Not sure what Schwartz is up to these days but here is what Larry Ellison had to say about him recently: "The underlying engineering teams are so good, but the direction they got was so astonishingly bad that even they couldn't succeed," said Ellison. "Really great blogs do not take the place of great microprocessors. Great blogs do not replace great software. Lots and lots of blogs does not replace lots and lots of sales."
http://abcnews.go.com/Business/wirestory?id=10630034&page=2 -James -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Chris Adamson Sent: Sunday, May 16, 2010 6:10 PM To: The Java Posse Subject: [The Java Posse] Re: H.264 explainers A couple of open-ended replies... hopefully getting off the specific points and on to bigger issues that I really hope people will consider. * Dick cites, with seeming skepticism, Jobs' stated preference for an open web. Of course Apple wants an open web -- when Windows has an 18- to-1 market share advantage over its nearest competitor <http:// marketshare.hitslink.com/operating-system-market-share.aspx?qprid=8>, anything closed on the web is going to tend to prefer Windows and marginalize Apple's (and everyone else's) platforms. This is the long- standing gripe of Mac owners about Flash, that its performance is wildly behind that of Windows (leading some to speculate that Flash's internals are tightly coupled to Windows, dooming ports to other platforms to inferiority). I do think Apple's idea openness stops at the web. That much seems obvious. Still, when Apple goes closed, they do so through native applications. For example, they created the iTunes Music Store not as a web storefront, but as part of the iTunes application (and later, as a stand-alone app for iPhone OS). * I don't doubt for a second that Lala will drop its cross-platform support, if Lala even continues to exist in its current form. That should be expected of a company that gets bought out: their mission changes. As an independent company, they tried to maximize revenue by getting the largest audience, and being cross-platform helped. As part of Apple, the employees and their technology will presumably be made to serve Apple's strategic goals. This is normal and happens all the time, even if it does suck for loyal customers who are left in the lurch by the change in direction. Surely Apple fans remember when Bungie stopped developing Mac games and started focusing on XBox once they were purchased by Microsoft. Still, I bet you won't see an Apple- only Lala web experience - the stuff will likely get assimilated into iTunes (cf., early point about preference for native apps for closed services) * I don't think it's an overstatement to say that Theora (or VP8 for that matter) is a "non-starter in the real world". It's essential to understand how total H.264's adoption by the industry -- not just Apple -- really is. When I checked some content-oriented websites, I found that DV magazine (which focuses on equipment) had 0 hits for theora, streamingmedia.com had 11 hits (mostly opinion pieces about codec politics and not reviews or tutorials) versus 248 for H.264, and legendary retailer B&H Photo and Video in NYC assumes that "theora" is a misspelling of "theory" (by comparison, they have 269 products that use H.264). Links at my blog <http://www.subfurther.com/blog/? p=1065>, which you should NOT read if you don't want to see me ripping on the FSF. And it's also important to put web video in context: while it is growing very quickly, online video viewing (as measured in minutes) is outnumbered by television viewing by a factor of more than 50-to-1 <http://blog.nielsen.com/nielsenwire/online_mobile/americans-watching- more-tv-than-ever/>. Even as that grows, it's not going to be big enough to be able to legitimize an H.264 rival before H.264 is simply displaced by the Next Big Thing * Another thing that a lot of people don't understand about video is that (very important) *an encoder is not a state machine*. By that, I mean that there is more to codecs that just the technical traits of the relevant algorithms... the ecosystem around the codec may actually be more important. The MPEG codecs have always standardized decoding, but not encoding. Any process that produces a bitstream playable by the reference decoder is a valid encoder. The idea of this is to foster competition among encoders (which is thought to be better able to fund R&D). The result of this can be rather amazing. According to the MPEG-IF (a separate entity from MPEG-LA), the bitrate required for "broadcast quality" encoding of MPEG-2 dropped from over 6 Mbps to 2 Mbps over just a few years <http://www.mpegif.org/public/documents/ vault/m4-out-20027.pdf>. This is a HUGE win: if quality is constant, as in web video, you send the same quality file in 1/3 the time. If bandwidth is constant, as in satellite TV, you get to cram more channels onto one transponder, or more videos onto your iPod. The takeaway is that the popularity and size of the ecosystem around a codec does matter: it produces better encoding software, more (and more knowledgeable) professional compressionists, etc. Choosing the little-used Theora and VP8 codecs opts out of these manifest advantages. * Yes, H.264 is patent-encumbered and requires royalties at several toll booths (always encoding and decoding, potentially large-scale distribution too... see <http://www.streamingmedia.com/Articles/ ReadArticle.aspx?ArticleID=65403>). But, I don't think a lot of readers here understand that the software community's distaste for patents is not shared by the world at large. Indeed, when I was in the television industry, we didn't think twice about adopting proprietary tools and formats. In fact, most of the stuff we worked with at CNN was known by their trademarked names: "chyron" (character generator), "avid" (video editor), "grass valley" (video switcher), "ultimatte" (still image compositing and effects), "betacam" (a Sony videocassette format), etc. The anti-patent advocates have done nothing to convince people outside of the software industry that patents are unconscionable and to be avoided; as a result, few video professionals know or care about H.264's patent encumbrances, as the licensing costs are so low (relative to the rest of the costs of being in the video biz at all), and the advantages of a patent-free format are irrelevant to them. Also, the objections to H.264 seem like rather selective outrage. Compact Disc, 802.11 wi-fi, and USB are all exactly like H.264 in that they are patent-encumbered, royalty-owing public standards established by industry bodies. Are there any software engineers who seriously don't use any of these? So, big points that I hope people will take away: * Codec and DRM are two totally different things. Theora wrapped in Apple FairPlay would be just as unplayable on Linux. * My original post speculated that moving the DRM to the network, through schemes like HTTP Live Streaming, offers some hope that DRM- crippled products might at least be able to support more platforms than client-side solutions like Flash and iTunes. Nobody replied to that part, so I guess we'll just hope for the best. * H.264 has absolutely, unequivocally won. There is no serious discussion about this among people whose lives and businesses depend on digital video. * The anti-patent community (and the free software crowd in general) really hasn't engaged the world beyond software professionals, as the flock to closed platforms like iPhone OS should illustrate. Actually, that's the optimistic reading. The pessimistic take is that they've tried, and failed. * Apple operates in its own self-interest. Most companies do. A few don't. Speaking of which, what's Jonathan Schwartz doing now? --Chris On May 16, 11:42 am, Dick Wall <[email protected]> wrote: > On May 15, 3:17 am, Chris Adamson <[email protected]> wrote: > > > > > In no way was comedy bit directed personally at you. I don't think > > you flog Ogg and/or Theora. I was really thinking more of the knee- > > jerk reaction we see from the Slashdot-type readership, the mob > > that's been trained to say this kind of thing because their peers > > all agree with it. Enlightened Linux fans realized long ago that an > > "everything should be in Ogg" argument is a non-starter in the real > > world (ESR was saying this like four years ago). > > I find a statement like "non-starter in the real world" just as > sweeping and overgeneralized as "ogg should be everywhere". The > standard used to distribute video is pretty unimportant, what matters > is that people can consume it where, when and how they want. If > YouTube switched to Ogg encoding, people would only care if it made > life harder for them, if it made life easier I believe it would go > without comment (maybe even some kudos, but people tend to complain > rather than praise). Sure, camcorders are probably unlikely to change > established standards, but since just about all video gets transcoded > in some way (usually for size) before it gets served over the net, > that's just a hardware detail. The point is that streaming and the web > are the distribution methods that are likely to be the most important > in the coming years, and who gains control of that (if anyone does) > will decide a lot about consumer rights or lack thereof. > > > As someone (maybe Daring Fireball?) pointed out, the story is that > > Adobe isn't open, but say they are. Apple isn't open, and don't > > claim to be. > > But this is the quote taken directly from Jobs' missive, and it's not > even taken out of context: > > "Apple has many proprietary products too. Though the operating system > for the iPhone, iPod and iPad is proprietary, we strongly believe that > all standards pertaining to the web should be open. Rather than use > Flash, Apple has adopted HTML5, CSS and JavaScript - all open > standards. Apple's mobile devices all ship with high performance, low > power implementations of these open standards. HTML5, the new web > standard that has been adopted by Apple, Google and many others, lets > web developers create advanced graphics, typography, animations and > transitions without relying on third party browser plug-ins (like > Flash). HTML5 is completely open and controlled by a standards > committee, of which Apple is a member." > > That's a fabulous statement if true, but if I even catch a whiff of a > video standard that Apple is pushing that will only play on Apple > blessed platforms, there is a massive I told you so coming! > > If Steve is as good as his word, these open standards that Apple > believes should pertain to the web will provide all platforms an equal > footing, right? That's certainly what it sounds like to me. Now, who > believes that will be the case versus what I suspect is a more > realistic fulfillment of the Apple vision, where it works perfectly > with MacOSX and on the iPhone and iPad, works somewhat less well on > Windows desktop machines (does anyone think that iTunes on Windows is > anything but slow, buggy and uncomfortable these days, it used to be a > lot better). > > But, enough speculation, I think we are likely to see some real > evidence on the openness vision pretty soon. Apple bought Lala - a > nice little music service I used (on Linux too) and really liked. They > just culled it, and rumor has it that they will be bringing out iTunes > live based on it. Anyone want to hazard a guess whether Linux, > Solaris, Android, Palm, Windows Mobile or anything other than Apple > products and windows will be supported by the new service? How > committed are Apple to open content really? I think we are about to > find out. > > Alternatively, does the openness stop at the web, and not include > other content like audio and video. Certainly this statement leaves > that question open - it masterfully implies one thing (openness) but > could go either way. If so, that question should be cleared up before > everyone jumps on the "flash should die" bandwagon. (Incidentally, > Lala used flash, certainly not a perfect solution, but it did work on > quite a few platforms and devices, and would have probably worked > nicely on the upcoming Android 2.2 with it's flash support). > > > > > > > > > I find it's very helpful to shift the analogy of Apple's products > > from being like other computers to being like console and hand-held > > videogame platforms. Actually, devices like XBox 360's and Nintendo > > DS's are far more tightly controlled than iPhones and iPads -- > > development tools are hugely expensive, licensing restrictions are > > far more severe, and the manufacturers rigidly control retail > > distribution (e.g., games can only be on the shelves for a certain > > number of months, after which they're pulled, unless they qualify > > for a "greatest hits" type release). Still, like Apple's mobile > > devices, consoles, and handheld games are also closed systems that, > > if they adopt any open standards, do so only for the sake of making > > the product more appealing for their own end-users. Nobody at Sony > > or Nintendo claims to be promoting standards or formats that are > > going to be of any use to Linux, or computer users of any stripe. > > Let's also note that all these devices are _very_ popular with end > > users. The simplicity and enjoyability of the end-user experience > > should not be underestimated. > > True, but neither are Sony, Nintendo, XBox etc. writing about how they > support open content on the web. Also, let's not underestimate how > widespread Windows became despite being inferior in user experience to > the Mac. Cost was probably one factor, but openness, particularly to > developers, was another that worked in its favor. > > > The H.264 codec is open to the degree that it was developed by MPEG, > > an ISO working group, and ratified through a public standards > > process. It's not free, because it contains patented technologies > > that require royalty payments. The DRM on DVDs, digital copies, and > > iTunes is a separate issue. > > It is a separate issue until that DRM makes it on to web content - > then it is no longer a separate issue, it is *the* issue - who > controls the access you have to online media. Right now Adobe has a > big hand in that, and it kind of sucks (particularly as you point out > later for BSD and other OSs that don't have flash video available - > including the iPhone and iPad of course). Fast forward to a future > with "open" H.264 video with DRM available only on Apple platforms and > Windows (going by Apple's current record, that is the extent of the > support outside of apple technologies) and has anything improved? I > think not. > > > Google's reach ends with the web. Of course, that's huge, but > > consider all the other communications media that deliver video: > > broadcast, cable, and satellite television, physical media (DVD, > > Blu- Ray), video telephony, personal devices (camcorders, Flips), etc. > > Since I came to Java from the cable TV world, I see web video as > > just part of the picture, and doubt that Google has the reach to not > > only flip web video, but to extend that into other media. > > Especially since the differences between the various > > current-generation codecs really aren't that great (they all have > > similar performance characteristics, because at the end of the day, > > they're almost all based off the same DCT algorithms, and then add > > in a lot of the same ideas to achieve additional compression, like > > motion estimation... they're already pretty close to the limits of > > information entropy). > > But, according to the Steve Jobs article, the web is what we are > talking about here. Apple already has its closed standards for video > delivery onto its platforms, what it is trying to do (apparently, this > is of course all my opinion) is extend that reach onto the web. As I > have said before, I would absolutely love to be proved wrong - it > would be easy enough, Apple could just push totally open video > standards that work everywhere on all platforms equally. I will then > stand there quietly while every Apple fan on this group tells me in > turn "I told you so" and you will not hear a peep out of me (mainly > because I will be happily watching videos on my Linux boxes). > > I don't see that as a likely outcome though. In the second response > you ask: > > > how would we feel about Flash if they were neglecting Linux like > > they did for the first half of the 2000s? > > Let's also note that their support for "many other platforms" only > > goes so far... Flash doesn't support other F/OSS operating systems > > like BSD or Haiku (nor do I suppose we really expect them to), they > > don't support Linux on any CPU architecture other than x86, and > > their official distro support > > <http://www.adobe.com/products/flashplayer/ > > systemreqs/#os> is limited to Red Hat Enterprise 5, openSUSE 11, and > > Ubuntu 7 and up. > > Well, it would suck, and I would be complaining about it loudly as a > Linux user (and I was back in the day, quite loudly). The point is, I > don't see this play by Apple as doing anything but making it worse! > It's simple - more platforms == more open == more better. Flash > already has x64 implementation beyond the 32 bit x86, but I am not > going to get distracted by that. What does it matter if H.264 players > can play on any platform out there, if the DRM to decode the video is > only available for 2 or 3 blessed platforms? If, as you I think > correctly point out, media will only be available from the rights > holders if it is DRMd, then who controls the DRM controls the media > (and pretty much the world) and this is, I believe, the play Apple is > trying to make under the guise of open standards. Is my analysis > wrong? Does someone want to make me happy and tell me it's going to be > alright for us Linux guys (not to forget Trond and his love for ... > > read more ยป -- You received this message because you are subscribed to the Google Groups "The Java Posse" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/javaposse?hl=en. -- You received this message because you are subscribed to the Google Groups "The Java Posse" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
