I submit that it's really not that big a deal. So you can expand or contract images to fit paper sizes without wasting any space. So what? Expanding and contracting isn't really that useful, and when it is useful the small amount of wasted paper is not important. I cannot see that switching my company's current workflows from US to metric paper sizes is going to either save money or open up new opportunities.
There are probably far more sheets of paper in the world that are US Letter size than any other size at all. And file folders to match. You want them all thrown away?
My wife and I travelled about 90 minutes to Lancaster, PA. (Specifically, we were here.) We walked in the little park in front of the building, looked up at Jupiter, and suddenly the ISS was about two degrees away from Jupiter and moving fast. Didn't have time to get the binoculars up, but I distinctly saw the two bright dots merge and separate. I think I was definitely in the path of totality, or at most no more than a quarter mile outside of it. ISS was the brightest damn thing in the sky. My binoculars couldn't make any details clear, still looked like a bright dot.
Since my birthday is tomorrow, I feel like the universe just gave me a birthday present.
Clipper 5.x for DOS, circa 1989. Multithreading is a red herring; all you need is a stack trace, and that only requires a decent VM, which Clipper pretty much had.
DNA was once considered
the sole repository of
heritable information.
But biologists are starting
to decipher a separate,
much more malleable layer
of information encoded
within the chromosomes.
Genetics, make way
for epigenetics
Four paragraphs from the first page:
A genome, the sum of heritable information that is held in
the chromosomes and that governs how an organism develops,
is not a static text passed from one generation to the next. Rather
a genome is a biochemical machine of awesome complexity.
Like all machines, it operates in three-dimensional space, and it
has distinct and dynamic interacting parts.
Protein-coding genes make up just one of those parts--and
often a small one at that, accounting for less than 2 percent of
the total DNA in each human cell. But for the better part of five
decades, those genes were enshrined by the central dogma of
molecular biology as the repository of heritable traits. Hence the
notion of the genome as a blueprint.
As far back as the 1960s, experimenters had uncovered important
information hiding elsewhere in the chromosomes.
Some was tucked among the "noncoding" DNA, and some lay
outside the DNA sequence altogether. The tools of genetic engineering
worked best on conventional genes and proteins, however,
so scientists looked hardest where the light was brightest.
In recent years, geneticists have been exploring the less visible
parts of the genome more thoroughly, in search of explanations
for anomalies that contradict the central dogma: illnesses
that run in families but pop up unpredictably, even differing
among identical twins; genes that switch on or off in cancers yet
harbor no mutations; clones that usually die in the womb. They
have found that these second and third layers of information,
distinct from the protein-coding genes, connect in surprisingly
deep and potent ways to inheritance, development and disease.
In the November issue of Scientific American, "The Unseen
Genome: Gems among the Junk" described those connections
for the second layer, which consists of myriad "RNA only"
genes sequestered within vast stretches of noncoding DNA. Science
had dismissed such DNA as the useless detritus of evolution,
because no proteins are made from it. But it turns out that
these unconventional genes do give rise to active RNAs, through
which they profoundly alter the behavior of normal genes. Malfunctions
in RNA-only genes can inflict severe damage.
The third part to the genomic machine, as fascinating as active
RNA genes and probably even more important, is the "epigenetic"
layer of information stored in the proteins and chemicals
that surround and stick to DNA. Epigenetic marks are so
named because they can dramatically affect the health and characteristics
of an organism--some are even passed from parent
to child--yet they do not alter the underlying DNA sequence.
DNA isn't everything. There are lots and lots of molecules in a cell besides DNA; some of them have persistent concentrations that can have effects even on the next generation.
9. astro-ph/0208010 [abs, ps, pdf, other]:
Title: Comparing WIMP Interaction Rate Detectors with Annual Modulation Detectors Authors: Craig J. Copi, Lawrence M. Krauss Comments: 4 pages, 2 figures. Submitted to Phys. Rev. Letters Journal-ref: Phys.Rev. D67 (2003) 103507
The last line, where it says Journal-ref, means it was accepted in Physical Review D (Particles, Fields, Gravitation, and Cosmology -15) -- a subjournal of PRL.
The point of this finding is that computation has an upper bound. K&S's work shows the best you can get, in a Friedmann universe with perfect computation results. Any other situation will do less computation.
Sure, the methods K&S propose are not practical. They're not even plausible. But if they were, they would achieve better results than any other possible method that has been thought of.
Some people have mentioned reversible computing. As others have pointed out to them, reversible computing requires you store all intermediate results. It reduces the number of bits you can actually make use of. And in an accelerating, expanding universe where you are losing access to energy every second, you may in theory be able to continue your reversible computing forever but your total storage cannot increase.
Therefore you have a finite number of possible states of the system -- which means that there's a finite limit to the thinking you can do before you have to repeat yourself.
Who knows what 1.5 * 10^220 bits of information processed is?
whereas the paper states
Information Processed <=
{complex math expression impossible to write in HTML} = 1.35 x 10^120.
I was nitpicking at julesh's number of 1.5 * 10^220, which did not appear in the paper. In fact, his number is over 100 orders of magnitude larger than the number in the paper.
The goal is not to replace pilots, but to allow air traffic controllers to make better use of their time. From improved ATC efficiency, they expect also expect to gain things like reduced "circling" time, better (more direct) flight plans, and reduced fuel usage.
I remember vividly when this happened (ten years ago, when "the Internet" usually meant USENET as opposed to the WWW). Before, "bad behavior" meant poor "netiquette"- crossposting to a dozen or so USENET groups. That was what pissed people off. But even the crossposters were flabbergasted by this. It seems trite now, but back in 1994, nobody had even dreamed of posting a message to every single USENET newsgroup in existence. The very idea was crazy. Posts were things you typed into newsreaders. You'd need to write a script to crosspost to every single newsgroup. Who would ever do that? It was just too incredible to believe.
Part of the outrage was that the spammers did not crosspost. Their script posted separately to each newsgroup. If they had crossposted, then the spam message would occupy a small amount of space on each server, but as separate posts, it occupied thousands of times as much. Some small sites with small retention were seriously hurt.
The real point is that merely saying "I won't act like a virus if the scanner thinks I am one" won't stop the scanner from thinking it is one. So there's no contradiction.
If the program that is running could call out to the is_a_virus() routine, that wouldn't be very useful to the anti-virus software, but...
Every antivirus program I've ever used has the ability for a user to scan a particular file at will. (It may be a simple command-line invocation, or it may mean navigating a dozen menus while holding down the alt key, but it's possible.) What a user can do, a program can do. Ergo a program can invoke the antivirus scanner on itself.
Just ran some samples using Excel. The rule of 72 is closest when the number of periods is 9: rule of 72 gives 8%, actual calc gives 8.006%.
When the number of periods is 4, the differences is almost a whole percentage point: Rule of 72 = 18%, actual = 18.921%.
After 9 periods, the rule of 72 starts giving results that are larger than the actual number, but less than a tenth of a percent different: 10 7.200% 7.177% 11 6.545% 6.504% 12 6.000% 5.946% 13 5.538% 5.477% 14 5.143% 5.076% 15 4.800% 4.729% 16 4.500% 4.427% 17 4.235% 4.162% 18 4.000% 3.926% 19 3.789% 3.716% 20 3.600% 3.526% 21 3.429% 3.356% 22 3.273% 3.201% 23 3.130% 3.060%
Some more interesting milestones are 180 and 360 periods (15 and 30 years): 180 0.400% 0.386% 360 0.200% 0.193%
Although the absolute difference is still shrinking and is now less than a 50th of a percentage point, the difference in the significant digits is very large.
Slashdot Mirror
gallons and quarts??? a gallon is four quarts, that's not 12.
To consider just one thing, can you imagine how many billions of dollars it would take to replace all the speed limit signs in the USA?
And what would it gain us? Tourists would be 0.004% happier seeing km/h on the road instead of mph.
NOT WORTH IT
I submit that it's really not that big a deal. So you can expand or contract images to fit paper sizes without wasting any space. So what? Expanding and contracting isn't really that useful, and when it is useful the small amount of wasted paper is not important. I cannot see that switching my company's current workflows from US to metric paper sizes is going to either save money or open up new opportunities.
There are probably far more sheets of paper in the world that are US Letter size than any other size at all. And file folders to match. You want them all thrown away?
Actually, if the moon is half a degree wide, then it was a lot closer than two degrees when I first spotted it. Maybe like 0.2 degrees.
My wife and I travelled about 90 minutes to Lancaster, PA. (Specifically, we were here.) We walked in the little park in front of the building, looked up at Jupiter, and suddenly the ISS was about two degrees away from Jupiter and moving fast. Didn't have time to get the binoculars up, but I distinctly saw the two bright dots merge and separate. I think I was definitely in the path of totality, or at most no more than a quarter mile outside of it. ISS was the brightest damn thing in the sky. My binoculars couldn't make any details clear, still looked like a bright dot.
Since my birthday is tomorrow, I feel like the universe just gave me a birthday present.
It's a shame the above wasn't written in English.
Clipper 5.x for DOS, circa 1989. Multithreading is a red herring; all you need is a stack trace, and that only requires a decent VM, which Clipper pretty much had.
Gibbs, W. Wayt. (2003). The Unseen Genome: Beyond DNA. Scientific American, 289(6), 106-113. (December 2003 issue)
Teaser line:
Four paragraphs from the first page:
DNA isn't everything. There are lots and lots of molecules in a cell besides DNA; some of them have persistent concentrations that can have effects even on the next generation.
Similar remarks for the other two cites you give.
The point of this finding is that computation has an upper bound. K&S's work shows the best you can get, in a Friedmann universe with perfect computation results. Any other situation will do less computation.
Sure, the methods K&S propose are not practical. They're not even plausible. But if they were, they would achieve better results than any other possible method that has been thought of.
Some people have mentioned reversible computing. As others have pointed out to them, reversible computing requires you store all intermediate results. It reduces the number of bits you can actually make use of. And in an accelerating, expanding universe where you are losing access to energy every second, you may in theory be able to continue your reversible computing forever but your total storage cannot increase.
Therefore you have a finite number of possible states of the system -- which means that there's a finite limit to the thinking you can do before you have to repeat yourself.
What I meant was that julesh asked (rhetorically)
whereas the paper states
I was nitpicking at julesh's number of 1.5 * 10^220, which did not appear in the paper. In fact, his number is over 100 orders of magnitude larger than the number in the paper.
In the 20s, when expansion was first detected by Dr. Hubble, the "Steady State" theory was advanced to explain it as an alternative to the "Big Bang" theory, which the late Sir Fred Hoyle found offensive. (By the way, he coined both phrases - Big Bang and Steady State.)
I read the paper. The equation says
Information Processed = 1.35 x 10^120 bits
Not sure where you got the other number.
a more useful purpose.
duh ... the story said that he was listening to cassette tapes that were copies of the R2R tapes.
Wrong. ICMP is at the network layer, UDP is the transport layer.
Then try gaim. It allows the same behavior, and is OSS to boot.
The goal is not to replace pilots, but to allow air traffic controllers to make better use of their time. From improved ATC efficiency, they expect also expect to gain things like reduced "circling" time, better (more direct) flight plans, and reduced fuel usage.
Part of the outrage was that the spammers did not crosspost. Their script posted separately to each newsgroup. If they had crossposted, then the spam message would occupy a small amount of space on each server, but as separate posts, it occupied thousands of times as much. Some small sites with small retention were seriously hurt.
Agreed. I was just nitpicking.
The real point is that merely saying "I won't act like a virus if the scanner thinks I am one" won't stop the scanner from thinking it is one. So there's no contradiction.
Every antivirus program I've ever used has the ability for a user to scan a particular file at will. (It may be a simple command-line invocation, or it may mean navigating a dozen menus while holding down the alt key, but it's possible.) What a user can do, a program can do. Ergo a program can invoke the antivirus scanner on itself.
Just ran some samples using Excel.
The rule of 72 is closest when the number of periods is 9: rule of 72 gives 8%, actual calc gives 8.006%.
When the number of periods is 4, the differences is almost a whole percentage point: Rule of 72 = 18%, actual = 18.921%.
After 9 periods, the rule of 72 starts giving results that are larger than the actual number, but less than a tenth of a percent different:
10 7.200% 7.177%
11 6.545% 6.504%
12 6.000% 5.946%
13 5.538% 5.477%
14 5.143% 5.076%
15 4.800% 4.729%
16 4.500% 4.427%
17 4.235% 4.162%
18 4.000% 3.926%
19 3.789% 3.716%
20 3.600% 3.526%
21 3.429% 3.356%
22 3.273% 3.201%
23 3.130% 3.060%
Some more interesting milestones are 180 and 360 periods (15 and 30 years):
180 0.400% 0.386%
360 0.200% 0.193%
Although the absolute difference is still shrinking and is now less than a 50th of a percentage point, the difference in the significant digits is very large.
20*20 - 20 = 20*19
20*19 - 19 = 19*19
For that matter, CR, LF, and even BS have an explicit typewriter/teletype meaning which only slightly survives in the text world.