Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 Internet speed test! ×
The Internet

Open Source Video Streaming Needed 239

Mike McCune writes "This article discusses how streaming video is controlled by three companies: Real, Microsoft and Apple. It discusses how open sourced video streaming software is needed. I looked around and found the start of some OS streaming software. There is a video streamer based on Darwin here and there is the start of several players here." But what about the codec patent problems? I have been told that they're the big holdup. [sigh]
This discussion has been archived. No new comments can be posted.

Open Source Video Streaming Needed

Comments Filter:
  • by Anonymous Coward
    Have you guys looked at vic? was designed for video conferencing, but it can't be that hard to hack it to stream from a file... http://www-mice.cs.ucl.ac.uk/multimedia/software/
  • by Anonymous Coward

    The Free Expression Project [free-expression.org], started by Dr Lynn Winebarger, is aiming to develop a copylefted set of streaming client+server software. Another interesting project hosted at the same site is a decompiler.

    A quote from their web page:

    This is the home page for a project to develop a copylefted suite of streaming media tools (server, client, encoders, and codec(s)). It is just getting started. This page is under pretty constant construction at the moment, but it does already provide some useful stuff for the motivated....

    The primary motivation in starting this project is to provide free tools for artists who want to use the Web as a distribution medium (See FreeSpeechTV and The Sync for examples), and to make Linux a viable platform for them to use.

  • by Anonymous Coward
    Just a thought, but if you look back a few weeks to the slashdot post about JPEG 2000 [slashdot.org] you'll find what I thought was a promising new technology that could be applied to streaming video.

    And since JPEG 2000 source should be available open source, it might not be too difficult to build a streaming video application atop it.
  • by Yarn ( 75 )
    Surely that's quick enough for you :)

    Besides, OpenSource doesnt necessarily mean Linux/X11.
  • by crayz ( 1056 )
    Intel is trying to, with USB 2. However
    1) USB 2 still isn't as good as FireWire for things like
    2) USB 2 isn't here yet
    3) by the time it is here, FireWire will be at 800Mbps

    And yeah, all these Lee-nooks hackers love to slam Apple for trying to charge for it. Sorry, but when you invest large amounts of money in R&D to develop something like FireWire, you charge for it.

    Also, Apple is in a consortium w/ other companies(like Sony), and the charge is now just 25, and with that you can have as many ports as you want, instead of the $1/port like it was before. And AFAIK, the money is split between 6 or 7 companies that all put money into developing FireWire.

    If FireWire dies, it will be because Intel has so much power that they can crush good technologies with FUD and vaporware.

    And if Intel doesn't crush FireWire, all us Mac users will laugh as we plug digital video cameras into our already year old Blue G3s.

    BTW, IIRC Compaq had the first machines w/ FireWire, but there are still very few PCs that have it. Some of Sony's laptop also include it. But since Apple owns the name "FireWire", Compaq calls it IEEE 1394, and Sony calls it iLink.
  • In the beginning it was controlled only by Real. Then Microsoft got in. Then Apple did.

    Apple has been taking marketshare away from MS. If they can do it, Open Source people can do it. You should stop complaining and start coding. These companies have market share because they make good products(well except for that Media Player shit), if you want to make a good product, people will use it.
  • www.apple.com/trailers/

    Check out the quality of the movie trailers there. I have a Mac and a cable modem, and Sorenson kicks ass big time. I don't know what you're talking about it having all these problems, it works flawlessly for me.

    Screw MPEG, yeah DVD and MP3s are nice, but for the best quality with the smallest file size, Sorenson is it.
  • I'm friends with people involved fairly heavily
    in providing streaming video, and I've discovered
    some disturbing things.

    Suppose I want to provide streaming video content.
    If I go the Real Networks route, I will have
    to pay $20,000 for a 200 stream server. Microsofts server costs me $0. This has
    been true for awhile.

    Now, however, MS has a new program: they will
    pay for the bandwidth of content providers as
    long as they exclusively use MS's media server.

    Now, in addition to the one time fee for the
    server software, I have to consider that
    MS will pay my $50,000/month bandwidth charge?

    How can an open source solution compete with
    this?

    -David
  • One answer: 24/7 live video. Check out www.hereandnow.net [hereandnow.net].
  • That's precisely my point. MS can afford to
    pay for bandwidth while streaming video
    is still young. Once their format has become
    the standard, they'll stop paying the bandwidth
    bills. By then it may be too late to break
    their stranglehold.

  • Yes, I know. They tried to charge too much at first. They backed off, and now it's a big 25 cents!

    Wow.


    They started off with a dollar a port, which would have made adding Firewire connectivity one of the most expensive parts of any system. They knew they had a great technology and--guess what--they blew it.

    The quarter license came too little, way too late, and now we're saddled with the horribly overstressed USB architecture.

    The general idea is that Apple would have made much more money actually selling video editing macs rather than talking about it for years on end and finally making a lone stand w/ Sony on the joys of home video editing. I'm sure the two companies, who thanks to apple couldn't even share the Firewire name(Is it Firewire? Is it i-Link? Is it IEEE-Gevalt), did pretty well. But that just can't compare to how much business they might have done if home video editing was The Big Thing. It could have been, if Apple hadn't been so stubborn. They could have guided the evolution of the industry in more ways than just blue plastic.

    Oh wait. Why am I responding to flamebait?

    Uhm, I don't know. Anyone who speaks kritikally of Apple is suddenly posting flamebait?

    Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com
  • Well, you'll never ever get Sorensen from Apple. It's not there's to give, it's a technology they license from Sorensen. As for why they don't build a Linux QuickTime client it would be that they don't see any return on the investment at this point in time. There also may be licensing reasons. You usually license a technology for a given market. So Apple licenses Sorensen for MacOS (obviously) and Windows (for additional market share)

    I think that if the right person or persons made an actual business case to Apple that may change, especially in light of the fact that they're going to have to port it to run over their BSD based MacOS X anyway.
  • This is probably correct. Did Sorensen approach Apple and ask or did they hide behind this clause? The real question is "without the provision would Sorensen have released its Codec in any form"? Given the lack of any of their other IP being freely available I think the answer is no.
  • I'm having difficulty getting past the spelling.
  • You don't get the point, I think. The codec is not the only problem: there's also the problem of where to find the streams ; DVDs use MPEG2, satellites use MPEG2, so I don't think we can go without MPEG2.
  • If you write an open source client/server that supports RN's plug-in API, you can use RN's plug-ins to get access to a bunch of codecs. Of course, you are still stuck with a bunch of proprietary codecs, but you can add some open source plug-ins and give people both.
  • ---
    From the guys who killed Firewire...
    ---

    ...how can a company kill FireWire and yet make it available in all of their higher-end machines (and some low ones)?

    I imagine within a year all Apple systems will have FireWire. Not quite the same definition of dead as I know of.

    - Jeff A. Campbell
    - VelociNews (http://www.velocinews.com [velocinews.com])

  • Oh dear.

    Moral philosophers will be kicking themselves the world over, having spent their lives searching for answers which they could have just picked up on a course.

    You daft bastard.

  • You brandish non-sequiturs with such ease that I wonder whether or not you have ever had the formal training to which you appear to allude [slashdot.org].

    Ownership of property, particularly intellectual property, is not a cut-and-dried issue. Neither is money an accurate way of measuring productivity.

    If you choose to declare your life as worthless, go right ahead. Seek death long enough and you'll find it I'm sure.

    You think that measuring the value of your life in terms other than fiscal equates with wanting to die? I feel very sorry for you.

    Hamish
  • In a related note, is there a consensus on the license for Apple's Darwin Streaming Server? Is it open enough?

    In a related note, does anyone think for themselves anymore? Read the license. If you think you can live with the terms, and it suits your purpose, it's open enough. Who cares what Bruce Perens or RMS or ESR or anyone else thinks? In the end, you are the one responsible for living up to the terms of the license, so make sure it meets your needs, not anyone else's.

  • by szyzyg ( 7313 )
    I rememebr at this time last year someone posted about shoutcast making audio streaming... and in my typical manner I posted a Been there, done that years ago response.

    Well... time to spoil your fun again - I've had video streaming using mpeg working Using icecast.

    The *big* problem right now is getting a low bitrate video codec which will work in real time.

    But if you're a user with one of those fancy mpeg hardware encoedrs and you can get it to work under linux then you can stream mpeg video across icecast with very few changes to the server.
  • I think that if the right person or persons made an actual business case to Apple that may change, especially in light of the fact that they're going to have to port it to run over their BSD based MacOS X anyway.

    Ahh that is true, and the OSX client is coming down the pipe pretty soo, so they say - maybe we will ending up seeing a quicktime player for Linux in the not too distant future after all.
  • Can't help ya with on the client side. Maybe Apple doesn't know you want it.

    Yeah, that's a good point - maybe they don't. I went to their complaint feedback page and was informed that their feedback database was down for maintenence. Anyone have a good email address to let them know that this is something we'd like to have?
  • They started off with a dollar a port, which would have made adding Firewire connectivity one of the most expensive parts of any system.


    Uh, lemme think now. Hard drive costs me $100. Video card costs me $75. CD drive costs me $50. Chip costs me between $75 and $400, depending on if it's a Celery or an Athlon. Obviously, what I pay at Fry's is much more than the OEM pays, but don't ask me to believe that something important to, say, Dell, or Micron, which would help them differentiate themselves from the other OEMs, is left off because it costs them a buck. They certainly will get more than a dollar from their customers in return for including it, hmm?
  • yes the codec problem needs to be solved, but what about simple outgoing bandwidth costs?

    witness slashdot, it's not even streaming media, and it needs advertising to support its bandwidth costs (i think?). streaming media chews huge amounts of bandwidth.

    and slashdot is in the US, where outgoing bandwidth is relatively cheap, especially now with DSL. here in australia, outgoing bandwidth can be 10 - 20 times as expensive.

    what it all means, is that the internet is great for anyone to publish any of their ideas as stories, as long as they don't get too popular. if they get popular, currently they might find they need to get ads, which some ideas and stories may not be compatible with.
  • The QuickTime file/stream format is also open. This means that anyone can write a compressor, decompressor, and/or player without going through QuickTime or a patented codec. Sure, you actually gotta write the code, but it *is* documented.
  • by Johnboy ( 15518 )

    MP3 = MPEG1 layer 3

    I.e., the audio component of MPEG1.

    Moderate the above post down, somebody.

  • wouldn't be that difficult if you compressed it lossy... lossless now, that would be sweet ;)
  • I have a build of the 0.3 release (with some enhancements) available at:

    http://www.pobox.com/~kostya/videod

    It has binary and source RPMs, some tools, etc.

    Currently looking to host the project somewhere (CVS, mailing lists, etc.).

    RealPlayer can use RTSP, and so it should be possible to add support for RealVideo to this code. It is not neceessary to figure out the actual encoding -- just the high-level file format so that the server can find frames.

    I agree that the codec and player problems are harder ones, but let's not forget the server. With some work, we could have a great streaming server for Linux based on these sources.

    -- Kostya Vasilyev

  • I'm just a grad student, not a Dr. The home page might be a little misleading, in that I'm working on an interactive programming system that makes the current streaming tools (servers, clients, codecs, etc) basic metaphors that can be easily combined in any weird and unexpected combination desired. Configuring any one of these tools should be easy, but the real fun will be constructing middleware that does different things (say scripting a "virtual switchboard" for an Internet talk show in a few hundred lines of Scheme code).

    Right now I'm working on a decompiler so people can more easily reverse engineer proprietary codecs and protocols, and to provide a more immediate benefit to projects like Icecast (who have more modest goals and actual software now).

    Lynn

  • OpenH.323, I believe, can act as a netmeeting server, and serve up H.323 voice, and soon video-conference(they may already have it) capabilities to Linux and Netmeeting hosts.

    This is rather off topic, but in some ways not, as video-conferencing is a bit different, and has a few other problems than straight streaming video (you don't have to handle someone interrupting you in the middle of a sentence in a straight 1-way streaming world).

    The discussion on free codecs has come up on the OpenH323 mailing list, and there are apparently a few out there that can be used, but as other posters have said, most of the REALLY good ones have some form of patent protection on them.

    Although both of these projects (an OSS Streaming Video Server/Client, and OpenH323 VideoConferencing) would benefit from the development of a really good Video Codec, until we can find a hardware manufacturer to put it in a DSP, we won't have an optimum solution (anyone who has done serious Videoconferencing work will tell you that for best results do you CODEC in hardware.

    McA
  • Well, I guess that we'll just have to implement a known codec in a country that doesn't have a patent office as stupid as the merkin one.

    It's been done before. There's LAME and there's BladeEnc; both would be patent licensing violations in the US, but since they are mostly distributed in non-US and non-Germany countries, there's little that FhG can do except sending bark letters.

  • 1) Use an extreamly viral license the prohibits mixed use of the patent with any patents not distributed under the same license, i.e. no company can use our patents unless the cross licence with the rest of the world via our license.
    I'm trying to promote something fairly close to that with the Open Patent License. [openpatents.org]

    2) Allow companies to get arround 1 by paying thorugh the nose and use the money to reward the researchers and pay for lawyers, i.e. compramize our principals occasionally.

    I would rather the researchers own the patents themselves, and merely license them under the OPL. If someone wants to use the patent other than as the Open Patent License allows, they can go to the patent holder, in this case the researcher or perhaps the university, to negotiate terms.

    Unfortunatly, the above dose require a lot of orginisation, a lot of conenctions, and a lot of work.

    I don't think these problems are insurmountable. Most companies who acquire software patents do so primarily for defensive reasons. I expect that after we get to a well-debugged version 1.0 of the license, and a few companies start agreeing to it, that more will quickly agree. I would expect most of the people for whom software patents are a net loss overall would eventually agree to the license.

  • I don't object to the organization helping someone obtain a patent, I just think it can be done somewhat indirectly, and without the need for the organization to end up owning it, (as long as it's licensed under the OPL), and that doing it indirectly can be the more stable solution in the long term.

    The strategy I have in mind is to put together either some sort of Idea Futures [gmu.edu] system as Robin Hanson has discussed, and/or a Castpoint system as Marc Stiegler discusses in Earthweb. [skyhunter.com]

    This way other people can end up sponsoring the sort of help you're talking about. As much as I'd like to cross-index all human knowledge myself, partially in order to patent inventions and license them under the OPL with an aim to freeing up knowledge, solving problems, and making money, but mostly because it would be fun, I still think that neither I nor a single organization would be as efficient at it as a system that encourages everybody to have a go. Also, the successful implementation of such a system would show that patents aren't necessary, and that their goals could be met using methods that don't restrict scientific and literary freedoms.

    Also, I'm lazy. I would rather PPI act more as a clearinghouse for developing solutions available under the OPL and let others do the legal work that would become necessary than somehow try to incorporate all of those legal functions within PPI through hiring or outsourcing. (I'm not a lawyer, so I try to avoid that sort of thing as much as possible. Obviously I have to go to an IP attorney to ask patent licensing questions and will have to ask for legal help in getting the license checked over and debugged, but all that's quite different from contemplating forming an organization that provides these sorts of legal services itself.)

    At the moment, however, I think it's more important to develop the license and get legal help in debugging it, and sell people and companies on the idea and the license. That has the possibility of freeing up a lot of patent encumberences quickly, so it's the first thing I'd want to do.

    For long term stability, I absolutely agree that we need implement something that helps researchers, for instance, patent their inventions to be licensed under the OPL in the cases where that's possible. (As opposed to cases in which, say, the university claims all rights.)

    But I consider selling the license to be the immediate goal. Solving the rest of the world's problems can wait until next year. :-)

  • I'm not sure I understand the emphasis on streaming formats.

    There are times when I want my text served up live, and in those cases I'll go on IRC at a pre-arranged time. But normally I want to read a piece of writing that has been finished and published, and that I can access whenever I want, so I browse the web. I think the same analogy holds for video content.

    Streaming makes sense for live broadcasts, since the timeliness and the possibilities for immediate interactivity compensate for the drops and stutters - but it seems to me that most video is more suitable for publication in a "download, then playback" format. I don't mind waiting for a download to complete if it means I can watch the result at my leisure and not have the experience compromised by network traffic problems.

    Streaming acts as one flimsy layer of protection against copying, but especially for non-establishment publishers, the main problem is getting the content to the viewer in the first place, not trying to enforce pay-per-view economic models whose plausibility in the digital age is somewhat dubious in any case.

    So to me, "broadcast" and "streaming" don't seem as important as "publication" and "download". Many of the arguments from the article still apply, but it is a slightly different game with a wider range of possible formats... so, is there something especially important about streaming video that I am missing?

    - Anthony.

  • Does QT Server stream not only Quicktime Encoded MPEG-1 streams, but also data that conforms to the standard *.MPG file format? It may! It may not! The general theme though is that while the QT player might be able to handle that variant of .MPG, no other player can, so you're stuck with Apple's ad, and inevitable lack of functionality.

    This is not a problem. Go right here [apple.com] for documentation for the QuickTime file format.

    --
  • Firewire is not an Apple technology. Apple was one of, IIRC, 8 companies that was involved in the development of Firewire, and one of 8 that holds patents on the technologies involved. The $0.25 is paid into a pool. What portion of this Apple get is unknown.

    USB 2 is an inferior technology. It isn't peer-to-peer, so it can't be used in consumer electronics, and it can't provide dedicated bandwidth, so it isn't good for digital video or other real-time applications. It also isn't technically a standard, like Firewire, AKA IEEE-1394 is. Perhaps most notably, it's total vaporware. USB 2 looks like standard Intel FUD designed to stop the adoption of Firewire, which is certainly in Intel's best interst; a peer-to-peer technology scares Intel, because it could cut the computer of of the picture, replacing it with smart appliances.

    --
  • can anyone give insight as to why they won't at least port and provide provide, say, a binary of their quick time player/plug in for Linux.

    QuickTime isn't very portable; it was ported to Windows by implementing quite a bit of the Mac OS API there. It would cost Apple quite a bit to port to Linux, and wouldn't help their quest to take over streaming media much because Linux isn't all that popular on the desktop.

    But the Quick Time file format [apple.com] is open, so you could write your own, if you could get the codecs from their respective owners.

    --
  • On a related note, is there a consensus on the license for Apple's Darwin Streaming Server? Is it open enough?

    It's basically GPL with a few clauses thrown in by the Apple legal team to allow Apple to remove code from distribution if someone decides to sue them over it. IIRC, there's also a clause that requires you to send Apple a URL for any publicly distributed changes you make.

    --
  • There's a project at Berkeley called MASH that has an OSS toolkit (also called MASH) for building streaming video applications. It uses open standards like RTP and MBONE.

    Recently the NSF agreed to fund the MASH Consortium, which will maintain and develop the MASH toolkit. I think it's going to be modeled after the X Consortium.

    For more info, see:

    MASH Project home page [berkeley.edu] The toolkit and several apps built with it are available now. (FYI, I am a grad student at Berkeley, but have never worked on the MASH project.)

  • motion jpeg (MJPEG) is also pretty open right now and can be streamed.
  • There's nothing stopping you from using a different player (IIRC, both Cisco's IP/TV and the Java Media Framework come with QuickTime-compatible players).

    I wasn't aware that IP/TV could parse QT-headered information. Very cool.

    This is one of the things I generally like I about Slashdot--chunks of knowledge that aren't composed of out-and-out flamage. You'd think I posted that people's mothers were spawns of satan or something.

    One thing I've begun to take very strong issue with is the presumption that it's acceptable to have a fleet of codecs required to play any single media file, with all the codecs wrapped in a single consistent wrapper(Quicktime/AVI wrapping Cinepak, Sorenson, MPEG-1, MPEG-2, AC-3, MP3, Metavoice, etc.) SDMI is planning to use this method, with the idea that "if one company's system is broken, there will still be 19 left."

    Talking about an open source streaming solution is empty without talking about the underlying protocol--hell, I've got an open source streamer right here(cat mystream.mpg | nc -l -p 5000). What? You want to use a custom UDP based architecture without any of that annoying Flow Control(poof goes the net ;-) and the ability to drop packets in favor of resyncing the media stream? Well, now you need to talk about the underlying format, now dontcha ;-)

    Experience has taught us that, even on the most compatible platform--I'll calm the flame war by not saying its name--so called "wrapping" architectures fail miserably with surprising regularity. Sure, Quicktime as a format is open, but Sorenson has gone on the record--no compatibility for Linux. Oops, now we ain't gonna be able to watch the Star Wars Preview...gotta go get a closed platform for that.

    Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com

    P.S. What the *hell* was I doing posting as sleep deprived as I was last night...I don't think I've ever woken up to as much of a Slashdot mess as I have tonight ;-) Mea culpa, tryin' to recover here!
  • MOV? All the pain of an AVI, with free delays while you deny Apple their cut. From the guys who killed Firewire...

    Ugh. [Dan slaps himself around a bit.]

    Firewire's launch has been botched--there's no other way to describe the bottom line that having one company call something Firewire while another calls it I-Link while others refer to an IEEE standard is just bad marketing sense that I'm sure some licensing scheme brought about.

    Too bad, too, because besides simply having the most awesome interface name of the last twenty years, Firewire pretty much is one of the more perfect external interfaces imaginable--though I don't think they've done the security wrangling that the SIO guys are doing. For those who don't understand security considerations of one bus uniting all devices,
    imagine the concept of a rootmouse that once plugged in issues calls directly to the hard drive retrieving critical files, all independant of the underlying operating system. That's the kind of worry you just don't have when your mouse is hanging off a 9600 bps UART.

    But overall, saying something like this was pretty much flamebait. Off hand, unsubstantiated, assuming that the rest of the audience took as obvious fact what is really a rather contentious issue--these are all things that pretty much guarantee you're gonna fuck something up, and as *ahem* numerous AC's felt free to "adjust my perspective", I fucked something up.

    There's very likely a good deal of hype streaming out of Intel against Firewire, and I fell right for it. Damnit.

    I was pretty exhausted when I wrote this post, but that's not really an excuse. I don't usually ask for moderation, but if someone wouldn't mind tossing a few points on this apology post, it might quell the flamage :-)

    Sorry, all.

    Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com

    P.S. Couple of you AC's expressed a problem with my writing in general? Email me, if you're not afraid of revealing yourself.
  • Around the time of the Star Wars Sorenson Compressed Trailer, one of the major Linux video player authors requested access to the Sorenson codec to play that trailer. Reports were that Apple refused to give the coder access to the codec.

    Go look back through Slashdot archives--I read it here.

    Yours Truly,
    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com

  • QT streaming server can serve any codec -- it doesn't really care what it is. Likewise, the quicktime player will play multiple codecs, including MP3 and mpeg2.


    We have not even begun to get confused. You've actually got more than just "the server" and "the codec" involved--you've also got the "wrapping architecture around the codec" to deal with. Does QT Server stream not only Quicktime Encoded MPEG-1 streams, but also data that conforms to the standard *.MPG file format? It may! It may not! The general theme though is that while the QT player might be able to handle that variant of .MPG, no other player can, so you're stuck with Apple's ad, and inevitable lack of functionality.

    I'm not crazy--there's a definite chance that this software plays nicely with whatever you throw at it. But it's honestly not a hard problem to stream MPEG, and it's generally just not a good idea to stream video when nobody has enough bandwidth to get an acceptable level of service.

    --Dan
  • ...It will be necessary to interface BGL low-level primitives to existing video and 2D APIs, high-level widget libraries and GUI builders. A feature reference that provides an example of the kinds of requirements for vector graphics is SVG (Scalable Vector Graphics), currently under specification at the W3C (see http://www.w3.org/TR/WD-SVGReq )...

    I was just reading about some opengl [opengl.org] and coding and spotted this article [opengl.org]. Looks like this idea is being looked at.

    links:
    http://www.opengl.org/News/Special/Features.html
    http://www.opengl.org/News/Special/BGL.html
  • ...doesn't make the GIMP less functional. Just because they use Photoshop doesn't magically make it better than the GIMP. Design agencies use what they've been using in the past- vis-a-vis, Photoshop. They use it because they "know" it- they recognize the name. It's the same thing that went before with IBM and is going on now with MS. Do a feature to feature comparison with Photoshop- I think one will find that there's something to the original poster's comments. The GIMP is better than Photoshop in at least some ways and comparable in most of the others. Just because someone doesn't use it, doesn't make the thing they use better. I know of a lot of people that use PowerPoint to do graphic arts work because they don't know that Photoshop or Corel exist. Does that make PowerPoint better than anything from Adobe or Corel? No. So why would that apply in this case?
  • especially in light of the fact that they're going to have to port it to run over their BSD based MacOS X anyway

    Err, umm, their "BSD-based" MacOS X

    1. will use a window system (their Display PDF-based window system) that's not the X Window System used on, I suspect, most desktop free-UNIX boxes;
    2. will have a MacOS API layer that most if not all free UNIXes have;
    3. will have what I suspect is a "Son Of NeXTStEP" API layer (Cocoa), although that might be implementable by the GNUStep folk.

    I.e., the mere fact that it has a bunch of Mach and BSD code at one layer is insufficient to make MacOS X so much like the free UNIXes of today to cause the availability of code for MacOS X to make a port to various free UNIXes a no-brainer.

  • They can, but you'd have to provide reason for them to believe it's worth the effort. Linux as yet still has a pretty small marketshare on the desktop, and Apple would want to recoup the costs by people registering QuickTime Pro - usually content creators, not an audience prone to using Linux.

    It's be nice, but don't hold your breath until Linux marketshare rises up a few points. It's hard enough to even get them to make a decent player for Windows.

    - Jeff A. Campbell
    - VelociNews (http://www.velocinews.com [velocinews.com])
  • I suppose we'll never get the code on Sorensen from Apple, but can anyone give insight as to why they won't at least port and provide provide, say, a binary of their quick time player/plug in for Linux. That's something I'm getting sick of - having to find a mac or windows box to check out a .mov file... at least I use linuxppc at home so I can just dual boot without having to move my butt.
  • As Linus Torvalds commented not too long ago, the starting point of the western calendar has a margin of error much greater than the one year we are all quibbling over (estimates of Christ's birth range from 7 BC to 6 AD depending on who you listen to), so calling 2000 the millenium, or 2001 the millenium, or 1994 the millenium are all equally correct, and equally incorrect.

    However, you are right. 2001 will be the 2000th year since year 1, and there (stupidly) was never a year 0 ... the monks in the dark ages setting up the Julian (and later the Gregorian) calendar didn't really get whole numbers, else they never would have stuck year 1 right after year -1, with no zero in between. We should dump the whole mess anyway and go to 13 months with 28 days each, and use the extra day as a holiday.

    In any event I will celibrate the start of the 3rd Millenium AD a second time in 2001 and get two great parties out of it ...
  • Have a look at the Free Expression Project (http://www.free-expression.org/ [free-expression.org]); these people are developing a copylefted suite of streaming media tools (server & client, encoders and codec(s)). They've been working on it for a while, and will be pretty grateful if other coders want to join their efforts.
  • I read the article and I don't understand how the lack of free software is going to prevent anyone from creating and distributing video content. Sony will happily sell me a video camera, VTR and other expensive gadgets without any restrictions on content. The software vendors will also sell me encoding software and video server software without any restrictions on content. So what's the problem? The television networks had power because they controlled the distribution system, not because they controlled the technology.
  • The creation of new algorithms is difficult. It's one of the things that cannot be done easily by the open source community, which has brought us a lot of very high-quality products in other areas, because you need a very specialized knowledge (signal processing, lossless compression etc.) that only few people have. There was a very well-written comment by Eric Scheirer on this on Slashdot months (a year?) ago, but I couldn't find it.

    This webnoise article [webnoize.com], linked on slashdot in may [slashdot.org] is probably what's being referred to here. The comment in question is toward the end of the article.

    Now, generally people say this sort of thing for two reasons. Most commonly, they actually believe that no one else on the planet is as smart/knowlegeable/experienced as they are, and couldn't possibly produce rocket science of the requisite level. This is despressingly common, and one of the psychological hurdles (like fear of being judged on work-in-progress as if it's the best one could do in a polished effort) that holds people back from adopting open development practices. This is another strain of what RMS called the Cathedral style [tuxedo.org] of development. It's what was said about writing compilers, operating system kernels and desktop applications, and I don't buy it here either.

    The other reason is when the authors are trying to create a aura of "professionalization" around their work, usually so they can remain well placed in their employer's finances. The arguments are similar: "We're trained professionals. Do not try this at home," but the motivation usually has more to do with greed (or at least comfort) than fear.

    Neither of these arguments are in line with the values of Open Source, and they amount to arrogant selfishness from that point of view. I'm not arguing that codec design doesn't require specialized knowledge, clever research, etc. Just that it's a fallacy that no one but Eric Schrier [mit.edu] &c. can do it.

    He did make some other points that a little harder to refute. For example, that the unfair 'fair licensing' patent situation is an historic artefact of the designing bodies being large corporations I can't dispute. He also makes the argument that allowing patented technology into the mpeg standard helps ensure that major corporate players use the standard, that the standard includes the best technology, and that the best technology be patented, rather than buried in an undisclosed proprietary standard. This appears to be well-reasoned, and is certainly consistent with his other claims.

    I differ about this being the best course, mostly based on what we've learned about network effects and life in an exponentially-growing market. And vis á vis the dvd stuff [opendvd.org], it appears that patents are more restrictive than trade secrets in an open source context.
  • Have you looked at the prices for decent video publishing software (not the home video stuff)?

    Granted, it's still cheaper to buy a Mac with lots of add-on software than to buy a television transmitter ans license. And right now, hardware and bandwidth are still as expensive as the software.

    But in the long run, the price of video publishing software is going to dominate the cost of creating and distributing video content, and free software can play an important role in making video publishing more widely available.

    Another consideration is that if we don't start developing more free video software, future "inventions" in that area will automatically be made by commercial companies and become patented. Useful little features that a free software author might spend 10 minutes on thinking up and implementing will be patented and locked up for the next 20 years if the field is left to the commercial software developers.

  • How did Apple kill FireWire? They invented it. It's in most of the shipping Macs, aside from the low-end iMac and the iBook. Did they kill it by asking for ONE DOLLAR PER PORT? Give me a break...

    The dollar a port wasn't the issue. It was Intel suffering from same Not-Invented-Here syndrome as Apple used to have.

    IEEE 1394 is open. Anyone can use it. Apple just wanted money for the use of the FireWire name... Isn't that aweful? A company spends hundreds of millions of dollars to create a wonderful new bus archecture and gets slammed for asking an extremely modest fee for it's use.

    Apple didn't kill Firewire... Intel did.
  • So let me get this straight... A dollar a port adds up the the most expensive part of the system???

    Intel Celerons run $100 or so
    Pentium III's run $400 or so

    The BX chipset is about $18-23

    Integrated SCSI adds about $120 to the price it seems....

    If intel would license firewire, chipsets would run you what? $25

    I think you've just fallen for the FUD that was perpetrated in order to get consumers uninterested in the technology, so Microsoft and Intel could retain full control of their platform.
  • That's got to be the worst yet... They didn't kill it by asking for money for the actual ports, but by asking for money for the NAME? No...

    One point that's been brought up is that Firewire is far from dead... but it would have been so much less so if there hadn't been so much Wintel FUD concerning it, when the truth of the matter is that Microsoft and Intel want to hear nothing about Apple possibly owning a piece of their platform. It's bad enough that Quicktime runs under Windows...

    Firewire could potentially kill off IDE, SCSI, USB, Serial and Parallel ports, Video capture hardware, and probably others as well... For instance if USB speakers could be workable, FireWire sound could have been as well. Imagine the simplicity the Wintel world could have had if not for Intels stubbornness...

    I'd suppose that even includes the what, Lintel? world as well... You'd basically have had Mac hardware, except with an x86 chip instead... All plug and play, supposing ISA can be killed off...

    I'll continue to say - blame Intel... And yes, we Mac users already have the benefits, because it's not dead... it just should have been bigger than it has been/
  • Not really... All video is is 30 pictures per second with sound on top of that.

    Most capture boards capture each frame individually. The only time interfame compression is used is right at the end of the creation process.

    An MJPEG-2000 codec could do wonders... And if it's open, you can add things like keyframing and lessen it's storage requirements even more.
  • By "wrapping architecture around the codec", do you mean "file/stream format". The QT file format is open as well. In addition, you could even port another codec to QT so you don't have to use Sorenson or one of the other kick-ass compressors out there.

    So, let's see... you could write an open QT compressor to compress a stream, use the open Apple streaming server, decompress with a client side codec, and then play using the standard Apple players (or write your own). QT is open and extensible at each stage in the pipe.

    As for "not enough bandwidth", I think you are seriously mistaken. My cable modem rocks, and the Sorensen/TPM trailer was very good quality. Streaming is an evolving technology.

    You are absolutely right on one point, though, you are getting very confused.
  • Apple killed firwire by putting restrictions on the ability to use the trademarked name "firwire".

    try selling a technology to consumers under 3 different names. Gee, is it FireWire, 1394, i.Link? or any of a couple OTHER names that companies have come up with.

    It's like Beta/VHS all over -- FireWire is faster (and I know, i do DV editing with it every day) but USB is open for anyone who wants it without worrying about liscensing issue. And it's "good enough" for the average consumer, so it wins...
  • I seem to recall 'www.free-expression.org' being involved with this.
  • The best solution is probable: control the viewer and server. If a codec maker dose not want to go OSS then we should make it hard for them to get users.. up to the point that they are really a hell of a lot better then anyhting we are using.. at which point we should compramize a little.

    I'm not sure why you think this is the best solution? The problem(I think) is "How do we create a situation in which *anyone* can distribute their own media?", something along the way in which HTML has opened content distribution to the masses. Server might be useful in doing so, in order to prevent large corporations from dictating who can and can't distribute media. Viewers, I suspect is *almost* a pointless battle, excepting that most corporations don't support Linux. In that respect, the battle is just to gain recognition and equality.

    The more powerful question is how to give people the capability to create content in the first place, and this is unavoidable; they have to own fairly high powered machines, they need some extensive video equipment, and they need the appropriate software.

    Whether you like it or not, the two best solutions I can think of are Apple's iMac DV and Sony's Vaio computer systems, with Firewire, software, and computing horsepower all bundled together in one fell swoop. All that's needed after that would be some tapes and a digital camcorder, which is still unfortunately in the $0.8k range.

    The battle over the clients just dictates who can see the media, not who can distribute it. The battle over the servers may be a moot point, what with Apple open sourcing their Quicktime server(but not their codecs) All that remains is for Apple to port a binary only Quicktime player for Linux, BeOS, etc.

    It is also worth mentioning that codec makers make money by licensing the authoring software (i think) or with crap attached to the codec, so there should probable be a push to implement OSS versions of the authoring software and codecs in countries where the patent dose not apply.. then make the OSS version the default, i.e. default RedHat xanim has no support for the codec so the user is forced to choose between downloading a single OSS xanim which is illegal or downloading a million codecs. This will cut into there proffit margin.

    Now you're getting somewhere; but somehow I suspect anything the OS community can come up with, the commercial houses will just come up with something better-one could always use the GIMP/Photoshop conflict as an analogy, and I really don't think Photoshop will be dying anytime soon. It supports too many professionals who just want to get their job done, and are willing to pay for it, rather than deal with the GIMP, much as a lot of people would rather deal with Mac or Windows over Linux, because both are rather incomplete in the useability department, though marvelous strides have been made.

    So I don't know too much about the Apple Quicktime Server they've open sourced; if it's anything like I imagine it, codecs aren't the issue, clients are the issue. If someone has content, then anyone can stream it because of this open source server, thus solving one problem with the media distribution system. I don't think one can get around the high cost of media content, if only because one still needs equipment and software. And clients, well, that something someone has to fight over the decoding codecs.

    -AS
  • The article seems to talk about how media content and distribution will be controlled by three corporations, Apple, Real, and Microsoft, as an analogy to CBS, NBC, and ABC. However, as far as I can tell, this analogy is crap. The three broadcasting networks control content, because up until now it was fairly expensive to create a video. They aren't winners because they control the video format(Beta, VHS, or DV), because they control the airwaves(they have to license out frequencies, not that they can own them!), or because they control the TVs(hardly! Sony, Panasonic, etc.), but because they have the funds to create content with actors, sets, special effects, scripts, and storylines, and they can attract advertisers.

    So in comparison, the format(codecs), the distribution(wires, cables, and lines), or the clients(Quicktime, Real, or Windows Media) is also similarly pointless. It's the production software, which from Adobe costs $600 for special effects and $900 for video editing), $1.4k for an iMacDV with firewire and additional software/hardware, $0.8k for a cheapo Sony Firewire enabled digital8 camcorder, and maybe even a USB/Firewire CD-R for backup/archival/storage purposes at $0.3k

    Getting actors, scripts, sets, etc, is a little harder, but Blair Witch proved it can be done. Most actors are starving/unemployed/checkout baggers, I guess, so it should not be too hard a problem =)

    So once everyone(if Apple can get an iMacDV into every household) has the capability to create video content, what's stopping people? Serving it? Viewing it? Isn't that what Apple's Open Source Darwin server is about?

    One thing I concede; Apple isn't yet supporting Quicktime for Linux. That sucks, but it's hardly the reason why media distribution will be controled by the big three; it's just short-sightedness on Apple's part.

    Correct me if I'm wrong, but I don't think this is a problem over codecs, streaming software, or even distribution; all that exists, it just remains up to the consumer to create content!

    -AS
  • This needs to be seen by everyone talking about writing codecs, formats, servers, clients, etc.

    There is, IIANM, the open source like Darwin server, and this tidbit tells us we can write our own codecs for Quicktime; the problem being that Apple hasn't released the general Quicktime player for Linux, etc. But if we can write our own codec and player, then the problem disappears!

    -AS
  • Wasn't it developed for audio compression?

  • by Dacta ( 24628 )
    Hence the question mark.
  • I'm not sure whether you have refuted Eric's theory of the hardness of creating new compression algorithms in the open source community. I'm not one of the PhD-level experts mentioned, so I think I'm relatively objective.

    It is, of course, not totally unlikely that somebody comes up with a new idea for a compression algorithm, the idea works, the algorithm can run on modern hardware, somebody creates a C library and públishes it under the GPL or a similar license, the inventor gives the algoritm to the public domain, avoiding patent issues. But how probable is this?

    If you create compilers or operating systems, you have a very complex project, but you have many smaller tasks that can be implemented by different people (simplifying a bit). Coming up with a new compression algorithm is a different thing.
  • MPEG-4 might be better for low bitrates. But the problems come with the patents. See this more audio-oriented FAQ [mit.edu] on MPEG-related patent issues. Especially interesting:
    18. If the open source community develops a new compression standard, can we use it without paying FhG?
  • Yes, you could of course improve (or just take) existing ISO reference source code for video encoding and decoding. But I guess ESR, RMS etc. won't be happy with that - they want something that is legal everywhere, and that might be a good idea in the long run.
  • The Geeks You Need Do Not Have The Hardware

    This is the constant by which I believe all Open Source software is developed. If the Geeks You Need have access to the hardware, The Software You Want will result, purely because we can't help it. We just do it, its cool, we scratch.

    I don't have a video capture card. Even if I did, I wouldn't have any content worth streaming. Maybe if a bunch of movie creating people came and begged me to do it and gave me a capture card I'd be inclined, but otherwise, wheres the point?

    I think you overestimate the algorithm problems. Video capture is one of those things that would inately suit the kind of thought processes Geeks have, polygons, optimisation over time, limits and granularity. These are not hard concepts, and believe it or not, such things could be broken down into small pieces for easy addition by a group of like minded people.

    Its a simple problem: With no content to drive it, there is no itch to scratch. mp3 players only became a big thing when cds were worth ripping. Until then, there was no content. Ripping a video simply isn't practical, and even if you did, there is no need to stream the results, when you can get better compression just using mpeg.

    mbone have done a good job of handling the video-conference end of things, leaving only realtime tv style stuff, exactly the kind of thing I got on the net to get away from. Ads piss me off, lack of pause pisses me off, the quality pisses me off. If I wanna see a movie, I'll hire it, nobody but corporates can afford content worth playing realtime, or the hardware to support it.

    Think Dawsons Creek, sure, if it were available and in decent quality, I'd watch it streamed, but what Open Source developer can afford to get hold of content which gains that kind of devotion? for just about everything else, download then play is fine..

    The rumour that this is some kind of big thing for the future is a misnomer, Net people won't stand for ads that get in the way of their content. Can you imagine geocities forcing you to sit for 2 seconds between each page click looking at an image? I think someone tried that once, it sank like a stone. Its too easy to go elsewhere.

  • Play one what? :)
    Play one badly recorded impromptue radio show put on by local geek? Play one song recorded on tape complete with ads off the radio? Play one slightly better but overplayed adless copy of a song from the library?

    No. Play one popular music track ripped digitally from a high quality medium.

    Remove One Popular Music Track, or High Quality and mp3 would never have come about. No itch, no scratch.

    And don't mock the 'Creek :)
  • I would rather the researchers own the patents themselves, and merely license them under the OPL. If someone wants to use the patent other than as the Open Patent License allows, they can go to the patent holder, in this case the researcher or perhaps the university, to negotiate terms.

    This would be much better if it is finatially realistic. I was under the impression (from things RMS has said) that patents were really too expencive for the researchers to obtain and defend.. which is why I suggested using a holding orginisation.

    The other problem I see is the researchers time. I am a graduate student in mathematics and I could very well see getting myself into the situation where I discovered things which were relevent, but I did not have the time to do the patent grunt work.. which to me includes actually figuring out what it might really be applicable to in addition to all the patent paper work.

    It might make sence to have a non-profit orginisation which did some of the research into the applicability of the research and actually applied for the patents in exchange for partial onership. I'm realy thinking of something where I can give them some realy raw shit and they will sorta check off a list of things that it could apply to and go patent them, i.e. use exactly the tactics which the big nasty co.s use, but since anyone can use it via the OPL we are the good guys.

    Hell, you could even implement it as start up headed for an IPO.. with partial onership in a crapload of patents and after winning some big lawsuits you could make yourself a lot of money.. and make the world safe for OSS in the process. There isn't really any reason to restrict it to patents relevent to software either.. do drugs patents and all sorts of other things.. just refuse to let drug co.s who wont OPL there patent use any of them without paying.

    I realize there is a little conflict of intrest here since this org. would need to descided between pushing a co. to OPL or just letting them pay up and use it, but the orginator of the idea would still own most of it and would have a lot of say in the process. Plus, the more money we extort the more likely we are to see real patent reform.

    Just a though..

    Jeff
  • > Another hack was done at one point so that it would stream meta-refreshing jpegs.

    Given that we just had an update [slashdot.org] on the status of the revision to the jpeg standard here on slashdot last week, and that it offered incredible quality wavelet based compression, couldn't icecast be modified to stream the new jpeg format when it is released? Wouldn't this be open source?
  • I think that even if you are not making any money from the implementation of a patented algorithm, you can be sued for damages related to the destruction of the POTENTIAL revenue stream that the company (or individual) might have made had they been able to use their patent.

    Hmmm...

    They must put sufficient description of the algorithm into the patent for anyone skilled in the field to be able to replicate the invention. The patent is published by the Patent Office. A piece of software that actually runs on a machine is just another description of the patented algorithm. It's only the use that infringes, not publishing a translation of the description (which happens to be clear enough that a computer can "understand" it, too).

    Seems to me that if they can sue the open-source coder, they should also be able to sue the patent office - or anyone who publishes a textbook that describes the patented technique. B-)
  • The list above only mentions the Compaq Presario 5600 series. I just bought a 5800 series machine (5888 - their current 700 Mhz Athlon box) and it also has firewire.

    So it looks like Compaq is sticking with it for a longer haul.
  • Unfortunately, most (all) of the good video compression algorithms are patented and cannot be used.

    I was under the impression that patents block the use of the patented item COMMERCIALLY - but that you can make one for your own personal, experimental use, provided you don't sell it or the product of its operation.

    Assuming the above is true - it would imply that one could make open-souce codecs and use them to compress and view movies, and that these codecs could be distributed, provided it was for free.

    There would be a violation if you showed, gave, or sold decoded movies to others (which is typically already a copyright violation), distributed video you'd encoded using the patented codec, or charged (even a media fee) for a distribution containing the codec source. So no servers and no including the codec in a commercial Linux distribution - the customer would have to download it from an open web/ftp site.

    Do we have a patent lawyer in the house, to check whether my understanding is correct?
  • Having an open source codec is only half the battle. If the program that you are interested in is only available in a proprietary format on an OS you prefer not to use, you are sol, no matter how good a competing codec may be. I already stopped listening to one radio program that I could only get online, because it switched from Realaudio to a Microsoft only format. The open source codec has to be sufficiently superior to entice content providers to use it in place of or in addition to the proprietary ones. Otherwise, it is the operating system wars all over again. Having a superior OS on your computer does you no good if the program you want/need to run is only available for the inferior but dominant one. The same thing for streaming video content.
  • While many of the codecs supported by xanim [pubnix.com] are not free (in both senses... still don't understand that analogy), I'm sure there are some in there. It's a good list of codecs to start from if you want to check all of them. I was unable to find a page with a list of free/non-free video codecs in the 10 minutes I spent on it.

    Does anyone know what codecs could possibly be used for this?
    ---
  • What I don't understand is why they can't just call it "IEEE 1394" and be done with it.

    "My computer has 2 1394 ports!" :-)
    ---
  • This has all happened long ago! When motion pictures were invented at the turn of the century, patents were used by Edison and several large Eastern industrialists to try to dominate the creation and distribution of moving pictures.

    It's rather odd, but the same Hollywood that now screams to expand it's protection via questionable IP laws, once was a group of lawless pirates ready to flee to Mexico if the Edison patent police showd up. This story [economist.com] in the millenial edition of The Economist [economist.com] talks breifly about Hollywood's checkered birth.

    When an open source video streaming player gets under development, hopefully the project won't have to be run out of Uzbekistan to escape the Real/Microsoft/Apple patent police. If it is, though, thank Hollywood for the inspiration!
  • An excellent primer on video codec technology may be found here [lycos.com].
  • IANAL, but I think that even if you are not making any money from the implementation of a patented algorithm, you can be sued for damages related to the destruction of the POTENTIAL revenue stream that the company (or individual) might have made had they been able to use their patent. (The amount is, of course, decided during the court case...)

    Basically, this means that you can implement any patent you want, as long as you don't have any effect in the marketplace.
  • What's most needed at this time (in my opinion) is an open source ILS server that is compatible with (sorry) Netmeeting.
    To start pulling more people over to linux, we need to start converting server platforms, allowing clients to use either M$ or linux software.
    If there's no good free ILS/netmeeting server software available, the server side stays tied to M$/non-free alternatives, wich is not what's wanted.
  • This reflects my thoughts... motion jpeg already exists, the new ISO JPEG2000 standard looks good on the surface, probably not much of a leap to code a server/player combo that takes these new wavelet-compressed jpegs and sequence/buffer them.

    As a webcaster, we primarily use RealMedia because of the installed base of players. We're getting a lot more requests to use Windows Media, which makes great business sense because I don't have to invest nearly as much capital for the Microsoft products because of the dramatic difference in server pricing between Real and MS.

    Seems to me that the new wavelet compression provides an opportunity to create a better streaming application, particularly to the bandwidth-challenged.

    Rick Berry
    President
    pressevent.com
  • by Anonymous Coward on Sunday January 09, 2000 @10:24AM (#1389560)
    What about LBL's vic [netbsd.org]? It's open source, runs on Unix, and includes both its own codec and its own delivery system.

    There's a link on the vic page to IVS [inria.fr], another very similar also-opensource tool.

    I haven't used either of these tools myself because I can't afford the hardware, but I have a feeling that the Linux/Wintel crowd is ignoring a lot of research that has already taken place. Research people are often uninterested or incompetent in packaging and marketing their tools, but they may well have already done the hard part for you. And it looks to me like Apple and Microsoft stole most of their ideas from these guys.

  • by Yarn ( 75 ) on Sunday January 09, 2000 @06:37AM (#1389561) Homepage
    ASF's are semi-open, at least MS publishes specs [microsoft.com]

    The CODEC's are protected like crazy tho.
  • by Unknwn ( 646 ) on Sunday January 09, 2000 @06:38AM (#1389562) Homepage
    icecast (see http://www.icecast.org) is a program for the streaming of mp3 audio. But changes have been made to the program at various points in its development to allow it to also stream MPEG video across the 'net. Another hack was done at one point so that it would stream meta-refreshing jpegs.

    Now for the obvious question -- if it can be modified to support this, why doesn't it support it out of the box? Relatively simple to answer. At present, streaming MPEG video takes up a shitload of bandwidth. I seem to remember that the internal network of the developers was strained when they streamed the video. So, some sort of better compression of video is needed. Secondly, multicast would be really useful. It's a feature on the near-future TODO list, but will involve lots of rewriting as well as updating clients to support (for just the audio; clients don't even exist really for streaming mpeg video :)

    --
    Jeremy Katz
  • by heroine ( 1220 ) on Sunday January 09, 2000 @06:38AM (#1389563) Homepage
    There's a good article on http://www.dv.com/magazine/2000/0100/johnson0100.h tml on why you might not want to jump on the streaming bandwagon. Streaming was originally developed so you wouldn't have to wait for the entire movie to download. What happened was that people didn't like constant interruptions from network conjestion so the purpose of streaming video became copyright protection. When you see those stream-only websites, the reason they do that is to prevent you from copying it and uploading it to your website. On the other hand, Linux has pretty good low-bitrate downloadable solutions. MPEG-2 with sound can go as low as 200kbit/sec. VALinux, which does all its multimedia on Win NT incidentally, hosts 300kbit streams.
  • by jetson123 ( 13128 ) on Sunday January 09, 2000 @09:52AM (#1389564)
    Creating algorithms isn't any more difficult than implementing a large sofwtare system. Even if you want to argue that basic video encoding ideas (motion compensation, etc.) at one point were patentable, the basic ideas are so old that they are in the public domain by now (or should be soon). Most of what protects MPEG-2 and similar standards are tweaks, tweaks that have alternative workarounds.

    But even if it were, standards bodies should insist that methods that are adopted into standards are available royalty free to anybody, or at the very least royalty free for open source implementations.

    In fact, they often do, and many standards have and will have free implementations as a consequence (JBIG and JPEG2000 being two of them). I suspect that with MPEG-2 and MPEG-4, it's the influence of the greed and money of the media industry and Hollywood that causes them to be governed by "patent pools". At the very least, corporate lawyers look at the money those companies are making and wondering: how can we get a slice of that.

    I have no problem with taking a cut from big media companies. The problem I have is that we still need a free format for free content, for socially important content, etc., something that allows people who don't have a lot of money to get their content across the Internet, using a free infrastructure.

  • by jetson123 ( 13128 ) on Sunday January 09, 2000 @10:52AM (#1389565)
    Several people have asked what the big deal is if the video delivery technology is proprietary: it's cheap and people can still use it to publish whatever they want to.

    While proprietary streaming video technology is a lot cheaper than a television station, it can nevertheless exclude a lot of voices from participating in the media. How?

    First, proprietary technology, adopted as a standard and protected by patents, still ends up being a lot more expensive than equivalent free technology, and the price alone keeps people from using it.

    Think not only of the cost of the encoder or a small streaming server, but of the cost of putting up a large streaming server that can reach a lot of people. Companies developing the streaming technology are going to charge for the use of their proprietary server technology based on commercial uses; maybe that ends up being only a fractional cent per connection, but that quickly adds up and becomes out of reach for non-commercial content providers.

    A second problem is that the technology will likely be licensed selectively when it comes to large service providers. The heavy duty streaming video servers may end up only being available from a few large hosting companies, and those may have restrictions on content similar to those of large television companies. That hasn't happened yet, but it is another way in which proprietary distribution technology can limit available content. And the analogy to the broadcast networks and their bland, low-quality content is quite close--in the case of the broadcast networks, it wasn't patents but broadcast licenses that were exclusionary and led to the dumbing down of America.

    And lastly, proprietary technology is self-perpetuating. If you leave the development of standards and technology to a few companies, they are going to develop and patent all the "innovations" and perpetuate their hold on the market. Many of those "innovations" may only be simple tweaks, ideas that would easily and naturally occur to any open source developer, but they will nevertheless be protected for decades to come. By playing the patent game right, companies can keep content in proprietary formats in perpetuity.

  • by Graymalkin ( 13732 ) on Sunday January 09, 2000 @05:38PM (#1389566)
    instead of an entirely new codec why not just use a version of MPEG-2 compression? MPEG-2 does not mean CSS, just in case anyone is flame happy for whatever reason. MPEG-2 is designed to be streamed, it's what DTV among other things are going to use. The bonuses of MPEG-2 is that a good number of newer video cards have hardware decompression and a good deal of those cards have X servers already written for them so a good deal of their specs are already known. MPEG-1 and 2 are also open formats and any Joe Programmer can makes an app using the codec. The hardware acceleration is a big bonus but even without it a smaller video frame wouldn't be too difficult for your CPU to decode. One of the best advantages of MPEG-2 is the good compression ratio, even lowbanders on 28.8s and 14.8s could get a decent quality picture and sound.
  • by flimflam ( 21332 ) on Sunday January 09, 2000 @06:38AM (#1389567) Homepage
    QT streaming server can serve any codec -- it doesn't really care what it is. Likewise, the quicktime player will play multiple codecs, including MP3 and mpeg2.

    Also, does anyone actually stream mpeg2? This is a serious question. I haven't seen it, but of course I don't have much time to look for content at the only place I have access to enough bandwidth to make streaming video worthwhile.
  • by harmonica ( 29841 ) on Sunday January 09, 2000 @06:34AM (#1389568)
    Unfortunately, most (all) of the good video compression algorithms are patented and cannot be used. Even if some may not like it, Microsoft's ASF (players and encoders, surprise, are both free; looks like someone has a lot of resources) is very good, it uses MPEG-4, suitable for low bitrate video. From my personal experience, the competitors don't come close.

    The creation of new algorithms is difficult. It's one of the things that cannot be done easily by the open source community, which has brought us a lot of very high-quality products in other areas, because you need a very specialized knowledge (signal processing, lossless compression etc.) that only few people have. There was a very well-written comment by Eric Scheirer on this on Slashdot months (a year?) ago, but I couldn't find it.

    Even if there would be a product, standardization is another large obstacle. Take image file formats: There are products superior to JPEG (like DJVU or LWF), but no website creator uses them, many people wouldn't know how to install the plug-ins (which are available for free). I guess even JPEG2000 will have a hard time once it's out. The success will depend on decoders being built into standard browsers by default. And they should work flawlessly (remember PNG)!
  • by Weezul ( 52464 ) on Sunday January 09, 2000 @07:50AM (#1389569)
    The article is correct in it's concern that only an OSS broadcasting solution can creat the future we all want.. where anyone can broadcast there video over the internet.. and patents are our obstical. The solution appears to be ``compramize our principals in the short term, but not in the long term.'' I will explain..

    A legislation solution (I think this is RMS's proposed solution) seems unlikely.. especially for Algorithms that really do require an understanding of a good bit of math to create.

    Many mathematicians are getting kinda sick of inventing some cool new thing to help sociaty only to see some company steal it's applications by putting one word like computer, audio, or video in front of it.. and many researchers could probable be talked into giving the open source community a list of possible applications when they invent something new.. so we could patent it first and use a viral patent license. Unfortunatly, this is considered too expencive by people in the know (like RMS). I personally feal this is workable if the patenting orginisation were less principaled then the FSF could pull something off by doing the following:

    1) Use an extreamly viral license the prohibits mixed use of the patent with any patents not distributed under the same license, i.e. no company can use our patents unless the cross licence with the rest of the world via our license.

    2) Allow companies to get arround 1 by paying thorugh the nose and use the money to reward the researchers and pay for lawyers, i.e. compramize our principals occasionally.

    Unfortunatly, the above dose require a lot of orginisation, a lot of conenctions, and a lot of work.

    The best solution is probable: control the viewer and server. If a codec maker dose not want to go OSS then we should make it hard for them to get users.. up to the point that they are really a hell of a lot better then anyhting we are using.. at which point we should compramize a little.

    It is also worth mentioning that codec makers make money by licensing the authoring software (i think) or with crap attached to the codec, so there should probable be a push to implement OSS versions of the authoring software and codecs in countries where the patent dose not apply.. then make the OSS version the default, i.e. default RedHat xanim has no support for the codec so the user is forced to choose between downloading a single OSS xanim which is illegal or downloading a million codecs. This will cut into there proffit margin.

    Jeff
  • by Anonymous Coward on Sunday January 09, 2000 @06:24AM (#1389570)
    While Microsoft and Real use proprietary streaming protocols, the Apple QuickTime software uses non-proprietary documented streaming protocols (RTP, RTSP). So while it is more difficult to create software that will interact with Microsoft and Real servers, it is less so to create one that would interact with a QuickTime server or create a new streaming server based around the same protocol.

    The real problem are the video codecs. Most of them are not even owned by the respective companies (e.g. most of the high-compression, high-quality codecs that Apple/MS use are licensed from other companies - Sorenson, Cinepak, Indeo). Note that for some of these it is possible to create freely-distributable binaries (a la Xanim) but it is highly unlikely that these companies will release the algorithms or source code any time in the near future.

    What would probably be the best bet is to scour the academic literature on video compression and/or bug people in the field for info and create an alternative codec that is freely distributable. You could then produce a plugin for QuickTime, VFW, etc. to encourage the general adoption of the codec. If a content-based company can run a server and deliver content without technology licensing fees, I'm sure they'd be interested.

  • by Anonymous Coward on Sunday January 09, 2000 @08:03AM (#1389571)
    MPEG-2 video looks reasonable at 3-4mbit/s. H263, which is what netmeeting uses, can be tweaked to be TV quality at 1mbit/s. MPEG-4 will probably look OK at less. "Talking head" video, for videoconf, is easier to encode and looks ok at 64kbit/s

    There are three things a video codec requires:

    • spatial transform, eg idct, which I think is patented. There are many wavelet types, some of which may not be spatial transformed are used to find the most significant information; the rest is chopped.
    • Coding scheme: huffman is unpatented, syntax-arith encoding is (but when does it expire?) Both require that lots of analysis be done on symbol frequency. This is tedious, not difficult.
    • motion compensatio: in oreder to make use of duplication between frames - again, not hard.
    . The open source community should not write off the idea of doing its own video codec. The most difficult part is the transform; if a decent unpatented one can be found, the rest is within the capabilities of many coders. Video codecs are EASIER to write than audio codecs, as the proportion of information that the eye actually percieves is lower. In order to do decent compression in audio, very complex models of the ear are required; this is not true of video. Oh, and the eyes colour resolution is lower than its luminance resolution
  • by Effugas ( 2378 ) on Sunday January 09, 2000 @06:28AM (#1389572) Homepage
    Lots of companies with large amounts of money invested in overly complicated streaming systems will complain, but there's a real bottom line:

    The standard video format is MPG, because MPG Just Works. Everywhere.

    AVI has failed. The general perception of an AVI file is one that might play, might not, might suddenly install a new codec, might not, who knows. No predictability.

    MOV? All the pain of an AVI, with free delays while you deny Apple their cut. From the guys who killed Firewire...

    RM. Realmedia ain't bad, but it just doesn't scale up too well. There's this common delusion that only people with broadband links should be able to view high quality video--in this paradigm, RealMedia can do OK, since relatively few people have consistently extreme high bandwidth links to the Net. But, ya know what? This paradigm leaves millions of people unable to view high quality video, except on television.

    Presume people can download clips and watch them later, and suddenly the stream-biased, bandwidth-capped format that is RealMedia suddenly looks stale and chunky.

    The bottom line, beyond quality issues, is that MPG has won for the same reason MP3 did: It Works. All the various copyright protection systems are obsessed with creating situations where the consumer tries to do something and It Doesn't Work. As I'm sure the consumer trials are showing, when Things Don't Work, consumers simply refuse to buy in. And that's the key--the investors may fund, the studios may create, but it's the consumer that pays for it all.

    MPG may not be a low bandwidth streaming format, by any means, but the general obsession of streaming--and streaming only--is short sighted at best, and suicidal at worst. It will be interesting to see how this pans out over the next few months.

    See y'all at the DVD trial...

    Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com
  • by phutureboy ( 70690 ) on Sunday January 09, 2000 @08:17AM (#1389573) Homepage

    RTP/RTSP is an open IETF standard, and is used by both RealNetworks [realnetworks.com] and Apple's Darwin Streaming Server [apple.com]. Real *used* to use a proprietary protocol called PNM, but no longer.

    The codecs are indeed the problem. Most are patented, and licensed only in binary format.

    From what I understand about it so far, it seems like MPEG4 is our best bet for an open codec. If I recall correctly:

    MPEG1 = Up to VHS quality. File-based format not at all suitable for streaming.

    MPEG2 = broadcast-quality video, very compute-intensive algorithms require expensive coding/decoding hardware on either end.

    MPEG3 = was abandoned during development, I forget why

    MPEG4 = open spec for encoding AV content, a new revision specifically designed for Internet streaming & multimedia is currently under development

    More info about MPEG can be found at http://drogo.cselt.stet.it/u fv/leonardo/mpeg/index.htm [cselt.stet.it].

    Also, the best introductory resource I've found is at http://www.nyquist-media.co .uk/streaming/streaming.html [nyquist-media.co.uk]. It's an excellent overall introduction to the technical standards and mechanisms of streaming video.

    On a related note, is there a consensus on the license for Apple's Darwin Streaming Server? Is it open enough?


    --

  • by Mononoke ( 88668 ) on Sunday January 09, 2000 @06:38AM (#1389574) Homepage Journal
    Yes, it's Apple's version of "open source", but it is available:

    Quicktime streaming source [apple.com]

    Can't help ya with on the client side. Maybe Apple doesn't know you want it.


    --

1 1 was a race-horse, 2 2 was 1 2. When 1 1 1 1 race, 2 2 1 1 2.

Working...