A few days ago, Google opened up a YouTube area on its product ideas website, so that people could submit their ideas about how to improve the ubiquitous video website. It turned out it was a bit of a can of worms Google opened there: everything focussed on HTML5 video support.
The top voted idea was adding support for HTML5 video and ditching Flash, which had twice as many votes as the second highest item. Soon enough, similar HTML5 video suggestions popped up, and nine out of ten top ideas were about HTML5 video. To make it all a little bit more frantic for Google, Slashdot picked up on it, further cementing HTML5 video support as the top-suggested feature.
Google has now given an official response, and is hinting that support for HTML5 video may indeed come:
We’ve heard a lot of feedback around supporting HTML5 and are working hard to meet your request, so stay tuned. We’ll be following up when we have more information. We’re answering this idea now because there are so many similar HTML5 ideas and we want to give other ideas a chance to be seen.
All HTML5 ideas have been moved to the “answered” section accordingly. Let’s hope this means we’ll see proper HTML5 support soon on YouTube, as the site remains one of the main reasons why you’d want Flash installed. It could mean a major blow for Adobe if YouTube ditched Flash.
The sooner we get a big player behind video in HTML5, the sooner we get rid of the abomination called Flash. Sure, it’ll take years but that’s just another reason to get started ASAP.
Flash maintained its 98% penetration for something like a decade before it started playing video. The video tag has great merit, but you’re going to be sorely disappointed if you think it’ll be the end of Flash. html5 video won’t even make a dent in Flash.
Even a dent of a dent is fine with me. Really, the only thing I care about with flash is video. And just on a couple sites, at that. Even if just 0.1% of the net moves over, I’ll be thrilled if it’s the 0.1% I actually care about. If a site’s using flash, for, say an intro or navigation menus it’s usually not one I want to go to in the first place.
You’re right that it isn’t going to end with the video tag but at the same time, with each piece of technology being developed there are less and less reasons to use Flash or Silverlight. I remember when I first started usign the internet over a decade ago. Flash and Java Applets were abused to buggery but as soon as new and exciting open standards came online there were less and less reasons to use them.
Microsoft has realised this and hence are starting to add new standards support to Internet Explorer 9 – they realise that it isn’t a matter of either/or – that there are some people who want to use Silverlight whilst others who don’t want to use it. What one might see eventually is a rise in the number of web standards based websites but at the same time people will still develop for Flash/Silverlight.
For me it isn’t a matter of dogma or religion – just simply the matter of the fact that there is no decent Flash plugin. The Flash plugin by Adobe is absolutely crap but if they turned around tomorrow and said that they were to open it up under a licence like CDDL or BSDL or some liberal licence, not only would the collective efforts of the open source community muck in and fix up long standing issues, there would be some real progress forward based on those who use it on a regular basis – endusers and so forth.
At least for YouTube, a Flash-less version is already possible. Install Greasemonkey (if you use FF) and then this script: http://userscripts.org/scripts/show/50771
It causes YouTube to embed the native MP4 files without Flash. You need either an MP4-compatible plugin installed or a browser that can play MP4 files using the <video> tag (AFAIK currently only Chrome and Safari, but I have no idea if those support user scripts).
Why not provide HTML5 for people with modern browsers, while providing a flash version for people stuck with outdated browsers (and a big prominent ad telling users to get a more modern browser).
http://camendesign.com/code/video_for_everybody#video-what
what they need to do is not launch 2 websites. and if your browser doesn’t work. then have a link to a new web browser. people need to upgrade their browsers anyway for security reasons. you tube would just give them another reason to do it now, and not 10 years from now.
While everyone might be jumping up and down with joy, I am not that thrilled.
You see, even though it has a lot of downsides, flash has some merits that are easily dismissed. Remember the time before Youtube and flash video? The videos distributed in the ‘net were encoded with various encoders and you never knew, even if you had the latest greatest codec packs or wmp plug-ins, if the video would play or if it needed a special codec (that may have even cost money). With flash video, everyone could watch videos without worrying about codecs or stuff like that because they were built into the platform. This unification receives a lot less credit than it deserves (kinda like Microsoft but let us not go there).
Now, step in html5 video. While it does relieve us from the curse of the Flash plugin, it also resurrects the nuisance of a gazillion codecs for actually displaying the videos. And here we go again with the video (and audio too, of course) spectacle:
* There is no de facto standard
* Numerous OSS\GNU/Free(tm) people climb out from under their rocks and start preaching about how their Free stuff should be used instead of technically superiour stuff that smart people created for profit or instead of stuff that some other guys put together on their free time in their basement but it’s not the right kind of Free(tm)
* Numerous commercial entities start aggressively pushing their products in attempt to create an aforementioned de facto standard and then cash in on it BIGTIME. This includes “spreading the facts” and other types of badmouthing.
* We restart dealing with malware-laden codec packs in attempt to view our favourite rick astley videos, and while doing so, greatly contribute to creating great and cool botnets for other people to capitalize on.
* Various tech columnists and other people blow a lot of hot air around discussing TEH BEST(tm) codec and also trying to make themselves look smart (kinda like I’m doing right now, but that’s beside the point) while others that follow news sources, that these “declarations of smarts” appear in, get headaches and need to get drunk to remedy it.
Long story short, I am not sure this is what we want. It’s like going back to the good old Win9x age – if YOU want to restart your pc after changing the desktop wallpaper then please, do so, but do not enforce this kind of behavior on me.
Well were I am living the most popular Browser is Firefox 3.5
OK, Theora is not as small as x264, but for Google bandwidth is essentially free and they have lots of space on their millions and millions of harddisks.
And Dirac (developed in the BBC basement) is on its way to be ready and it is in many ways more advanced than x264, although chips need to be faster (wait 18 months)
MS lost the browser wars. Even German authorities nowdays advise citizens to NOT use IE.
Just include the free codecs and x264 and then no codecs are needed. Web developers use what most people have.
Knock yourself out!
All doom and gloom are we?
I see only positive things in this, because on the internet Youtube sets the standards.
“This site is best viewed in Netscape Navigator 4.0 and greater”
“Viewable only in Internet Explorer 4.0 or later”
Mighty useful if the most used codec in the video tag turns out to be SuperVideoSoft NextGenSuperVideoCodec(c) available for free (mind you, not Free(tm)) from the SuperVideoSoft website, but not for redistribution.
Oh I will, believe me, but this doesn’t remedy the situation a bit.
He was talking about the codec. There is no de facto codec standard for HTML5. It’s a draft standard and they weren’t able to reach an agreement on the codec.
Their bandwidth isn’t free, it costs them millions per year.
Google has noted that there’s no way the company could serve up YouTube’s billions of streams using the much less efficient Ogg Theora codec, saying it would consume the world’s Internet bandwidth due to its less sophisticated compression
http://www.appleinsider.com/articles/09/07/06/ogg_theora_h_264_and_…
Firefox = Theora = Standard
AND
Youtube Bandwidth $ = 0
http://www.wired.com/epicenter/2009/10/youtube-bandwidth/
Edit: And theora might be a bit worse than h264 but it certainly wouldn’t eat the worlds bandwidth, that is just stupid FUD.
Edited 2010-01-16 21:02 UTC
Most people don’t use Firefox so I’m not seeing a standard.
There’s no such thing as free bandwidth. That title is misleading. It just states that Google buys their own infrastructure and they *may* have enough to just trade with other ISPs. It’s speculation on the part of a single researcher not connected to google. It also says nothing about investment or maintenance costs of all that fiber.
Ohh yeah, Flash is so much more secure than my own trusted set of codecs, it’s even better, to me, that it is a great single exploitable target (under heavy fire right now), making it even easier to create thousands of botnets with an ever growing number of users, instead of two or three using tampered “Mittey” Pirate’s mng codecs.
I’m going to have a go at this FUD also.
Actually, Flash also does not specify the codec. Most of YouTube once was H.263 (which Theora can beat). Lately YouTube has moved to h264. After the end of this year h264 will no longer be costless, and Google will be using something else. They will possibly move to VP6, VP7 or VP8 (having bought On2 technologies), but Theora (which is effectively enhanced VP3) is also on the cards one would think.
The de facto standard would be the one that is most used. Right now, that would be either h263 or h264 delivered via Flash.
h263 delivered via Flash (which is a good portion of the video currently on the web) is most decidedly inferior to Theora 1.1. h264 is marginally better that Theora 1.1, but not by much, and delivering it via Flash burdens it. HTML5 using Theora 1.1 would be better and cheaper than h264/Flash. You got it backwards.
We will get badmouthing, of that there is no doubt, as your own post shows. meanwhile, when h264 starts costing a bomb, and considering HTML5/Theora is readily available, already implemented in Firefox, yeilds equivalent performance and is absolutely free … the choice will become a no-brainer.
If we get HTML5/Theora, the codecs can be open source. No malware.
Its kinda hard to beat “same performance but free”. Theora will win in a no-brainer contest.
Why is it acceptable to you up until now to have been required to install a problematic Flash plugin in order to watch most video on the web, but now that you can be freed of that need you start complaining?
You are making no sense here.
Edited 2010-01-16 14:34 UTC
Flash doesn’t specify a SINGLE codec, but it does specify 3 or 4. Which still means you can create one file and be reasonably-confident that most users will be able to view it (in contrast to the “codec wild-west” that manjabes describes).
While I can’t speak for manjabes’ intentions, his post doesn’t give me the impression of “FUD” or “badmouthing. Those terms imply that he is either attempting to hold up the status quo as “the way things ought to be,” or that he’s going out of his way to attack the alternatives.
My impression is that he was simply pointing out the reality of the current situation – rather than trying to portray the current situation as something positive or desirable (a subtle, but important distinction). Realists and fanboys aren’t do sometimes come to the same conclusions, but that doesn’t make them interchangeable.
“Can” being the operative word. It’s possible, but by no means guaranteed or even particularly likely.
If HTML5 video does catch on, my fear is that web devs & content producers will just take the proverbial path of least resistance (AKA go with codec/container combo that with largest installed base) rather than moving en masse to a more open format. And today, that would mean Windows Media or Quicktime (given that choice, I’d take Flash any day of the week).
For techies and hobbyists, yes. But put yourself in the shoes of a commercial producer/distributor of online video content (content aimed at a general, non-techie audience). In that situation your primary goal is to get as large an audience as possible, so a primary concern is “how much effort will Typical Joe have to go through in order to view my content.”
That’s because there is a well-documented inverse relationship between the amount of effort it takes to view content & the number of people who will actually undertake the effort. Fifteen years ago, that meant obsessively optimizing image files to keep page load times down – today, that means providing A/V content using technology & formats that most users already have installed (Flash).
So you’re (hypothetically) free of the need to install a single Flash plugin, in favour of the need to install 3 or 4 different plugins instead (unless Apple and Microsoft adopt Theora – not something I expect, to say the least).
Most users are not going to see that as a step up.
I’m sorry, but you should educate yourself. Google already pays licenses for AVC: http://www.mpegla.com/main/programs/AVC/Pages/Licensees.aspx (#223)
Just like Apple moved to PowerPC after buying PA Semi?
No. Google more likely just wants a better performing AVC encoder. Getting experienced staff in-house is a good way to achieve that.
Do you seriously believe that? In low bitrates it’s somewhat true, because 5MB vs. 6MB is not much in absolute numbers (while it is in relative numbers), but try delivering HD content (YouTube recently adopted 1080i) in AVC High Profile and Theora and you’ll see the difference very clear.
The higher the resolution, the worse Theora’s performance. Dirac is more competitive at such resolutions. On top of that, unlike Theora, Dirac is actually standardized (as VC-2).
You are completely right. It’s been so long that I had forgot about codec hell on the internet. Quicktime, AVI’s 100 different versions, mpeg, Real (they’re still out there, probably waiting with venture capital for html5 video). I do not look forward to that, and it will be a huge step backward. And since html5 backed down from specifying a standard video format, we’re all basically moving back to the embed tag!
Shit.
Yeah, f–king opensource! Just like Firefox ruined the internet for everyone, when we were so content with IE! .. sense the irony? ok, here’s a line smartass: “Stick your binary plugins where the sun does’nt shine!”
The sooner we get rid of binary plugins, like Flash and Silverlight the better!
Your views are unrealistic. You should take a crash course on codecs before attacking FOSS and royalty free codecs. I would have no problem if the standards were made for free for FOSS and paid via the price of the product. for example MPEG4 encoders decoders at the software level cannot touch the container if a licence is not bought. That is why we need
a. Free codecs
or
b. Free standards paid on device purchase per consumer.
If b does not hold a is preferable.However if vendors followed a and disclosed the HW API (NOT HOW THE DEVICE IS MANUFACTURED) then you could see a jump in quality.
theres no way google would use theora. why? because of bandwidth. bandwidth usage is growing and growing on youtube. if they use theora the amount of bandwidth use would rise considerably. even if it’s just a kilobyte larger, how many videos do you think youtube serves a day? a week? a year?
if you want google to use theora, you better get working on making it more efficient at smaller bitrates than h.264. because theora might be cheaper for firefox to use, but h.264 is cheaper for google.
They won’t, if their purchase of On2 is finalized.
They’ll be pushing VP7 or VP8 instead (Theora is based on, although also a significant improvement over, VP3).
That’s why I would really love to see Google use Dirac. It’s just as free as Theora, but just as good, if not better, than H.264. It would save them a lot of bandwidth. The only downside it that it’s rather CPU-intensive on the client side (but on YouTube quality movies, it shouldn’t be too bad).
If Google implemented it in Chrome and Firefox implemented it too, it easily could become the standard codec for the video tag.
They won’t use Dirac, they’ll use VP7 or VP8… or whatever they end up developing now that they’ve bought Ontoo. They did buy them for a reason, after all. Of course, they could always just stick with H.264. youtube videos themselves aren’t encoded in flash video, that’s just the mechanism used to display them in a web browser. Their videos are actually encoded as H.264 outside of an flv container. Since most people are going to have H.264 anyway, the only reason I could see Google changing away from it is the up and coming robbery, excuse me, licensing changes in H.264. That is a pretty big reason though, and they must have seen the writing on the wall or they probably wouldn’t have purchased Ontoo in the first place.
Right, I forgot about that. That probably would make more sense for Google. If they made the codec free to use, it could become the standard for the video tag. How does VP8 compare to H.264 and Theora?
VP8 is claimed to be better than h264 and Theora (which are about level with each other).
However, the people who claim that are On2 technologies themselves. There is quite some controversy over this claim.
However, the people who claim Theora is as good as H.264 are Theora and Open Source supporters themselves. There is quite some controversy over this claim.
(Note: I personally do think that Theora would be good enough for YouTube, but you have to admit that H.264 is definitely better if used properly, i.e. not how YouTube uses it.)
And which browser except Chrome would play VP8 videos?
It required the strength of the entire media industry to wrestle MS’ Windows Media formats down and force MS to adopt MPEG-4 AVC and AAC.
MS won’t support VP8 in Internet Explorer 9 (which will support the <video> tag) just because Google asks nicely.
Correct me if I’m wrong, but I was under the impression that Firefox supports exclusively Theora for HTML5 video….
Edited 2010-01-17 09:17 UTC
You think that Firefox users are the majority?
Last time I checked, IE still had a market share of 60-70%. Microsoft now also ships AVC codecs for all its platforms (Windows Mobile / Zune, Xbox 360, and Windows). IE9 will also support the <video> tag — therefore likely also AVC.
Apple supports AVC in Safari/QuickTime on desktop platforms and uses it exclusively on iPhones.
Google ships a AVC decoder with Chrome.
IIRC Safari and Chrome have a combined market share of roughly 9-10%.
That means that AVC has a market penetration of roughly 70-80% and that’s not even counting all those cell phones that ship with MPEG-4 and no Theora support.
You think that Firefox users are the majority? Last time I checked, IE still had a market share of 60-70%. Microsoft now also ships AVC codecs for all its platforms (Windows Mobile / Zune, Xbox 360, and Windows). IE9 will also support the tag — therefore likely also AVC. Apple supports AVC in Safari/QuickTime on desktop platforms and uses it exclusively on iPhones. Google ships a AVC decoder with Chrome. IIRC Safari and Chrome have a combined market share of roughly 9-10%. That means that AVC has a market penetration of roughly 70-80% and that’s not even counting all those cell phones that ship with MPEG-4 and no Theora support. [/q]
The (worldwide) trends are very clear. In some countries, Firefox has indeed overtaken IE:
http://thenextweb.com/europe/2009/12/02/congratulations-mozilla-fir…
It won’t take that long before IE is distinctly in second place behind Firefox worldwide.
Google Chrome also supports HTML5 with Ogg Theora codec. So too now does Opera.
That puts native (no plugin required) support for HTML5/Theora web videos (along with other capabilities such as SVG, DOM2, CSS3 and a fast and correct JIT compiler for ECMAscript) enabled in over half of the web browser in use.
No it isn’t that clear. The few countries that have majority Firefox share are outliers. Firefox has actually lost global share in the last few months:
http://gs.statcounter.com/#browser-ww-monthly-200812-201001
IE could easily remain in first for years to come with Chrome taking users from Firefox.
Well, that is a point, kind of. Chrome looks quite attractive until you realise that it doesn’t allow for blocking of ads (it only allows for ads to be hidden, but not blocked from downloading). I would expect that Chrome wanes in popularity once people realise this, but for the moment it is true that Firefox has a challenger.
Not that this has any impact on the main point. The FOSS version of Chrome, which is Chromium, is also able to be run on any platform, using any OS on any device. Chromium also supports HTML5/Theora. So even if Chrome/Chromium starts to take a significant number of former firefox users, the overall trend of HTML5/Theora-capable browser is still the same. Firefox/Chrome/Chromium/Opera will still overtake IE.
It is a point, but it is a moot point in the overall context of this discussion.
http://en.wikipedia.org/wiki/Chromium_(web_browser)
I can run Chromium on my Arch Linux install right now, BTW, if I want to.
http://www.archlinux.org/packages/?sort=-last_update&arch=&repo=&q=…
It supports HTML5/Theora, but it doesn’t support blocking the downloading of ads. For the moment, I’ll stick with Firefox. If Chromium improves, I may switch.
Edited 2010-01-18 03:28 UTC
While I personally would like to see IE losing further market share in favor of a FOSS browser, your claim is pure speculation.
Speculation based on remarkably consistent trends.
The long term trend is perfectly clear (look at the slope of the graphs):
http://thenextweb.com/europe/2009/12/02/congratulations-mozilla-fir…
http://www.electronista.com/articles/09/12/21/chrome.safari.make.sm…
Firefox is already ahead of IE in about 15 countries now (I can’t find a link at the moment, sorry).
The bias in one’s statistics gathering methods only determines the date when one expects that Firefox will overtake IE globally.
It is, after all, the world wide web.
Are you mad? I just provided a link that showed how Firefox has gone into a slight downslope worldwide.
http://gs.statcounter.com/#browser-ww-monthly-200812-201001
It is and while Europe has been a supporter of Firefox Asia is still a stronghold for IE.
http://gs.statcounter.com/#browser-af-monthly-200812-201001
South America has greater support for IE than North America.
http://gs.statcounter.com/#browser-sa-monthly-200812-201001
If you want to cheer on Firefox then that is fine but there is no clear trend that ensures it will overtake IE worldwide. The alternative browsers could battle amongst each other for another decade without a single one being able to gain majority share.
As I said … Not that this has any impact on the main point. The FOSS version of Chrome, which is Chromium, is also able to be run on any platform, using any OS on any device. Chromium also supports HTML5/Theora. So even if Chrome/Chromium starts to take a significant number of former firefox users, the overall trend of HTML5/Theora-capable browsers is still the same. That is to say, Firefox/Chrome/Chromium/Opera will still overtake IE.
IE9 isn’t going to be out in time for it to impact Google’s decision. And Firefox is too big of a minority for Google to ignore.
What I actually think might happen is that Google, after completing the acquisition of On2, will release their codecs as open-source and submit them to a standardizing body, and the discussion about an AVC replacement will thus be ended.
So Google will just ignore IE, but not FF? That does not make sense.
But hey, let’s just see the pure economic standpoint: Google won’t drop Flash support any time soon, because it serves as a technology for legacy browsers.
Google already hosts MPEG-4 AVC versions of all YouTube videos, because Flash can read them.
That means, Google will not be harmed by offering those MP4 files via the <video> tag.
Flash does not read Theora files. That means, Google would need to duplicate YouTube’s entire video collection for Theora conversation for a (and I quote you) “minority”.
Yes, it will. From next year on, Google will be required to pay a royalty in order to be able to distribute files encoded in MP4.
Format conversion will cost Google less than paying codec royalties to MPEG LA. Small sites such as Dailymotion and tinyvid.tv have already done it, it isn’t a big deal.
Educate yourself. The license fees are capped at 5 million per year. That’s peanuts for Google which makes billions per year.
BTW, Google is already paying AVC license fees, because it’s distributing AVC codecs with Chrome and Android/Nexus1 and Google does not care at all about those fees.
At this time, AFAIK Theora 1.1 out-performs Dirac and is almost the same performance as h264. This was decidedly not the case with Theora 1.0 or earlier, but Theora 1.1 has seen a very significant improvement in Theora.
Edited 2010-01-16 14:13 UTC
That myth is a result of a wrongly conducted test.
You have got that story slightly wrong.
The original wrong result of testing Thusnelda reported that Thusnelda was better than h264.
http://news.slashdot.org/article.pl?sid=09/05/07/2352203
This turned out to be wrong. The test was not applied correctly.
Eventually, it was found that Theora 1.1 was slighly behind h264 quality, not slightly ahead as originally reported.
http://web.mit.edu/xiphmont/Public/theora/demo7.html
Theora 1.1 is a slight improvement over the Thusnelda results (Thusnelda was development code for Theora 1.1), but Theora 1.1 never actually overtook h264 quality in the end, but it got very close.
You can check it out for yourself here if you like:
http://people.xiph.org/~greg/video/ytcompare/comparison.html
Same filesize, same bit-rate, almost imperceptible difference in video quality.
This is why I reported it as I originally did:
Not a myth at all. Sorry to burst your bubble.
Edited 2010-01-17 22:48 UTC
Dude, you are constantly confusing format with encoder.
There is no single h.264/AVC.
In case of Theora both are more or less the same, because there are no independent Theora encoders (at least I’m not aware of any — all encoding apps I know are based on Xiph’s libtheora).
There are many AVC encoders out there. The comparisons you posted only compare to x264, not DivX HD / MainConcept, not Nero Digital, not Apple, not Elecard, not Thomson / ProCoder, etc. pp.
How can you seriously claim that Theora is better than AVC when the only comparison is made with x264 with default settings and does not even process HD video content?
Even worse, http://web.mit.edu/xiphmont/Public/theora/demo7.html does not even have a single image comparison with x264. Only numbers and graphs. Those mean nothing.
True, they give some indications about quality, but actual quality cannot be measured with software, but has to be conducted by living people in blind tests (that is a person rates the quality without knowing which software produced the video, to be unbiased).
I’ve seen 1h45m of 720p live-action video squeezed into only a 1.4GB AVC file with astounding quality (I don’t know which encoder was used). I’ve never ever seen Theora matching something like that.
Show me a Theora video matching that and I’ll change my mind about Theora.
Until you prove me wrong by showing me awesome Theora performance with HD video, I’ll keep my opinion:
Theora is OK, but not great for low-res video.
AVC is great for any resolution.
If either of them implement a codec we are getting nothing but needless bloat. They should, as is the case with Safari, implement video support using the native media framework of the platform. If that has a Dirac codec, it is supported.
If the Theora support in Firefox isn’t implemented as install the codec if needed but is built into the app then it is a stupid step to have taken.
Luckily at least Opera got it and adopted GStreamer.
Thing is, as soon as you take this approach you are pretty much back to the way it was with the <embed> tag. If you say “we support any codec, as long as it is installed on the system”, you are just encouraging the kind of codec hell that most of us are thankful to Flash video for banishing…
Btw, the Flash plugin does indeed bundle its own codecs instead of using the system codecs. And codecs don’t require much disk space. The complete set of Xiph decoders takes up less than a megabyte, in fact. I don’t consider that “needless bloat”, I consider that “justified bulge”
.
It can be fixed but if not Broadcom and friends could help (and give some Open Source friendly HW API).
The latest version of Theora (version 1.1 ) uses the same bandwidth as h264.
http://people.xiph.org/~greg/video/ytcompare/comparison.html
Same file size, same bit-rate, imperceptible differences in video quality. Try it for yourself.
While it is fine and dandy to have a discussion, please, lets just stick to the facts hey, and not just make stuff up.
After the end of this year, MPEG LA will start charging sites like Google for their use of h264.
http://www.streaminglearningcenter.com/articles/h264-royalties-what…
It will be a very small charge per video served, but you said it yourself: “how many videos do you think youtube serves a day”. Precisely the point. After this year, h264 will no longer be cheap for Google. This is precisely why Google are looking right now for an alternative to h264.
Edited 2010-01-16 14:04 UTC
For high res Dirac pwns Theora PERIOD
But Schrödinger still needs a bit of work, but there is no doubt in my mind that Dirac is the future.
Good analogy
Theora 1.0 = Ext3
Theora 1.1 = Ext4
Schrödinger= BtrFS
That’s only a comparison to YouTube’s current and unknown encoding setting with a single file. However it does show that Theora has improved in the last year.
I’m not sure why you were modded down when Google has already made statements about this.
Google has noted that there’s no way the company could serve up YouTube’s billions of streams using the much less efficient Ogg Theora codec, saying it would consume the world’s Internet bandwidth due to its less sophisticated compression.
http://www.appleinsider.com/articles/09/07/06/ogg_theora_h_264_and_…
Oh wait I know why you were modded down. FOSS advocates can’t stand dissenting views. This place has become worse than Slashdot. They already debated this and acknowledged Theora’s limitations.
Although it might be true, what you give here is not really a reliable source. There is not even a proper quote of some Google statement or anything. The site simply says that Google has said this. Please if you provide evidence to support your view at least give something where we can see that there was some sort of statement from someone at Google.
Google open source programs manager Chris DiBona is skeptical, however, and articulated the search giant’s concerns about Theora’s compression efficiency during the debate on the WHATWG mailing list.
“If [YouTube] were to switch to Theora and maintain even a semblance of the current YouTube quality it would take up most available bandwidth across the internet,” DiBona said. “
http://arstechnica.com/open-source/news/2009/07/decoding-the-html-5…
So can the parent be modded back up now? Funny how if you make claims that support FOSS positions you aren’t asked to provide any sources. As I said this place has become worse than Slashdot.
This may no longer be true from this year – AFAIK the Divx licence did not charge royalties til now on a per video watched/downloaded basis, but this either has changed on December 31 2009, or will on December 31 2010, when the cost of legally serving h.264 videos will rise.
There have also been tests done by some people to show similar levels of encoding with similar file sizes between youtube and theora – the real problem may be hardware. There are hardware h.264 encoders and decoders, but the theora hardware is less common.
I thought it was already well-discussed?
Greg Maxwell from Xiph has shown, that you can get quite resonable results with Theora+Vorbis combo instead of what YouTube offers in lo-fi ( http://people.xiph.org/~greg/video/ytcompare/comparison.html ), while Maik Merten did the same for hi-fi ( http://people.xiph.org/~maikmerten/youtube/ ).
I haven’t heard if those results were challenged, so my best knolwedge up to know is that there’s nothing particulary prohibitive in using Ogg Theora+Vorbis as a direct replacement for H.264+AAC (hi-fi) and H.263+MP3 (lo-fi). Feel free to update me. =}
Siverlight FTW
Yawn.
Wake me up when its a true cross-platform technology, fully spec’ed out in an open standard, *including* the DRM.
Until then it merely has all the disadvantages, and none of the advantages, of Flash.
It IS cross-platform. It will never be “open” as in open-source. Flash isn’t “open” either.
Silverlight has plenty of features that Flash will never have. Flash and Java are yesterday’s technology. Silverlight and .Net are tomorrow’s technology.
Edited 2010-01-16 13:45 UTC
Silverlight has plenty of features that Flash could never will never have. Flash and Java are yesterday’s technology. Silverlight and .Net are tomorrow’s technology.
huh? .Net and Silverlight are just a rehash of old technologies. Essentially just a Java clone better in some cases worse in others.
Edited 2010-01-16 13:49 UTC
C# syntax may look very similar to Java’s, but that doesn’t make it a clone. If you start developing in C# and .Net you realize how much more advanced and flexible they are compared to Java.
Just without Java’s xplatform nature.
Would you like to name exactly what other platforms Silverlight runs on apart from the different windows platforms.
Then compare it to the same examination done for Flash.
When (and only when) the Silverlight list is comparable to the Flash list can you properly state that it is cross platform.
Oh, by the way, don’t give us this ‘moonlight’ rubbish. I’ve tried to use it and many sites just redirect you to the MS website to download Silverlight. Utter crap. Not fit for use etc.
I’m no flash apologist. I wish it would got away. But using Silverlight is most certainly not the answer.
OS X. Microsoft has a native implementation for Silverlight that is released exactly at the same time the windows version is released.
And can disappear at any time MS chooses for it to disappear, and there would be nothing Apple or its users could do about it.
Mac is just that ‘token other’ that MS provides just to try to convince some people that Silverlight isn’t Windows-only, but anyone who knows the history of MS Office on the Mac knows the real deal.
And we still haven’t gotten to all the *nixes out there.
Moonlight, for the *nixes.
Moonlight doesn’t have Silverlight’s DRM, and never will. Wanna watch Netflix? You need that DRM.
Try again.
A situation that HTML5 would not change unless there was a similarly suitable DRM technology on *n*x that Netflix could use. Had there been greater TPM proliferation / platform adoption, it may have been easier to enable DRM for scenarios such as this on open source OSes because the TPM would provide secure key management.
DO you want to Space Alien replant as well?
Seriously… you have a huge chip on your shoulder. MS offers it for OS X… it is in Microsoft’s best interests to continue to offer it and to get it in use in as many places as possible.
While we are on the topic of things that could be taken away from you…. The Government could decide tomorrow to revoke everyone’s drivers license. Do you get an ulcer worrying about that possibility as well?
platforms covered by Silverlight – OSX and Windows (for x86 only) – are not the story by a long way for web clients. They are not even the story for desktops systems.
They cover 98% of the computing market. Sorry that the 2% are getting POed
No they don’t, not even close. They don’t even cover 98% of the desktop market, and they cover only a small part of the rest of the computing market.
Sorry… Didn’t know I needed to count the Nokia N900.
News Flash… Flash is not available on phones… oh…. but silverlight is!!
98.2%
microsoft made program for mac os users and after a few years of no updates they canceled them
they will do the same thing with silverlight
(the first thing that will get no updates is drm)
Edited 2010-01-17 15:10 UTC
maybe on windows not on linux and other oses
No way! We already had Microsoft try to set their own standards on the web, remember ActiveX? You really want them to try again?
Given a choice between a proprietary Microsoft standard which works, and a jumbled mess of so-called “open standards” for codecs that nobody can agree on, I’m leaning toward the former.
There is only one open codec being touted for HTML5 at this time, and that is Ogg Theora. It is a royalty free codec, able to be implemented by anyone for any platform, without any fee. It is the only mature open codec available, it is the only cross-platform codec available, and it has entirely competitive performance with other proprietary codecs being used for web video at this time.
What jumbled mess?
One open codec is a jumbled mess? On what planet?
You mean there is only one codec being touted by FOSS advocates. Apple doesn’t support Theora and though Google supports it in Chrome their actions have made it clear that they are not going to ditch Flash.
At this point it is looking like the codec won’t be specified which means web publishers will just use flash instead of building multiple video files to ensure compatibility with all the browsers.
No, I mean just as I said, that there is only one open codec being touted by anybody.
Apple doesn’t support any open codec.
I disagree. Flash is restricted in the platforms it covers (not as restricted as Silverlight, but still restricted). This whole thread is about Google’s moves towards HTML5.
Over half of the web browsers in use now support HTML5/Theora out of the box, without requiring a plugin. At least one of these browsers is truly cross-platform and implemented on a vast range of devices and operating systems, including mobile and handheld devices.
It has possibly become moot now. Websites will perhaps offer HTML5/Theora by default, and perhaps Flash with some codec or another as a fallback. From the end of this year, MPEG LA have said they will charge websites for using h264. This alone will possibly be enough to see the death of h264 used on the web.
Edited 2010-01-18 00:29 UTC
But the codec isn’t going to be defined as part of HTML5 which allows for the possibility of multiple open source codecs. We’ve already seen people here bring up the Dirac codec as an alternative to Theora. It could end up in a situation where everyone is trying to get you to download their favorite codec.
No they don’t which is a big part why Theora is not going to be the default codec.
You disagree that Google is supporting Flash with their actions? It doesn’t say much that they have kept YouTube in Flash and that they partnered with Adobe to bring Flash to their phone? Yea Google is really on the warpath against Flash (eye roll).
And Flash currently has a 94% penetration rate so that doesn’t mean much. Even if IE and Safari eventually support Theora that still leaves a lot of users with IE-1.
Why would they build multiple files and write additional browser detection scripts when they can just use Flash? I’m not seeing a good motivator here. I can see eventual adoption but it is going to be at a glacial pace unless Google converts YouTube and forces Flash users to upgrade or install a plug-in.
Look I hate Flash as much as anyone here but I am also aware that “good enough tech” that is widely installed can take a long time to replace. You really have to cut off access to users if you want rapid change.
Because .h264 files will cost to distribute next year, and OGG won’t. A website like YouTube is hardly going to be converting to OGG _by hand_, it’ll be automated, so having multiple files is completely moot. Disk space is cheap.
http://camendesign.com/code/video_for_everybody
No detection needed.
MPEG-4 AVC is open. ffmpeg offers a FOSS decoding solution, while x264 offers a FOSS encoding solution for AVC.
Software patents only apply in the US (and Japan IIRC). Therefore AVC is royalty-free in the rest of the world as well. And besides Mozilla no one really cares about Theora.
GIF was patented a very long time as well — that didn’t hinder its adoption.
And why should the rest of the world suffer from an inferior, non-scalable codec, just because the American people only vote for pro-patent parties?
Edited 2010-01-18 02:45 UTC
No it is not. MPEG LA is the organisation that holds the patents for MPEG-4.
http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC
http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Patent_licensing
No claim of MPEG-4 AVC being “open” is made.
This matters not in the slightest when MPEG LA start charging YouTube and any other site for distibuting video data encoded in h264, regardless of what software was used to encode and decode it at either end.
FOSS licenses require that the code being offered is able to be distributed to anyone. Therefore, Theora meets this requirement and MPEG-4 AVC does not. Therefore, Theora is many times more preferable than MPEG-4 AVC.
Oh yes it did. Patents on GIF were the entire reason for the development of PNG. PNG is now more widely used, even after the patents on GIF are expired.
http://en.wikipedia.org/wiki/Portable_Network_Graphics
A reasonable question if it were true that Theora was actually appreciably inferior. Since it isn’t, the question is moot.
Edited 2010-01-18 03:01 UTC
Your claims are completely ridiculous and by repeating your false claims over and over again does not make them true.
AVC/h.264’s specifications are available completely free of charge: http://www.itu.int/rec/T-REC-H.264-200903-I/en
Everybody is allowed to implement those specifications. There are no strings attached when you implement them.
Oh, and by the way: Where is the guaranty that Theora is actually free of patents? Does Xiph offer any kind of legal insurance against unexpected lawsuits?
Don’t answer, my questions are rhetorical. I know that Xiph neither offers a guaranty that Theora is 100% free of patents nor offers legal insurance in case of a lawsuit.
And regarding your repeated claims that Theora is supposed to be an actually good codec… As I wrote in http://www.osnews.com/thread?404514 (where you refused to answer): Show me Theora managing to compress 720p HD content with the length of 1h45m into mere 1.4GB without visible quality impact. Do it. Don’t waste your time comparing video resolutions that were state of the art 10 years ago. 720p is the very least, a video codec has to perform very well these days.
Edited 2010-01-18 12:24 UTC
Your claims are completely ridiculous and by repeating your false claims over and over again does not make them true.
AVC/h.264’s specifications are available completely free of charge: http://www.streaminglearningcenter.com/articles/h264-royalties-what…
After 2011 MPEG LA will charge a fee for transmission of h264-encoded video data, regardless of who wrote the encoder software that was used.
Surprisingly, yes. Theora is absed on the VP3 codec from On2 technologies. On2 have a patent awarded to them specifically for this codec. According to the patent Office, there is no technology that VP3 infringes, because VP3 itself is the patented technology. On2 have made an arrangement with Xiph.org such that Xiph.org may use and develop VP3 (as Theora) and distribute the resulting code as they please, as long as they please, royalty-free if they like.
http://www.theora.org/faq/#VP3
Au contraire … Xiph.org Foundation has negotiated an irrevocable free license from On2 to the patented VP3 codec. It is fully licensed by Xiph.org.
http://www.theora.org/benefits/
Why don’t you show me such a comparison, where h264 spanks Theora?
I have watch 720p and higher videos encoded in Theora. It works very well.
I have already provided a link in this thread to a comparison paper where h264 and Theora are seen to produce almost-identical quality output for the same filesize and bitrate.
http://people.xiph.org/~greg/video/ytcompare/comparison.html
In denial of this direct evidence, you claim this is not right … so the ball is clearly in your court. You now have the burden of proof, not I. Show me. Put up or shut up.
Edited 2010-01-18 13:04 UTC
Yes, it does.
That is not the software’s fault, but the US patent law’s fault.
Are you a US citizen? Then fix your laws instead of bitching against better codecs.
Parts of VP3 are patented by On2, not every single used function. On2 and Xiph guess that the other used technologies’ are not patented by anyone else, but considering how broken the US patent system is, it’s impossible to tell.
Easy:
http://etill.net/projects/dirac_theora_evaluation/include/halbach-2…
“The experiments showed that the reference system H.264 outperforms the two codecs [Theora and Dirac] signiï¬cantly, both in the objective and the subjective/visual evaluation.”
That one is an actual scientific comparison and I found it using a simple Forestle search.
You OTOH just link to comparisons by Xiph people showing that the Xiph codec is supposed to be better.
Yeah, right…
Microsoft’s studies also show that Microsoft’s products are better. So what?
Xiph is biased towards its own product, just as Microsoft is biased towards its own products.
1.) No HD resolutions –> Useless.
2.) Biased, because a Xiph product is measured by Xiph. –> Useless.
Ah, OK… you post biased comparisons without HD content over and over again and now it’s my turn. You sound like an “Intelligent Design” proponent who refers to unscientific, biased “proofs” over and over again.
But OK, I did my own comparison.
A scene of X-Men Origins in h.264/AVC: http://www.flickr.com/photos/19616885@N00/4284748443/sizes/o/
Same scene with Theora (default settings): http://www.flickr.com/photos/19616885@N00/4284748597/sizes/o/
If you want, I can upload the short clip I made in both formats.
No, it doesn’t. Sorry, but you don’t get to define open.
It doesn’t matter where the fault lies for the software to be restricted and not-open, all that matters is if it is, or is not restricted.
Nope, Aussie. Doubtless, h264 is not patented in Australia. However, the previous government was witless enough to sign away Australia’s IP independence in order to try to sell some beef to the Yanks at pittance prices, so I am still impacted by American greed and ripoff practices when it comes to their ridiculous notion of IP.
Yet no-one has launched a lawsuit against it, because VP3 is itself patented, and other codecs could fall foul to its patented functions. There is little to be gained in a patent Armageddon, and all of the Free Software arsenal of defensive patents would be brought to bear against anyone who tried.
That was Theora 1, from 2006.
From the linked PDF:
Yes, Theora I was worse than h264, we knew that.
Now try it again for Theora 1.1, which is what is being touted now, in 2010, as a possible codec standard for use with HTML5. That, you will find, is a totally different kettle of fish, performance wise.
Only Theora 1.1 is claimed to be competitive with h264.
There is no claim from Xiph.org that Theora is better than h264 performance-wise. Merely comparable, almost as good, but not better.
Your paper was a nice try, but no cigar. It references a 3-year-old version of Theora. The ball is still in your court.
Microsoft EULA licenses specifically disallow benchmarking and side-by-side performance comparison of its products and other products (especially those running on another OS). So what?
We can still run competing video codecs on the same hardware (at least those codecs from software vendors who can stand a little competition). When we do, we find that Theora 1.1 is almost on a par with h264.
Why do you find this so hard to understand?
At least it is a comparison of the correct versions of the codecs in question, so it is less useless than the paper that you quoted.
My goodness, what did you do to get such a poor result?
Let me guess … Theora version 0.9 or something?
Theora 1.1 can do waaaaaaaay better than that.
h264:
http://people.xiph.org/~greg/video/ytcompare/bbb_youtube_h264_499kb…
Thusnelda (development version of Theora 1.1):
http://people.xiph.org/~greg/video/ytcompare/bbb_theora_486kbit_366…
Open them each on a separate tab, and flip between the tabs. You won’t notice any difference.
Same filesize, same bitrate, same image size … imperceptible difference in image quality.
Deal with it.
Edited 2010-01-19 11:28 UTC
No. I used the Theora library as offered in PackMan’s openSUSE repo (don’t mind the German text — “Version” is the same word in German and English anyway):
I just used the default settings, just as in your comparisons, just as the Xiph people used x264 with default settings. The video has a bitrate of 2389kb/s. As you can see from the image dimensions, the video is HD content (not exactly 720p, because it’s more widescreen than 16:9, but still HD).
Oh, btw: http://web.mit.edu/xiphmont/Public/theora/demo7.html doesn’t give any specifics about the used x264 version either except “patched ffmpeg”.
You aren’t doing it right, somehow.
Just about everybody on the planet gets waaaaaay better results than you do.
Warning, this is a very large file, 87.2 MBytes, as it is a movie trailer (Sherlock Holmes Trailer 2) from tinyvid.tv
http://media.tinyvid.tv/1s2sj33y1ozc3.ogg
You might be able to see it via HTML5 directly in your (non-IE) browser here:
http://tinyvid.tv/show/1s2sj33y1ozc3
According to SMPlayer:
I think this is 720p, is it not? That is the type of result that you SHOULD be getting. Indistinguishable from h264 at the same filesize and bitrate.
Edited 2010-01-19 12:17 UTC
Wow! Simply… Wow. I keep hearing how Theora has improved and that it is supposedly comparable to H264 these days but I had yet to see a convincing video encoded with it that backs up this claim. Until now. That clip has a surprisingly crisp image compared to the screenshot posted by the person that you’re arguing with, which is closer to what I recall of Theora encoded videos that I’ve seen some time ago.
If this is the result of the encoder with some reasonable defaults, I guess that it is more than enough for the proposed use case which leaves just the bandwidth question: how much smaller would be the same video encoded with H264, if at all? There is also the fact that there are lots of hardware solutions out there that use H264 specifically but I wonder what will happen if the MPEG-LA starts asking for royalties fees starting next year as some other people already pointed out? Is it going to somehow slowdown the adoption of H264 and give room for a competing and hopefully freer codec?
And at this point, I agree that choosing Theora or Dirac would be a nice compromise for a standards body to chose as some sort of “fallback” solution as it is the only solution that will truly leverage the play field for all the industry players. Leave Apple and Microsoft free to pursue their plans if needed be but choose Theora or Dirac to be the reference. That sounds reasonable to me.
Edited 2010-01-19 23:08 UTC
RE: Theora filesize
From the Sherlock Holmes trailer example, we get the following calculation:
89000 KBytes (in the trailer)
150 secs (duration of the trailer)
593.33 KBytes/sec
35600 KBytes/min
120 Mins/movie
4272000 KBytes/movie
That is to say, approximately 4.3 GB for a two-hour movie @ 720p resolution, image quality equivalent to h264.
Par for the course isn’t it?
Edited 2010-01-20 00:34 UTC
For those of you who might be curious about the quality of the sample video of Theora @ 720p resolution, but who are not inclined to download the entire large file, here are a few screenshots of the running movie trailer as taken by SMPlayer on my system:
http://ourlan.homelinux.net/qdig/?Qwd=./Theora_720p&Qiv=name&Qis=M
Really? You post one ill-conducted, biased test after another, but I do it wrong? Riiiggghhhttt….
Answer me this, sunshine: how exactly can the objective quality of a 720p video clip, encoded in Theora, be “biased” in any way?
Here are a few screenshots of the running movie trailer as taken by SMPlayer on my system:
http://ourlan.homelinux.net/qdig/?Qwd=./Theora_720p&Qiv=name&Qis=M
And here is a snapshot of the SMPlayer window, playing the clip.
http://ourlan.homelinux.net/qdig/?Qwd=./Theora_720p&Qif=SMplayer.pn…
What are these, exactly? Am I supposed to have fudged it? Are you suggesting that I magicked it somehow, and used a better Theora encoder than the Theora encoder? And then I posted it on the tinyvid.tv site? Just to trick you?
Hmmmmmmm?
You quite rudely ask me for HD content, not from Xiph.org, like so:
So when I post exactly what you requested, that being HD content encoded in Theora from an independent source, all you can say is that I am biased, and that your own poor result (Heaven only knows how you stuffed it) is the true result?
We are supposed to believe you when you didn’t even realise that Theora 1.1 had undergone significant improvement from the 1.0 version circa 2006, about which you posted an old research paper?
Pffft.
You make no sense.
Edited 2010-01-20 12:20 UTC
Easy: You use high-quality, but slow settings for one encoder and faster, lower quality settings (=”defaults”) for the other.
Blurry crap. By far no match for good AVC encoders.
Good AVC encoders can easily squeeze a full-length 720p movie in 1.4GB without visible quality loss.
moonlight the patent trap that doesn’t work
mono the patent trap that doesn’t get windows .net programs to work outside ms windows
google will never switch youtube to silverlight
I think google really should focus on this subject. All in all it’s google that claims to support open standards.
Unfortunately, the HTML5 Video codec problem remains unsolved, and I suspect it’s going to get worse before it gets better.
We currently have two main codecs to worry about – h.264 and Theora.
Right now, the “obvious” choice is h.264. It has higher encoding quality than Theora, or alternatively uses less bandwidth for the same quality. There’s a wider variety of encoders available, both hardware and software, including real-time encoders suitable for live streaming. Some platforms have hardware accelerated h.264 decoding, and a software-only codec may be undesirable (laptops) or impossible (iPhone). The format is more widely known and understood than Theora, with better tool support, mostly through QuickTime.
Theora could not possibly win against that. It should be comparable to h.264, but the encoder is not very mature. Even if it were clearly better, the lack of hardware acceleration support on platforms like the iPhone completely rules it out. Some companies (Apple and Nokia) claim concerns about submarine patents too, so it’s doubtful they’d ever use it.
The problem with h.264 is that neither Mozilla nor Opera could support it themselves. The format is patented up to the eyeballs, and shipping a decoder requires payment of royalties to the MPEG-LA, up to $5,000,000 per year. Since Firefox can be freely redistributed, they have no way to track how many copies they distribute within a year, and no way to collect royalties on behalf of the MPEG-LA. I imagine Opera would be in a similar situation.
Firefox and Opera could defer to system codecs like Safari does with QuickTime. The codecs available through QuickTime on a Mac aren’t the same as the codecs available through DirectShow on Windows. h.264 is available on Windows 7, but not on Windows XP or Windows Vista unless you install some extra codecs. Good luck persuading anyone to do that
Should Microsoft ever chose to implement the video tag, they’ll almost certainly do it through DirectShow. Unless they ship a h.264 codec with IE, that means that it will support neither h.264 nor Theora on Windows XP and Vista. The only usable codec included with Windows XP and Vista is WMV, so you’d have to add that into the mix.
There’s also the licensing uncertainty surrounding use and distribution of h.264 encoded video.
Currently, distributing h.264 encoded content over the internet doesn’t require royalty payments, but it might starting next year. The MPEG-LA still haven’t announced their licensing for 2011, but have stated that it won’t cost more than equivalent free-to-air TV broadcasts. Of course, those are described in terms like broadcast area, which make no sense at all over the internet.
Chances are that the free ride is over at the end of the year. The MPEG-LA are likely to ignore small video providers for now, and go straight for the likes of YouTube. As soon as you get large enough to pass the MPEG-LA’s minimums (assuming there are any), you’re going to get hit with royalties which were designed to extract as much profit as possible from YouTube, Hula, and so on.
The worst part – it’s not even possible to invent some new codec that stomps all over h.264 yet has no royalty requirements, and use that instead. The installed base of devices with h.264 hardware decoders basically prevent any other video codec from being used, no matter how good it is.
That means that Dirac’s out of the question too, I’m afraid.
I wouldn’t be surprised if Google had a go at it though. They must have purchased On2 for a reason, and it surely wasn’t to improve Theora.
Royalty break down for clarification:
http://www.mpegla.com/main/programs/avc/Documents/avcweb.ppt
AVC/H.264 License Terms:
Codec Manufacturers
<ul>
<li>Products sold to end users and OEM for PC but not part of OS (decoder, encoder or product consisting of one decoder and one encoder = “unitâ€)
<ol>
<li>0 – 100,000 units/year = no royalty (available to one legal entity in an affiliated group)</li>
<li>US $0.20 per unit after first 100,000 units/year</li>
<li>Above 5 million units/year, royalty = US $0.10 per unit</li>
<li>Enterprise cap: $3.5 M/year 2005-2006, $4.25 M/year 2007-08, $5 million per year 2009-10</li>
</ol></li>
<li>An Enterprise selling branded OEM for PC OS may pay for its customer:
<ol><li>0 – 100,000 units/year = no royalty (available to one legal entity in an affiliated group)</li>
<li>US $0.20 per unit after first 100,000 units/year</li>
<li>Above 5 million units/year, royalty = US $0.10 per unit </li>
<li>Enterprise cap: $3.5 M/year 2005-2006, $4.25 M/year 2007-08, $5 million per year 2009-10 </li>
</ol></li>
<li>Includes make, sell and limited right of consumer use by or between end users (e.g., in connection with a video teleconference or mobile messaging)</li>
<li>Royalties begin January 1, 2005</li></ul>
AVC/H.264 License Terms:
Participation Fees
<ul><li>Where End User pays for AVC Video<ul><li>
Subscription (not limited by title) – 100,000 or fewer subscribers/yr = no royalty; > 100,000 to 250,000 subscribers/yr = $25,000; >250,000 to 500,000 subscribers/yr = $50,000; >500,000 to 1M subscribers/yr = $75,000; >1M subscribers/yr = $100,000</li>
<li>Title-by-Title – 12 minutes or less = no royalty; >12 minutes in length = lower of (a) 2% or (b) $0.02 per title </li></ul>
<li>Where remuneration is from other sources
<ul>
<li>Free Internet Broadcast (not title-by-title, not subscription) – no royalty during first term (through 2010; not greater than Free Television thereafter)</li>
<li>Free Television – (a) one-time $2,500 per transmission encoder or (b) annual fee starting at $2,500 for > 100,000 HH rising to maximum $10,000 for >1,000,000 HH</li></ul>
<li>Enterprise cap: $3.5M/yr 2006-07, $4.25M/yr 2008-09, $5M/yr 2010</li>
<li>Royalties begin January 1, 2006</li>
</ul>
Yeah, parts of AVC are patented in the US, but that’s not the format’s fault, but the fault of America’s citizens who don’t vote for parties that want to ban software patents. As a European/African/Asian citizen, why should I care about America’s broken legal system?
Mandriva has a “US version” of its Linux distribution without any patented codecs and an international version with ffmpeg and stuff.
Mandriva never had problems because of that. Mozilla could do the same with Firefox.
Take a look at: http://www.mpegla.com/main/programs/AVC/Pages/PatentList.aspx and http://www.mpegla.com/main/programs/avc/Documents/avc-att1.pdf
I am seeing patents for h264 for US, Canada, Germany, France, Great Britain, South Korea, Australia, Japan and more, and that is on the first page.
I hope they don’t support both Flash and HTML5.
Lets see a hard switch.
If you want to watch YouTube, you’ll need to use Chrome or FireFox.
People would leave Internet Explorer in droves.
It would temporarily screw devices where you can’t change the browser and need flash, like the PS3… but an update would come soon enough.
I think that’s being optimistic at best. If Google really did a hard switch, Microsoft would quickly push out an update to IE8 that would include support for the video and/or audio tags. It’d probably be a rough job and, in typical Microsoft fashion, would be buggy for a while but they would do it. They’d do so grudgingly, perhaps, but if it meant keeping IE relevant you can bet they would.
You’re dreaming, they wouldn’t want to see that kind of a drop in their ad revenue. A lot of people sit and watch youtube videos at work where IE is the only option. At the very least Google would provide a plug-in.
Yeah, i’m sure your manager would have a huge problem with you having to work instead!!
Edited 2010-01-16 22:06 UTC
Lol I work from home so I play xbox when I take break.
If you work at home on your own computer then install Chrome or Firefox.
Wow I can install Chrome or Firefox? I didn’t know that, this is my first time here.
Try reading the whole thread next time before jumping into an alternative browser sales pitch. I said people watch youtube videos on IE6 at work. People like you remind me of evangelists that ask you if you have heard of Jesus even if you are clearly American and over 18. Jesus Christ? No I have never heard of him. Is that somehow related to Christmas?
P.S. I wrote this from Chrome, douchebag.
Google isn’t even actively supporting IE6 any longer: http://www.techcrunch.com/2009/07/14/youtube-will-be-next-to-kiss-i…
That shows a warning about cutting support in the future.
It might mean that they will simply show IE6 users a message about their browser being not supported but will still give them access.
Google will possibly offer access to IE6 users through Google Chrome Frame.
http://www.techcrunch.com/2009/09/22/google-turns-internet-explorer…
It is the easiest way. Google Chrome Frame supports HTML5/Theora, and it is a Google solution (as opposed to requiring and being dependant on and beholden to software from some other company).
Edited 2010-01-18 00:36 UTC
Not only that, but it would also look anti-competitive. Google doesn’t need that since they are coming under increasing anti-trust scrutiny in the EU and US.
the could block IE6 and say it because of security problems with that version
I wish all websites would block IE6 but it wouldn’t make financial sense for many of them.
Ha, thats the lease anti-competitive thing they could do. Right now they only support platforms that Adobe deems important. If they went to HTML5 video tags, they would be supporting many more platforms. Its not Google’s fault that IE is in the stone age.
If Google were to block their main rival’s browser and in effect funneling people to either their browser or the outside browser that they bankroll (at least for a little while longer) it would be the definition of anti-competitive.
Also as stated earlier, Google’s advertisers would hit the roof. It would be wrong and they would need to come up with a new motto.
Edited 2010-01-18 05:38 UTC
EDIT: This comment seems to be duplicated for some reason, and editing one edits the other as well…
A complete block of IE, yes, that’s a ridiculous idea that I’m sure Google would never even contemplate.
However, they did come up with that neat Chrome-inside-IE plugin a few months ago. There’s really not that much of a difference between requiring the user to install that plugin to view video rather than the Flash plugin, so I do think it’s a possibility.
Much more likely, though, is to have support for the HTML5 video tag while leaving the Flash plugin as a fallback for browsers which need it. That’s very easy to do, so the only real question is whether they have the HTML5 video in Theora or limit it to just their existing h264 streams. As much as I think Firefox users will complain, I can see them sticking with h264 at least for a while, and recommending Chrome (or Safari) to use it.
Edited 2010-01-18 06:15 UTC
Ha, while people may watch youtube at work, is Google getting and ad revenue from them? Are they clicking through and buying things at work too?
Why would switching browsers be easier than switching video hosting sites?
I am hopeful that Google might do this “don’t be evil” kind of thing. Please, get rid of flash. Think of the kittens!
I’m sure everyone at adobe is going to chime in–like their jobs depend on it : )
Edited 2010-01-16 16:37 UTC
What is so bad about Flash? Well, I mean aside from the fact that it’s not RMS-approved Free Software. From my vantage-point, that seems to be the only problem with Flash. “It won’t run on obscure Linux distro running on ARM!” Who cares. Linux folks should consider themselves very lucky that Adobe spent time making Flash a first-class citizen on Linux when Linux has about 1% marketshare on the desktop and probably fewer than those 1% use Flash (based on the rabid postings I see here and elsewhere). Oh yeah, and Linux users don’t pay a dime for it either.
Flash solves a lot of problems with an implementation that continues to improve. It’s available everywhere that counts. As another poster pointed out, before Flash, we had dozens of codecs and players, which a user had to struggle with to get any given video to play. With Flash video, you go to a website and the video plays. Fancy that! Of course, I guess that’s too simple for the zealots, who would rather users struggle with getting codec-du-jour to work on their machine as they contemplate the moral and ethical issues surrounding libre vs. gratis vs. proprietary software and formats.
Edited 2010-01-16 19:44 UTC
Hmm, how’s about the fact that it makes my Core2 Duo CPU jump up to 90% use just to play a video? That’s good enough reason to hate it in my case. Flash is the *only* thing that does that except for compiling software, and no other video playback does anything like Flash when it comes to usage.
Also, flash a first class citizen on Linux? You’ve obviously never tried to really use the Linux flash plugin or you’d know just how bad it is by comparison to the Windows version. It’s worse even than the OS X version, and that’s bad enough. Not that I blame Adobe, with Linux’s incompatible and constantly changing subsystems I’m amazed they bothered to target it at all.
It sucks on Windows, too, in terms of CPU. We get it. That can be fixed. It’s not a fundamental flaw in the Flash idea, it’s an implementation problem with current versions of Flash.
That the Linux API, such as it is, makes it hard for Flash to take advantage of hardware acceleration, etc. (assuming any is actually available, what with the state of drivers) is the sadder issue. It’s sadder because Flash isn’t the only loser here. Because X has failed to get its act together and because other parts of the desktop stack are in constant states of flux, incompleteness and bugginess, writing software for Linux that does interesting things is unnecessarily difficult, if not impossible. So the real problem is not Flash so much, but the Linux desktop.
Never said it couldn’t. However, given that only Adobe holds the complete specifications for Flash and, given that they don’t seem to give a single care about the problem, means that a deficiency in the implementation still reflects upon Flash as a whole. At the moment, there is only one true implementation of Flash (Gnash and Swfdec don’t count as full implementations). Which means that at the moment, whether I hate the implementation of Flash or the idea behind it is irrelevant. There is only one Flash, and it sucks. For the record though, it’s the implementation that I hate.
Yes but Adobe is in charge of the current implementation and that is the one that spikes my cpu over a 2″x2″ 15 frame ad. I’m sure everyone has had a case where they noticed their cpu was maxed and it just turned out to be a couple Flash ads.
Sure they are improving but they move too slowly. I’d rather see a fresh codebase.
There’s also the security issue.
I’m sure having it open source would be much better. Instead of one crappy, but functional implementation, we’d have 12 very crappy and barely functional implementations which are nevertheless free as in speech and therefore perfect.
No need to worry since if there is a bunch of competing codecs then web publishers will just use flash.
Who said anything about open source? Didn’t I already say that Gnash and Swfdec aren’t full implementations? Adobe’s implementation of their own standard is a bug-ridden piece of bloatware. That’s just a fact. They don’t care about fixing it. That’s another fact. Nowhere did open sourcing of flash come into this until you brought it up.
And thus you perpetuate the same lie as you did in previous Flash related articles – by chance are you an employee of Adobe? vested interest in the status quo? making plenty money prolonging the problem than addressing it?
Microsoft has a test suite which Moonlight is tested against; there is NOTHING and I mean ABSOLUTELY NOTHING stopping Adobe making a similar test suite so that third parties can test their implementations against.
There is NOTHING stopping 12 different implementations to be all compatible with each other – it is people like you, who have vested in the status quo, who would be shown up for pathetic and inept programmers if the 12 implementations ended up producing better results than the company who invented the damn thing in the first place.
I don’t work for Adobe and never have. But I do thoroughly enjoy the logical leaps that false premise allowed you to make. Keep up the good work.
The rest of us are enjoying your asinine comments.
Yes, I’m really out there in left field defending software that, for the most part, just works. I know that’s a hard concept for some in the FOSS community, but users really do value software that they can actually use without having to read a man page or navigate through disparate dialogs spread across a variety of programs and operating system GUI panels. I certainly enjoy being able to go to YouTube and just, you know, watch a video without having to deal with all that crap. I think people in the real world generally agree.
Siride, what makes you think that HTML5 video is going to devolve into a huge codec mess. So far, the only browser where that’s even a possibility is Safari. Firefox, Chrome, and soon Opera will all support the same codec OOTB and for the most part you can’t add more video codecs even if you try. That sounds a lot like the current situation with flash, if you ask me. The only current problem is Safari, and IE, and if HTML5 video really takes off I don’t expect that to continue to be an issue. In fact, I’d say that Flash just won’t ever be dropped if it is.
The reasons to get rid of Flash aren’t anything political for me, it’s just that it simply doesn’t work very well for me no matter how much you yell about how great flash is.
It’s slow, bloated, and responsible for the majority of browser crashes, which is why I can’t wait for firefox to add out of process plugin support in a few months. Why do I have to run an entire runtime when all I want is to display a video with a few simple controls? That’s just asking for security problems, and it’s no surprise that Flash is a major attack vector when you consider that there is no automatic updating for all the security bugs found every month.
Firefox and HTML5 just works. It “just works” on sites like these, without requiring any additional plugins for video:
http://openvideo.dailymotion.com/en
http://tinyvid.tv/
It “just works” on quite a range of platforms and web clients:
http://en.wikipedia.org/wiki/Features_of_Mozilla_Firefox#Cross-plat…
http://en.wikipedia.org/wiki/Firefox_Mobile
When Google listen to their users response in this poll, and implement HTML5 in YouTube, Friefox/Fennec will “just work” out of the box for YouTube for more people on more devices than any other browser. Full stop.
There is no man page for firefox.
Enjoy.
PS: If you go to YouTube you will not be able to enjoy watching any videos (it won’t “just work” to start with) until you install a flash plugin. For many web client devices in current use, there is no flash plugin, so it will never “just work” for those users.
Edited 2010-01-17 10:28 UTC
There is also NOTHING stopping distros from getting together and working on package compatibility but it has a snowball’s chance in hell of happening.
There was also NOTHING stopping the open source world from creating something like Silverlight but it never happened.
It’s a legitimate concern to be wary of the FOSS movement and its preference for open source over quality and standardization. It also has a history of overestimating its ability to compete with proprietary offerings.
Backwards.
HTML5/SVG/DOM2/CSS3/JIT ECMAScript/SMIL are better than Silverlight.
Just because it never happened in IE doesn’t mean that it never happened.
You hugely overestimate the capability of proprietary offerings.
Edited 2010-01-17 09:53 UTC
Backwards.
HTML5/SVG/DOM2/CSS3/JIT ECMAScript/SMIL are better than Silverlight.
Just because it never happened in IE doesn’t mean that it never happened.
You hugely overestimate the capability of proprietary offerings. [/q]
That is subjective not fact. There are plenty of things about Silverlight that make it superior to some of those things you mentioned. Just off the top of my head is the ability to use C# and Boo (python) as well as JS. The only real drawback is that its an add-on to the browser and runs in its own sandbox as most of these plugins do. The other is that its closed and proprietary. Not a bad thing in itself but that means that porting it is reliant on the company that owns it, and/or a 3rd party implementation that is alway behind by a mile.
The most important failings of Silverlight are these:
(1) It depends on proprietary .NET technology, only part of which has been offered by the technology author as open standards,
(2) It includes proprietary extensions which work only on one platform,
(3) For video, it requires a proprietary codec which is available only from one software vendor, who has released it as a closed binary blob only for use on three desktop platforms for x86/x86_64 only and for no other architectures, devices or platforms.
These are all facts, there is no subjectivity there.
Any one of those failings is a severe problem for consideration of Silverlight as a public access standard for the web.
The fact that Silverlight suffers from the three severe problems listed above makes it totally unsuitable for use on the web. As a “web standard”, it is utter garbage.
Rejected out of hand. Wouldn’t even be considered by any sane standards body.
You listed some web technologies, some of which are in draft, none of them are as feature rich as Silverlight when it comes to providing interactive web interfaces, but whatever, I have a better question for you anyways.
Why didn’t we have a basic video player plug-in alternative 10 years ago? Forget HTML5, where was a FOSS alternative to Flash like Quicktime when we needed one? Need I mention all the delayed and buggy releases of Flash in Linux? Remember Flash 9?
Actually I have underestimated the viability of proprietary offerings like Flash. 10 years ago I considered Flash to be a temporary annoyance with the expectation that it would be replaced by an open source alternative.
Now that is being tied to hardware decoding I can see it being around another 10.
You are getting confused here. Flash is not h264, and flash is not a codec. Flash is not supported in some hardware, h264 is. Hardware support for video includes a lot more than just the codec. Everything related to video support in hardware following the codec can be used by video streams encoded with a different codec.
No I’m not confused, you’re just unaware of how Flash is being tied to portable devices.
Flash hardware support is being built into phones that won’t know how to provide hardware decoding for any other type of file.
No, you are definitely confused. When they say “flash hardware support”, they actually mean “hardware video support”. “Hardware video support” includes a lot of functions, optionally including the decoding part of the video codec. The video decoding support being installed on hardware today is often indeed h264, but that will be moot when supply of h264-encoded streams over the internet dries up due to MPEG LA charging after the end of this year.
Apart from the codec function, though, the remainder of the “hardware video support” can be used with any codec.
None of this has to do with Flash video, other than the fact that Flash video typically uses h264. The wrapper that is Flash is not supported by the hardware. One could replace the Flash wrapper with something else entirely, and it would mean squat as far a support in hardware goes.
Sorry, but you did get confused.
The Windows version isn’t even equal across all browsers. The IE plug-in is a separate build and better optimized for cpu usage. This is why you are better off using IE if you are trying to maximize battery life.
or just adblock until adobe gets its act together.
I refuse to accept ads until they stop killing my battery
So Google just released the Nexus, if they going to go with Theroa or anything else, don’t you think they would have included a hardware decoder for it?
Every other mobile phone since the iPhone came out has had a h264 decoder hardware, i just can’t see them shooting themselves in the foot by going with anything else.
But as I said in my “Why I hate HTML 5” blog entry:
http://my.opera.com/deathshadow/blog/2010/01/09/why-i-hate-html5
HTML5 on the whole pisses me off because most of the so called semantics are the exact opposite, and for the most part things like “audio” and “video” are a step backwards from all the progress “STRICT” gave us.
The whole original idea for moving forward was to have LESS tags, simpler tags, not to tack on five dozen new ones that may or may not mean what you want, or to have fifteen tags for content that could be handled by one.
Dig into the specifications, and moving forward “object” was SUPPOSED to replace IMG, EMBED, APPLET, etc, etc… So now instead of riding microsofts ass about not supporting it properly (now that IE8 actually does) we backpedal on that simplification to add a whole bunch of tags to do things we can already do? Even more ***tarded is HTML 5 legitimizing EMBED as a tag… just trying to make it even more of a jumbled half-assed mess?
I swear, it’s like the entire idea of HTML5 was hijacked by people who missed the entire point a few years ago – the train wreck of new unnecessary and redundant tags coming off as worse than the escalation of goofy proprietary tags during the original browser wars.
I suspect the people who never embraced the separation of presentation from content or semantic markup pissing in the pool – the same people my recently departed friend Dan Schulz was referring to when he said “The people who wrote fat bloated endless nested tables abusing TD’s for no good reason today just write fat bloated endless nested DIV’s abusing DL’s for no good reason”.
Oh, and **** theora. Buggy, bloated, slow, no hardware support. The only people who give a flying **** about it are the ‘free as in freedom’ whackjobs. The same ones who throw around the word freedom without actually understanding what it means.
Edited 2010-01-16 23:38 UTC
Bah, past the edit limit…
Just occurred to me:
>> The whole original idea for moving forward was
>> to have LESS tags, simpler tags
CISC (HTML 5) vs. RISC (HTML 4 STRICT)?
What HTML 4 Strict?
You mean XHTML 1.0/1.1 Strict.
You’re forgetting that XHTML was oppressively limited and thus relying on massive custom classes and id selectors to do a lot of work that is too commonly reused and could more easily be consolidated into standard element tags, thus the move back to more elements while adding more CSS with 3 to override if you want.
Edited 2010-01-17 12:14 UTC
Don’t take offense to this, but spoken like someone who doesn’t know what XHTML 1.0 is, or what TRANSITIONAL/STRICT means. From what you said it almost sounds like you’ve never even HEARD of HTML 4 Strict.
The ONLY difference between HTML 4 STRICT and XHTML 1.0 STRICT is the need for XML style closings. The only differences between STRICT and TRANSITIONAL, be it HTML or XHTML is that some tags and attributes you shouldn’t be using in the first damned place if you’re going to have CSS are ‘deprecated’ (aka don’t use them!).
That’s NOT “oppressively limited” nor does it require massive use of classes if you take the time to understand how the “cascading” part of CSS actually works. While certainly you have the complete **** retards who write up code like this throwing classes at EVERYTHING:
<div class=”nav”>
<ul class=”navList”>
<li class=”navItem”><a href=”/” class=”navAnchor”>Home</li>
<li class=”navItem”><a href=”/” class=”navAnchor”>FAQ</li>
<li class=”navItem”><a href=”/” class=”navAnchor”>Links</li>
</ul>
</div>
That’s not the fault of strict or transitional, XHTML or HTML, it’s the fault of nimrods who don’t get how CSS works – since NONE of those classes are neccessary… and the DIV around the UL is probably also not neccessary.
<ul id=”mainMenu”>
<li><a href=”/”>Home</li>
<li><a href=”/”>FAQ</li>
<li><a href=”/”>Links</li>
</ul>
Being all that’s probably needed for that. You’ll excuse me if I don’t consider NAV to be a giant savings over that:
<nav>
<li><a href=”/”>Home</li>
<li><a href=”/”>FAQ</li>
<li><a href=”/”>Links</li>
</nav>
When that’s just another tag for people to remember. Given that people already can’t be bothered to use FIELDSET, LEGEND, LABEL, CAPTION, TH, THEAD, TBODY, and can’t be bothered to learn that they shouldn’t be using CENTER, FONT, ALIGN, BORDER, BGCOLOR, etc, etc – much less learn how to use heading tags in the proper **** order… I don’t see tacking on a bunch more tags as an improvement.
Or asshat nonesense like this ****
<p class=”contentP”>Some text</p>
<p class=”contentP”>Some More text</p>
<p class=”contentP”>Some text</p>
<p class=”contentP”>Some text</p>
If all your content paragraphs are going to have the same class on them, what the *** are you putting a class on them for?!? Style them off the **** parent of make that the default behavior for all of them. Rule #1 of using classes and ID’s – only use them on elements that should be DIFFERENT from their kin.
The whole idea of STRICT was to be a lead-in to the next version of HTML, said version following the trend of having less tags so people might actually use the RIGHT tags for things. UL and OL supplanting MENU, DIR, etc… OBJECT supplanting IMG, APPLET and the proprietary EMBED – HTML5 seems to be completely undoing all that progress and in many cases adding tags that do the SAME *** THING some of the deprecated tags do. ‘NAV’ – WTF? We HAD a tag called MENU, both basically do the exact same thing as a UL, so why make a new tag instead of just bringing back the old one that ALREADY WORKS in every browser?!? Whiskey tango foxtrot!
Gets worse since that ends up being MORE markup since if you have one in the footer and one in the ‘header’, or multiple ‘navigation’ elements in the header (like a small top menu before the main menu) you’ll end up throwing classes or ID’s at them ANYWAYS. Look at google’s home page – there are two menus in that header up top, so you’ll end up resorting to a class or ID to do that same layout ANYWAYS.
That and certain tags like ‘header’ are just encouraging people to throw even more useless unneccessary wrappers around elements. We have HEADING tags, we don’t need a header tag. Footer I could ALMOST see, but really that’s not all that semantic for markup since header implies before, footer implies after, and as such that’s presentational… and presentation has no **** business in the markup!
Edited 2010-01-17 15:38 UTC
You’re an asshat.
All my CSS adheres to strict separation of Interface from Structure.
Your assumption I don’t grasp the concept of strict XML well-formedness as a kludge called XHTML is absurd.
They overreached and under-achieved with XHTML. They went hog wild with XML and after a decade of oops we used the wrong tool for the wrong job sensible development came along and is shaping HTML 5 into what HTML 4 should have been.
You jumping the gun on what you expect HTML 5 to ultimately become when it is labeled a Recommendation makes you look like a whiny twit.
Thank you so very much for showing that:
1) you apparently had NO CLUE what I was talking about when I said STRICT, as in STRICT and TRANSITIONAL which has NOTHING to do with if it’s ‘well formed’ XML or not. HTML can be poorly formed XML, and still have STRICT since HTML STRICT just means don’t use the tags deprecated or obsoleted in that version of HTML. You may want to familiarize yourself with:
http://www.w3.org/TR/REC-html40/sgml/dtd.html
So it doesn’t sound like you don’t know anything about HTML, or what STRICT means in reference to both it and XHTML 1.0. (or that XHTML 1.1 doesn’t even HAVE transitional or strict!)
2) That proving once again people who cannot defend their viewpoints usually resort to posts that are nothing more than a personal attack with nothing to actually add to the discussion. If you’re going to resort to the attack, back it up with some meat not some whiney little “Is not”.
Learn about a subject before opening mouth and inserting foot. There IS such a thing as HTML 4.01 Strict, HTML 4.0 STRICT and XHTML 1.0 STRICT… and news flash, there’s no such thing as XHTML 1.1 STRICT, since XHTML 1.1 has no Transitional! If you knew all your valid doctypes, you’d actually know that… If you knew ANYTHING meaningful about HTML, you’d probably know that…
Otherwise, you don’t know enough about HTML/XHTML to even be contesting what I said.
Seriously:
Really, then what’s this?
http://www.w3.org/TR/html4/appendix/changes.html#h-A.1.1.14
1.1 Strict, right…
http://www.w3.org/QA/2002/04/valid-dtd-list.html
Nope, not right. I don’t see XHTML 1.1 strict there – BECAUSE THERE’S NO SUCH THING!
But in BOTH HTML and XHTML we have STRICT, TRANSITIONAL and FRAMESET.
Get it now?
Edited 2010-01-17 20:05 UTC
It’d be great if Google goes ahead and adds HTML5 support to YouTube. Not only would it be a good example of what HTML5 can do, but it would also expose any limitations with it (which would hopefully be fixed in future). Nothing like giving software a good hammering to find any problems with it.
It could also be positive in another way. It may force Adobe to (further) open up the Flash format.
and pay Google oodles of cash to keep YouTube a Flash-based site.
I don’t like what the vudeo tag will bring for the future of cross platform video. Like others have pointed out there is no guarantee that site makers will use an open codec (or whatever Google chooses) over their own proprietary solution or existing ones like Quicktime or WMV. The issue is with DRM an open format and HTM5 does not resolve that issue and that is an extremely important issue that needs to be resolved. The other issue is convenience, having a user install Quicktime or WMP anytime they want to watch a video is going backwards, just like in 1999 when you had to have the right codec, the right player, the right version of the player in-order to play video . Flash made that a thing of the past. One plugin (which most people already had) is all you need and its used almost universally by content providers.
I’m not a fan of flash. I don’t understand why my core2duo MBP sounds like an airplane taking off every time I watch a silly video clip yet I can watch 1080p video in quicktime without barely a cpu fan murmur. The issue is the implementation of Flash by Adobe (really macromedia). Adobe released a standard but not a testing suite so that implementors can test with. What Adobe needs to do is take a long hard look at the code base and start to deprecate functionality that is no longer relevant and start all over again. However there is nothing wrong with the technology behind Flash, its a brilliant idea that just has a very bad implementation.
Two words for you: Hardware Acceleration
… and see why quicktime sucks donkey on Windows, since it’s hardware support there is a *** joke.
Functioning as a browser plugin has traditionally meant all blitting of the video files had to be handled by the browser – and browsers generally aren’t designed to handle datastreams that fast. It’s actually nothing short of a miracle it’s practical at all, and if it is, most of the time that’s ENTIRELY because it’s 100% CPU bound.
If you download those buggy/slow flash videos and play them in a native player, you can see what a difference just having hardware blitting can make even without hardware decoding. Shoving large amounts of memory like that from RAM to video memory just isn’t something a browser is set up to do – and why as a plugin Flash is nowhere near as good as it is when used as a standalone.
Which is why the next-generation beta of Flash (at least for windows) kicks some serious ass. It can leverage hardware accelleration for decoding, and has better hardware support to bypass the browser and do the blitting using whatever is present for 2d speedups.
As I mentioned in my post, it takes the Atom 1.6 which is not capable of HD video playback, and makes it able to handle 1080p HD video as flash smooth as silk even in the browser thanks to it seeing and using things like the nVidia ION.
Of course, it could just also be that you’re talking about a “Fisher Price my first computer”.
http://www.cad-comic.com/cad/20050608
Edited 2010-01-17 15:53 UTC
I don’t get the joke in that comic. Is it supposed to be funny? As for flash. Silverlight on the same machine barely breaks a sweat, so all the blitting nonsense while true still doesn’t explain why Flash runs like absolute garbage on any platform other than windows (it doesn’t run great there either, but cpu usage is not as drastic).
Annoying as the resource hog may be, it still does things that we can’t have elsewhere for the time being (ok, Shockwave, but…): Sync audio and client-side rendered vector graphic animations.
This is *very* powerful, and until someone writes an alternative VM for the browser that does the same and gets 99% browser coverage *and* open-sources the multi-million $ effort (yes, it costs money to develop software), Flash is what I’ll use.
Google could easily mess with Adobe and produce such a VM — they have the manpower, the experience amongst their engineers (or they could hire Nicolas Cannasse), but we all know that actually solving the problem costs a lot of money which nobody is prepared to shell.
The greater question is, why do we look to them to solve our problems? I don’t see many of the people complaining about Flash busy implementing an alternative. The answer is, of course, it is easier to complain than to shell out the dough.
The resource hogging *could* be solved quite easily, btw: allow the user to choose a maximum fraction of CPU cycles for the AVM and a priority level.
But that would make for a less smooth experience (audio and video is inherently soft real-time and thus need fine-grained scheduling or an advanced caching scheme) and the same users would be complaining about their sub-optimal experience instead.
Flash has been subjected to increasingly more critique:
http://www.osnews.com/comments/21901
http://www.osnews.com/comments/21680
This is embedded in a context of growing antipathy against Adobe. Can it be pinned down on “arrogance”? Maybe to some extent, but I think the main issue here lies in “strategy”, or “focus”.
I used to be a great fan of Adobe and their products. Granted, that was in the days that things were simple and straightforward; the days of non-bloated PDF document formatting, readers and additional software; the days before the acquisition of Macromedia (Flash, Shockwave, Dreamweaver).
1. Then (2006) there were the fights between Apple and Adobe revolving around the Intel move;
2. Then (both in 2006 as well) there was disagreement between Microsoft and Adobe, making Flash under IE rather unpleasant for a while and with Adobe being upset about not being mentioned on the list of software approved for Vista;
3. Gradually the nimble Acrobat Reader turned into the sluggish Adobe Reader, packaged with unnecessary software;
4. Now the latest version of the Adobe Reader fails to run on my Mac;
5. Now Flash frequently crashes in Safari and stalls Firefox;
6. Now Adobe has stated not to care for a 64-bit Shockwave…
Just like Microsoft has recently announced to return to their core businesses (a move long overdue), perhaps the time has also come for Adobe to reassess its position and cut off the excess. For years, the focus has been on features and possibilities. Consequently, the documents composed with Acrobat or Creative Suite applications offer many features, though some are unnecessary, indeed making it all a bit bloated.
A strange in strategy should among others lead to:
– a discontinuation of Shockwave;
– stripping Flash, to the extent that functionality better taken care of by other means (e.g., scripting languages, video solutions) is removed;
– PDF should be more clearly divided into domains, each with their dedicated functionalities (a black and white text&image scientific paper has different requirements from a perfectly styled iPod magazine).
Adobe still goes strong with their famed Creative Suite and Acrobat family. I am not suggesting this should at all change. I do, however, observe that Adobe is losing its reputation due to unfortunate and unnecessary problems revolving around part of their product line, that (a) has seen major developments over the past year that went faster than Adobe was able to handle, and (b) are better off being discontinued or drastically rethought to save face.