Windows XP = lean Windows Vista = fat Windows 7 = leaner than Vista = Windows XP
I must say that "bloat" is about the least information-laden phrase I hear bandied about:).
What's a consensus defintion of what it means? Wasteful use of RAM? Any additional use of RAM? Does hard drive space count? What if it's for optional non-RAM loaded stuff like templates?
Is is bloat for Vista to include a lot of printer drivers in the default image? It wasn't good for Netbooks with small SSD drives, but didn't impact system performance. And I remember lots of complaints about the full install size of Office back in the day, even though that was mainly templates that didn't need to be installed.
I think it'd be useful if we all were a little more specific about that.
Back in the SP1 days, you could run XP acceptably on 256MB of RAM, and pretty decently on 512. Today, 512 feels cramped, and is the bare minimum I'd recommend for running SP3 and all the security patches.
1GB is a more reasonable minimum if you actually want to use apps. Firefox 3 is hungry enough that it'll use up 100-300MB if you have a lot of tabs open, so you really do need at *least* 1GB to run the OS plus just a web browser, which is really pretty minimal in terms of applications.
Haven't you just described 3rd party apps getting using more RAM? Unless you have some actual data comparing working set between XP RTM and XP SP3, that seems like a more obvious difference here.
Also, given the price of RAM, isn't it appropriate for apps to be using more memory to provide better performance, roughly tracking increases in average user RAM? Sure, I could run Netscape Navigator 3 on my 20 MB Quadra 700, but on the aggragate I think I'll take the last 12+ years of browser improvements!
Since we deliver the codec pack, that doesn't need to wait for a Moonlight implementation, so it'll hopefully be faster. It's too early to talk about specific schedules, though, as we haven't even announced a release date for the H.264 version of Silverlight.
For example, the current Moonlight 1.0 beta codec pack includes all of the SL2 codecs (specifically WMA 10 Pro which wasn't included in Silverlight 1.0).
This isn't just about a stream, but a media experience that can incorporate rich media and interactivity. For example, the pop-up stream switcher that's still available in full-screen mode.
Silverlight offers a single plugin that provides XAML + managed code + media support. It's a great fit for developing media player applications that can live in a browser that have features beyond what a classic media player can handle. And then delivering them without having the security issues of installing or running native code, or requiring a user to authenticate or install.
Silverlight 3 will be adding MPEG-4 support, this year, FWIW.
While I am happy that they made an open source port so quickly however. It is just a hack for a one time occasion. For people with different needs will still be left out in the cold.
No, it was a hack because the current beta is Silverlight 1.0 compatible (no managed code, just JavaScript for control logic) while the main player was Silverlight 2, using managed code.
So what Novell did in an inspirational rush was reimplement a version of the player using JavaScript.
Now that this model works, it'll be availble for other events. Of course, it won't be needed once Moonlight gets to Silverlight 2 parity.
Good news everyone. After burning some midnight oil, we (Microsoft) and Novell have pulled together a Silverlight 1.0 compatible verison of the inauguration player that's compatible with both Moonlight Beta 1 and Silvelight 1.0 on PowerPC Macs.
It's a bad article. There's nothing intrinsically infringing about MKV + H.264. And there's defintely nothing Blu-ray specific about this, unless they're saying that it can handle a MKV that encapsulates the unrecompressed original Blu-ray bitstreams. Is anyone actually using THAT for piracy? I thought most download HD was still 720p.
DivX has a very different useage model than Flash (downloads with a player app instead of streaming to a browser plugin). They're hardly competing head to head.
As for container support, Flash is just MPEG-4. The Adobe model is to use "dumb" media formats and control high-level display functions via the Flash application itself.
That said, a MKV parser could be written for Silverlight via managed code, including caption overlay support, multilanguage audio, etcetera. The Silverlight MediaStreamSource API allows for parsers and protocols to be written in managed code, while still using the native decoders built into Silverlight.
I asked the digital media guys if there were any new features for encoding in Win 7, and they said that it supports up to 8-way threading for WMV encoding! The limit in Vista and XP was 4-way threading.
This requires either 8 cores or 4 hyperthreaded cores, and the resolution of the encoded file be at least 480 pixels tall.
I bet this will get used and abused by benchmarkers. Hopefully they'll use an easy-to-decode source file so that the encoder doesn't get bound by source decode speed.
If anyone's curious about what's going on under the hood, I did a blog post a bit ago about how Netflix does their VC-1 encoding for the service and how it's evolved over time.
For those talking about quality, a few points I want to make:
It won't ever look better than the source; some stuff is only availble in lower quality SD masters. Stuff like "The Office" that comes in as HD looks spectacular at the highest bitrate.
HD (up to 3800 Kbps) quality today is only on devices, while the desktop version only goes up to SD (1500 Kbps).
Those bitrates are averages, with the peaks higher. You'll generally need to be able to sustain 1.5x the listed bitrate in order to reliably get the maximum bitrate band for that content. Today, the desktop version is SD only
re's lots of ways to handle this other than full virtualization that offer a better user experience.
Great, then bring them up, because what MS has been doing isn't it.
Perhaps you haven't been reading Chris Jackson's blog discussing the myriad ways that Windows balanaces compatibility of older applications with allowing improvements to OS:
And here's all the stuff about the shims that let old apps that want to do stuff like write to system.ini still run without actually writing anything to system.ini:
It's hard to imagine any rating system that's more screwed up than the system used in most companies to rate the output of their programmers. If they were consciously trying to sabotage the software, they probably wouldn't come up with any schemes nearly as effective as judging programmers by lines of code produced.
Fully agreed. The other post's anecdote about having trouble getting a review by makign code smaller is quite telling.
Anyway, that kind of measurement certainly isn't done here at Microsoft (although plenty of productive programmers write a lot of lines of code). By example, great stuff like the speed/quality improvements in video playback and scaling in Silverlight 2 didn't really add much or any lines of code; it was about superior design. The guy who wrote that stuff may not rate that highly in KLOC/week, but he's certainly recognized as a rock star.
Theres a bunch of different streams it switches between dynamically. Which one you get can be limited by:
Available bandwidth (we'll send as many bits as won't cause buffering) Available CPU speed (we don't want you to drop a bunch of frames) Resolution of the media player (no need to send 1280x720 when the video window is scaled down to 320x240)
All of the above can be switched every couple of seconds, so if you have a heavy CPU process in the background, the data rate will drop during it, and then go back up again after.
To make sure you're not blocked by frame size, you can alt-Enter in IE to take the browser full screen.
Also, to see your current stream, mouse over on the little horizontal bars in the lower right corner.
Lastly, bear in mind this is a pre-beta here. We'll have big improvements by the time it goes 1.0 next year.
Yes, all targeted tax collection is simply forced "wealth transfer" hidden behind yet another name.
Is there a non-targeted form of tax collection you're offering as an alternative:).
Taxation is necessary to pay for public good and services that would otherwise be unfunded. Taxation is also appropriate to fund activities that are more efficiently done by government than without it.
Given that, there's a couple of different axes to figure out: what's the minimum level of taxation to provide a balanced budget over the business cycle with appropriate spending (governments should run a surplus in good times and a deficit in bad times), and what's the right combination of taxes that provides that revenue with a minimal amount of friction to economic growth and other goals. That's why cigarette and gas taxes are a good thing, as they provide revenue while discouraging behavior we want to discourage. If you need to get revenue from somewhere, that's a whole lot better than many alternatives.
An example of a really bad tax would be a gross receipts tax, since it's painful to administer, and hugely distorting as it massively rewards vertical integration, driving out smaller innovative companies.
Inheritance taxes are another good one. Lots of people seem to hate them for some reason, but it seems like the dead will find being taxes more than most, and enabling the decedents of wealth to remain wealthy without any economic input of their own for generations seems the wrong incentive for future wealth generation.
Anyway, there's often a knee-jerk attitude that all taxes are equally bad, and no tax should ever be added or raised, even to offset lowering other taxes. But we need taxes, and there's differences between them, so we should develop as optimal a tax system as we can.
Microsoft developed the codecs used by *both* HD-DVD and Blu-Ray. Microsoft had no significant interest regarding which format took hold --- in either scenario, the players would be running Microsoft software. Microsoft's only interest was for the format war end quickly.
Oh, the codec side of things was relatively minor. We developed VC-1 and related tools, but have a patent position and a lot of involvement in H.264 as well, and that side of things always supported VC-1 on BD as well.
The bigger effort and team was focused on building the interactive players for the Toshiba and Xbox 360 players. The whole HDi layer was jointly developed by Microsoft and Disney.
As a XML markup + scripting code-behind, it was a lot like a subset of Silverlight, actually.
I don't think Microsoft wanted either format to gain critical mass - wide and early adoption is a threat to Microsoft's goal of 'services', including pay per view and digital downloads. Microsoft set HD video back by a year, that's all they got and that's all they wanted.
I worked on the HD DVD team back then, and we manifestly wanted HD DVD to win, and we invested quite a lot in it. However, we didn't bet the Xbox 360 on it the way Sony bet the PS3 on BD (which appears to have been a good choice from the console business perspective). In the end, Sony was willing pay to whatever cost it took for BD to win.
Our interest was much more in delivering great video experiences than in which particular substrate thickness of polycarbonate imaged with a blue-violet laser won in the end.
This is a sample of what I've been working on these days:
As an Oregon resident, I'll state my preference for a higher gas tax for just these reasons.
A gas tax simply aligns with the public externalities of motor vehicles a lot better than just milage, since bigger cars cause more wear. There's no incentive for buying less damaging vehicles this way. Also, gas taxes are easy to collect, while this is more complex. Net revenue will be reduced by the cost of monitoring, plus there's the initial capital cost of getting the whole thing set up.
And while all taxes cause some distortion in the market, it's best to pick ones where the distortion is the least painful or disruptive, or otherwise aligned with society goals. Reducing petroleum imports and carbon emissions are both clear public goals. If consumption is going down, the tax is doing what it should, and so the best thing to do is to raise it to maintain the incentive to get smaller, more efficient vehicles that we saw last summer.
Since governments at all levels need funding, higher gas taxes seem like one of the best options. And a high tax sets a minimum on gas prices, and so a floor for how inefficient a vehicle people are willing to take. A $0.50 gallon tax, split evenly between states and the fed, would pay for a whole lot of economic recovery, give a stable floor to the value of alternative energy, and still be way cheaper than it was a few months ago. Right now, we're seeing state governments cutting services and payroll at the very time we need an expansionist policy nationwide to avoid deflation. The net effect is the federal government will need to borrow and spent even more money to balance out the state cuts before we can even start climbing out of the hole (if state payrolls drop by 500K, that means the fed employment target from the stimulus plan needs to be 3.5M, not 3.0M, to have the same effect).
I'd much rather see our governor recommend raising the gas tax by $0.25, drop this milage/GPS nonsense, and restore funding to education, get the new I-5 bridge started, etcetera.
Guttmann's article hasn't ever been useful for anyone as far as I can tell. Reading it again, I can't think of a single falsifiable prediction he's made that hasn't been falsified:).
I know it's been quoted and linked widely for years, but I can't see any of the predictions of problems he made pre Vista launch that haven't been disproven by reality since.
Open up a reasonable number of apps, with windows scattered around the screen. For extra credit, have them being actually animating something (video playback, whatever)
Open up Task Manager and look at the CPU utilization bars.
Turn off Aero Glass
Grab a big foreground window and shake it like crazy over your other windows.
Turn Aero Glass back on
Repeat shake
Note that without Aero Glass you get a huge CPU spike due to all the rendering that doesn't get offloaded to the CPU, while with Aero Glass you won't see a similar spike in CPU activity.
Yep. H.264 met its goals well. It's just that high performance software playback wasn't one of them:). Baseline profile isn't too bad, but Main and HIgh allow the CABAC entropy coding mode, which isn't amenable to either parallel or GPU processing (unless the video was encoded as independent slices).
I'm biased, but I like VC-1's mix of performance and quality. It's about half the complexity of H.264 (and hence about twice that of MPEG-2), but is within 15% of the bandwidth efficiency of H.264 even at very low bitrates (and converges at moderate-high bitrates).
Well, hard to speculate without you saying what your location is:).
I've worked with digital content distribution in a variety of markets.
But Hulu absolutely needs to do georestriction, since all content licensing contracts with the studios are for specific regions (USA for Hulu, of course). And don't blame this on the studios themselves; the movies are financed by partners in different region pre-paying for exclusive rights. So in many cases the USA-based studios don't have any right to distribute that content in other territories. Hulu's contracts with the studios absolutely will have specific requirements for how they handle georestrictions.
The TV networks took a good 10-15 years to work out the contracts to let them distribute current shows on the internet. In the past, content was created by indepdendent production companies, who licensed it for first broadcast to a network, who then distributed it to their corporate-owned and independently-owned affiliates, and then after a few years the original production company sold them into syndication directly. Now we see the networks funding their programming entirely or in partnership, so they're able to get the rights determined in advance. But it can be a big legal quagmire for older shows created before the current market existed.
It's frustrating how long it's taking, certainly, but these are thorny problems that require a whole lot of stuff to be renegotiated. And it's not just the video; music rights to popular music on the soundtracks can also be a huge challenge as well.
Slashdot Mirror
Windows XP = lean
Windows Vista = fat
Windows 7 = leaner than Vista = Windows XP
I must say that "bloat" is about the least information-laden phrase I hear bandied about :).
What's a consensus defintion of what it means? Wasteful use of RAM? Any additional use of RAM? Does hard drive space count? What if it's for optional non-RAM loaded stuff like templates?
Is is bloat for Vista to include a lot of printer drivers in the default image? It wasn't good for Netbooks with small SSD drives, but didn't impact system performance. And I remember lots of complaints about the full install size of Office back in the day, even though that was mainly templates that didn't need to be installed.
I think it'd be useful if we all were a little more specific about that.
Back in the SP1 days, you could run XP acceptably on 256MB of RAM, and pretty decently on 512. Today, 512 feels cramped, and is the bare minimum I'd recommend for running SP3 and all the security patches.
1GB is a more reasonable minimum if you actually want to use apps. Firefox 3 is hungry enough that it'll use up 100-300MB if you have a lot of tabs open, so you really do need at *least* 1GB to run the OS plus just a web browser, which is really pretty minimal in terms of applications.
Haven't you just described 3rd party apps getting using more RAM? Unless you have some actual data comparing working set between XP RTM and XP SP3, that seems like a more obvious difference here.
Also, given the price of RAM, isn't it appropriate for apps to be using more memory to provide better performance, roughly tracking increases in average user RAM? Sure, I could run Netscape Navigator 3 on my 20 MB Quadra 700, but on the aggragate I think I'll take the last 12+ years of browser improvements!
Even though its a different stream, the fact that the website was willing to put up support for other systems sets an excellent precedent.
Actually, both players consumed the exact same stream. The difference was in the Silverlight application that plays the stream
Stream details:
http://on10.net/blogs/benwagg/The-Obama-Inauguration-coming-to-Linux-and-PowerPC-Macs-Plus-compression-details/
Since we deliver the codec pack, that doesn't need to wait for a Moonlight implementation, so it'll hopefully be faster. It's too early to talk about specific schedules, though, as we haven't even announced a release date for the H.264 version of Silverlight.
For example, the current Moonlight 1.0 beta codec pack includes all of the SL2 codecs (specifically WMA 10 Pro which wasn't included in Silverlight 1.0).
This isn't just about a stream, but a media experience that can incorporate rich media and interactivity. For example, the pop-up stream switcher that's still available in full-screen mode.
Silverlight offers a single plugin that provides XAML + managed code + media support. It's a great fit for developing media player applications that can live in a browser that have features beyond what a classic media player can handle. And then delivering them without having the security issues of installing or running native code, or requiring a user to authenticate or install.
Silverlight 3 will be adding MPEG-4 support, this year, FWIW.
http://on10.net/blogs/benwagg/H264-and-AAC-support-coming-in-Silverlight/
While I am happy that they made an open source port so quickly however. It is just a hack for a one time occasion. For people with different needs will still be left out in the cold.
No, it was a hack because the current beta is Silverlight 1.0 compatible (no managed code, just JavaScript for control logic) while the main player was Silverlight 2, using managed code.
So what Novell did in an inspirational rush was reimplement a version of the player using JavaScript.
Now that this model works, it'll be availble for other events. Of course, it won't be needed once Moonlight gets to Silverlight 2 parity.
Details from my blog:
http://on10.net/blogs/benwagg/The-Obama-Inauguration-coming-to-Linux-and-PowerPC-Macs-Plus-compression-details/
Good news everyone. After burning some midnight oil, we (Microsoft) and Novell have pulled together a Silverlight 1.0 compatible verison of the inauguration player that's compatible with both Moonlight Beta 1 and Silvelight 1.0 on PowerPC Macs.
Miguel's Blog: http://tirania.org/blog/archive/2009/Jan-20.html
My Blog: http://on10.net/blogs/benwagg/The-Obama-Inauguration-coming-to-Linux-and-PowerPC-Macs-Plus-compression-details/
The updated player should be up around 6 am EST if you want to validate your installs.
http://pic2009.org/
http://on10.net/blogs/benwagg/First-MediaStreamSource-example-is-up/
Well, not all from Blu-ray, though :).
It's a bad article. There's nothing intrinsically infringing about MKV + H.264. And there's defintely nothing Blu-ray specific about this, unless they're saying that it can handle a MKV that encapsulates the unrecompressed original Blu-ray bitstreams. Is anyone actually using THAT for piracy? I thought most download HD was still 720p.
QuickTime has supported streaming H.264 since well before Flash, and it's coming in the next version of Silverlight as well.
http://on10.net/blogs/benwagg/H264-and-AAC-support-coming-in-Silverlight/
DivX has a very different useage model than Flash (downloads with a player app instead of streaming to a browser plugin). They're hardly competing head to head.
As for container support, Flash is just MPEG-4. The Adobe model is to use "dumb" media formats and control high-level display functions via the Flash application itself.
That said, a MKV parser could be written for Silverlight via managed code, including caption overlay support, multilanguage audio, etcetera. The Silverlight MediaStreamSource API allows for parsers and protocols to be written in managed code, while still using the native decoders built into Silverlight.
http://on10.net/blogs/benwagg/First-MediaStreamSource-example-is-up/
That'd be a cool project to see!
I asked the digital media guys if there were any new features for encoding in Win 7, and they said that it supports up to 8-way threading for WMV encoding! The limit in Vista and XP was 4-way threading.
This requires either 8 cores or 4 hyperthreaded cores, and the resolution of the encoded file be at least 480 pixels tall.
I bet this will get used and abused by benchmarkers. Hopefully they'll use an easy-to-decode source file so that the encoder doesn't get bound by source decode speed.
Nerdy details at my blog:
http://on10.net/blogs/benwagg/8-way-multithreading-in-Windows-7/
If anyone's curious about what's going on under the hood, I did a blog post a bit ago about how Netflix does their VC-1 encoding for the service and how it's evolved over time.
http://on10.net/blogs/benwagg/Netflixrsquos-Neil-Hunt-shares-encoding-workflow-info/
For those talking about quality, a few points I want to make:
It won't ever look better than the source; some stuff is only availble in lower quality SD masters. Stuff like "The Office" that comes in as HD looks spectacular at the highest bitrate.
HD (up to 3800 Kbps) quality today is only on devices, while the desktop version only goes up to SD (1500 Kbps).
Those bitrates are averages, with the peaks higher. You'll generally need to be able to sustain 1.5x the listed bitrate in order to reliably get the maximum bitrate band for that content.
Today, the desktop version is SD only
re's lots of ways to handle this other than full virtualization that offer a better user experience.
Great, then bring them up, because what MS has been doing isn't it.
Perhaps you haven't been reading Chris Jackson's blog discussing the myriad ways that Windows balanaces compatibility of older applications with allowing improvements to OS:
http://blogs.msdn.com/cjacks/default.aspx
And here's all the stuff about the shims that let old apps that want to do stuff like write to system.ini still run without actually writing anything to system.ini:
http://blogs.msdn.com/cjacks/archive/tags/Shims/default.aspx
It's hard to imagine any rating system that's more screwed up than the system used in most companies to rate the output of their programmers. If they were consciously trying to sabotage the software, they probably wouldn't come up with any schemes nearly as effective as judging programmers by lines of code produced.
Fully agreed. The other post's anecdote about having trouble getting a review by makign code smaller is quite telling.
Anyway, that kind of measurement certainly isn't done here at Microsoft (although plenty of productive programmers write a lot of lines of code). By example, great stuff like the speed/quality improvements in video playback and scaling in Silverlight 2 didn't really add much or any lines of code; it was about superior design. The guy who wrote that stuff may not rate that highly in KLOC/week, but he's certainly recognized as a rock star.
http://on10.net/blogs/benwagg/Silverlight-2-RC0-is-out/
http://on10.net/blogs/benwagg/Demo-of-Silverlight-2-scaling-quality-improvements/
it is long past the time that MS should be leaving bad code in new OSes just to claim 'Backward Compatibility' when it is totally unnecessary.
It seems like seamless backwards compatibility is a feature that doesn't require a detailed defense :).
Bear in mind that being able to run ancient, crufty, no-source-available line-of-business apps is pretty darn important for many Windows customers.
There's lots of ways to handle this other than full virtualization that offer a better user experience.
Here are some benchmarks right over at tomshardware that show that the "SlashDot world" in this case is accurate (amazing!).
From January 2007.
Hence very early drivers and pre SP1.
Not relevant to a comparison today.
Wow, lots of snark, but I don't see any posts to the actual answer yet...
Embarassing, but not catastrophic. A bug relate to the last day of a leap year, as was speculated above. Will resolve itself automatically on Jan 1st.
ahref=http://forums.zune.net/412486/ShowPost.aspxrel=url2html-1800http://forums.zune.net/412486/ShowPost.aspx>
Theres a bunch of different streams it switches between dynamically. Which one you get can be limited by:
Available bandwidth (we'll send as many bits as won't cause buffering)
Available CPU speed (we don't want you to drop a bunch of frames)
Resolution of the media player (no need to send 1280x720 when the video window is scaled down to 320x240)
All of the above can be switched every couple of seconds, so if you have a heavy CPU process in the background, the data rate will drop during it, and then go back up again after.
To make sure you're not blocked by frame size, you can alt-Enter in IE to take the browser full screen.
Also, to see your current stream, mouse over on the little horizontal bars in the lower right corner.
Lastly, bear in mind this is a pre-beta here. We'll have big improvements by the time it goes 1.0 next year.
Yes, all targeted tax collection is simply forced "wealth transfer" hidden behind yet another name.
Is there a non-targeted form of tax collection you're offering as an alternative :).
Taxation is necessary to pay for public good and services that would otherwise be unfunded. Taxation is also appropriate to fund activities that are more efficiently done by government than without it.
Given that, there's a couple of different axes to figure out: what's the minimum level of taxation to provide a balanced budget over the business cycle with appropriate spending (governments should run a surplus in good times and a deficit in bad times), and what's the right combination of taxes that provides that revenue with a minimal amount of friction to economic growth and other goals. That's why cigarette and gas taxes are a good thing, as they provide revenue while discouraging behavior we want to discourage. If you need to get revenue from somewhere, that's a whole lot better than many alternatives.
An example of a really bad tax would be a gross receipts tax, since it's painful to administer, and hugely distorting as it massively rewards vertical integration, driving out smaller innovative companies.
http://en.wikipedia.org/wiki/Gross_receipts_tax
Inheritance taxes are another good one. Lots of people seem to hate them for some reason, but it seems like the dead will find being taxes more than most, and enabling the decedents of wealth to remain wealthy without any economic input of their own for generations seems the wrong incentive for future wealth generation.
Anyway, there's often a knee-jerk attitude that all taxes are equally bad, and no tax should ever be added or raised, even to offset lowering other taxes. But we need taxes, and there's differences between them, so we should develop as optimal a tax system as we can.
Microsoft developed the codecs used by *both* HD-DVD and Blu-Ray. Microsoft had no significant interest regarding which format took hold --- in either scenario, the players would be running Microsoft software. Microsoft's only interest was for the format war end quickly.
Oh, the codec side of things was relatively minor. We developed VC-1 and related tools, but have a patent position and a lot of involvement in H.264 as well, and that side of things always supported VC-1 on BD as well.
The bigger effort and team was focused on building the interactive players for the Toshiba and Xbox 360 players. The whole HDi layer was jointly developed by Microsoft and Disney.
As a XML markup + scripting code-behind, it was a lot like a subset of Silverlight, actually.
I don't think Microsoft wanted either format to gain critical mass - wide and early adoption is a threat to Microsoft's goal of 'services', including pay per view and digital downloads. Microsoft set HD video back by a year, that's all they got and that's all they wanted.
I worked on the HD DVD team back then, and we manifestly wanted HD DVD to win, and we invested quite a lot in it. However, we didn't bet the Xbox 360 on it the way Sony bet the PS3 on BD (which appears to have been a good choice from the console business perspective). In the end, Sony was willing pay to whatever cost it took for BD to win.
Our interest was much more in delivering great video experiences than in which particular substrate thickness of polycarbonate imaged with a blue-violet laser won in the end.
This is a sample of what I've been working on these days:
http://smoothhd.com/
Still pre-beta, but I don't think that optical media will be the hard or the interesting part of HD video delivery much longer.
As an Oregon resident, I'll state my preference for a higher gas tax for just these reasons.
A gas tax simply aligns with the public externalities of motor vehicles a lot better than just milage, since bigger cars cause more wear. There's no incentive for buying less damaging vehicles this way. Also, gas taxes are easy to collect, while this is more complex. Net revenue will be reduced by the cost of monitoring, plus there's the initial capital cost of getting the whole thing set up.
And while all taxes cause some distortion in the market, it's best to pick ones where the distortion is the least painful or disruptive, or otherwise aligned with society goals. Reducing petroleum imports and carbon emissions are both clear public goals. If consumption is going down, the tax is doing what it should, and so the best thing to do is to raise it to maintain the incentive to get smaller, more efficient vehicles that we saw last summer.
Since governments at all levels need funding, higher gas taxes seem like one of the best options. And a high tax sets a minimum on gas prices, and so a floor for how inefficient a vehicle people are willing to take. A $0.50 gallon tax, split evenly between states and the fed, would pay for a whole lot of economic recovery, give a stable floor to the value of alternative energy, and still be way cheaper than it was a few months ago. Right now, we're seeing state governments cutting services and payroll at the very time we need an expansionist policy nationwide to avoid deflation. The net effect is the federal government will need to borrow and spent even more money to balance out the state cuts before we can even start climbing out of the hole (if state payrolls drop by 500K, that means the fed employment target from the stimulus plan needs to be 3.5M, not 3.0M, to have the same effect).
I'd much rather see our governor recommend raising the gas tax by $0.25, drop this milage/GPS nonsense, and restore funding to education, get the new I-5 bridge started, etcetera.
Guttmann's article hasn't ever been useful for anyone as far as I can tell. Reading it again, I can't think of a single falsifiable prediction he's made that hasn't been falsified :).
I know it's been quoted and linked widely for years, but I can't see any of the predictions of problems he made pre Vista launch that haven't been disproven by reality since.
Get Vista running on a DX9 capable machine.
Open up a reasonable number of apps, with windows scattered around the screen. For extra credit, have them being actually animating something (video playback, whatever)
Open up Task Manager and look at the CPU utilization bars.
Turn off Aero Glass
Grab a big foreground window and shake it like crazy over your other windows.
Turn Aero Glass back on
Repeat shake
Note that without Aero Glass you get a huge CPU spike due to all the rendering that doesn't get offloaded to the CPU, while with Aero Glass you won't see a similar spike in CPU activity.
Yep. H.264 met its goals well. It's just that high performance software playback wasn't one of them :). Baseline profile isn't too bad, but Main and HIgh allow the CABAC entropy coding mode, which isn't amenable to either parallel or GPU processing (unless the video was encoded as independent slices).
I'm biased, but I like VC-1's mix of performance and quality. It's about half the complexity of H.264 (and hence about twice that of MPEG-2), but is within 15% of the bandwidth efficiency of H.264 even at very low bitrates (and converges at moderate-high bitrates).
Well, hard to speculate without you saying what your location is :).
I've worked with digital content distribution in a variety of markets.
But Hulu absolutely needs to do georestriction, since all content licensing contracts with the studios are for specific regions (USA for Hulu, of course). And don't blame this on the studios themselves; the movies are financed by partners in different region pre-paying for exclusive rights. So in many cases the USA-based studios don't have any right to distribute that content in other territories. Hulu's contracts with the studios absolutely will have specific requirements for how they handle georestrictions.
The TV networks took a good 10-15 years to work out the contracts to let them distribute current shows on the internet. In the past, content was created by indepdendent production companies, who licensed it for first broadcast to a network, who then distributed it to their corporate-owned and independently-owned affiliates, and then after a few years the original production company sold them into syndication directly. Now we see the networks funding their programming entirely or in partnership, so they're able to get the rights determined in advance. But it can be a big legal quagmire for older shows created before the current market existed.
It's frustrating how long it's taking, certainly, but these are thorny problems that require a whole lot of stuff to be renegotiated. And it's not just the video; music rights to popular music on the soundtracks can also be a huge challenge as well.