On Thu, 17 Jun 2004 23:52, PrussianSnow wrote:> >Paul E Wrote > > [snip] > > >Windows 2000 (although it's the same for all of them) doesn't read in > >any mime-type to correctly identify a file, it just uses the extension. > > [snip] > > What if, at least for Windows, someone wrote a simple application which > when associated with .ogg files and an .ogg file was opened simply read > enough of the ogg stream to identify the type of content and then launched > an appropriate (and selectable) application?You'd have to associate media software to handle the particular codec types by hand, rather than just choose whether to retain current mappings or clobber them with some app at install time. It would suck. People would whine about it. And it's some app someone would have to write and maintain - the code itself would be straight forward, but the media association gui design, installer, source code and release management, website and all the other trappings of a software project would end up being a millstone around someone's neck. And you'd have to have one for gnome, kde, mozilla... Or, in a pinch, we could just call the video ones .ogv or .ogm or something. John
<200406180031.18792.jwm@eslnz.co.nz> Message-ID: <3143C89E-C061-11D8-8350-00039317863E@suespammers.org> On Jun 17, 2004, at 08:31, John Morton wrote:> Or, in a pinch, we could just call the video ones .ogv or .ogm or > something..ogm is already used for videos, btw. I think it normally is Ogg audio with AVI video of some sort.
From: "PrussianSnow" <1edm@qlink.queensu.ca>> >Windows 2000 (although it's the same for all of them) doesn't read in > >any mime-type to correctly identify a file, it just uses the extension. > > What if, at least for Windows, someone wrote a simple application whichwhen> associated with .ogg files and an .ogg file was opened simply read enoughof> the ogg stream to identify the type of content and then launched an > appropriate (and selectable) application?Windows ME and XP do that for .AVI files. So it can show video properties (rather than launching seperate programs.) And when it encounters a broken one, it can take a minute or two for it to finally decide that it doesn't recognise it. During that time, you can't delete the file, you can't rename it, you can't do anything with it. All because Windows actually looks inside the file for properties instead of just accepting the file extension. This is considered to be an extremely annoying bug with people who work with video. The point is, problems like that can easily occur in routines that examine files & broken files. And the OS is definetly not the place for those kinds of bugs. The simpler the better.> This would give users their expected operation at least on launch. ItI don't think most users expect a single extension to launch different programs. I don't think any user expects that. One extension always launches one specific program.
<007301c45480$89aa1a30$86389c3f@computername> Message-ID: <40D1C6CA.70600@sympatico.ca> noprivacy@earthlink.net wrote:> .oga for audio only and .ogv for audio and video? Close enough to'.ogg' to> be recognised. > > But still, it might be nice to have some extension that is somewhat > descriptive of Theora video. .OGT perhaps?Using an extension named after a specific codec seems like a bad idea for a generic container format. There's no reason to believe that Theora will be the only video codec to use an ogg container, just as Vorbis is not the only audio codec that uses ogg. -- Jon Doda
<40D1C6CA.70600@sympatico.ca> Message-ID: <001001c45489$680a3a50$79649c3f@computername> From: "Jon Doda" <jdoda@sympatico.ca>> noprivacy@earthlink.net wrote: > > But still, it might be nice to have some extension that is somewhat > > descriptive of Theora video. .OGT perhaps? > > Using an extension named after a specific codec seems like a bad idea > for a generic container format. There's no reason to believe that > Theora will be the only video codec to use an ogg container, just as > Vorbis is not the only audio codec that uses ogg.Yes, I know. I was a bit hesitant to even suggest that, but I was thinking that: 1) it might be a good idea to get it into the minds of users that this isn't DivX or such with a funny extension. 2) As a way to boost recognition (and awareness) of an open source video codec called Theora. Except for those two reasons, I agree with you that it might be a bad idea.
<40D1C6CA.70600@sympatico.ca> Message-ID: <001c01c49ce0$c39425d0$a720a8c0@NESCAFE>> Using an extension named after a specific codec seems like a bad idea > for a generic container format. There's no reason to believe that > Theora will be the only video codec to use an ogg container, just as > Vorbis is not the only audio codec that uses ogg.I think you're missing the point. It ain't a codec problem, it ain't a container problem, it's a media problem. It's a problem about associating a specifie extension for all files including video, and another for all audio only files. A different extension means for us, Windows-explorer-like people, a different icon making a simple and quick distinction between file according to their content, a different player association for audio (often light and minimised) and video (with a lot of tweaking options), no matter what codec is embedded inside. No matter if it's a hacked .ogm with XviD inside, or a brand new Theora encoded video, we like our movies showing up with a movie-like icon, opening with a video capable player. By the way, why does it bother so much some of you people ? If you've got a "magic scanning program that finds out which data is in your files" or if you've got a "magic mime-like type attached", why would you care for different files extensions ? You won't make Windows user switch to *nix, and you won't push Ogg based media any further either. English not being my mother tongue, forgive my rudeness and approximative grammer.
<001001c45489$680a3a50$79649c3f@computername> <40D1DEBA.5090905@ellisfoundation.com> Message-ID: <000401c454a9$355dea50$05649c3f@computername> From: "Paul Ellis" <public@ellisfoundation.com>> I think the extensions need to specify the TYPE of media that it is, not > the specific content. Do most users care if the files is Theora or Divx > or Xvid? Not really.No, not really. As I said, I was hesitant to even suggest that. It's just that due to the OGM extension being Divx & Ogg, people may start associating video with Theora as DivX. Thinking that maybe Theora is just a new version of Divx, like XVid, or such. To most people, there are only a few types of video. Microsoft. Quicktime MOV. Real .RM Mpeg-1. And DivX. I was just thinking out loud that it might be helpful if people realized that this is a new video format and not an old one with a funny extension. Something they would automatically associate Theora. More of publicity and recognition.> video, even thought they can have different codecs inside them. Same > with .ogm too, it doesn't have to be Divx and Vorbis, it can be any AVI > compatible video codec I think.Right. OGM is just the ogg container with a different extension. But it's been done as only DivX now for quite a while. I think it's a little late to try and put Theora in there in people's mind.> the most sense to do .oga .ogv (or something like that), but > unfortunately .ogg has already caught on quite strong for audio. So > perhaps it should stay audio, and .ogv should be adopted for video. ItYeah. That is a bit of a problem. And I do agree that it's probably too late to change .ogg to mean anything but audio. And that there isn't really any need to do so. Still, it can be done. Encoder apps can automatically suggest using the .oga extension for Vorbis audio. People are getting used to the idea of .MP4 being the same way, with the result .m4a being audio. The same could be done with .ogg But there isn't any real need to do so.> is the simplest solution, nobody would have to maintain special > programs, it would be compatible with any OS (not just XP, Gnome, & KDE) > which is what open standards are about, the files could easily be sortedYup. It would be the best solution. Apps just get programmed for that extra extension. And since there aren't any Theora video player or editing apps yet, it's really no big deal. When they do end up being written, the authors just add .OGV or such as the preferred extension. Smart people will know that they are the same exact format. And that in reality the .ogg extension is the official prefered generic extension. And that it's really a container format that can hold any codec. But users wont. Ogg will be Vorbis (and only vorbis) audio and .ogv etc. will be ogg video.
<001c01c49ce0$c39425d0$a720a8c0@NESCAFE> <40D1EF69.4030405@ellisfoundation.com> Message-ID: <001601c49ced$002f5260$a720a8c0@NESCAFE> From: "Paul Ellis" <public@ellisfoundation.com> To: "Nescafe" <nescafe@fr.fm>> If I understand you correctly you wonder why some of us would want the > extension changed even if there was a program to determine the file's > contents right ?That was the exact opposite. I was wondering why those non-Windows user who claim not to rely on file extension would care about Xiph promoting different extensions for Windows-like file browsing system users...> Right now I could send an .ogg file to many people and they could > double click it and Winamp would play it because it grabs that > extension. But we'd have a problem with Theora. The person would > have to install another program and then go in and configure it. > It sounds simple, but that would break most people.And all Ogg files would look the same with the same damn icon whatever their content. To those that know Tortoise, it's a CVS client for Windows that manages, through some still mysterious to me API, to change a file's icon according to its status. I know it is possible, but that would simply be overkill. My point of view is : - Windows users like me need some standard Xiph supported extension to make a difference between an Ogg containing audio only or a video. - The codec used inside absolutely doesn't matter. I am talking about media type. - Other OSes / file explorers users with stronger file identification modules do not care about several different file extensions. Except maybe for the fact they don't like Windows.
<001c01c49ce0$c39425d0$a720a8c0@NESCAFE> <40D1EF69.4030405@ellisfoundation.com> <001601c49ced$002f5260$a720a8c0@NESCAFE> Message-ID: <20040617194522.GA17559@xiph.org> On Fri, Sep 17, 2004 at 09:31:59PM +0200, Nescafe wrote:> > And all Ogg files would look the same with the same damn icon whatever their > content. To those that know Tortoise, it's a CVS client for Windows that > manages, through some still mysterious to me API, to change a file's icon > according to its status. I know it is possible, but that would simply be > overkill.And who draws the line between "audio" and "audio+subtitles"? What about "audio of a brand new codec which I cant decode"? What if your favorite media player only supports Vorbis, but you have a Ogg FLAC file which is supported by a media player which you like less, and want it to open in a media player which you most like? The problem cannot be solved in the extension. Your OS is the problem, or prehaps just your file manager. If you /want/ different icons and different programs to open files depending on the file's _content_, not just it's _type_, then you need something more advanced to fill this niche. Ogg is the file format. Wanting different extensions depending on it's content is like wanting different text file extensions depending on if the content is personal or work related, etc.
<007301c45480$89aa1a30$86389c3f@computername> <20040617173110.GA19750@derobert.net> Message-ID: <000501c454a9$366dfac0$05649c3f@computername> From: "Anthony DeRobertis" <anthony@derobert.net>> On Thu, Jun 17, 2004 at 10:26:52AM -0500, noprivacy@earthlink.net wrote: > > > If so, then that's pretty well established and it might be better to use > > a > > different extension indicate a Xiph version. Just to seperate Theora > > from > > being Divx. > > Why? ogm is video in an ogg container format. No reason why there should > be a different extension based on codec.This has already been answered repeatedly, but you may have missed those messages. Or maybe I accidently sent a few directly to the people, instead of to the list (since that requires explitly adding the vorbis@xiph.org address.) 1) ogm is already pretty well established as being DivX and Vorbis. Trying to put Theora in there in people's mind probably wouldn't work too well. 2) the Theora specific extension would be more promotion etc. Something that regular users would automatically associate with, rather than some generic container that might contain no telling what and require no telling what codec to play. That's one of the problems that effect the .AVI extension. If you do video from the web or p2p stuff, you gotta have a couple dozen codecs installed because there's no telling what codec might be used. I've downloaded video clips off p2p stuff. Tv bloopers off live tv, or a commercial that was only seen in a specific countery, etc. A lot of times it wont play and wont tell what codec is needed. You gotta use some other program to examine the video and see what codecs were used, and then go search the web to see what it is and where to get it, and hope that it's still available for XP and not just Win9x, etc. And Linux users are in worse shape, of course. It's a mess, all because a generic container format can hold so many different types of codecs. Generic containers have their good points, but they have bad points too. Using some codec specific extension would mean the regular user would be able to recognise it and know what codec is needed to play it. Not very future-proof, no. But it would help the average user. Still, it could be done similarly if Xiph basically says ".ogv is the Xiph only codec video extension and .ogm is the third-party codec video extension." That way if the user sees .ogv, they know to go one specific place to get all the codecs (as a single package?) that would be used in a file with that extension. If it's .ogm, then they have to search the web, just like they do with .avi files.
<20040617194522.GA17559@xiph.org> Message-ID: <000701c49cf0$95067910$a720a8c0@NESCAFE>> And who draws the line between "audio" and "audio+subtitles"?The file extension.> What about "audio of a brand new codec which I cant decode"? > What if your favorite media player only supports Vorbis, but you > have a Ogg FLAC file which is supported by a media player which > you like less, and want it to open in a media player which you > most like?The problem is not really about which player will manage to open your file. It's about quickly identifying which one is audio, which one is video, according on the oldest and simplest file-system metadata I heard about : File extension.> The problem cannot be solved in the extension. Your OS is the problem, > or prehaps just your file manager. If you /want/ different icons and > different programs to open files depending on the file's _content_, not > just it's _type_, then you need something more advanced to fill thisniche. Or just a simple hack which has worked for years and years. File extension is simply a guess. Of course a JPG file can be named with a AVI extension, but it's a first clue that is almost always enough. It enables the use of simple file managers like Windows', and doesn't break the more advanced ones. I see nothing against that.> Ogg is the file format. Wanting different extensions depending on it's > content is like wanting different text file extensions depending on if > the content is personal or work related, etc.The birth of the OGM extension speaks for itself. There was a need, it was filled. I just hope Xiph will officially support this *kind* of simple tricks, if not *that* one. M1V, MPG, M2V, MP1, MP2, MP3 do it well. Maybe these extensions aren't self-explanatory enough, but they are quite enough for millions of people to distinguish between audio, video and muxed.
<001c01c49ce0$c39425d0$a720a8c0@NESCAFE> <40D1EF69.4030405@ellisfoundation.com> <001601c49ced$002f5260$a720a8c0@NESCAFE> <20040617194522.GA17559@xiph.org> Message-ID: <20040617200205.GA25434@reric.net> On Thu, Jun 17, 2004 at 12:45:22PM -0700, Arc Riley wrote:> And who draws the line between "audio" and "audio+subtitles"? What > about "audio of a brand new codec which I cant decode"? What if your > favorite media player only supports Vorbis, but you have a Ogg FLAC file > which is supported by a media player which you like less, and want it to > open in a media player which you most like? > > The problem cannot be solved in the extension. Your OS is the problem, > or prehaps just your file manager. If you /want/ different icons and > different programs to open files depending on the file's _content_, not > just it's _type_, then you need something more advanced to fill this niche. > > Ogg is the file format. Wanting different extensions depending on it's > content is like wanting different text file extensions depending on if > the content is personal or work related, etc.Yet the unfortunate reality is that 95% of end users will get irritated that there's no easy way to get their audio file to load in their audio player, and their video file to load in their video player. Then they'll go use somebody else's codec who's not so hard-headed. You imagine a world where some vast number of users have all variety of unsolveable problems, despite the fact that there's mostly one really, really big problem, and it's very solveable. It's technically unappealing, but serious enough that it could hinder adoption of every Xiph codec because it won't "just work" for many end users. It's not like having two or three common filename extensions makes the situation any worse than it is now. If you have to build a "switcher" app it can just as easily handle all three. And if the day ever comes where OSes will be smarter about this, then wonderful, things will work even better. It it smart for Xiph to risk everything on forcing OS developers to fix the file-extension problem right now? Bikeshed. Sorry.
<000701c49cf0$95067910$a720a8c0@NESCAFE> Message-ID: <20040617202339.GB17559@xiph.org> On Fri, Sep 17, 2004 at 09:57:37PM +0200, Nescafe wrote:> > And who draws the line between "audio" and "audio+subtitles"? > > The file extension.<SARCASM> Great idea! Ok, here's a proposal: .oggvorbis # Just Vorbis .oggwrit # Just Subtitles .oggvorbiswrit # Vorbis w/ Subtitles .oggflac # Just FLAC .oggflacwrit # FLAC w/ Subtitles .oggspeex # Just Speex .oggspeexwrit # Speex w/ Subtitles .oggtheora # Just Theora (no audio) .oggtheorawrit # Theora w/ Subtitles (no audio) .oggtheoravorbis # Theora w/ Vorbis .oggtheoravorbiswrit # Theora w/ Vorbis and Subtitles .oggtheoraflac # Theora w/ FLAC ... You get the idea. This goes on for several hundred more as we add MNG subtitles, which are overlays not textual subtitles, and double the size of this list everytime you add an additional codec. I should add: .oggvorbisflac # Vorbis and FLAC chained together .oggvorbisspeex # Vorbis and Speex chained together etc... Since it's completely legal to have an Ogg file with multiple bitstreams chained end-to-end of different codecs. Ah, I can hear you now. "Several thousand different extensions is just too much to handle! We just need to lump these into TYPES, not CODECS" Ok. .oggaudio .oggtext .oggoverlay .oggvideo .oggaudiotext .oggaudiooverlay .oggaudiovideo .oggtext .oggtextoverlay .oggtextvideo .oggoverlayvideo This list doubles everytime a new TYPE of codec is created... uh-o! What about MIDI? What catagory is MIDI in? It's kinda audio, but it's scripted audio! Is it a script type? Or is it something specific on it's own? After all it doesn't output PCM, it outputs notes, like the difference between textual and graphical subtitles. What about if you want different icons for when it's midi or vorbis? Or Vorbis and Speex? Better go back to the thousands of different extensions, as above... </SARCASM> No. Absolutly not. The file TYPE is Ogg. It's .ogg for a reason. Use something else to tell your files apart. Not extension.> The problem is not really about which player will manage to open your file. > It's about quickly identifying which one is audio, which one is video, > according on the oldest and simplest file-system metadata I heard about : > File extension.You're right, it is the "simplest". It's also wholly ususable for this purpose, as it is unusable for most purposes. Use something else.> It enables the use of simple file managers like Windows', and doesn't break > the more advanced ones. I see nothing against that.Yes, it actually does break "the more advanced ones". When you upload such files to webservers the admins need to add the extension to the mime-types file so it can properly tell the receiver what mime-type it is.> The birth of the OGM extension speaks for itself. There was a need, it was > filled. I just hope Xiph will officially support this *kind* of simple > tricks, if not *that* one..ogm is not Ogg. It's something that somebody else wrote as a cross between AVI and Ogg. It's also been abandoned, as a project. If you want to solve this problem on Windows, write something that plugs into the preview thing on the file manager (if you even have one) that gives you the codecs which are used, prehaps also with play length etc. Or whatever else you want. We are not going to impose a bunch of stupid rules and exceptions that serve to make Windows users lives a little easier at the expense of the time and confusion of every non-Windows user.
<20040617194522.GA17559@xiph.org> Message-ID: <40D2070D.7020707@slothy.com> Subtitles on a song with no video? To me that just seems like a silly example. See, here's the problem with this debate: there definitely IS a gray area about it, so pointing out that there are edge cases doesn't invalidate the argument. "It's too hard to do right" should not be an excuse. I suspect that if people start doing crazy things like subtitles on an audio track that people will probably still only care that it's an audio track and name it that (for instance if someone has the subtitle data in the id3v2 tag and this becomes commonplace, people will still call them .mp3 files). Don't worry about what to name each file, just define something to say when it's audio and when it's video and let the users figure out which slot to put their files in. Sorting a directory of files based on their file type is pretty freaking essential. Just because 1% of users use an OS that can do that without changing the extension doesn't mean that's the best choice. Why bother making a Winamp plugin if nobody cares about making Ogg Vorbis successful for the masses? This is just a continuation of that. I think you can point to the fact that .ogm has already been successful for videos as proof that Windows people want this. You might not - hey, maybe the linux filesharing apps can automatically remap the filename so you don't have to stoop so low. Either there needs to be a push for "We need to try to make this as successful as possible, even if we have to compromise our lofty techie ideals", or someone needs to say "success be damned, it would hurt my feelings to have two file extensions for the same container format!" Maybe we could turn the argument around a little, because I have to admit that I don't really understand the all-.ogg viewpoint yet. What is the net loss for having two file extensions for the same container format? For example, how would things be worse if divx had decided to call it .divx instead of .avi? Obviously they did it like that to piggyback on the native support for avi, but ignoring that (since .ogg and .ogm both universally have almost no native OS support), what would be the negative aspects of this? Jon Arc Riley wrote:>And who draws the line between "audio" and "audio+subtitles"? What >about "audio of a brand new codec which I cant decode"? What if your >favorite media player only supports Vorbis, but you have a Ogg FLAC file >which is supported by a media player which you like less, and want it to >open in a media player which you most like? > >The problem cannot be solved in the extension. Your OS is the problem, >or prehaps just your file manager. If you /want/ different icons and >different programs to open files depending on the file's _content_, not >just it's _type_, then you need something more advanced to fill this niche. > >Ogg is the file format. Wanting different extensions depending on it's >content is like wanting different text file extensions depending on if >the content is personal or work related, etc. > >
<20040617194522.GA17559@xiph.org> Message-ID: <004b01c454b0$a9704120$05649c3f@computername> From: "Arc Riley" <arc@xiph.org>> And who draws the line between "audio" and "audio+subtitles"? WhatWhy does there need to be line drawn between those? Seriously.... Why would you need to get that bogged down into that much detail? Most users are only going to care about easily and quickly telling the difference between audio and video. Both on their own system and when they see it elsewhere, such as on the web or within a p2p program during a search. They are already accustomed to being able to do that by the extension with mp3 vs. mpg, and wmv vs. wma. And .ra vs. .rm. Sure, knowing what codec etc. can be nice (unless you write the code buggy! Like on XP), but generally, their first concern is knowing whether it's audio or video. Even if it's wrong 1 time in a hundred thousand (like finding a .mp3 as a .wav) then that's not too bad. That's still good enough for the web, and good enough for most personal situations (because when you do find out, you can fix it on your system and you wont have that problem with that rare file again.)> about "audio of a brand new codec which I cant decode"? What if your > favorite media player only supports Vorbis, but you have a Ogg FLAC file > which is supported by a media player which you like less, and want it to > open in a media player which you most like?Those are valid problems. And that's a normal problem with any generic container format that can hold arbitrary codecs. That's why people end up with third party file examiners. Even having OS extensions added aren't going to help the cases where it's a new format. However, this is an *entirely* seperate problem than how the average user is going to tell the difference between audio & video on a web site or such. This is much more along the lines of Microsoft's attitude when they programmed in the .AVI extension handler in WinXP. And wrote it buggy, so that now every WinXP user dreads getting a damaged avi file, and says some very unprintable words when they do encounter them. And after a while, it eventually occurs to them to try and find a way to disable that 'feature'.> The problem cannot be solved in the extension. Your OS is the problem, > or prehaps just your file manager. If you /want/ different icons and > different programs to open files depending on the file's _content_, not > just it's _type_, then you need something more advanced to fill thisniche. That's starting to sound like a very small niche. Simply differentiating between audio and video solves 99% of that. Sure, if you want greater info, then an OS handler could do it. And it might even be nice to have. But that's a seperate issue than helping users tell the difference between audio & video. I remember I once played around with an extension for Win9x that could show in the icon what bit rate the mp3 was. It was kinda fun for a while, until I started having too many problems with it and the file listings took too long because it opened each and every file and examined it. Then I removed it.
On Thu, 17 Jun 2004, Jon Doda wrote:> noprivacy@earthlink.net wrote: > > .oga for audio only and .ogv for audio and video? Close enough to '.ogg' to > > be recognised. > > But still, it might be nice to have some extension that is somewhat > > descriptive of Theora video. .OGT perhaps? > Using an extension named after a specific codec seems like a bad idea > for a generic container format. There's no reason to believe that > Theora will be the only video codec to use an ogg container, just as > Vorbis is not the only audio codec that uses ogg.Assuming applications can deal with multipart extensions or longer extensions: file.theora.ogg, file.theora-ogg, file.t.ogg, file.t-ogg file.vorbis.ogg, file.vorbis-ogg, file.v.ogg, file.v-ogg just some ideas. -Dan
On Thu, 17 Jun 2004, Arc Riley wrote:> We are not going to impose a bunch of stupid rules and exceptions that > serve to make Windows users lives a little easier at the expense of > the time and confusion of every non-Windows user.so basically youre sacrificing windows usability in favor of non windows. and people wonder why vorbis isnt more broadly accepted? -Dan
<20040617194522.GA17559@xiph.org> Message-ID: <40D21148.3090707@sympatico.ca> Arc Riley wrote:> And who draws the line between "audio" and "audio+subtitles"? What > about "audio of a brand new codec which I cant decode"? What if your > favorite media player only supports Vorbis, but you have a Ogg FLAC file > which is supported by a media player which you like less, and want it to > open in a media player which you most like?That's a straw man argument. Different audio codecs hardly require a completely different UI, therefore there is no reason any audio player can not support all of the ogg audio codecs. An audio player can never support the ogg video codecs though, and that's the point. If your current audio player only plays Vorbis but you have a lot of FLAC files than you'll clearly have to switch to a different player. If you only have a few FLAC files than it's not a huge problem to have to open those files in a player that can handle them. In any case, this is a much rarer problem than the one the new extensions would solve, and therefore the fact that the extensions doesn't solve it is not a valid argument against them.> The problem cannot be solved in the extension. Your OS is the problem, > or prehaps just your file manager. If you /want/ different icons and > different programs to open files depending on the file's _content_, not > just it's _type_, then you need something more advanced to fill this niche.That's may be true, but if it is, it's a problem with every major OS and filemanager in existance. It would probably be easier to allow for some new extensions rather than try and "fix" dozens or hundreds of other programs.> Ogg is the file format. Wanting different extensions depending on it's > content is like wanting different text file extensions depending on if > the content is personal or work related, etc.Only if you needed to open work and personal files of the same format in different programs. If you did, you probably would have to use different extensions. -- Jon Doda
<KBEGKJCPEBDLMBOHODAAMEIACBAA.1edm@qlink.queensu.ca> Message-ID: <40D21661.3040907@matroska.org> PrussianSnow wrote:>>Paul E Wrote > > [snip] > >>Windows 2000 (although it's the same for all of them) doesn't read in >>any mime-type to correctly identify a file, it just uses the extension. > > [snip] > > What if, at least for Windows, someone wrote a simple application which when > associated with .ogg files and an .ogg file was opened simply read enough of > the ogg stream to identify the type of content and then launched an > appropriate (and selectable) application? > > This would give users their expected operation at least on launch. It > doesn't address the 'sort by type' issue but I think some Windows 'shell > extensions' could help this at least in XP where the 'details' view of a > folder can include things like 'artist', 'album', 'bitrate', etc. > > PrussianSnowThis app is already existing, is called Ogg Prak, was ordered by Emmett Plant and coded by Ludovic 'Blacksun' Vialle, the developer of The Core Mediaplayer. It should be in the Xiph CVS still, if they didnt delete it .... Christian matroska project admin
<001c01c49ce0$c39425d0$a720a8c0@NESCAFE> <40D1EF69.4030405@ellisfoundation.com> <001601c49ced$002f5260$a720a8c0@NESCAFE> <20040617194522.GA17559@xiph.org> Message-ID: <303907827.20040618081545@moz.geek.nz> Arc said:> The problem cannot be solved in the extension. Your OS is the > problem, or prehaps just your file manager.The obvious solution is for xiph to persuade all the various OS writers to change their file management to be conformant with the vorbis standard. Stop laughing, dammit, that is *exactly* the solution you are proposing. Just because you would like to be able to dictate to Microsoft and Linux doesn't mean that predicating acceptance of your standard on so doing is a sensible choice. I think it's better to go with the way almost everyone works right now, while leaving open the option of upgrading later. -- Moz Save homo ped plumbium - they're driving themselves to extinction.
<001001c45489$680a3a50$79649c3f@computername> <40D1DEBA.5090905@ellisfoundation.com> <000401c454a9$355dea50$05649c3f@computername> Message-ID: <20040617231837.GM23020@griffon> noprivacy@earthlink.net (noprivacy@earthlink.net) wrote:> Right. OGM is just the ogg container with a different extension. But it's > been done as only DivX now for quite a while. I think it's a little late to > try and put Theora in there in people's mind.I'm not so sure. *.avi files can have many different video codecs, can't they? The actual codec isn't important when it comes to this issue. I don't see any reason at all why *.ogm couldn't be used for Theora-plus-Vorbis as well as DivX-plus-Vorbis and Xvid-plus-Vorbis (as it is today).> Yup. It would be the best solution. > > Apps just get programmed for that extra extension. And since there aren't > any Theora video player or editing apps yet, it's really no big deal.No. You had it right the first time -- it's _not_ about the applications. It's about the files, and the way human beings relate to them. If I unpack a source tree, I see *.c files, and *.h files, and *.txt files, and *.html files, and maybe *.y files or *.cxx files, and so on. _I know what every one of these files is_ just by looking at the name. I know which ones to read and which ones not to read yet. When the compile blows up, I know how to interpret the source code files based on which language each of them is written in, up to my own personal limitations in each such language. When I look at files under my /export/music hierarchy, I see *.mp3 files and *.ogg files and *.flac files. I know what each of _those_ files is just by looking at the name. Now, that's not a big deal because they're under the /export/music directory, so I know they're music files -- i.e., audio only. So they all get fed to xmms anyway. But I also have files that are downloaded by giftd, and they end up in ~gift/.giFT/completed/ -- and they're all mixed in together, regardless of type. When I go through that directory to sort things out, I see *.mp3 files and *.ogg files and *.mpg files and *.avi files and so on. I know what each of _those_ files is, too, because nobody in their right _mind_ would share a movie on a P2P network and name it *.ogg. It's just crazy. (And if someone did, I'd rename it as soon as I got around to watching it.) Being able to tell how to deal with a file, in a qualititave way, by its name, is _fundamental_ to how people interact with computers. When I see *.ogg I know it's an audio file, so I'd use xmms or ogg123 to play it. If it's a *.mpg file, then I use mplayer. Seeing a movie labeled *.ogg would be like seeing a package of hamburger in the vegetable aisle. It's cognitive dissonance. It's unnecessarily confusing. But what we _don't_ need is for Xiph to get involved in this issue (at least not at this time). This is between the users and the application developers. A standard is going to emerge -- either *.ogm or *.ogv, would be my guess. When that standard finally settles down, then Xiph can advocate it, make the reference encoder generate it, and so on. -- Greg Wooledge | "Truth belongs to everybody." greg@wooledge.org | - The Red Hot Chili Peppers http://wooledge.org/~greg/ | P.S. -- If this makes me an old-school hardcore command line curmudgeon who just "can't relate" to the new way of using icons instead of names, and clicking on things to open them in some automatically "associated" helper application, then so be it. I don't care. P.P.S. -- I can write a similar rant about web pages that give you an HTML page when you "click on" a URL that ends with .tar.gz (*cough* sourceforge *cough*). Names are _important_. Fucking with standard names makes people confused and angry. Not all of us use our computers the way your grandmother does. Some of us have more than one finger, and more than just a mouse, and more than one window open at a time. You've been warned. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Digital signature Url : http://westfish.xiph.org/pipermail/vorbis/attachments/20040617/c89c647a/attachment.pgp
<001c01c49ce0$c39425d0$a720a8c0@NESCAFE> <40D1EF69.4030405@ellisfoundation.com> <001601c49ced$002f5260$a720a8c0@NESCAFE> <20040617194522.GA17559@xiph.org> <40D2070D.7020707@slothy.com> Message-ID: <20040617233050.GN23020@griffon> Jon Shiring (slothy@slothy.com) wrote:> Sorting a directory of files based on their file type is pretty freaking > essential.Bingo!> I think you can point to the fact that .ogm has already been successful > for videos as proof that Windows people want this. You might not - hey, > maybe the linux filesharing apps can automatically remap the filename so > you don't have to stoop so low.I'm about as hardcore a Unix-head as you're likely to find. I don't think Windows. I don't use Windows, whenever I can help it. I don't think GUI. Right now I'm looking at a workspace with 9 rxvt (terminal emulator) windows open in it. I don't think "object oriented". OO is the worst disease the computer industry has ever experienced. I think command line. I think syntax. I think procedures. I think verbs, not nouns. I think about what I want to _do_, not what I want someone else to _see_. I think active, not passive. And I can't understand how anyone -- _anyone_, user or developer, GUI or CLI, Windows or Mac or Amiga or Unix -- can use the same extension for both audio and video files. That's lunacy. So, just so you know: it's not just for the Windows users. -- Greg Wooledge | "Truth belongs to everybody." greg@wooledge.org | - The Red Hot Chili Peppers http://wooledge.org/~greg/ | -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Digital signature Url : http://westfish.xiph.org/pipermail/vorbis/attachments/20040617/96695cdb/attachment.pgp
<20040617231837.GM23020@griffon> Message-ID: <006401c454ca$be843980$6a649c3f@computername> From: "Greg Wooledge" <greg@wooledge.org>>noprivacy@earthlink.net (noprivacy@earthlink.net) wrote: >> Right. OGM is just the ogg container with a different extension. Butit's>> been done as only DivX now for quite a while. I think it's a little lateto>> try and put Theora in there in people's mind.>I'm not so sure. *.avi files can have many different video codecs, >can't they? The actual codec isn't important when it comes to this issue. >I don't see any reason at all why *.ogm couldn't be used for >Theora-plus-Vorbis as well as DivX-plus-Vorbis and Xvid-plus-Vorbis >(as it is today).There have been a lot of messages lately, but I think I was trying to talk about avoiding the .AVI problem. With .avi, you have to play 'guess the codec'. Figure out what codec the avi container is wanting, and then try to hunt it up and hope that it's still available. That's why there are so many of the warez "codec packs" for Windows. They contain dozens of warez copies of codecs for AVI files that you might run into. With OGM, it's been DivX, so Theora would be competing with it in its own container format for user recognition and understanding. People might begin thinking that Theora is just an open source version of DivX, like XVid. I was thinking along the lines that if you are going to do a special video extension, then you might as well make (encourage) it to be just Xiph codecs. Makes it a whole lot easier for users to find the required codec if they don't already have it. And OGM could be for third party codecs. In that case, they can play "Hunt the codec" like they normally do for AVI files. I think in that particular message (which started with a 'thinking out loud' message), I was trying to come up with a way so that users wouldn't have to do that with Theora. Even though that wouldn't solve the other Audio vs. Video subject. (This was a sub-subject)>> Yup. It would be the best solution. >> >> Apps just get programmed for that extra extension. And since therearen't>> any Theora video player or editing apps yet, it's really no big deal.>No. You had it right the first time -- it's _not_ about the applications. >It's about the files, and the way human beings relate to them.At that point, a lot of people had been going off and saying that you needed all sorts of OS handlers written, and this requirement and that requirement, etc. I was saying that none of that was needed. Just a matter of chosing a video extension that the user could easily recognise, and that using it would be just a matter of a few applications deciding to use it. No OS handlers etc. needed. We are agreeing here. It's just that particular message you replied to was probably less than clear due to it being one of several sub-subjects going on.>But what we _don't_ need is for Xiph to get involved in this issue (at >least not at this time). This is between the users and the application >developers. A standard is going to emerge -- either *.ogm or *.ogv,I think that only Xiph can suggest the official standard extension to use for video. Otherwise people are going to be making their own choices. A standard might end up being chosen, but it would take longer for it to get accepted than if Xiph makes it an official recommendation.>would be my guess. When that standard finally settles down, then >Xiph can advocate it, make the reference encoder generate it, and so on.And take credit for it retroactively.
On Thu, 17 Jun 2004, Arc Riley wrote:> Using your example, how would one catagorize a file that was part > audio-only, part video-only? What if it contained almost all audio but > had a single still-frame attached to it? Does that change the > extension? What if the user wants this included in their audio > directory, but because of some "standard rule" it became a video file > because of a single frame of MNG or a little (optional) text?This is likely to be so rare as to be a non issue. We are talking about the most common case, and the current barrier to wider adoption / ease of use of ogg/vorbis/theora in existing systems, frameworks, and software. You seem to be still missing the big picture, obsessing over contrived situations and ignoring the most common/most likely situations. Lets worry about the 99.99999% common case instead of hassling over the .00001% case. Nothing is preventing *you* from removing *all* extensions from all of *your* files since you are so confident in the file metadata to identify it and always using whatever tools you write to parse them always for every trivial file and directory operation. Everyone else would rather have easily identifiable content at a glance with 'ls' without having to pipe everything through a program which reads the ogg container of every file and parses them (not to mention the enormous performance hit this would entail). We're just talking about naming conventions after all. -Dan
<001001c45489$680a3a50$79649c3f@computername> <40D1DEBA.5090905@ellisfoundation.com> <000401c454a9$355dea50$05649c3f@computername> <40D225BF.2050701@ellisfoundation.com> Message-ID: <006501c454ca$c05f8930$6a649c3f@computername> From: "Paul Ellis" <public@ellisfoundation.com>> Do you really think most people know what video they are really > playing? I would venture to guess that most people only know a video isMany don't, no. Many do, however. For example, they realize that Divx is better than mpeg-1. They may not know why, but they can see it and have learned. (And how do they know Divx... because it usually says so in the file description or in the filename. And because the Divx decoder adds the divx logo when it plays the file.) I was trying to think of a way to avoid the whole AVI mess. Where you have to guess what codec is going to be needed. Using an already existing extension used by DivX didn't sound like a good way to seperate Theora from Divx. But as I've said many times, that started as just a "thinking out loud" message.> If anyone wants to check ask a non-geek friend of yours to open a video > player and see what happens. I'm blown away when I tell my Dad (in hisNot every computer user is like that. Yes, many are. No question about it. But the majority do have more experience with it and understand things better. They may not understand as much as us, but they've figured out a few things.> Anyway, the point is that average users know very little, and that is > all we can expect out of them. Changing their computer to work withThose kinds of 'average users' aren't really the target for Vorbis or Theora. They will probably never encounter either format. They probably use Windows Media Player because it came pre-installed on their computer when they bought it from "Dude, you bought a Dell!" They probably still get tickled when they log into AOL and it says "You've got mail!". And they probably actually read the spam they get and seriously wonder whether they need to buy the product or send money to that poor fellow in Nigeria. The reality is that most of those stupid people are probably so stupid about computers they will never be involved in a .ogg file, a vorbis audio file, or a Theora video file. So it doesn't really matter. But for the millions of people who are smart enough to use a web browser, or use a p2p program, etc., then those people do matter. And they can figure out a few things.> some identifying program is as hard for them as changing oil in a car is > for most people. Sure it's really easy, maybe 5-steps, but if you don't > know how to do it, you wouldn't even know where to start.[shrug] In the case of a special ogg video format, that'll be taken care of when they install the player or the codec. Then they wont have to think about it.> What would we gain in switching to .oga? Nothing. I want Ogg, Vorbis,Very little. I didn't say we would or should. The message I was replying to said that. I was just pointing out that if Xiph really did want to make ogg generic and change it to .oga, it would be possible. If they had done that in the beginning, it might be better than it is now, but they didn't. And there's not much point in doing it now.> Now if Monty, Segher, or any of the other > hackers involved are in the 1% then maybe things should be different.They do seem to be in the 1% who feel that users should be deleriously happy with a single file extension.
<000401c454a9$355dea50$05649c3f@computername> <20040617231837.GM23020@griffon> Message-ID: <200406180241.39823.haxe@pansensack.de>> Being able to tell how to deal with a file, in a qualititave way, by > its name, is _fundamental_ to how people interact with computers.Completely correct. This is an important fact.> But what we _don't_ need is for Xiph to get involved in this issue > (at least not at this time). This is between the users and the > application developers. A standard is going to emerge -- either > *.ogm or *.ogv, would be my guess. When that standard finally > settles down, then Xiph can advocate it, make the reference encoder > generate it, and so on.Yes, a standard extension for ogg video definitely _will_ arise, and it certainly won't be *.ogg, no matter if the Xiph people like that or not. That said, it should be Xiph people to choose a standard extension in advance, and propagate it as the default extension NOW. That would save some time of confusion. I personally would prefer *.ogv (not *ogt or again *ogm), but that's indeed not important. It is only important that it differs from *ogg and from other extensions that are not likely to be opened with a video player application. To all you Xiph people who may have influence on that: This is really a very important and necessary discussion. Making a good decision about the filename extension of ogg files containing "mainly video" will have a huge effect on how many people will use ogg video. Please propagate a special filename extension (other that ogg) now, before other people will do (and thus cause even more confusion). If not, this could one day be considered a big historical mistake. Some footnotes:> Is there still any great reason to require just three letter > extensions? Probably not too many DOS, Win3 or Win95 users around > anymore. ?Any OS or app problems? ?Maybe, because people have > habbitually used 3 letters for a long time.No. Today, more than three letters are perfectly fine. But I think it's not necessary. *.ogv would serve well.> file.theora.ogg ... > file.vorbis.ogg ...This is indeed a common scheme to name files that can carry varying types of content. I have seen many files named somevideo.divx.avi, obviously to indicate the used codec. But for the _big_ distinctions that more than 95% of the users feel to be important (because they use different applications for the different types), like the distinction between "mainly audio" vs. "mainly video", there will always be a top-level distinction between the filename extensions. There is a reason why *.mp3 is called *.mp3 and not *.audio3.mpg Hauke Hachmann
<000401c454a9$355dea50$05649c3f@computername> <20040617231837.GM23020@griffon> <200406180241.39823.haxe@pansensack.de> Message-ID: <20040618004942.GK26977@patrick.wattle.id.au> Haxe wrote: | I personally would prefer *.ogv (not *ogt or again *ogm), but that's | indeed not important. It is only important that it differs from *ogg | and from other extensions that are not likely to be opened with a video | player application. [...] | No. Today, more than three letters are perfectly fine. But I think it's | not necessary. *.ogv would serve well. I think that truncating the "ogg" bit in the extension is kind of ugly and not really necessary these days. How about ".oggv" or even ".oggvid"? Not that it matters ;-) Cameron.
I drafted a slightly warm email this morning, but recalled it. Let's try to be a little more cool-headed over this. We all want Ogg to succeed, and are just trying to find the best way forward. Xiph have come up with two ideas: a container format for various media, and a bunch of free codecs for various media types. The first idea is interesting, and may be useful given sufficient software support. The second idea is useful, and is being used by the general public right now in the form of Vorbis audio in files with a .ogg extension. With the advent of a second Ogg codec, Theora, the first idea is hindering the second. The general public, with their Microsoft operating systems, simply isn't ready for a single file format that can contain audio and video. Maybe they never will be. Right here and now, the general public is just saying "I want my MP3", and it's hard enough to get them to take any alternative seriously, let alone confuse them about whether a given file is music or video. "Container utopia", where everything is an ogg and the software helps the user to determine how to play it, may be just around the corner, but the world isn't ready for it yet. Saying "Their OS is broken" isn't helping. Some of them know it, most of them don't care because right now they can double-click on their .MP3 or .WMA and it plays the music, and they can double-click on their .MOV or .WMV or .AVI and it plays video. Asking them to put all their media into .OGG files is just asking for failure. Practicality has to win over ivory-tower perfection. Yes, there are grey areas, such as audio with subtitles, but I think the answer there is that it is audio, with subtitles. It's primarily audio. If your audio player can handle subtitles, then it'll work, no problem, and if you're aurally impaired, then you will have to get an audio player that can handle subtitles. Pathological cases, such as a speex intro, followed by some slides, with a vorbis soundtrack, then a theora clip, followed by some more speex, can just be called ".OGG". By the time someone actually creates something as complex as this in a real-world case, there must by definition be software that can open ogg containers and process whatever they contain, and it will be able to handle legacy files such as ".OGV" or ".OGGV". Personally, I keep my music and my videos in separate directories anyway. I don't play music on my PC, and I don't have a portable media player that plays video, so this issue doesn't actually impact me in any practical sense. I wonder what iRiver, Neuros, and Rio would have to say about this issue, as they will also have to deal with the consequences. Phil Hibbs Capgemini Aston, UK Our name has changed. Please update your address book to the following format: "recipient@capgemini.com". This message contains information that may be privileged or confidential and is the property of the Capgemini Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message.
<000401c454a9$355dea50$05649c3f@computername> <20040617231837.GM23020@griffon> <200406180241.39823.haxe@pansensack.de> <20040618004942.GK26977@patrick.wattle.id.au> Message-ID: <20040618150135.GB31221@vervet.localnet> On Fri, Jun 18, 2004 at 08:49:42AM +0800, Cameron Patrick wrote:> I think that truncating the "ogg" bit in the extension is kind of ugly > and not really necessary these days. How about ".oggv" or even ".oggvid"? > Not that it matters ;-)Yea, I like that. .ogv is not as obviously ogg as .oggv - so you can at least apply the "market ogg by picking a good extension" argument (as was previously mentioned to promote theora vs DivX). For the record, I've heard people ask "what codec do I need to play a .ogm file?" (when they already had DivX) - advertising theora in the filename is not important, those that do video encoding are those who matter, and they know all about codecs, they won't be confused in thinking ".ogm" is divx... My apologies for not threading nicely - I've replied to more than just the email I've just replied to... Hugo
k> References: <993BE7B40B41D31193C50008C75F68F207E980DB@exast02.capgemini.co.u k> Message-ID: <200406190323.08597.jwm@eslnz.co.nz> On Sat, 19 Jun 2004 02:12, Hibbs, Phil wrote:> I drafted a slightly warm email this morning, but recalled it. Let's try > to be a little more cool-headed over this. We all want Ogg to succeed, and > are just trying to find the best way forward.Indeed. Mailing lists are poor environments for reaching consensus; they're more likely to generate flamage. I suspect that the key Xiph developers have long since killfiled this thread. John
[Resend: sent to Cameron rather than the list by mistake, as I hadn't noticed that the list no longer sets reply-to.] On Fri, Jun 18, 2004 at 08:49:42AM +0800, Cameron Patrick wrote:> Haxe wrote: > > | I personally would prefer *.ogv (not *ogt or again *ogm), but that's > | indeed not important. It is only important that it differs from *ogg > | and from other extensions that are not likely to be opened with a video > | player application. > [...] > | No. Today, more than three letters are perfectly fine. But I think it's > | not necessary. *.ogv would serve well. > > I think that truncating the "ogg" bit in the extension is kind of ugly > and not really necessary these days. How about ".oggv" or even ".oggvid"? > Not that it matters ;-)CD-ROMs and standalone audio/video players. Strict ISO9660 compliance means you have to stick to the 8.3 convention. If a player doesn't understand either Rock Ridge or Joliet extensions, some mangling will have to happen in the CD-ROM authoring side. I agree that something should be at least suggested by Xiph now, before a de facto standard comes from the users, which may be something technically worse than what Xiph is objecting to at the moment. .oga and .ogv seem perfectly reasonable suggestions to me. Purists can still use .ogg for everything. I suggest using: .oga for content that has a majority audio content (ie. any other content is supplementary to the audio). This includes Vorbis, FLAC, Speex, MIDI, Vorbis+subtitles. .ogv for content that has a majority visual component, eg. Theora, Theora+Vorbis, Tarkin, Dirac, a slideshow, etc. -- Paul Martin <pm@zetnet.net> (work) <pm@nowster.zetnet.co.uk> (home)
<000401c454a9$355dea50$05649c3f@computername> <20040617231837.GM23020@griffon> <200406180241.39823.haxe@pansensack.de> <20040618004942.GK26977@patrick.wattle.id.au> <20040618150135.GB31221@vervet.localnet> Message-ID: <20040620044702.GA3364@ghostscript.com> On Fri, Jun 18, 2004 at 05:01:35PM +0200, Hugo van der Merwe wrote:> Yea, I like that. .ogv is not as obviously ogg as .oggv - so you can at > least apply the "market ogg by picking a good extension" argument (as > was previously mentioned to promote theora vs DivX).Why not just use .theora then? -r
With regards to using *.ogg for everything (audio, video, or both), new tools for detecting the content of ogg files is definitely needed for Windows. This is probably a trivial task as ogginfo probably can report that a certain ogg file has a second stream other than audio. One tool that comes to mind (for windows) is VorbisExt at http://vorbisext.sourceforge.net I think this is one useful app where you can get info and edit tags of ogg files through the properties menu in Windows. A slight addition to this program to detect the presence of a video stream and what kind of video it is will be a first step for Windows users to manage their ogg files. Secondly, I think we could do with an Ogg organiser program or Ogg classifier, that basically classifies and categorises your ogg files as audio only, video only, or audio+video, in a windows explorer like interface. Also, it should have its own executing function where double clicking on each file would start the appropriate player. Basically what I am saying is we should *ditch Windows explorer* and have our own "Ogg explorer" (hey, that's another name to try :) ) which will open the appropriate program (which can be set via preferences) when the file is clicked, whether it is Ogg Vorbis (starts winamp or foobar2k), or Ogg Theora (starts up RealPlayer). If we make Ogg Explorer downloadable (and supported) by Xiph.Org, and boast about it its efficiency in organising your ogg files via content, then I'm sure the average user will want to give it a try (when they visit vorbis.com), given that it doesn't require a 20 MB download, spray 1000s of DLLs onto their system, and use up 50 MB of RAM on startup. :) :) :) Best regards, Steve.
On Mon, 21 Jun 2004, Stephen So wrote:> Some setup will be needed on their part though. I mean, after a fresh > install of Windows XP, I don't think double clicking ogg (or oga, ogv, > ogm) files does anything useful. :) In which case, the user will say > to themselves, "I need to get a tool to open and organise these files".You're forcing them to install 2 separate tools to deal with ogg files. A metadata handler, and a media player. Whereas with .mp3 .wmv etc they only need one (a media player). You're making ogg more hassle for them, not less. This is not the path to wide adoption. -Dan
>Paul E Wrote[snip]>Windows 2000 (although it's the same for all of them) doesn't read in >any mime-type to correctly identify a file, it just uses the extension.[snip] What if, at least for Windows, someone wrote a simple application which when associated with .ogg files and an .ogg file was opened simply read enough of the ogg stream to identify the type of content and then launched an appropriate (and selectable) application? This would give users their expected operation at least on launch. It doesn't address the 'sort by type' issue but I think some Windows 'shell extensions' could help this at least in XP where the 'details' view of a folder can include things like 'artist', 'album', 'bitrate', etc. PrussianSnow
How about we don't go wrangling about with half a million extensions to fix stupid problems with short-sighted software? Fix the short-sighted software and insist on half decent implementations of the OGG format. As an example, how many applications that purport to read/write WAV files can handle the full complement of RIF tags? Probably zero; but this was understood and the rules for reading and writing WAV files were formulated to allow programs to modify WAVs without mutilating the tags they didn't understand. YET, if it has a WAV extension 90% of programs simply assume it's raw PCM data and chug happily on destroying cue points, section markers, silent sections, comments, compressed audio and any other 'esoteric' optional features. Relegating these useful transparent, optional features of the 'standard' to the effective status of proprietary extension. Let's not repeat that with OGG. PrussianSnow