No. The codec in Flash is Spark, which is derived from H.263. It's very different from Sorenson Video 3, and their decoders are radically different.
Among other things, Spark was designed for the decoder to be fast (even on StrongARM and other non-desktop processors), small, and portable. Sorenson Video 3 was designed for high compression efficiency. Sorenson Video 3 looks a lot better than Spark at moderate data rates because it didn't have to make the same tradeoffs. However, Flash MX will play in all kinds of places that QuickTime won't.
If you're getting bad color, you're probably using Sorenson Video 2 (just called "Sorenson Video" in the QuickTime UI). The Basic, free version of Sorenson Video 3 (called "Sorenson Video 3") built into QuickTime since 5.0.2 doesn't have this problem, and is much, much higher quality and much, much faster than the old free codec.
For those curious about the details of the technologies in question, here goes. FWIW, I was a beta tester for both codecs, have taught classes with them, and cover them both extensively in my forthcoming book.
Sorenson currently sells two different codecs, Sorenson Video 3.1 Pro, and Spark Pro, both bundled with versions of their Squeeze encoding tool.
Sorenson Video 3.1 Pro is an advanced version of an encoder/decoder built into QuickTime. It's an excellent codec, with good compression efficiency, a B-frame mode that dramatically improves QuickTime streaming, and many other groovy features. All versions of Sorenson Video are QuickTime only.
Sorenson has also had a MPEG-4 codec in beta for forever (I did the first public demo of it back at QuickTime Live 2000). MPEG-4 is a superset of "baseline" H.263 (an older standard codec, designed for video conferencing), and any MPEG-4 decoder is required to also play back baseline H.263. Sorenson's MPEG-4 encoder includes a baseline H.263 encoder as well, so you can use the codec to make files compatible with H.263 decoders as well (like the Java Media Framework).
The Spark codec, which Sorenson licensed to Macromedia, and Spark Pro, the advanced encoder version included in Sorenson's Squeeze for Flash MX encoding tool, are derived from H.263, based on Sorenson's work with the MPEG-4 codec. Spark Pro is enormously better than the plain Spark incoder built into Flash - that one doesn't even let you specify a data rate.
I haven't read Apple's complaint, but I'd guess that they're alleging that parts of Sorenson Video were used to develop the Sorenson MPEG-4 codec, and which in turn wound up in Spark, which was licensed to Macromedia. I have no idea if this actually happened, or whether or not it would be permitted under their contract if it did.
Both codecs do have a number of features in common, like a configurable threshold for automatic keyframe insertion, an optional image smoothing (deblocking) filter on decode, and 2-pass VBR encoding.
Anyway, knowing as much as I do about these codecs, I feel completely unqualified to have an opinion on the legal merits of this case.
Actually, both MPEG-1 and Sorenson Video 3 use Y'CrCb (aka YUV) 4:2:0 color, where there is one color sample for each 2x2 block of pixels.
The older Sorenson Video 1 & 2 used YUV-9, which has one color sample for each 4x4 block of pixels. This isn't nearly enough, and caused quality problems.
If you're implying that we only use some percentage of our brain, than you're repeating one of the most long-lived and completely erroneous memes of popular neuroscience.
The regions of our brain are rather specialized. So while each part gets used some of the time, we don't use all of it all the time. About the only time where all the brain is active at once is in a seizure, which certainly doesn't help chess playing at all.
Given the massive evolutionary sacrifices required for our big brains (painful, dangerous labor and extremely dependent infants compared to other animals), there was clearly a correspondingly strong evolutionary pressure for big brains. If it was possible to have done it with only 10% of the volume, we'd either have much smaller heads, or be a heck of a lot smarter.
You can also get smaller file sizes with GIF and PNG for many kinds of content, even though they're lossless.
I've had full 800x600 MacOS 9 screen shots compress down to 20K, losslessly, with PNG. JPEG even at Q=0 couldn't make a file that small, and it still looked gawdawful.
While the distinction is often described as natural images versus synthetic images, 3D rendered stuff should be treated as natural images.
The real issue is whether there are large areas of EXACTLY the same color, and sharp edges. In a typical screen shot, you might have several 100x100 blocks of exactly the same color. PNG and GIF do a wonderful job with those. They also do a great job with sharp edges, while the Discreet Cosine Transformation of JPEG causes a lot of artifacts with those, or requires a lot of bits to encode them accurately. So something simple as a black line on a white background is quite mathematically complex for JPEG to render.
Overall, PNG will give slightly smaller files, and more importantly provides for more than 256 colors. I look forward to when it is ubiquitously supported by all browsers.
that since they're actually water, they can't hurt you!
This is a big advantage over untested herbal supplements which may or may not do anything, and may or may not interact with other drugs.
Conspiracy theory for the day: homeopathy is A Cunning Plan to keep the softheaded away from untested NON-placebo stuff.
Re:The problem with all these equations...
on
Rare Earth
·
· Score: 2
Of course, the big problem with the "Earth's life was seeded from Beyond the Stars!" is that it begs the question of how the life that seeded us evolved. It'd be interesting to know, but wouldn't answer any fundamental questions.
Of course, the same could be said about Intelligent Design arguments. "So, where did God come from, and why did he want to create life? Would other possible gods have behaved differently? And why are you building that man-sized bonfire, and walking towards me with that rope?"
1. Not a good metaphor, to the point I don't understand what you're trying to say. Would having more different accounting systems be a good thing, instead of one, clear, internationally recognized one?
2. Economists spend a lot of time trying to recognize and quantify this kind of externality. The whole "emmissions trading" concept is a way to address and limit how economic activity can degrade the environment. Tarrifs aren't a good solution, since they don't impose costs for goods not traded internationally. I'd like to see these costs levied at production, and then having those goods freely traded after that. The tarrif system would just give an economic incentive to pollute one's own country. And I'm not sure what the alternative is here. It's not like non-capitalist countries do better, or even as well as capitalist nations at enviornmental protection (although this is arguably more of an effect of free speech). Russia has much, much worse environmental damage than the US.
3. I don't think anyone is proposing it as a single international government. It does involve a number of international bodies, and greater international integration.
4. I don't know what the heck you mean here. Even if you argue the potential size of the economy is finite (hard to even imagine what that would mean - a hard limit to innovation?), with free trade, we could get closer to the finite potential than without it.
5. The UN has been pretty successful in reducing international conflict (although with much less success at civil wars). It hasn't lived up to its lofty goals, but I think we've clearly been better off with it than without it.
As for Keynes, his analysis of demand in the economy has arguably prevented a return of real depressions (but not recessions), but giving powerful tools to help reduce the business cycle. Again, it isn't perfect, but clearly a lot better than what we had from 1929 on.
Of course, as the cost of transport goes down, the importance of where labor is matters less and less. So free trade gets us most of what free movement would (bringing jobs to those who can do them most productively).
While I'm all for immigration, making it absolutely free would cause a lot more social disruption than the gains it would offer.
From the US perspective, we're arguably close to the right balance. The illegal immigrants we're getting are the kind of folks who could accumulate $5K in CHINA, and are willing to risk life and limb, and say goodbye to their former life in order to have a shot at making a life in America. I'm inclined to think anyone who was able to sneak in is someone we're likely going to want to have as part of our society. They might have to drive a cab, but their kids are going to be doing incredible things.
Actually, George Soros has been a huge philanthropist, and spends a lot more of his time spending money than making it now. As of 1999, he'd already given away $1B of his own money, and is still at it. He also lost a huge amount of money in Russia in the '90s, trying to help build an open society there.
By all indications, he's aiming to die broke, but is busy trying to figure out the best bang for his (enormous) buck along the way.
Generosity isn't really the right metaphor. The current term of art is social investment, and that makes more sense to me. The idea is building the institutions and structures to help build sustaninable, healthy societies. This runs the gamut from AIDS prevention to environmental preservation to microloan programs to clean government iniatives.
Food aid, while crucially important (humans don't deserve mass die-offs, no matter how lousy their governments), doesn't really solve the underlying problems of these countries. The underyling problem is why there isn't food in the first place.
Important hint: democracies don't have famines. Malnutrition, yes, but not large-scale famines.
You get rich in two ways - by having more money, and by having money buy more. I'd rather make $50K a year in a world with a wide variety of imports than $100K if everything I used had to be produced in the US. I'm way richer this way, and so is everyone else on average.
The irony is that corporations don't actually improve their bottom line much by exporting jobs - they merely preserve their bottom line, since dropping prices erode much of the theoretical benefit. Overall corporate profits aren't substantailly different decade-by-decade than they were 20 or 50 years ago.
The big winners of real globalization are individuals, because it enables them to do the work they are most productive at, and reduces the costs of all goods, since everyone is able to produce good more cheaply.
Indeed. While dumping can happen, it's far from clear it was happening either with softwoods or steel. In the case of the steel industry, it was especially dubious, since minimills are doing just great. It's the old-line integrated steel producers that are having trouble. But this is largely because they signed terribly expensive deals with their unionized workers, including massive unfunded pension and health care deals for retired workers. This is one of the reasons why 401(k) deals are so good - you aren't screwed if your employer goes out of business!
The big problem is, of course, that the highly paid steelworkers aren't able to add as much value per $ of salary as global competitors (or even local competitors from nonunionized mini mills). Thus, those jobs are inevitably going to be lost in a free economy.
And as I said earlier, saving these jobs means we're going to lose three times as many jobs in other parts of the economy that rely on cheap steel. And we're going to be paying a heck of a lot more for cars, appliences, etcetera. In essence, tarrifs are taxing ourselves, but with a very, very stupid tax.
I've always been unclear why those who most advocate a wealth transfer to the third world in the form of massive forign aid are so unwilling to lose a few US jobs in exchange for more jobs in countries that need them a lot more than we do. The third world needs jobs AND aid.
The US economy should be based on the things we do best. Free trade is a great way to find out what you're good at with great accuracy and low latency. Nothing's worse for an economy than a government industrial policy.
Yes. One of the charming ironies of economics is that it takes a lot of government intervention to keep a free economy. If monopolies were legal, it'd be every corporation's goal to become a monopoly.
Trade barriers don't really help this problem - if anything they make local monopolies much more common. Check out all the national champion airlines and utilities. It's really hard (but not impossible, viz Microsoft) to maintain a global monopoloy, just because there are so many different places to have to exert pressure, and so many different markets that a disruptive competitor could find a niche into.
The real solution is powerful, independent, and apolitical antitrust agencies, on both the national and international levels.
Absolutely. Many biggest problems that most underdeveloped countries are due to insufficient globalism, not to globalism itself.
Corruption: One of the big factors of success in Western economies is transparency and lack of bribary. While we're far from 100% free of these, we're doing a lot better than most of the world. Look at the Enron scandal - we agree that it's inapprpriate for an auditor to go easy on the audits in order to gain more consulting work. The richest heads of state in the world are in third world countries. While rich people often hold elected office in the US, they're typically poorer instead of richer after it's all done (elections are expensive!).
Opaque and bribary-ridden societies mean inefficient uses of resources. If funcionaries see a primary goal of their employment being to maximize graft, the net effect is that anything that involves government approval will take as much time and require as many participants as possible. It also increases the power of the well connected over the competant, so companies who tend to get government contracts tend to be lousy and them.
Clear title to assets: A lot of the land in poor countries has unclear title, which means they can't be used as assets for loans, sold, or otherwise be treated as a form of wealth. The reasons for these are often due to a combination of a corrupt government (too many bribes to get the paperwork), and well intentioned but misguided attempts to enforce a more communal rural economy.
Free trade: This is a problem in rich and poor countries, although rich countries tend to at lesat give lip service to free trade. If everyone is free to trade goods across national borders, the net effect is that everyone can do what they're most efficient at. Protectionism enables local producers to be less efficient, with the inevitable effect of empoverishing the many for the benefit of the few. For example, in the new US steel tariffs are estimated to cost at least three jobs in steel consuming and shipping industries for every steelworker job they save. Not only is it a bad thing for our trading partners, it's bad for ourselves. Everyone loses in protectionism.
Among other things, most JPEG files use Y'CrCb 4:2:0 color space, which only includes one each chroma sample per 2x2 blocks of pixels. Thus there are only 12 bits to encode each pixel uncompressed.
However, PNG does offer the ability to encode in index colors, like 8-bit mode, and it is enormously better at encoding synthetic graphics, like screen shots. Under MacOS 9 (which has a very simple color scheme) I've had 1024x768 screen shots be only 20K. No JPEG can't come close for that kind of content.
Re:JPEG does have a lossless mode
on
JPEG2000 Coming Soon
·
· Score: 2, Informative
MWright had a good technical introduction, so I'll just outline a few of the practical areas where wavelets make JPEG2000 rock.
First, compression efficiency. Although lossless and near-lossless quality isn't hugely better, data rates for "good enough" quality (defined as where the image is understandable and artifacts aren't too distracting) are much, much lower. Unlike the old Discreet Cosine Transformation (DCT) method of JPEGs, which get blocky at high compression, wavelets get softer, which is much less obvious. So, this might not help things with pro digital cameras much (which were lightly compressed in the first place), it will help a lot with web images and such, or consumer digital cameras.
The other nice thing about wavelets is that they are constructed in bands. First, the base image is encoded at a low resolution. Then this is used as prediction for the next resolution, and that band only has to store how the image is different from the prediction.
This is groovy, because you can decode the individual bands as they're transmitted, giving a low-resolution proxy image once only a few percent of the file is transmitted, getting progressively higher quality over time. While progressive/interlaced JPEG/GIF gets the same effect, wavelets do it more efficiently.
Many years ago, Intel created a video file format that used these properties, IVF. Didn't ever get any market traction. You can still get the tools from Ligos.
The standards all go well beyond just images in web pages. I extensively use PNG today as a video codec instead QuickTime for doing lossless compression of RGB images.
And JPEG is of course used in all kinds of places, like the Motion-JPEG standard used for the majority of Pro video editing systems.
While they certainly offer a lot for the web, PNG and JP2 will be important in a lot of other vertical markets. That's one of the wonderful things about standards, is they can be used in so many ways in industries where having to develop technology from scratch just doesn't make sense.
I expect we'll see a similar effect with MPEG-4, as the basis for all kinds of vertical technologies that end users might never see.
Actually, the 480x320 might have been where they went wrong with the Newton. The older versions were powerful enough, but too big. Instead of using new technologies to make the Newton more powerful yet, they should have made it smaller.
Instead of doubling the screen resolution by increasing pixel density, they should have kept the resolution 320x240, and cut the size of the unit in half. Instead of going to the full 162 MHz processor, they should have used a slower, more power efficient one, so fewer batteries were needed.
It still would have been somewhat bigger than a Palm, but with all that Newton software to use...
Something to put up on the virtual "what could have been" shelf with my dual processor PowerPC NeXT system...
Indeed. We sacrifice our long-term political interests for silly short-term political issues in swing states all the time.
Another example: high steel tarrifs don't get us anything we actually want - we're going to lose four jobs in steel-consuming industries for every one we save. AND we're PISSING OFF our allies at the time we need the most. Why? Because steel workers have a good lobby, but people who buy cars and work in dockyards don't. And because Detroit will vote Democratic whatever happens, but Pennsylvania could go either way.
And heck, why are family farms so sacrosanct? It's not like anyone in government is trying to save family restaurants, and we actually need those in our cities - most people couldn't care if they're eating produce from Florida or Chile.
Of course, I don't think we're worse at this kind of stuff than most other countries. It's just that with our greater power, our mistakes have greater consequences. We need to hold ourselves to a higher standard than we hold others.
I was pleased to see Bush announced a 50% increase in our foreign aid budget, targeting countries with good government. This is an excellent start.
Your point is excellent - inheritance taxes are a good idea.
Assuming the government is going to spend and hence raise a fixed amount of money, I'd rather they take it from dead people than living people. And as a prep-school alumnus, I can say from personal experience that the expectation that one will eventually become rich without any effort on one's own part is corrosive. Some folks adopt noblesse oblige and do just fine, but far too many tend to waste their lives waiting for their money to come, instead of finding something meaningful to do.
And, of course, rich people hate paying taxes, so they give their money away to foundations and such if the think it was going to be taxes. There would be less of this as inheritance taxes go up.
And yes, I'm speaking as someone who is likely to inherit a meaningful amount of money. I'd rather pay lower taxes now, and have less in the future. And heck, I'd rather my wealthy relatives live forever - they're a lot more fun than money (money can't tell stories about being blown out of a jeep by a panzerfaust).
Re:Next: Is Globalism good or evil ?
on
Globalism Post 9/11
·
· Score: 3, Interesting
That's like asking if weather is Good or Evil. To the extent we're going to have a multinational economy (and who doesn't want to have fresh strawberries in December, and laptops made in Taiwan), globalism, as strictly defined, is the way this all works.
Globalism gives us many wonderful things we all want. The question isn't Globalism: yes or no. The question is how best to encourage its benefits and deal with its disadvantages. No one has proposed any alternative to Globalism that isn't much worse than things are today, let alone how good they could be.
One of the charming ironies of global capitalism is that it achieves the "from each according to their abilities" part of the socialist creed far better than pure socialism ever did. Of course, a pure capitalism doesn't address the "to each according to their needs" part at all, which is why there aren't any pure capitalist societies in the world, and why we need to strengthen and improve the World Bank and IMF, not eliminate them.
What we want is a global economy where people can compete ferociously on production, but where there is properly integrated environmental protection, and a safety net where none starve, and all can get a good education irrespective of the economic success or failure of their parents.
Global capitalism means we can all get goods and services as the lowest cost available for the quality we need. Remember, the two ways you get richer are through wages going up and prices going down. Those who want high steel tariffs want to enrich the few at the expense of the many.
For those curious about the details, check out David Ricardo.
Engagement is messy - the Spiderman paradox
on
Globalism Post 9/11
·
· Score: 2, Insightful
Folks,
I find it remarkable how contradictory the two streams of advice about how the US should change its behavior are. One one hand, there are those who say we need to be less involved in the rest of the world - no global policing, no liberation of Kuwait, leave Iraq alone, leave Bosnia alone, no more support for Israel etc. Others say we need to be more involved - impose a Palestinian/Israeli settlement, neutralize Iraq, increase and improve foreign aid, make freeer trade, etc.
Neither side is essentially wrong - what they all want is for us to do the "right" things, and not do the "wrong" things. Of course, which is which is far from obvious.
As in all things, Stan Lee said it best: "With great power comes great responsibility." Spiderman's greatest failure wasn't what he did, but what he didn't do, allowing a criminal to go free and kill his Uncle Ben. As semi-facetious as the argument may sound, the US has been bit by that radioactive spider, and we need to accept the fact of our power, and figure out how best to use it.
Due to a combination of excellent geography, a successful culture built by waves of immigration, and dumb luck, the US has the ability to Make Things Different in the world. We can make changes. But while we're powerful, we aren't prescient, and can't predict the long-term ramifications of what we do with any accuracy.
But choosing not to act is still an action - was not intervening in Rawanda any better than invading Iraq would be? Our power means we have to make these decisions, and sweat them, and argue about them endlessly on talk shows, and then, gut clinched, try to do the right thing and hope it all works out.
Many of the examples of past US evil are from the cold war. I think we can agree that a good number of the things we did were not only not effective for our goals, but hurt them in the long term. But do you think that, with what they knew at the time, the people who made those decisions knew which were which? Does this mean we should have let the Soviet Union have a free hand all over the world? No.
The US needs to be engaged with the world, and we have to know it's going to be a messy business that will make us enemies. So we need to do it in as smart a way as possible, with a long view, with clear-eyed compassion, and with as little attention to our trivial domestic politics as possible.
For a recent example of us NOT doing this right, we failed to drop textile tarrifs for Pakistan. Pakistan's leadership, whatever their faults, have been a surprisingly good partner in our current conflict. And they're dirt poor, which helps terrorists, and hurts the development of a healthy society. Pakistan makes a lot of textiles, and could important a lot more to the US, helping develop a better, non-aid-based economy. But, in order to not risk the Carolinas in the mid-term elections later this year, and his reelection in 2004, Bush refused to lower tarrifs or increase Pakistan's quota for textiles. I know it sounds like a small deal, but getting this kind of stuff right could help enormously.
Anyway, there is nothing we can do that will keep people from hating us. We have the power to pick winners and losers, and so we'll be resented for what we don't do just as much as for what we do. So we just have to do the right thing as best we can, and give the world as few VALID reasons to hate us as possible.
QuickTime provides is own "keyframe every" option. In VP3's codec-specific dialog, it provides its own, more advanced implementation (which lets you set a minimum number of frames between keyframes, not just a maximum). Thus, most folks turn off QuickTime's keyframe insertion and let VP3's do it. I believe this is what the tester did, based on the description.
Leaving the keyframe interval blank may, depending on the tool, force EVERY frame to be a keyframe. Definitely not what you want.
The "Allow Dropped Frames" command actually controls the interplay of data rate versus frame rate, not image quality. When dropped frames are allowed, the codec will reduce frame rate in order to hit the target data rate.
One objection you didn't mention is the use of the "Fast Compression" mode. The tester may have gotten better quality (and much slower encoding) if that had been on.
This brings up an interesting question for the free/open source community. Under MPEG-4, it's perfectly acceptable to make an encoder or decoder than is open source. However, the patent license is $0.25 each per encoder and decoder, with a cap of $1M/each per company per year.
So, would folks be willing to cough up a quarter for this codec, if it remained open source?
Is there time for a new license: free speech, but not free beer?
Slashdot Mirror
No. The codec in Flash is Spark, which is derived from H.263. It's very different from Sorenson Video 3, and their decoders are radically different.
Among other things, Spark was designed for the decoder to be fast (even on StrongARM and other non-desktop processors), small, and portable. Sorenson Video 3 was designed for high compression efficiency. Sorenson Video 3 looks a lot better than Spark at moderate data rates because it didn't have to make the same tradeoffs. However, Flash MX will play in all kinds of places that QuickTime won't.
If you're getting bad color, you're probably using Sorenson Video 2 (just called "Sorenson Video" in the QuickTime UI). The Basic, free version of Sorenson Video 3 (called "Sorenson Video 3") built into QuickTime since 5.0.2 doesn't have this problem, and is much, much higher quality and much, much faster than the old free codec.
Folks,
For those curious about the details of the technologies in question, here goes. FWIW, I was a beta tester for both codecs, have taught classes with them, and cover them both extensively in my forthcoming book.
Sorenson currently sells two different codecs, Sorenson Video 3.1 Pro, and Spark Pro, both bundled with versions of their Squeeze encoding tool.
Sorenson Video 3.1 Pro is an advanced version of an encoder/decoder built into QuickTime. It's an excellent codec, with good compression efficiency, a B-frame mode that dramatically improves QuickTime streaming, and many other groovy features. All versions of Sorenson Video are QuickTime only.
Sorenson has also had a MPEG-4 codec in beta for forever (I did the first public demo of it back at QuickTime Live 2000). MPEG-4 is a superset of "baseline" H.263 (an older standard codec, designed for video conferencing), and any MPEG-4 decoder is required to also play back baseline H.263. Sorenson's MPEG-4 encoder includes a baseline H.263 encoder as well, so you can use the codec to make files compatible with H.263 decoders as well (like the Java Media Framework).
The Spark codec, which Sorenson licensed to Macromedia, and Spark Pro, the advanced encoder version included in Sorenson's Squeeze for Flash MX encoding tool, are derived from H.263, based on Sorenson's work with the MPEG-4 codec. Spark Pro is enormously better than the plain Spark incoder built into Flash - that one doesn't even let you specify a data rate.
I haven't read Apple's complaint, but I'd guess that they're alleging that parts of Sorenson Video were used to develop the Sorenson MPEG-4 codec, and which in turn wound up in Spark, which was licensed to Macromedia. I have no idea if this actually happened, or whether or not it would be permitted under their contract if it did.
Both codecs do have a number of features in common, like a configurable threshold for automatic keyframe insertion, an optional image smoothing (deblocking) filter on decode, and 2-pass VBR encoding.
Anyway, knowing as much as I do about these codecs, I feel completely unqualified to have an opinion on the legal merits of this case.
Hope this helped clarify things slightly.
Actually, both MPEG-1 and Sorenson Video 3 use Y'CrCb (aka YUV) 4:2:0 color, where there is one color sample for each 2x2 block of pixels.
The older Sorenson Video 1 & 2 used YUV-9, which has one color sample for each 4x4 block of pixels. This isn't nearly enough, and caused quality problems.
If you're implying that we only use some percentage of our brain, than you're repeating one of the most long-lived and completely erroneous memes of popular neuroscience.
The regions of our brain are rather specialized. So while each part gets used some of the time, we don't use all of it all the time. About the only time where all the brain is active at once is in a seizure, which certainly doesn't help chess playing at all.
Given the massive evolutionary sacrifices required for our big brains (painful, dangerous labor and extremely dependent infants compared to other animals), there was clearly a correspondingly strong evolutionary pressure for big brains. If it was possible to have done it with only 10% of the volume, we'd either have much smaller heads, or be a heck of a lot smarter.
The Snopes page is quite informative:
You can also get smaller file sizes with GIF and PNG for many kinds of content, even though they're lossless.
I've had full 800x600 MacOS 9 screen shots compress down to 20K, losslessly, with PNG. JPEG even at Q=0 couldn't make a file that small, and it still looked gawdawful.
While the distinction is often described as natural images versus synthetic images, 3D rendered stuff should be treated as natural images.
The real issue is whether there are large areas of EXACTLY the same color, and sharp edges. In a typical screen shot, you might have several 100x100 blocks of exactly the same color. PNG and GIF do a wonderful job with those. They also do a great job with sharp edges, while the Discreet Cosine Transformation of JPEG causes a lot of artifacts with those, or requires a lot of bits to encode them accurately. So something simple as a black line on a white background is quite mathematically complex for JPEG to render.
Overall, PNG will give slightly smaller files, and more importantly provides for more than 256 colors. I look forward to when it is ubiquitously supported by all browsers.
that since they're actually water, they can't hurt you!
This is a big advantage over untested herbal supplements which may or may not do anything, and may or may not interact with other drugs.
Conspiracy theory for the day: homeopathy is A Cunning Plan to keep the softheaded away from untested NON-placebo stuff.
Of course, the big problem with the "Earth's life was seeded from Beyond the Stars!" is that it begs the question of how the life that seeded us evolved. It'd be interesting to know, but wouldn't answer any fundamental questions.
Of course, the same could be said about Intelligent Design arguments. "So, where did God come from, and why did he want to create life? Would other possible gods have behaved differently? And why are you building that man-sized bonfire, and walking towards me with that rope?"
1. Not a good metaphor, to the point I don't understand what you're trying to say. Would having more different accounting systems be a good thing, instead of one, clear, internationally recognized one?
2. Economists spend a lot of time trying to recognize and quantify this kind of externality. The whole "emmissions trading" concept is a way to address and limit how economic activity can degrade the environment.
Tarrifs aren't a good solution, since they don't impose costs for goods not traded internationally. I'd like to see these costs levied at production, and then having those goods freely traded after that. The tarrif system would just give an economic incentive to pollute one's own country.
And I'm not sure what the alternative is here. It's not like non-capitalist countries do better, or even as well as capitalist nations at enviornmental protection (although this is arguably more of an effect of free speech). Russia has much, much worse environmental damage than the US.
3. I don't think anyone is proposing it as a single international government. It does involve a number of international bodies, and greater international integration.
4. I don't know what the heck you mean here. Even if you argue the potential size of the economy is finite (hard to even imagine what that would mean - a hard limit to innovation?), with free trade, we could get closer to the finite potential than without it.
5. The UN has been pretty successful in reducing international conflict (although with much less success at civil wars). It hasn't lived up to its lofty goals, but I think we've clearly been better off with it than without it.
As for Keynes, his analysis of demand in the economy has arguably prevented a return of real depressions (but not recessions), but giving powerful tools to help reduce the business cycle. Again, it isn't perfect, but clearly a lot better than what we had from 1929 on.
Of course, as the cost of transport goes down, the importance of where labor is matters less and less. So free trade gets us most of what free movement would (bringing jobs to those who can do them most productively).
While I'm all for immigration, making it absolutely free would cause a lot more social disruption than the gains it would offer.
From the US perspective, we're arguably close to the right balance. The illegal immigrants we're getting are the kind of folks who could accumulate $5K in CHINA, and are willing to risk life and limb, and say goodbye to their former life in order to have a shot at making a life in America. I'm inclined to think anyone who was able to sneak in is someone we're likely going to want to have as part of our society. They might have to drive a cab, but their kids are going to be doing incredible things.
Actually, George Soros has been a huge philanthropist, and spends a lot more of his time spending money than making it now. As of 1999, he'd already given away $1B of his own money, and is still at it. He also lost a huge amount of money in Russia in the '90s, trying to help build an open society there.
By all indications, he's aiming to die broke, but is busy trying to figure out the best bang for his (enormous) buck along the way.
Generosity isn't really the right metaphor. The current term of art is social investment, and that makes more sense to me. The idea is building the institutions and structures to help build sustaninable, healthy societies. This runs the gamut from AIDS prevention to environmental preservation to microloan programs to clean government iniatives.
Food aid, while crucially important (humans don't deserve mass die-offs, no matter how lousy their governments), doesn't really solve the underlying problems of these countries. The underyling problem is why there isn't food in the first place.
Important hint: democracies don't have famines. Malnutrition, yes, but not large-scale famines.
You miss the VERY important corallary to this:
"And lets us buy cheaper shoes."
You get rich in two ways - by having more money, and by having money buy more. I'd rather make $50K a year in a world with a wide variety of imports than $100K if everything I used had to be produced in the US. I'm way richer this way, and so is everyone else on average.
The irony is that corporations don't actually improve their bottom line much by exporting jobs - they merely preserve their bottom line, since dropping prices erode much of the theoretical benefit. Overall corporate profits aren't substantailly different decade-by-decade than they were 20 or 50 years ago.
The big winners of real globalization are individuals, because it enables them to do the work they are most productive at, and reduces the costs of all goods, since everyone is able to produce good more cheaply.
Indeed. While dumping can happen, it's far from clear it was happening either with softwoods or steel. In the case of the steel industry, it was especially dubious, since minimills are doing just great. It's the old-line integrated steel producers that are having trouble. But this is largely because they signed terribly expensive deals with their unionized workers, including massive unfunded pension and health care deals for retired workers. This is one of the reasons why 401(k) deals are so good - you aren't screwed if your employer goes out of business!
The big problem is, of course, that the highly paid steelworkers aren't able to add as much value per $ of salary as global competitors (or even local competitors from nonunionized mini mills). Thus, those jobs are inevitably going to be lost in a free economy.
And as I said earlier, saving these jobs means we're going to lose three times as many jobs in other parts of the economy that rely on cheap steel. And we're going to be paying a heck of a lot more for cars, appliences, etcetera. In essence, tarrifs are taxing ourselves, but with a very, very stupid tax.
I've always been unclear why those who most advocate a wealth transfer to the third world in the form of massive forign aid are so unwilling to lose a few US jobs in exchange for more jobs in countries that need them a lot more than we do. The third world needs jobs AND aid.
The US economy should be based on the things we do best. Free trade is a great way to find out what you're good at with great accuracy and low latency. Nothing's worse for an economy than a government industrial policy.
Yes. One of the charming ironies of economics is that it takes a lot of government intervention to keep a free economy. If monopolies were legal, it'd be every corporation's goal to become a monopoly.
Trade barriers don't really help this problem - if anything they make local monopolies much more common. Check out all the national champion airlines and utilities. It's really hard (but not impossible, viz Microsoft) to maintain a global monopoloy, just because there are so many different places to have to exert pressure, and so many different markets that a disruptive competitor could find a niche into.
The real solution is powerful, independent, and apolitical antitrust agencies, on both the national and international levels.
Absolutely. Many biggest problems that most underdeveloped countries are due to insufficient globalism, not to globalism itself.
Corruption: One of the big factors of success in Western economies is transparency and lack of bribary. While we're far from 100% free of these, we're doing a lot better than most of the world. Look at the Enron scandal - we agree that it's inapprpriate for an auditor to go easy on the audits in order to gain more consulting work. The richest heads of state in the world are in third world countries. While rich people often hold elected office in the US, they're typically poorer instead of richer after it's all done (elections are expensive!).
Opaque and bribary-ridden societies mean inefficient uses of resources. If funcionaries see a primary goal of their employment being to maximize graft, the net effect is that anything that involves government approval will take as much time and require as many participants as possible. It also increases the power of the well connected over the competant, so companies who tend to get government contracts tend to be lousy and them.
Clear title to assets: A lot of the land in poor countries has unclear title, which means they can't be used as assets for loans, sold, or otherwise be treated as a form of wealth. The reasons for these are often due to a combination of a corrupt government (too many bribes to get the paperwork), and well intentioned but misguided attempts to enforce a more communal rural economy.
Free trade: This is a problem in rich and poor countries, although rich countries tend to at lesat give lip service to free trade. If everyone is free to trade goods across national borders, the net effect is that everyone can do what they're most efficient at. Protectionism enables local producers to be less efficient, with the inevitable effect of empoverishing the many for the benefit of the few. For example, in the new US steel tariffs are estimated to cost at least three jobs in steel consuming and shipping industries for every steelworker job they save. Not only is it a bad thing for our trading partners, it's bad for ourselves. Everyone loses in protectionism.
Among other things, most JPEG files use Y'CrCb 4:2:0 color space, which only includes one each chroma sample per 2x2 blocks of pixels. Thus there are only 12 bits to encode each pixel uncompressed.
However, PNG does offer the ability to encode in index colors, like 8-bit mode, and it is enormously better at encoding synthetic graphics, like screen shots. Under MacOS 9 (which has a very simple color scheme) I've had 1024x768 screen shots be only 20K. No JPEG can't come close for that kind of content.
MWright had a good technical introduction, so I'll just outline a few of the practical areas where wavelets make JPEG2000 rock.
First, compression efficiency. Although lossless and near-lossless quality isn't hugely better, data rates for "good enough" quality (defined as where the image is understandable and artifacts aren't too distracting) are much, much lower. Unlike the old Discreet Cosine Transformation (DCT) method of JPEGs, which get blocky at high compression, wavelets get softer, which is much less obvious. So, this might not help things with pro digital cameras much (which were lightly compressed in the first place), it will help a lot with web images and such, or consumer digital cameras.
The other nice thing about wavelets is that they are constructed in bands. First, the base image is encoded at a low resolution. Then this is used as prediction for the next resolution, and that band only has to store how the image is different from the prediction.
This is groovy, because you can decode the individual bands as they're transmitted, giving a low-resolution proxy image once only a few percent of the file is transmitted, getting progressively higher quality over time. While progressive/interlaced JPEG/GIF gets the same effect, wavelets do it more efficiently.
Many years ago, Intel created a video file format that used these properties, IVF. Didn't ever get any market traction. You can still get the tools from Ligos.
The standards all go well beyond just images in web pages. I extensively use PNG today as a video codec instead QuickTime for doing lossless compression of RGB images.
And JPEG is of course used in all kinds of places, like the Motion-JPEG standard used for the majority of Pro video editing systems.
While they certainly offer a lot for the web, PNG and JP2 will be important in a lot of other vertical markets. That's one of the wonderful things about standards, is they can be used in so many ways in industries where having to develop technology from scratch just doesn't make sense.
I expect we'll see a similar effect with MPEG-4, as the basis for all kinds of vertical technologies that end users might never see.
Actually, the 480x320 might have been where they went wrong with the Newton. The older versions were powerful enough, but too big. Instead of using new technologies to make the Newton more powerful yet, they should have made it smaller.
Instead of doubling the screen resolution by increasing pixel density, they should have kept the resolution 320x240, and cut the size of the unit in half. Instead of going to the full 162 MHz processor, they should have used a slower, more power efficient one, so fewer batteries were needed.
It still would have been somewhat bigger than a Palm, but with all that Newton software to use...
Something to put up on the virtual "what could have been" shelf with my dual processor PowerPC NeXT system...
Indeed. We sacrifice our long-term political interests for silly short-term political issues in swing states all the time.
Another example: high steel tarrifs don't get us anything we actually want - we're going to lose four jobs in steel-consuming industries for every one we save. AND we're PISSING OFF our allies at the time we need the most. Why? Because steel workers have a good lobby, but people who buy cars and work in dockyards don't. And because Detroit will vote Democratic whatever happens, but Pennsylvania could go either way.
And heck, why are family farms so sacrosanct? It's not like anyone in government is trying to save family restaurants, and we actually need those in our cities - most people couldn't care if they're eating produce from Florida or Chile.
Of course, I don't think we're worse at this kind of stuff than most other countries. It's just that with our greater power, our mistakes have greater consequences. We need to hold ourselves to a higher standard than we hold others.
I was pleased to see Bush announced a 50% increase in our foreign aid budget, targeting countries with good government. This is an excellent start.
Your point is excellent - inheritance taxes are a good idea.
Assuming the government is going to spend and hence raise a fixed amount of money, I'd rather they take it from dead people than living people. And as a prep-school alumnus, I can say from personal experience that the expectation that one will eventually become rich without any effort on one's own part is corrosive. Some folks adopt noblesse oblige and do just fine, but far too many tend to waste their lives waiting for their money to come, instead of finding something meaningful to do.
And, of course, rich people hate paying taxes, so they give their money away to foundations and such if the think it was going to be taxes. There would be less of this as inheritance taxes go up.
And yes, I'm speaking as someone who is likely to inherit a meaningful amount of money. I'd rather pay lower taxes now, and have less in the future. And heck, I'd rather my wealthy relatives live forever - they're a lot more fun than money (money can't tell stories about being blown out of a jeep by a panzerfaust).
That's like asking if weather is Good or Evil. To the extent we're going to have a multinational economy (and who doesn't want to have fresh strawberries in December, and laptops made in Taiwan), globalism, as strictly defined, is the way this all works.
Globalism gives us many wonderful things we all want. The question isn't Globalism: yes or no. The question is how best to encourage its benefits and deal with its disadvantages. No one has proposed any alternative to Globalism that isn't much worse than things are today, let alone how good they could be.
One of the charming ironies of global capitalism is that it achieves the "from each according to their abilities" part of the socialist creed far better than pure socialism ever did. Of course, a pure capitalism doesn't address the "to each according to their needs" part at all, which is why there aren't any pure capitalist societies in the world, and why we need to strengthen and improve the World Bank and IMF, not eliminate them.
What we want is a global economy where people can compete ferociously on production, but where there is properly integrated environmental protection, and a safety net where none starve, and all can get a good education irrespective of the economic success or failure of their parents.
Global capitalism means we can all get goods and services as the lowest cost available for the quality we need. Remember, the two ways you get richer are through wages going up and prices going down. Those who want high steel tariffs want to enrich the few at the expense of the many.
For those curious about the details, check out David Ricardo.
Folks,
I find it remarkable how contradictory the two streams of advice about how the US should change its behavior are. One one hand, there are those who say we need to be less involved in the rest of the world - no global policing, no liberation of Kuwait, leave Iraq alone, leave Bosnia alone, no more support for Israel etc. Others say we need to be more involved - impose a Palestinian/Israeli settlement, neutralize Iraq, increase and improve foreign aid, make freeer trade, etc.
Neither side is essentially wrong - what they all want is for us to do the "right" things, and not do the "wrong" things. Of course, which is which is far from obvious.
As in all things, Stan Lee said it best: "With great power comes great responsibility." Spiderman's greatest failure wasn't what he did, but what he didn't do, allowing a criminal to go free and kill his Uncle Ben. As semi-facetious as the argument may sound, the US has been bit by that radioactive spider, and we need to accept the fact of our power, and figure out how best to use it.
Due to a combination of excellent geography, a successful culture built by waves of immigration, and dumb luck, the US has the ability to Make Things Different in the world. We can make changes. But while we're powerful, we aren't prescient, and can't predict the long-term ramifications of what we do with any accuracy.
But choosing not to act is still an action - was not intervening in Rawanda any better than invading Iraq would be? Our power means we have to make these decisions, and sweat them, and argue about them endlessly on talk shows, and then, gut clinched, try to do the right thing and hope it all works out.
Many of the examples of past US evil are from the cold war. I think we can agree that a good number of the things we did were not only not effective for our goals, but hurt them in the long term. But do you think that, with what they knew at the time, the people who made those decisions knew which were which? Does this mean we should have let the Soviet Union have a free hand all over the world? No.
The US needs to be engaged with the world, and we have to know it's going to be a messy business that will make us enemies. So we need to do it in as smart a way as possible, with a long view, with clear-eyed compassion, and with as little attention to our trivial domestic politics as possible.
For a recent example of us NOT doing this right, we failed to drop textile tarrifs for Pakistan. Pakistan's leadership, whatever their faults, have been a surprisingly good partner in our current conflict. And they're dirt poor, which helps terrorists, and hurts the development of a healthy society. Pakistan makes a lot of textiles, and could important a lot more to the US, helping develop a better, non-aid-based economy. But, in order to not risk the Carolinas in the mid-term elections later this year, and his reelection in 2004, Bush refused to lower tarrifs or increase Pakistan's quota for textiles. I know it sounds like a small deal, but getting this kind of stuff right could help enormously.
Anyway, there is nothing we can do that will keep people from hating us. We have the power to pick winners and losers, and so we'll be resented for what we don't do just as much as for what we do. So we just have to do the right thing as best we can, and give the world as few VALID reasons to hate us as possible.
This is a little geeky.
QuickTime provides is own "keyframe every" option. In VP3's codec-specific dialog, it provides its own, more advanced implementation (which lets you set a minimum number of frames between keyframes, not just a maximum). Thus, most folks turn off QuickTime's keyframe insertion and let VP3's do it. I believe this is what the tester did, based on the description.
Leaving the keyframe interval blank may, depending on the tool, force EVERY frame to be a keyframe. Definitely not what you want.
The "Allow Dropped Frames" command actually controls the interplay of data rate versus frame rate, not image quality. When dropped frames are allowed, the codec will reduce frame rate in order to hit the target data rate.
One objection you didn't mention is the use of the "Fast Compression" mode. The tester may have gotten better quality (and much slower encoding) if that had been on.
This brings up an interesting question for the free/open source community. Under MPEG-4, it's perfectly acceptable to make an encoder or decoder than is open source. However, the patent license is $0.25 each per encoder and decoder, with a cap of $1M/each per company per year.
So, would folks be willing to cough up a quarter for this codec, if it remained open source?
Is there time for a new license: free speech, but not free beer?