If I understand the popular sentiment correctly, it is that: a) the foundation is committed to preserving knowledge it gathers in open formats, and b) knowledge should be presented by foundation projects using exclusively open formats Some people assume that a) and b) are inseparable. I would beg to differ. Let's: - allow uploads only of ogg files, or other open formats, and - allow downloads of these ogg files AS WELL as transcode into flash-presentable formats We say we want to encourage adoption of open formats. That's fine and dandy -- let's only allow open formats in uploads. But a large part of our mission is delivering content and I feel that we've been burying our head in the sand by pretending that there's no problem with access. I would venture to say that the vast majority of our users cannot watch or hear any multimedia files off of our websites. I would also suggest that they are not likely to, in the next several years at best. I suggest that if we change our attitude and allow alternative, actually-workable presentation of our media, we will be more correctly fulfilling our mission. Let's allow Flash for audio and video. -ilya haykinson On 9/28/06, SJ <2.718281828(a)gmail.com> wrote:

Better yet, let's allow uploads of material in any format -- and offer a conversion service, storing the result in a free format. Then allow downloads in that base free format and (at least temporarily) in other formats as well. The current system, under which almost no video is uploaded or downloaded, is surely not the best one... for wikipedia readers or for xiph.org. --SJ On 9/29/06, Ilya Haykinson <haykinson(a)gmail.com> wrote: -- ++SJ

David Gerard wrote: Probably Iran... -- Alphax - http://en.wikipedia.org/wiki/User:Alphax Contributor to Wikipedia, the Free Encyclopedia "We make the internet not suck" - Jimbo Wales Public key: http://en.wikipedia.org/wiki/User:Alphax/OpenPGP

It's fine to me having automated ogg2flash for clients, similar to what we already do providing "adapted" images. I don't really know how youtube and such webs internally handle videos, but i think the most important bit is not if they provide the proper "software" (being flash) but that users choose to play it and it plays. If i choose on commons to play a file, i need to wait until my browser downloads the whole file to get Videolan started. On the other hand, if i ask videolan to open it, it can start playing it almost inmediatly, as the same program is getting and playing. It's the streaming part. A flash ogg player (what we ideally'd need, though probably deserves its own project) must be able to show as it receives, otherwise it will be "too slow". And it is able to do it as it is embedded on the browser. Optionally making the urls with another pseudo-protocol such that the browser passes the url to the target app instead of a downloaded file would also be nice, even if users had to install another app for it. (there's no need for another app, should bug the mozilla guys with this...) Finally, my opinion of such "alternative" and "embedding options" for videos is that it must be up to the user. So we wouldn't embed into our web a java applet nor a flash player (ugh!) to show it, but give them a box "Here there's XXXX.ogg {video|sound} You can [[media:download it]], [[play it on your broser]] (needs xyz plugin) or [[play it with flash]] ^{[[Help:flash|help]]}." The "play it" links would activate javascript in order to convert the box into the proper code to insert it (yes, we're adding another requeriment: javascript but i really doubt un-javascriptted users being able to use an activex flash plugin, they're fine with the download link). Our users are smart enough to try each of it until find one suitable for their system. And they can even install them (most of them probably won't care on what plugin they install, the internet is full of sites installing crap...). But from the smart user, we give the user the options; our web is fast-to-download (extra stuff is on demand), and we don't ask for extra plugins (how's that plugins are required to see wikipedia?? / what has launched the java VM??). Note 1: Yes, we still need the 'magic solution'. Note 2: I have recently found some ogg audio files made with audacity that not all ogg players correctly played (none or part of it). Any comment?

[snip] It's funny... I added a java (GCJ compatible) audio player on enwiki a while back and I've been mostly unable to get people to even comment it (although it has been used by tens of thousands)... It makes me sad to see this thread where so many people are willing to expend energy talking about how we need to compromise our commitment to free content in free formats when I can't even get people to help out with simple free solutions which don't require much compromise. I guess it's just another example of the meta-pedian illness of talk over action, and that I probably should just ignore this thread... But I do want to point out, for the record: 1) The majority of video formats used on the web require some degree of installation, although some of them can be autodownloaded into Windows Media Player or use a popular client which many users installed long ago (Real, for example). Our instructions (on enwiki at least) for installing video support are effectively a single step, and if there is room for improvement... {{sofixit}} 2) I'm not aware of any computer make thats ships with flash installed. So all these flash video sites do require an install.. although it's one install providing for many forms of webcruft. :) 3) There are free software java codecs for both Ogg/Theora and Ogg/Vorbis. I haven't worked too much with the Theora plugin because it was a bit buggy under GCJ six months ago, but the java vorbis codec is very mature and robust. We couldn't ask for a better framework from which to build a video player (and there is even a SOC project to deeply integrate this support with MediaWiki).

On 9/30/06, Ilya Haykinson <haykinson(a)gmail.com> wrote: Why don't we just offer something that is both easy to use and which does not lock the content in encumbered formats? Ease of use is something that everyone should have, not just people who can afford commercial propritary systems. The effort spend arguing that we should use propritary software X could be better applied into finding a better solution that doesn't require such a compromise.

On 9/30/06, Walter Vermeir <walter(a)wikipedia.be> wrote: [snip] It's already setup on enwiki, linking to a modified copy on toolserver. I set it up weeks ago and have been able to gather fairly little interest in it... thus my frustration with people insisting that we must use proprietary formats for ease of use. Hours to talk abour unfree tools but not a second to build something real...

David Gerard wrote: The Vietnamese-language projects currently link to JOrbisPlayer on all its OGG description pages. [1] We use a hack in order to make this work: we have [[MediaWiki:Fileinfo]] call [[MediaWiki:Playfile]], passing the MIME type as a parameter. If anyone creates another OGG player on the Wikimedia Toolserver, we'd be happy to link to that as well. [1] See, for example, the JOrbisPlayer link at: http://vi.wikipedia.org/wiki/H%C3%ACnh:SCA_approach.ogg -- Minh Nguyen <mxn(a)zoomtown.com> [[en:User:Mxn]] [[vi:User:Mxn]] [[m:User:Mxn]] AIM: trycom2000; Jabber: mxn(a)myjabber.net; Blog: http://mxn.f2o.org/

So at last it is here? Congratulations! However... *It doesn't work for me (start & stop buttons do nothing). *I can't find how the applet gets the file to play. I guess the above problem has to do with the applet may need to get passed the file location on jorbis.player.play.0 instead of http://tools.wikimedia.de/media/wikipedia//d/d4/ *The link to http://en.wikipedia.org/wiki/Media_help. is not very useful, it redirects to [[Wikipedia:Media_help]] where there's no reference to it. Neither from [[Wikipedia:Media]] *The page (/~gmaxwell/jorbis/JOrbisPlayer.php as per Minh Nguyen msg) shows too Beta. Not like a finished and usable tool (well, it seems it isn't). Thanks fo the reference. Java is quite the same as the Flash solutions.

On 9/29/06, Artur Fijałkowski <wiki.warx(a)gmail.com> wrote: While I'm sure you have the best of intentions, when given the option between "not using free tools" and "being forced to use it" most people will opt for the former. Enforcing a de-facto "only computer geeks are welcome" sort of rule is not going to really help the long-term interest of our projects. I really love MediaWiki's support of SVG -- you can upload and download an open format, but it always renders it to be viewed in any old browser. I think we should take this as something of a model -- the goal should be to combine maximum usability with maximum freedom, and as has been pointed out these are not and should not be incompatible. We shouldn't have to "force anyone" to learn anything to use our resources, if freedom is really a goal. FF

There's already code in branch from one of our Summer of Code projects for this, using the Java player from Fluendo and optionally also a Firefox extension / IE plugin based on the VLC player. If you'd like to help Michael continue work on this, great. -- brion vibber (brion @ pobox.com)

On 10/1/06, Fastfission <fastfission(a)gmail.com> wrote: On which step of the Media help page (http://en.wikipedia.org/wiki/Wikipedia:Media_help which should be linked next to almost every mediafile on enwiki) did you encounter problems? I look forward to when media in general works without installation for most users.. but that isn't the case for Video at least: Flash, Realvideo, Quicktime, etc.. all require an installation on most PCs out there.. And in most of those cases the installation is more complex than following our media help. Even with WMV a huge amount of the video out there is encoded with codes which Windows does not naively include... although Windows Media does have an autoinstlation facility which makes this painless for most users. Our content would not be free if we were distributing it inside proprietary and patented formats. Fortunately free and easy to use are not at all mutually exclusive. ... although carrying out these silly arguments and getting actual word done are mutually exclusive .. so...

On 10/1/06, peter green <plugwash(a)p10link.net> wrote: I'd rather not. 1) The time it would take to code the downsampling and switching could be better spent on other parts of the code. 2) As a contributor I would be upset that users are being given a poor quality version of my contribution rather than being asked to install improved software. Or put another way, *I* won't be working on that. :) You can if you really think it would be useful. I've also found that about 18% of the people who try the play in browser link don't have any form of Java installed. I think it's clear that it isn't a magic bullet in any case. One of the nice things about the Michael Dale's SOC project is that it will magically use the best method available, only falling back to java if there is no VLC plugin available. This way we can ask users to perform the installation who can (which is the vast majority of the users) but still provide java for some of the ones who can't.

On 10/2/06, Brion Vibber <brion(a)pobox.com> wrote: Last I checked it worked without sound in javasound less VMs. I don't personally have access to anything with that VM I don't think so. I've been aggravated, and I think rightfully so, when someone complained about an image I spent hours getting right being low quality when it really was just an artifact of our image handling. This happens all the time due JPEG artifacts, which are often visible even on the image page depending on the users's display... or the image appearing being blurry because straight filtered downsampling can be unflattering. It is perfectly reasonable to adapt to the users environment to an extent, but it can be carried too far: For example, we could 'help' uses with browsers without image support by converting images to text using [[AAlib]]. I think t we'd be helping users more by spending our efforts on: 1) Having a really good set of instructions for one click installation of support software. 2) Encouraging browser vendors to ship support with their browsers. (The reference Theora and Vorbis codecs are BSD licensed, so there shouldn't be too much of a license-political issue there). ... instead of dishing out a hacked up player that will have a much lower audio quality on their system than the flash based sites. Some food for thought: Latest stats (from the last days logs) from the toolserver java audio player (linked from enwiki and other places): 27,828 unique IPs loaded the PHP page 23,327 unique IPs loaded the JAR 16% have no java enabled. 17,789 unique IPs loaded an Ogg 19% have Java but it failed Overall it worked for about 64% of the IPs. As far as I know, lack of javasound API (i.e. old MS JVM) is the only significant cause of loaded but failed right now. These numbers are a bit different from when I first introduced the player.. Perhaps client side caching is influencing the results. ... It also could be because folks without Java have either installed it, or they have learned not to try the player. Hm. The above numbers to make a greater argument for extending support to uses with the old JVM.

"Gregory Maxwell" &lt;gmaxwell(a)gmail.com&gt; escribió en el mensaje news:e692861c0610011437ydb0f8f1yd39e62e79a988fdc@mail.gmail.com... I get that *exactly location* no file appended to it. I went to http://vi.wikipedia.org/wiki/Hình:SCA_approach.ogg as per Minh Nguyen mail, there i followed teh link so i arrived to http://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.php?play=http://com… Page source is: <applet code="JOrbisPlayer.class" codebase="http://tools.wikimedia.de/~gmaxwell/jorbis/" archive="JOrbisPlayer.jar" width="600" height="35" align="baseline"> <PARAM NAME="jorbis.player.play.0" VALUE="http://tools.wikimedia.de/media/wikipedia//d/d4/"> <PARAM NAME="jorbis.player.icestats" VALUE="no"> <PARAM NAME="jorbis.player.playonstartup" VALUE="yes"> </applet>Then i tried url http://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.php?play=http://upl… day's featured multimedia)Page source has:<applet code="JOrbisPlayer.class" codebase="http://tools.wikimedia.de/~gmaxwell/jorbis/" archive="JOrbisPlayer.jar" width="600" height="35" align="baseline"> <PARAM NAME="jorbis.player.play.0" VALUE="http://tools.wikimedia.de/media/wikipedia//d/d4/"> <PARAM NAME="jorbis.player.icestats" VALUE="no"> <PARAM NAME="jorbis.player.playonstartup" VALUE="yes"> </applet>...and no soundhttp://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.php?play=http%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2Fb%2Fb5%2FMonteverdi_-_non_si_levav%2527ancor.oggproduces equal output and silence.If i call simply the .php http://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.phpit is<applet code="JOrbisPlayer.class" codebase="http://tools.wikimedia.de/~gmaxwell/jorbis/" archive="JOrbisPlayer.jar" width="600" height="35" align="baseline"> <PARAM NAME="jorbis.player.play.0" VALUE="http://tools.wikimedia.de/media/wikipedia//d/d4/"> <PARAM NAME="jorbis.player.icestats" VALUE="no"> <PARAM NAME="jorbis.player.playonstartup" VALUE="yes"> </applet>i.e. the ?play= parameter is being clearly ignored. I tried to check if it was indeed the correct parameter, but as i said, there's no doc reference to it.There isn't any a text on the java console.It is not a Java problem as on the jcraft page the same works after two warnings: 'Application will run without usual security restrictions' and 'certificate has expired'. http://www.jcraft.com/jorbis/player/JOrbisPlayer.php?play=http%3A%2F%2Fuplo… works fine (not www.jcraft.com/jorbis/player/JOrbisPlayer.php?play=http://upload.wikimedia.… ) As jcraft works, it's not it. I'm not using msjvm neither. (......) Well, it's a better aproach, but there's nowhere saying "You can point users to this, it works like that". I was expecting docs about the java, not about how to install an ogg player. Seems i misunderstood. I find references to "continually improved", "if it doesn't work try in a few days" a bit frustrating. But hell, this may be me ;-) How did they do? :-) I am using Versión 1.5.0 (build 1.5.0_06-b05) by Sun. Setting the console output to 5 (all) it seems the error may came because your applet is unsigned and thus JavaSound is directly denied? I paste you the output (spanish output): basic: Modalidad de receptor registrada basic: Referencia a cargador de clases: sun.plugin.ClassLoaderInfo@1dfafd1, contador de referencia=1 basic: Receptor de progreso agregado: sun.plugin.util.GrayBoxPainter@998b08 basic: Cargando miniaplicación... basic: Inicializando miniaplicación... basic: Iniciando miniaplicación... network: Conectando http://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.jar con proxy=DIRECT basic: Carga de http://tools.wikimedia.de/~gmaxwell/jorbis/JOrbisPlayer.jar desde la antememoria basic: No certificate info, this is unsigned JAR file. I was waiting to catch you on irc, easier to explain and feedback, instead of filling a long bug on the ml, but well... here you have it. PS: I was going to propose filling a bug into hemlock's bugzilla but it seems it disappeared / is unreachable.

"Gregory Maxwell" wrote: To have the vi folks know? So i can know they are calling it badly? ;-) Well, as i don't go too much by en: i didn't know. BTW: Do you know that if you follow (on the template which link to your free and gnu-compilable sound tool) the [[beta]] link referring to it you get a Vista screenshot? Thanks. This sounds² much better. :-) Bugging you again: If i stop a music while it is being played, the sound which was hearing keeps vibrating... i needed to kill java (i.e. the browser). Seems like sending "Start note" and -as stopping it- missing "Stop note" Platonides ² On both senses :D

On 13/10/06, Jason Spiro &lt;jasonspiro4+news(a)gmail.com&gt; wrote: By, through a global popular revolution, stripping Microsoft of their intellectual property rights and, upon copylefting Windows Media Player, making it compatible with open and free formats. -- Oldak Quill (oldakquill(a)gmail.com)

Cortado (http://www.flumotion.net/cortado/) is a GPLd implementation of Vorbis/Theora in Java. It would be useful to have this on all .ogg pages. Ed. Gregory Maxwell wrote: ...