admit that the requirement for inclusion is a huge issue with the GFDL, and so significant that projects like Wikinews have decided to abandon the license entirely in part because of this issue. For projects like Wikibooks, it is only a minor annoyance that only takes up a couple of extra pages when distributed elsewhere.
It would be nice to have, in a future version of the GFDL, some optional clause that would allow for smaller works to simply provide a network location to a copy of the GFDL, instead of requiring the complete text, or something alone similar lines of thought. Unfortunately, this is the wrong mailing list to try and get those sort of changes implimented.
I'm not really thinking about semi professional or professional films but well in my case narrated instructional slide shows. So the number of people involved isn't a problem. The problems asside from incudeing the lisence are:
Has to be GFDL only since I can't figure out how the almagimation stuff is meant to apply to videos. Free art are and CC are out (thank goodness for Wildebeests).
Ok I've included the lisence now how long do I have to include it for? 10 sec? 1 min?
This is a very good question. In theory, one video frame might be all that is legally required, but that hardly seems to be fitting. Clearly this is one area that the GFDL does fall flat on its face as it is not very clear what is needed here.
-- Robert Scott Horning
foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
On 1/3/07, Gregory Maxwell gmaxwell@gmail.com wrote:
Still, it seems very premature to shape our decisions around bandwidth concerns which do not yet exist.
It seems very reasonable to me to shape any policy decision around estimates of its potential impact. In this case, we're not only talking about storage/backup and bandwidth requirements, but also about the question of focus for the WMF. All these depend highly on a number of other variables: - what scope of videos we allow - how broadly we include these videos in Wikimedia projects - whether we want to be able to stream all of these videos, or a subset thereof, in real time.
Since we haven't settled answers for any of these questions, it's hard to predict the impact, but it's easy to predict that under an inclusive policy, with the videos being highly visible and easy to access, we'd be significantly exceeding current storage and bandwidth use quickly, even without reliable real-time streaming. This raises the question whether this is an area the Wikimedia Foundation wants to expand its focus to.
How much traffic are we talking about? If we don't know, how could the internet archive fairly accept it?
Since the Archive appears to be getting significantly more contributions than we do: http://jnana.wikinerds.org/index.php/Form_990-PF_for_Internet_Archive_%28200... .. and since they specialize on archiving, I think it's safe to say that they're in a better position to host these files than we are. That said, we should ask them -- and other potential providers, such as Kennisnet -- whether they'd be willing to help us out.
We have far less video, so we do see less video collaboration..
Video collaboration does not require video hosting in the same place. See Annodex/Metavid.
Last I checked, many of the videos that we'd consider hosting which that have would not fit our copyright guidelines. A lot of the full length movies for example, are only free in the sense that the movie itself was not renewed. The music, plot, and other elements are still copyrighted.
Certainly, and while Wikimedia strives for free content purity, that desire has always been balanced with a healthy dose of pragmatism. A film is a very large aggregate of pictures and sound; achieving purity within such an aggregate gets more difficult the larger it gets. It is still desirable but shouldn't, in my opinion, be the defining exclusionary criterion. In other words, just because a film contains some fair use music at some point, or a company logo, it is not necessarily significantly less worthy of being featured or used within a Wikimedia project.
wikimedia-l@lists.wikimedia.org