Anyone know what the Wii hardware is capable of for a video experience. As a SD device, it could do 480p60 or 576p50 at best. But the processor is basically a semi souped-up 800 MHz G3, right? My old 800 MHz G4 couldn't play back 480p30 High Profile H.264 and the AltiVec SIMD that the Wii lacks is a big help for that.
Perhaps the ATI video card inherited some DXVA features?
There's some DVD playback, so we know MPEG-2 works, and I could imagine VC-1 or MPEG-4 part 2 (divx/xvid) working for 480p24. But unless there's some dedicated hardware in there, H.264 Main or High profile seems pretty unlikely.
While there's certainly no lack of FPS games on Xbox 360, there's also plenty of accessible family-friendly titles. We've been having great fun over the holidays with the new Banjo-Kazooie: Nuts and Bolts. The platforming aspects don't appeal to all of the older set, but everyone really gets into the vehicle design aspect of it. Particularly the middle-aged and up men who grew up working on cars.
Rock Band/RB2 is also a big hit, and the singing position is great for those intimidated by controllers.
Xbox.com shows 347 Xbox 360 games rated "E (Everyone)" compared to 181 "M (Mature)" FWIW. If you include the 91 "E10" games, that makes more than twice as many family-friendly titles as Mature.
Yep, that's a whole different workflow when you can do it all together.
Kudos if you can do all steps of the process well. That can result in some absolutely incredible work when a single vision can inform the whole process.
Myself, I've done enough graphic design (production manager for the college student paper, that kind of thing) to realize I'm simply not that good at the aesthetic aspects of it. I was happy to have my words in the hands of a professional.
Hmm. I'm quite sure that Word was capable of doing the above at least as far back as Word '97, when going from.doc to.doc. Sometimes "Smart Quotes" had to be fiddled with to make sure it didn't get autocorrected as it was typed, but that was pretty easy to do. Certainly the Mac made it pretty easy to insert any character in a font like that; I hadnt used WinWord much back then.
My book was a few years later than that, but I definitely used Word and defintely had some text formatting complexities that went fine in the workflow. For example, in video Y' means luma, but Y with a closed single quote doesn't mean anything.
Perhaps your publishers were running some kind of autocorrect macro that messed things up or something (which would be a fine argument for your "never work with them again" plan the second time it happened).
And I've got a couple due in 2009 for different publishers, so this has been much on my mind.
Based on a lot of the other comments, people are really focusing on the formatting aspects of the workflow: Latex, FrameMaker and all that. But if you're writing a book for a standard tech publisher, you likely will never even have a direct conversation with whomever does the layout. You turn in structured text and figured to an editor, when then passes it off to layout after editing.
And if it's any kind of a series, they'll be doing formatting according to a well defined template and style that'll map to the styles in the document you give them.
So, the actual workflow is that you get a Word template, and write everything in there. The key thing is to follow the Styles religiously - every paragraph should have one as you type it. Think writing in old school HTML, or XML to someone else's Schema.
Also, try not to even think about formatting; there's no saying what goes on what page based on Page Preview in Word or alternative. If you want a new section, use a section break. This is object-oriented writing, where you're really trying to get the content into the right structure for easy processing later on.
I recommend working in Outline and Normal/Draft mode only, since that's where you see the structure of what you're doing. Personally, I'm a born again believer in outlining. I outline a chapter, and then jump in and write the part of it I'm thinking about at the moment. With the outline there, it's easy to realize I need to introduce a concept earlier in the chapter and then jump there and do a quick sketch of it, since the earlier section already exists in the structure. The act of writing an outline also helps define all the stuff you didn't know you needed to figure out.
But don't be a slave to the outline as it exists; structure can need editing as much as prose. Don't be afraid of moving sections and chapters around as helps you communicate better. That's a lot easier to do early in the process.
The basic problem was that MCP's editors (I guess copy editors initially) loaded the text I gave them into Microsoft Word (I assume, I can't remember if they confirmed this). It immediately "corrected" all the punctuation. Since the book was about Unix, there was an abundance of single and double quotes, backticks, and so forth. They all got totally screwed up. On proof reading, I spotted these, fixed them and sent the corrected text back. Then of course they loaded the text into Word again and broke everything a second time.
Well, it doesn't sound like you picked the wrong publisher as you weren't using the correct tool.
Every publisher I know has a whole style sheet and a Word template that tells you exactly how to do all your formatting. And if you get the formatting write in there, then they aren't going to convert it to anything and all your formatting would stay intact. By "load text into word" do you mean you sending were them a.txt file?
The style guides tend to be very clear on what you can and can't do about formatting, and it's really not worth trying to outthink them as that'll mess up the later production process like you saw here.
Unless they were so out of it that they actually didn't tell you anything about how you were supposed to send your stuff in advance, in which case, yes, avoid them like the plague:)!
Theora is a codec, equivalent to VC-1 and H.264. It's also a lot less efficient, so you'd get lower quality given a particular bitrate.
AVI is a wrapper, equivalent to AVI, ASF, or MPEG-4. If you're talking about alternatives to AVI for Theora, you probably mean Ogg
But Silverlight itself isn't a media format, but a rich application runtime for browsers based on.net. Silverlight 2 mainly uses WMV as its media format, but it will be addding MPEG-4 and H.264 soon.
I have no idea what the OP meant about AVI solving all the problems, honestly. It's not a very good wrapper; its main value is simplicity.
Re:FTA: "Silverlight 2 is a cross-platform...."
on
Silverlight 2.0 Released
·
· Score: 2, Informative
Note that Flash on mobile is very different than Flash on the desktop. Flash Lite 2.0 is really a cut-down version of Flash 7, and so no ActionScript 3.0, very different video support, etcetera. It's not like a dekstop Flash 9.swf is going to run unmodified on a phone.
Also, Silverlight for Windows Mobile and Symbian (Nokia phones) has been announced and is in development.
I'm sure that Microsoft kindly shared the specs for SilverLight 2.0 with Mono/Novell during the development so that the Mono project would not have to play catch-up once 2.0 came out. Right?
Are you looking for something like publishing the XAML spec under the Open Specification Promise?
But even the best DHTML/JS tree control, tabs, slider panel, etc run slower then native widgets, Silverlight, Flash/Flex.
Isn't Flex just a development environment on top of the normal Flash ActionScript 3 implementation? I don't know why Flex would offer any better performance than Flash. The.NET CLR/DLR in Silverlight is enormously faster than either.
Re:I thought the Olympics was a silverlight failur
on
Silverlight 2.0 Released
·
· Score: 1, Interesting
My impression was that the amount of Olympics streaming using Silverlight was less than YouTube during the same time period. If so, it doesn't seem like much of a success to me.
If the bar for success for video on the web is deliver more content than YouTube, than there has not been a success in web video since YouTube launched:). 9.9 million hours of video in 17 days is a whole lot of video.
Some better metrics for success might be: Was it profitable for NBC? Did viewers get a good experience? Did it innovate anything new in video delivery?
My biased opinion is "yes" in all three categories.
(Calling it a success because people installed silverlight isn't much. Afterall, the same people would have probably installed a rootkit and trojan in order to watch the Olympic streaming. They just don't care.)
People don't care what software they install as long as it delivers what they want? Probably true, but that sounds more like a feature of Silverlight, not a bug.
If a consumer is always aware what technology is in their media player, the player is probably too obtrusive. The user should be mainly aware of the awesome experience.
Likewise Microsoft aren't helping so they don't want it either; in fact they are actively hindering with their usual technique of making Silverlight a moving target.
Eh? It's an in development technology! How exactly could it NOT be a moving target? There's been breaking changes documentation available for quite a while indicating what was different from Beta 2 to the final version.
In more relevant news, there's now Microsoft-funded Eclipse support for Silverlight coming http://www.eclipse4sl.org/
Except that a plugin is not available on Linux. MS touts Moonlight as a nearly complete port of Silverlight to Linux but in fact it's very far from being usable - even Gnash is light years ahead of Moonlight when compared with their closed source versions.
Bear in mind Moonlight is GPL'ed. The appropriate response on Slashdot is, I believe, "if you want it, start writing some code":).
A good post on this topic from my colleague Alex Zambelli:
I think it's important to remember in this discussion that BD discs are expensive and difficult to manufacture, with constrained production for the dual-layer 50 GB discs. The production lines to make the discs are also more expensive and produce fewer discs per day, so the amortization of that takes up a bigger chunk of the price of each BD disc.
On of the big advantages of HD DVD was that it could use upgraded DVD replication lines, and so the cost per disc was lower.
Very broadly discussed on AVS Forum a few years ago.
You are right and wrong. If I want to keep MPEG-LA happy, i need to pay quite a bit of money if there are more than 100K downloads IIRC. For a open source project that is *not* selling the code it is a lot of money. Most OS projects are not done with money, but with time.
But if you're just distributing the source, you wouldn't need to pay the patent. If it's a real product, or if users are compiling it into one, is doesn't seem to be an infringement on "freedom" in the classic RMS definition to wind up paying a fee. You still have full control over the code and technology used.
But here is the real rub. Even if you pay the fees, they give *no* guarantee what so ever that you are not infringing some other patent.
Theoretically true, although that hasn't happened much in practice, at least in this space. And Theora and Dirac are in the same legal position, and don't have the market effet of lots of companies looking for patents to assert to get a share of the MPEG-LA revenue.
Also these non discriminatory terms are crap. If the fee is high enough, it discriminates plenty. Add to that the bulk of these patents are utter rubbish. I would say there are not with the spirit of anything free, freedom, beer, or otherwise.
Have you actually all of the H.264 or VC-1 patents? If so, you've got more patience than I do!
Also, it's not ND, but RAND - REASONABLE and non-discrimintory. If anything H.264 and VC-1 are much more reasonable than MPEG-2, since they're a lot cheaper per unit and have an annual cap for high-volume products.
Anyway, feel free to hate software patents, but it seems to me like that's a different issue than open source.
Except that Theora not only start being behind the then-emerging H.264 and VC-1, but its implementations are launch were quite a bit weaker than even Xvid of the era, and have essentially stagnated.
And since then, it's fallen even further behind; the implementations of standardized codecs has been improving a lot more each year than Theora, as have proprietary codecs like the later entries in On2's VPx series (Theora was forked from VP3; On2 just announced VP8).
There's been some interesting work in the last year or so, but very much research, not development. The new screen shots showing improvement a few months ago didn't even include rate control, and weren't close to being usable in production.
In the same six months, x264 has made a variety of dramatic improvements that are already usable in the real world.
So, question: why so little progress on Theora for so long? For all the focus of the industry on "we must have a non-patent encumbered codec!" it seems like the open source efforts have all gone into xvid and x264. Anyone know what was blocking progress on Theora for so long?
This came up in yesterday's discussion of the Canonical codec pack.
Standardized codecs, like VC-1 and H.264, have full open specifications and typically even reference source code implementations that can be reused in a variety of ways.
However, they also require patent fees depending on use and jurisdiction.
The issue of free software has always been asserted to be about "speech, not beer" but it seems like there's an assumption that it has to be free as in "speech AND beer." I'm sure all kind of arguments can be made that it should be that way, but everything I've read recently sort of begs that question by conflating the issues of closed source and patent licensing. When to me they look like pretty orthogonal issues; all of the reasons why people say they like open source are still delivered if the source is open, irrespective of whether a patent fee is paid. And MEPG-LA makes patent administration and payment pretty straightforward under RAND terms.
I think it's a lot more than a semantic difference!
So, the classic statement around here is that it's not free software if it's free as in beer, but not as in speech.
But standardized-but-patented is actually free as in speech, but not as in beer. You can see the source code, modify it, share the modifications, etcetera. But, depending on juristiction, you've also got to pay the patent fee for some uses.
This seems to involve a very different set of concerns and objections than the normal ones given to commercial software.
No. I'm sure they're making a pretty penny here. The highest single codec license fee I know if is MPEG-2, which was $2.50 last I checked. VC-1 and H.264 are less than a dollar each.
Lots more about codec licensing than you'd ever care to learn can be found at http://www.mpegla.com/.
These aren't propritary codecs, they just have patents on them.
The codecs themsleves are standardized by entities like ISO and SMPTE, and typically have full reference source code freely available.
"open but with royalty bearing patents" is a very common circumstance, really.
Also, whether or not you're authoring with the codec or decoding is immaterial. There are different patents for encoding and decoding, and a product that does both typically has to pay more than a product that only does one.
As it comes to universality, Flash has broader coverage than any other distribution mechanism in the real world today, and therefore Flash would be the appropriate choice for Web distribution.
Often stated, but never verified. The published methadology for the Flash numbers is a consumer survey, and thus doesn't capture enterprise desktops, which are more likely to not have Flash installed, or have older versions with much weaker video support.
WMV is almost certainly more broadly supported, since it's on 100% of all Windows machines, plus a big chunk of Macs via Flip4Mac and Silverlight. Open source tools like VLC and (soon) Moonlight have WMV support as well. It's certainly much more supported than Flash versions that support H.264, since that's been out less than a year.
Slashdot Mirror
Anyone know what the Wii hardware is capable of for a video experience. As a SD device, it could do 480p60 or 576p50 at best. But the processor is basically a semi souped-up 800 MHz G3, right? My old 800 MHz G4 couldn't play back 480p30 High Profile H.264 and the AltiVec SIMD that the Wii lacks is a big help for that.
Perhaps the ATI video card inherited some DXVA features?
There's some DVD playback, so we know MPEG-2 works, and I could imagine VC-1 or MPEG-4 part 2 (divx/xvid) working for 480p24. But unless there's some dedicated hardware in there, H.264 Main or High profile seems pretty unlikely.
While there's certainly no lack of FPS games on Xbox 360, there's also plenty of accessible family-friendly titles. We've been having great fun over the holidays with the new Banjo-Kazooie: Nuts and Bolts. The platforming aspects don't appeal to all of the older set, but everyone really gets into the vehicle design aspect of it. Particularly the middle-aged and up men who grew up working on cars.
Rock Band/RB2 is also a big hit, and the singing position is great for those intimidated by controllers.
Xbox.com shows 347 Xbox 360 games rated "E (Everyone)" compared to 181 "M (Mature)" FWIW. If you include the 91 "E10" games, that makes more than twice as many family-friendly titles as Mature.
Yep, that's a whole different workflow when you can do it all together.
Kudos if you can do all steps of the process well. That can result in some absolutely incredible work when a single vision can inform the whole process.
Myself, I've done enough graphic design (production manager for the college student paper, that kind of thing) to realize I'm simply not that good at the aesthetic aspects of it. I was happy to have my words in the hands of a professional.
Hmm. I'm quite sure that Word was capable of doing the above at least as far back as Word '97, when going from .doc to .doc. Sometimes "Smart Quotes" had to be fiddled with to make sure it didn't get autocorrected as it was typed, but that was pretty easy to do. Certainly the Mac made it pretty easy to insert any character in a font like that; I hadnt used WinWord much back then.
My book was a few years later than that, but I definitely used Word and defintely had some text formatting complexities that went fine in the workflow. For example, in video Y' means luma, but Y with a closed single quote doesn't mean anything.
Perhaps your publishers were running some kind of autocorrect macro that messed things up or something (which would be a fine argument for your "never work with them again" plan the second time it happened).
But I don't think Word itself was the problem.
I wrote a book a few years ago.
That was in Office v.X on an old PowerBook (I even started in the original "can't print" beta of Word v.X).
http://www.amazon.com/Compression-Great-Digital-Video-Techniques/dp/157820111X/
And I've got a couple due in 2009 for different publishers, so this has been much on my mind.
Based on a lot of the other comments, people are really focusing on the formatting aspects of the workflow: Latex, FrameMaker and all that. But if you're writing a book for a standard tech publisher, you likely will never even have a direct conversation with whomever does the layout. You turn in structured text and figured to an editor, when then passes it off to layout after editing.
And if it's any kind of a series, they'll be doing formatting according to a well defined template and style that'll map to the styles in the document you give them.
So, the actual workflow is that you get a Word template, and write everything in there. The key thing is to follow the Styles religiously - every paragraph should have one as you type it. Think writing in old school HTML, or XML to someone else's Schema.
Also, try not to even think about formatting; there's no saying what goes on what page based on Page Preview in Word or alternative. If you want a new section, use a section break. This is object-oriented writing, where you're really trying to get the content into the right structure for easy processing later on.
I recommend working in Outline and Normal/Draft mode only, since that's where you see the structure of what you're doing. Personally, I'm a born again believer in outlining. I outline a chapter, and then jump in and write the part of it I'm thinking about at the moment. With the outline there, it's easy to realize I need to introduce a concept earlier in the chapter and then jump there and do a quick sketch of it, since the earlier section already exists in the structure. The act of writing an outline also helps define all the stuff you didn't know you needed to figure out.
But don't be a slave to the outline as it exists; structure can need editing as much as prose. Don't be afraid of moving sections and chapters around as helps you communicate better. That's a lot easier to do early in the process.
The basic problem was that MCP's editors (I guess copy editors initially) loaded the text I gave them into Microsoft Word (I assume, I can't remember if they confirmed this). It immediately "corrected" all the punctuation. Since the book was about Unix, there was an abundance of single and double quotes, backticks, and so forth. They all got totally screwed up. On proof reading, I spotted these, fixed them and sent the corrected text back. Then of course they loaded the text into Word again and broke everything a second time.
Well, it doesn't sound like you picked the wrong publisher as you weren't using the correct tool.
Every publisher I know has a whole style sheet and a Word template that tells you exactly how to do all your formatting. And if you get the formatting write in there, then they aren't going to convert it to anything and all your formatting would stay intact. By "load text into word" do you mean you sending were them a .txt file?
The style guides tend to be very clear on what you can and can't do about formatting, and it's really not worth trying to outthink them as that'll mess up the later production process like you saw here.
Unless they were so out of it that they actually didn't tell you anything about how you were supposed to send your stuff in advance, in which case, yes, avoid them like the plague :)!
What are you looking for in terms of API or codec info?
Silverlight will be adopting H.264 in its next major version: http://on10.net/blogs/benwagg/H264-and-AAC-support-coming-in-Silverlight/
The VC-1 codec is already a SMPTE spec: http://en.wikipedia.org/wiki/VC-1
The Silverlight Xaml vocabulary spec for Silverlight 2 was published in September: http://download.microsoft.com/download/0/A/6/0A6F7755-9AF5-448B-907D-13985ACCF53E/%5BMS-SLXV%5D.pdf
Theora is a codec, equivalent to VC-1 and H.264. It's also a lot less efficient, so you'd get lower quality given a particular bitrate.
AVI is a wrapper, equivalent to AVI, ASF, or MPEG-4. If you're talking about alternatives to AVI for Theora, you probably mean Ogg
But Silverlight itself isn't a media format, but a rich application runtime for browsers based on .net. Silverlight 2 mainly uses WMV as its media format, but it will be addding MPEG-4 and H.264 soon.
http://on10.net/blogs/benwagg/H264-and-AAC-support-coming-in-Silverlight/
I have no idea what the OP meant about AVI solving all the problems, honestly. It's not a very good wrapper; its main value is simplicity.
Note that Flash on mobile is very different than Flash on the desktop. Flash Lite 2.0 is really a cut-down version of Flash 7, and so no ActionScript 3.0, very different video support, etcetera. It's not like a dekstop Flash 9 .swf is going to run unmodified on a phone.
Also, Silverlight for Windows Mobile and Symbian (Nokia phones) has been announced and is in development.
I'm sure that Microsoft kindly shared the specs for SilverLight 2.0 with Mono/Novell during the development so that the Mono project would not have to play catch-up once 2.0 came out. Right?
Are you looking for something like publishing the XAML spec under the Open Specification Promise?
http://blogs.windowsclient.net/rob_relyea/archive/2008/10/14/ms-slxv-silverlight-xaml-vocabulary-2008-specification-v0-9-published.aspx.
Otherwise, Microsoft would be releasing a technology that will only work reliably on Windows and shun the other major platforms.
Hum... I wonder why they just don't do like Adobe or Sun and release a version for Linux, Mac and Windows?
All versions of Silverlight have shipped day-and-date Mac/Win. Mono is doing a Linux version called Moonlight.
http://www.microsoft.com/silverlight/resources/install.aspx#sysreq
Platforms for Silverlight 2:
Windows Vista (including Windows Server 2008)
Windows XP SP2
Windows 2000
Windows Server 2003
Mac OS 10.4.8+ (Intel only)
Browsers:
Internet Explorer 6 and 7
FireFox 1.5, 2, and 3
Safari
But even the best DHTML/JS tree control, tabs, slider panel, etc run slower then native widgets, Silverlight, Flash/Flex.
Isn't Flex just a development environment on top of the normal Flash ActionScript 3 implementation? I don't know why Flex would offer any better performance than Flash. The .NET CLR/DLR in Silverlight is enormously faster than either.
My impression was that the amount of Olympics streaming using Silverlight was less than YouTube during the same time period. If so, it doesn't seem like much of a success to me.
If the bar for success for video on the web is deliver more content than YouTube, than there has not been a success in web video since YouTube launched :). 9.9 million hours of video in 17 days is a whole lot of video.
Some better metrics for success might be:
Was it profitable for NBC?
Did viewers get a good experience?
Did it innovate anything new in video delivery?
My biased opinion is "yes" in all three categories.
I've got this blog post with some more details about Silverlight and the Olympics:
http://on10.net/blogs/benwagg/Final-Olympics-numbers/
(Calling it a success because people installed silverlight isn't much. Afterall, the same people would have probably installed a rootkit and trojan in order to watch the Olympic streaming. They just don't care.)
People don't care what software they install as long as it delivers what they want? Probably true, but that sounds more like a feature of Silverlight, not a bug.
If a consumer is always aware what technology is in their media player, the player is probably too obtrusive. The user should be mainly aware of the awesome experience.
Likewise Microsoft aren't helping so they don't want it either; in fact they are actively hindering with their usual technique of making Silverlight a moving target.
Eh? It's an in development technology! How exactly could it NOT be a moving target? There's been breaking changes documentation available for quite a while indicating what was different from Beta 2 to the final version.
In more relevant news, there's now Microsoft-funded Eclipse support for Silverlight coming
http://www.eclipse4sl.org/
Except that a plugin is not available on Linux. MS touts Moonlight as a nearly complete port of Silverlight to Linux but in fact it's very far from being usable - even Gnash is light years ahead of Moonlight when compared with their closed source versions.
Bear in mind Moonlight is GPL'ed. The appropriate response on Slashdot is, I believe, "if you want it, start writing some code" :).
A good post on this topic from my colleague Alex Zambelli:
http://alexzambelli.com/blog/2008/08/16/moonlight-20-help-wanted/
I think it's important to remember in this discussion that BD discs are expensive and difficult to manufacture, with constrained production for the dual-layer 50 GB discs. The production lines to make the discs are also more expensive and produce fewer discs per day, so the amortization of that takes up a bigger chunk of the price of each BD disc.
On of the big advantages of HD DVD was that it could use upgraded DVD replication lines, and so the cost per disc was lower.
Very broadly discussed on AVS Forum a few years ago.
You are right and wrong. If I want to keep MPEG-LA happy, i need to pay quite a bit of money if there are more than 100K downloads IIRC. For a open source project that is *not* selling the code it is a lot of money. Most OS projects are not done with money, but with time.
But if you're just distributing the source, you wouldn't need to pay the patent. If it's a real product, or if users are compiling it into one, is doesn't seem to be an infringement on "freedom" in the classic RMS definition to wind up paying a fee. You still have full control over the code and technology used.
But here is the real rub. Even if you pay the fees, they give *no* guarantee what so ever that you are not infringing some other patent.
Theoretically true, although that hasn't happened much in practice, at least in this space. And Theora and Dirac are in the same legal position, and don't have the market effet of lots of companies looking for patents to assert to get a share of the MPEG-LA revenue.
Also these non discriminatory terms are crap. If the fee is high enough, it discriminates plenty. Add to that the bulk of these patents are utter rubbish. I would say there are not with the spirit of anything free, freedom, beer, or otherwise.
Have you actually all of the H.264 or VC-1 patents? If so, you've got more patience than I do!
Also, it's not ND, but RAND - REASONABLE and non-discrimintory. If anything H.264 and VC-1 are much more reasonable than MPEG-2, since they're a lot cheaper per unit and have an annual cap for high-volume products.
Anyway, feel free to hate software patents, but it seems to me like that's a different issue than open source.
Except that Theora not only start being behind the then-emerging H.264 and VC-1, but its implementations are launch were quite a bit weaker than even Xvid of the era, and have essentially stagnated.
And since then, it's fallen even further behind; the implementations of standardized codecs has been improving a lot more each year than Theora, as have proprietary codecs like the later entries in On2's VPx series (Theora was forked from VP3; On2 just announced VP8).
There's been some interesting work in the last year or so, but very much research, not development. The new screen shots showing improvement a few months ago didn't even include rate control, and weren't close to being usable in production.
In the same six months, x264 has made a variety of dramatic improvements that are already usable in the real world.
So, question: why so little progress on Theora for so long? For all the focus of the industry on "we must have a non-patent encumbered codec!" it seems like the open source efforts have all gone into xvid and x264. Anyone know what was blocking progress on Theora for so long?
This came up in yesterday's discussion of the Canonical codec pack.
Standardized codecs, like VC-1 and H.264, have full open specifications and typically even reference source code implementations that can be reused in a variety of ways.
However, they also require patent fees depending on use and jurisdiction.
The issue of free software has always been asserted to be about "speech, not beer" but it seems like there's an assumption that it has to be free as in "speech AND beer." I'm sure all kind of arguments can be made that it should be that way, but everything I've read recently sort of begs that question by conflating the issues of closed source and patent licensing. When to me they look like pretty orthogonal issues; all of the reasons why people say they like open source are still delivered if the source is open, irrespective of whether a patent fee is paid. And MEPG-LA makes patent administration and payment pretty straightforward under RAND terms.
Did I just miss some prior discussions?
I think it's a lot more than a semantic difference!
So, the classic statement around here is that it's not free software if it's free as in beer, but not as in speech.
But standardized-but-patented is actually free as in speech, but not as in beer. You can see the source code, modify it, share the modifications, etcetera. But, depending on juristiction, you've also got to pay the patent fee for some uses.
This seems to involve a very different set of concerns and objections than the normal ones given to commercial software.
No. I'm sure they're making a pretty penny here. The highest single codec license fee I know if is MPEG-2, which was $2.50 last I checked. VC-1 and H.264 are less than a dollar each.
Lots more about codec licensing than you'd ever care to learn can be found at http://www.mpegla.com/.
Microsoft does not support DVD playback for free but most PCs sold with DVD drives have playback software installed by the OEM.
Vista Home Premium and Ultimate also include Microsoft's own DVD playback for WMP and Media Center on-disc.
These aren't propritary codecs, they just have patents on them.
The codecs themsleves are standardized by entities like ISO and SMPTE, and typically have full reference source code freely available.
"open but with royalty bearing patents" is a very common circumstance, really.
Also, whether or not you're authoring with the codec or decoding is immaterial. There are different patents for encoding and decoding, and a product that does both typically has to pay more than a product that only does one.
We'll, it's really a math question.
WMV: Windows + Mac * (Flip4Mac OR Silverlight OR VLC) + Linux * VLC
Flash: Windows * Flash + Mac * Flash + Inux * Flash
If you're think of H.264 enbled Flash, I think even Dobe only states a 80% consumer market share, and it'll be corresoningly lower in Enterprise.
All VC-q needs is a 6 year old version of WMP, while H.264 needs a less than a year old version of Flash.
As it comes to universality, Flash has broader coverage than any other distribution mechanism in the real world today, and therefore Flash would be the appropriate choice for Web distribution.
Often stated, but never verified. The published methadology for the Flash numbers is a consumer survey, and thus doesn't capture enterprise desktops, which are more likely to not have Flash installed, or have older versions with much weaker video support.
WMV is almost certainly more broadly supported, since it's on 100% of all Windows machines, plus a big chunk of Macs via Flip4Mac and Silverlight. Open source tools like VLC and (soon) Moonlight have WMV support as well. It's certainly much more supported than Flash versions that support H.264, since that's been out less than a year.