This machine will cost a lot of money, but how much will determine if it will get built. While Japan might be a great place to do it to keep symmetry, we need to learn from the SSC failure. It was (partially) built in Texas in spite of the fact that it would have been much cheaper at Fermilab. If this machine will be built cost considerations must be foremost. That probably means Fermilab again is the only realistic place.
I agree with your primary thesis. As evidence I offer my experience of many more years than I care to admit pursueing a doctorate on three continents all the while acknowledging that the degree was only an admission ticket to the higher ranks of academia. Now that I'm working, I much prefer having to know multiple fields and actually getting things done.
Unicode would have adequate space for almost all scripts, if not all if Chinese could be resolved. This could include the Hangul pictographs and the Japanese Kanji. This leaves two alternatives. Either allow both Chinese governments to dictate the full extent of their character sets (and sacrifice the workability of Unicode) or have non-Chinese dictate a solution that fits into the allotted space. I personally prefer the later.
You know, this is actually the one topic that I am probably best versed on discussing. My info sci masters advisor was on the committee which established ASCII and my linguistics masters was on medieval Chinese dictionaries. Plus, I used to live in Japan.
There are _slightly_ more than 2000 kanji in Japanese, but Japanese printers, like my wife's father, don't use more than 2100 absolute tops.
Chinese characters obey Zipf's law on a near perfect logarithmic scale. As in, the first ten characters make up about 60% of written text. For each unit of ten up from that include about 60% of what is left. At 10,000 characters you have all but about 2.5% of most newspaper text. The few thousand extra that I spoke of covers mostly proper names.
Chinese most certainly can be written satifactorily in this manner.
The 64,000 should suffice. Ideographic scripts, like Chinese are were the problem arises. The number of characters in Chinese is not fixed, unlike the number in most alphabets. I have a Chinese novella which was written in just 300 characters. 10,000 would be a good place to start, a few thousand more would cover all but specialized texts. Japanese could fold into Chinese, since there are only 2000 kanji characters and a few hundred kana.
Throw in Arabic, Cryllic, Sanskrit, Dravidian, Hangul (Korean) and Navaho and you still add only a few thousand. The odd European characters (the 'ss' in German, the extra Danish vowels, . ..) add a few hundred tops. Even the special linguist marks and punctuation don't add much.
If you have to double the Chinese, now you run into trouble. Its classical characters vs. simplified. The later is for the PRC. If you also bloat the number of characters required so that specialized religous characters are required, now you start to push the system. 64K would be fine if a special marker character could be used which signify's that the next character is from the special table. Unicode has resisted this effort.
NASA is an arm of the US federal government. It operates on a 15 to 18 year cycle of high employment because otherwise its employees would start to acquire pension rights. Hence assuming that its operations will move at a faster rate is unrealistic. I think Goldin is being much more realistic than anyone is giving him credit for.
Don't believe me? Fine, consult any nuclear physicist. The problems with conventional nuclear plant design have to deal with a) transfer medium, b) shielding and c) control. Liquid lithium in a gravity fed tank deals with all three. That the programs existed and were canceled under the Clinton administration are objects of public record. No trust in me required.
You're actually thinking of breeder reactors, a diffferent animal. They took plutonium in, produced power and more plutonium, some even being capable of purifying it slightly. You are correct with your timelines for breeders. I was referring to a different type of reactor which didn't produce commercial quantities of power. It was called a nuclear furnace. It's sole function was burning long half life material. Basically, bombard it with a heavy enough dose of radiation and the material transforms from long half-life material to short half-life material. The energy dissipated is too unstable to produce power from commercially.
Back in 1992, there were two programs running at Argonne National Labs, near Chicago. The first was a fail-safe nuclear plant. It used liquid lithium as the transfer medium, with a gravity fed tank. The lithium protects against leaks, you can't cause a meltdown and when it retires, the reactor seals itself in. The second was a reactor that "burned" long half-life radioactive material and reduced it all to 50 year or less half-life material. That is MUCH easier to store, not requiring million year storage.
What happened to both programs? The Clinton administration killed both of them, supposedly at Al Gore's insistence. I was asked to refute the VP's evidence. In one word, his reasoning was crap. Working prototypes of both reactors existed, they performed as advertised and were truly safe. But they were "nuclear" so Al Gore wanted them killed. Pure lip-service environmentalism. I was there, I performed the analysis, I reviewed the documentation.
If the neural net takes linkages as input then Disney could find themselves on the list of blocked sites, a somewhat ironic development to say the least. (If you dont understand this, then you obviously haven't poked around these sites very much.)
The Hubble is by far the most powerful telescope advance in the history of astronomical observations. And your estimate of "not only did it cost billions of dollars before it was even launched" is just plain wrong. While expensive, it cost no more than the Palomar observatory cost in constant dollars. It's operation costs are very reasonable and any astronomer who has ever seen pictures from it will inform you that it is the best investment we could have made. While Hubble initially did initially have a problem with its focus, it was still working and the array of instruments on board allowed observations to be made continuously.
The Hubble has been an incredible success, is an exemplar of government sponsored science at reasonable cost.
This factor would hold if the distribution of the elements in the galaxy were uniform. With different molecular weights, gravity and centripetal force will guarentee that any circular, rotating galaxy will have a band, a certain distance from the center where the concentration of the heavier elements is appropriate for life. Most of the rest of the galaxy is a lifeless void.
Re:Sagan Spinning In His Grave?
on
Explaining SETI
·
· Score: 2
Now, I have to disagree with you on the pi issue, or lack thereof. I thought the dialog between the bureaucrats about how long the blank tape lasts was a suitable replacement and a lot less technical for the average movie-goer.
Sagan's normal arguements on faith were two pronged a) disprove an interventionist god, b) apply the Principle of Parsimony (Occam's Razor) to refute the need for God. I thought the movie paralleled this M.O.
Re:Contact - not my favorite
on
Explaining SETI
·
· Score: 2
While Sagan's book is really good about the SETI side of things, I thought it really, really came up short on other issues. Sagan helped me get into his alma mater, but we squared off on the roles of both capitalism and religion. He basically thought they had little or none in the modern world (or at least the modern mind.) These deep beliefs pervaded both his book and the movie. Then again, I think he would have regarded the movie as a decent adaption of the better book.
There are two sets of odds which must be considered when calculating the probability of SETI working. First, the chance of life developing elsewhere. Second, the chance of elsewhere developed life becoming intelligent. The first can be relatively common (or more specifically, only rare, not extremely rare) but the second be extremely rare and SETI still won't work. And this looks like the case.
Personally, I'm all for SETI. As long as it's zero success rate proscribes it's receiving massive funding. I know that when scientists have to scrounge, have to be imaginitive, they are often doing their best work. Should Aricebo be used for SETI? Occasionaly, but there is plenty of other science to do as well.
This does sound like a Bond villian scheme, specifically like Octopussy. I do think it bears pointing out the other advantage of this scheme. You use 15 small nukes to get the equivalent of 15 very large nukes. If you know this is coming you can covertly prepare to protect your government. Whilst, most US citizens would rather see their government severely disrupted, this may not be true for the rest of the world. (No, they'd all like to see the US government disrupted too.)
Actually, even though I don't agree with filters for general usage, the filter industry couldn't get better news. Well designed filters would be able to handle such a simple end run. The challenge is now, be well-designed or hang it up.
Fundamentally, the arguements in this come down to: does a different media require a different contract or fee? Selling the screenplay version of a popular book, apart from the print rights is one side. An encylopedia article, which appears in the print version, on CD and even on the website would represent the other side.
And if this set of government is a representation then perhaps we should rescind the Declaration of Independence and rejoin the British Commonwealth. I understand Australia may be giving up their seat at that.
As a former junior faculty member, I always strove to keep assignments different enough between years and between students that collaboration couldn't be cheating (at least in my mind.) Finding out how someone else solved a previous problem is learning. Passing this off as original doesn't merit a higher grade.
In all fairness, I was the exception with changing my assignments from year to year and giving different students different assignments for each and every mark. Most faculty just don't want to work that hard at teaching.
Several comments have been made about the quality of the beer at a perspective geek hangout bar. Would good beer really make a difference? I work with databases during the day but brew beer during the evenings. Good beer could be made for just slightly more than standard (about $3 per pint) if a steady market could be locked in. Marketing, bottling, decor and entertainment lock up a big portion of the required capital of a brewpub.
I agree with you that it does get said too often. My opinion is that a) a lot of work can get done by people not at the acme of electronic interconnectivity and b) while these have their usefulness and some jobs (now) virtually require them (sales and troubleshooting folks come to mind) generalizing this ability to everyone in the company just isn't appropriate. There was an old fortune in the UNIX fortune file, "One reason computers can get more work done than humans is that computers don't have to stop and answer the phone."
I carry no pager. I don't have a mobile phone. I walk away from my desk and go outside to get some fresh air and resolve the issues that come up during my programming duties. I talk to people in other departments to find out what the issues there are and how my work should help solve them. These are all things that make me successful. The last thing I want is the whole building following me around.
At its most effective, hacktivism will not cripple internationalization or globalization, it will merely raise the costs of doing it. This would have the reverse of the intended effect; only those capable of reaping immense profits will do so, thus fostering a nastier bunch of corporate mercenaries.
Slashdot Mirror
This machine will cost a lot of money, but how much will determine if it will get built. While Japan might be a great place to do it to keep symmetry, we need to learn from the SSC failure. It was (partially) built in Texas in spite of the fact that it would have been much cheaper at Fermilab. If this machine will be built cost considerations must be foremost. That probably means Fermilab again is the only realistic place.
I agree with your primary thesis. As evidence I offer my experience of many more years than I care to admit pursueing a doctorate on three continents all the while acknowledging that the degree was only an admission ticket to the higher ranks of academia. Now that I'm working, I much prefer having to know multiple fields and actually getting things done.
Unicode would have adequate space for almost all scripts, if not all if Chinese could be resolved. This could include the Hangul pictographs and the Japanese Kanji. This leaves two alternatives. Either allow both Chinese governments to dictate the full extent of their character sets (and sacrifice the workability of Unicode) or have non-Chinese dictate a solution that fits into the allotted space. I personally prefer the later.
You know, this is actually the one topic that I am probably best versed on discussing. My info sci masters advisor was on the committee which established ASCII and my linguistics masters was on medieval Chinese dictionaries. Plus, I used to live in Japan.
There are _slightly_ more than 2000 kanji in Japanese, but Japanese printers, like my wife's father, don't use more than 2100 absolute tops.
Chinese characters obey Zipf's law on a near perfect logarithmic scale. As in, the first ten characters make up about 60% of written text. For each unit of ten up from that include about 60% of what is left. At 10,000 characters you have all but about 2.5% of most newspaper text. The few thousand extra that I spoke of covers mostly proper names.
Chinese most certainly can be written satifactorily in this manner.
The 64,000 should suffice. Ideographic scripts, like Chinese are were the problem arises. The number of characters in Chinese is not fixed, unlike the number in most alphabets. I have a Chinese novella which was written in just 300 characters. 10,000 would be a good place to start, a few thousand more would cover all but specialized texts. Japanese could fold into Chinese, since there are only 2000 kanji characters and a few hundred kana. .) add a few hundred tops. Even the special linguist marks and punctuation don't add much.
Throw in Arabic, Cryllic, Sanskrit, Dravidian, Hangul (Korean) and Navaho and you still add only a few thousand. The odd European characters (the 'ss' in German, the extra Danish vowels, . .
If you have to double the Chinese, now you run into trouble. Its classical characters vs. simplified. The later is for the PRC. If you also bloat the number of characters required so that specialized religous characters are required, now you start to push the system. 64K would be fine if a special marker character could be used which signify's that the next character is from the special table. Unicode has resisted this effort.
Computers were certainly interconnected more than a decade ago. I courted my first wife via DARPA and she's been gone for more than a decade.
NASA is an arm of the US federal government. It operates on a 15 to 18 year cycle of high employment because otherwise its employees would start to acquire pension rights. Hence assuming that its operations will move at a faster rate is unrealistic. I think Goldin is being much more realistic than anyone is giving him credit for.
Don't believe me? Fine, consult any nuclear physicist. The problems with conventional nuclear plant design have to deal with a) transfer medium, b) shielding and c) control. Liquid lithium in a gravity fed tank deals with all three. That the programs existed and were canceled under the Clinton administration are objects of public record. No trust in me required.
You're actually thinking of breeder reactors, a diffferent animal. They took plutonium in, produced power and more plutonium, some even being capable of purifying it slightly. You are correct with your timelines for breeders. I was referring to a different type of reactor which didn't produce commercial quantities of power. It was called a nuclear furnace. It's sole function was burning long half life material. Basically, bombard it with a heavy enough dose of radiation and the material transforms from long half-life material to short half-life material. The energy dissipated is too unstable to produce power from commercially.
Back in 1992, there were two programs running at Argonne National Labs, near Chicago. The first was a fail-safe nuclear plant. It used liquid lithium as the transfer medium, with a gravity fed tank. The lithium protects against leaks, you can't cause a meltdown and when it retires, the reactor seals itself in. The second was a reactor that "burned" long half-life radioactive material and reduced it all to 50 year or less half-life material. That is MUCH easier to store, not requiring million year storage.
What happened to both programs? The Clinton administration killed both of them, supposedly at Al Gore's insistence. I was asked to refute the VP's evidence. In one word, his reasoning was crap. Working prototypes of both reactors existed, they performed as advertised and were truly safe. But they were "nuclear" so Al Gore wanted them killed. Pure lip-service environmentalism. I was there, I performed the analysis, I reviewed the documentation.
If the neural net takes linkages as input then Disney could find themselves on the list of blocked sites, a somewhat ironic development to say the least. (If you dont understand this, then you obviously haven't poked around these sites very much.)
The Hubble is by far the most powerful telescope advance in the history of astronomical observations. And your estimate of "not only did it cost billions of dollars before it was even launched" is just plain wrong. While expensive, it cost no more than the Palomar observatory cost in constant dollars. It's operation costs are very reasonable and any astronomer who has ever seen pictures from it will inform you that it is the best investment we could have made. While Hubble initially did initially have a problem with its focus, it was still working and the array of instruments on board allowed observations to be made continuously.
The Hubble has been an incredible success, is an exemplar of government sponsored science at reasonable cost.
This factor would hold if the distribution of the elements in the galaxy were uniform. With different molecular weights, gravity and centripetal force will guarentee that any circular, rotating galaxy will have a band, a certain distance from the center where the concentration of the heavier elements is appropriate for life. Most of the rest of the galaxy is a lifeless void.
Now, I have to disagree with you on the pi issue, or lack thereof. I thought the dialog between the bureaucrats about how long the blank tape lasts was a suitable replacement and a lot less technical for the average movie-goer.
Sagan's normal arguements on faith were two pronged a) disprove an interventionist god, b) apply the Principle of Parsimony (Occam's Razor) to refute the need for God. I thought the movie paralleled this M.O.
While Sagan's book is really good about the SETI side of things, I thought it really, really came up short on other issues. Sagan helped me get into his alma mater, but we squared off on the roles of both capitalism and religion. He basically thought they had little or none in the modern world (or at least the modern mind.) These deep beliefs pervaded both his book and the movie. Then again, I think he would have regarded the movie as a decent adaption of the better book.
There are two sets of odds which must be considered when calculating the probability of SETI working. First, the chance of life developing elsewhere. Second, the chance of elsewhere developed life becoming intelligent. The first can be relatively common (or more specifically, only rare, not extremely rare) but the second be extremely rare and SETI still won't work. And this looks like the case.
Personally, I'm all for SETI. As long as it's zero success rate proscribes it's receiving massive funding. I know that when scientists have to scrounge, have to be imaginitive, they are often doing their best work. Should Aricebo be used for SETI? Occasionaly, but there is plenty of other science to do as well.
This does sound like a Bond villian scheme, specifically like Octopussy. I do think it bears pointing out the other advantage of this scheme. You use 15 small nukes to get the equivalent of 15 very large nukes. If you know this is coming you can covertly prepare to protect your government. Whilst, most US citizens would rather see their government severely disrupted, this may not be true for the rest of the world. (No, they'd all like to see the US government disrupted too.)
Actually, even though I don't agree with filters for general usage, the filter industry couldn't get better news. Well designed filters would be able to handle such a simple end run. The challenge is now, be well-designed or hang it up.
Fundamentally, the arguements in this come down to: does a different media require a different contract or fee? Selling the screenplay version of a popular book, apart from the print rights is one side. An encylopedia article, which appears in the print version, on CD and even on the website would represent the other side.
No taxation without representation.
And if this set of government is a representation then perhaps we should rescind the Declaration of Independence and rejoin the British Commonwealth. I understand Australia may be giving up their seat at that.
As a former junior faculty member, I always strove to keep assignments different enough between years and between students that collaboration couldn't be cheating (at least in my mind.) Finding out how someone else solved a previous problem is learning. Passing this off as original doesn't merit a higher grade.
In all fairness, I was the exception with changing my assignments from year to year and giving different students different assignments for each and every mark. Most faculty just don't want to work that hard at teaching.
Several comments have been made about the quality of the beer at a perspective geek hangout bar. Would good beer really make a difference? I work with databases during the day but brew beer during the evenings. Good beer could be made for just slightly more than standard (about $3 per pint) if a steady market could be locked in. Marketing, bottling, decor and entertainment lock up a big portion of the required capital of a brewpub.
I agree with you that it does get said too often. My opinion is that a) a lot of work can get done by people not at the acme of electronic interconnectivity and b) while these have their usefulness and some jobs (now) virtually require them (sales and troubleshooting folks come to mind) generalizing this ability to everyone in the company just isn't appropriate. There was an old fortune in the UNIX fortune file, "One reason computers can get more work done than humans is that computers don't have to stop and answer the phone."
I carry no pager. I don't have a mobile phone. I walk away from my desk and go outside to get some fresh air and resolve the issues that come up during my programming duties. I talk to people in other departments to find out what the issues there are and how my work should help solve them. These are all things that make me successful. The last thing I want is the whole building following me around.
At its most effective, hacktivism will not cripple internationalization or globalization, it will merely raise the costs of doing it. This would have the reverse of the intended effect; only those capable of reaping immense profits will do so, thus fostering a nastier bunch of corporate mercenaries.