There are very few revolutionary discoveries now - and fewer groundbreaking ideas - most are evolutionary. We are at a point where science is getting smaller and less accessible to the average reader. Consequently, this doesn't make good reading, so we're still reduced to reading about warp cores and wormholes.
It would make the universe a poorer place to think that we had discovered all of the fundamental ways to manipulate it, and that all we can look forward to is ever more expensive ways of fiddling further and further to the right of the decimal point. The problem with truly revolutionary discoveries is that they are, by their very nature, impossible to search for directly -- you can't know what it is that you do not know. All we can do is keep pushing out the limits of what we do know until someone looks at a set of results for a hitherto-predictable test and says, "That shouldn't be happening", and then follows it out to find out why. Over and over again.
What makes Spider Robinson a commentator on SF, Sci-Fi or anything else other than pablum?
You mean besides winning a Locus award for Best Critic? Besides being book reviewer for Galaxy, Analog and New Destinies magazines for nearly a decade, and continuing to write occasional book reviews and a regular Op-Ed column, "Future Tense," for The Globe and Mail, Canada's national newspaper.? Nothing, I guess...
And as for the 'Speculative Fiction', well, he isn't a writer of that either.
The people who voted to award him three Hugo awards (science fiction's top honor), a Nebula award, the John W. Campbell Award for Best New Writer, the E.E. "Doc" Smith Memorial Award (Skylark), the Pat Terry Memorial Award for Humorous Science Fiction, and a second Locus award for Best Novella would appear to disagree with you. But you can always define 'speculative fiction' to be whatever you want, and set up your definition to exclude what he writes.
I don't read science fiction for exactly this reason. What meaning could science fiction possibly have when science is constantly going beyond our imaginations already?
You might want to look at some of Vernor Vinge's work. Much of his work has centered around the concept of 'the Singularity' -- a point at which we create, through technology, an entity of greater than human intelligence (whether machine intelligence, technologically-augmented humans, or bioengineered humans, or something else entirely), which results in an exponential runaway in the rate of technological advance, such that we can no longer even imagine what advances the future will bring. Dr. Vinge has written a paper on his view of the Singularity, and there is a critical discussion of the concept.
The point of science fiction, with the rate of technological advance going beyond where a writer can hope to have the background to predict in even the broadest terms over more than a short time, isn't about the technology itself, the way much of the space-opera of the '40s and '50s focussed on the 'gosh-wow' hardware, but on the way people deal with living in a world where it has become impossible to understand how things work -- and you begin to approach the embodiment of Clarke's Law: "Any sufficiently advanced technology is indistinguishable from magic." When everybody has a device you can pour dirt into and have it produce anything that can be made by recombining the atoms in the dirt, it changes the focus of people's drives -- but not the existence of those drives.
Do we really know that the computers tried to take over in the Dune universe? Maybe I'm misremembering, but I thought Herbert always left the problem of intelligent machines unstated, just alluding to fact that there had been problems. It could have been that humans lost all their will in a world where machines did everything, even the thinking or perhaps something else entirely happened.
Recently out in paperback by Brian Herbert and Kevin J. Andersen is The Butlerian Jihad, fourth of the 'Dune prequels'. In the book (presented as background, so I'm not giving away plot), the advances in robotics created worlds where humans fell into idleness with no drive. One individual tried to turn humanity aside from this, and eventually collected a small group of people who renamed themselves after mythological figures, setting out to conquer the worlds to direct them outward again. Eventually this group chose to have their brains placed in life-support systems, allowing them to control robotic bodies, becoming 'cymeks' with no life-span limits. One of the cymeks, through lack of foresight, turned over too much of the operation of his domain to his AI servitors, which had been programmed with a desire for expansion. When he returned from a conquest, he found that the AI had taken over his capitol world and spread to other worlds; the cymeks were subjugated to the great AI Omnius by sheer force of numbers, and the AI conquered more and more worlds (sending out periodic memory updates to the AIs on other planets to keep them from diverging too far). It's at this point that The Butlerian Jihad starts, with the free humans defending themselves against the forces of the Synchronized Worlds, with the war continuing in The Machine Crusade (to be released in hardcover in about a week).
I was going to posit the exact opposite. If you look at most Science Fiction from the 50s or 60s, you see that people believed that technology would improve much more quickly than it did.
It's not that they believed that technology would advance much more quickly than it did, but that they misjudged the shape of technological advances, resulting in gross mismatches between the predicted technology and what we actually have.
For example, look at Heinlein's novel Starman Jones. In it, ships jump between stars at 'congruencies', where applying the right amount of thrust at the right time along the right vector causes you to jump 'around' folded threespace to another location lightyears away. But the navigation crew who guide the ship take manual readings from instruments, set up calculations using these readings by hand, use large 'secret' books that are essentially decimal-to-binary conversion tables to translate them into the format the computer understands, enter them into the computer, then translate the binary results from the computer back into decimal. We don't have the torch drive, or any of the sources of power described in the book -- but modern computer technology would be hooked directly into the navigation instruments, have the congruency locations stored in non-volatile memory, and all the 'navigator' would have to do is pick a congruency and specify what thrust to use getting there; the computer would handle all of the navigation functions itself.
If you think that's bad, you should see the state of science-fiction fans! PEEEEEE-U!
Unfortunately, 'fench' (a portmanteau word merging 'fen', the fannish plural of 'fan', and 'stench') has been with us for decades; it's not a recent change.
"Ahh, the heady air of a con." "Euuw. What's that stench?" "Smells like... gamers." -- Frank Cho, "Liberty Meadows"
Maybe geek shows are cancelled because they frequently cost as much or more per episode as mainstream show (due to special effects or whatever else), while having a lower viewership.
Also, the good geek shows actually expect the viewer to think about the plot and the characters, rather than just opening the top of your skull and pouring in the predigested 'entertainment experience'. The vast majority of viewers don't want to have to think about their entertainment, and are turned off by the show because of it. Also, shows like that are also more 'difficult' (read 'expensive') to write, because the production house's stable of writers just churn out the same pap with a new coat of paint, so they either pay for better writers, or use the schlock writers and watch the viewer rating tank because the scripts are dreck. And paying for better writers reduces their profits, even when they don't hogtie them by having them make changes to the plots and characters to turn the series into another schlocky clone of all the other programs...
I say, it was a good show. Not an outstanding show, but watchable, and IMHO it could have become a niche favorite a la Buffy had the PHBs at Fox given it the chance.
Firefly had the potential to be a good show. Unfortunately, like other shows that have relied on character development and plot, rather than flashy action or T&A, it was going to live or die on the strength of its scriptwriting... and Fox didn't let it tell a coherent story, then killed it even before lame scriptwriting could ruin it -- it was still trying to deepen the characterization, and hadn't been far enough along to slip into lame 'Old West in outer space' recycled plots.
Now you can not only call 911, you can send them a _picture_ of the guy that's mugging you and taking your expensive cellphone!
Laugh not; a recent kidnapping attempt (quick Google search for news story about the incident) had the teenager being accosted pull out his cellphone, take pictures of the attempted kidnapper and the license plate of the car he was driving, and ran off; the pictures were used by the police to identify the kidnapper.
And pens won't work either, because you can easily photocopy a signature and trace over it.
Actually, the anti-forgery algorithms that have been developed can tell when you're doing that; the better algorithms don't look for the outline of the letters you write (sign your name five times, and there will be differences between them), but instead look for the patterns of movement and speed while you're writing your name; if you're tracing a signature, your pen movements are going to be very different than the original signer's was.
1: Most domains have more than one MX. While the mailhost you connect to to verify may not have mail to send you, one of the other ones, or a private mailserver that is used by users but is not a gateway MX, may actually be trying to send mail to you.
The host that is trying to send mail is the host that gets verified; if there is a private mailserver trying to send mail out, it would need to have a resolvable hostname for the receiving mail server to use to connect back to it to fetch the mail it has.
2: Spammers could just use long, short lived, subdomain/host combinations. OK, you'd know what IP address the message came from, but you can generally figure that out now, and in fact, the spammer could add forged prior relay headers in front of their own anyway, just like they do now.
While a spammer could set up a complete system as a mailhost, it would still need to have a registered FQDN in order to be a valid mail forwarding host. What I proposed eliminates the problem of J. Random Spamhost connecting to a mailhost, claiming to be, say, mail.yahoo.com, and dumping several hundred thousand mail messages onto the Net; there still will have to be some verification process to keep the backtrail through prior systems from being forged.
The vast majority of spam is sent with some form of false address. Developing a way to be able to trust the origin of email is the way to end the spam crisis.
It's going to be functionally impossible to fix the problem of spammers opening an account and pumping email through it until it gets closed, but the transmission of email could be hardened by changing the SMTP protocol from 'call-up' to 'call-back'.
The SMTP protocol is set up to allow a host to contact another host and dump mail to it; there's no validation that the originating host is who it claims to be in the SMTP transaction. If you change the setup for the mail transfer connection to use the following mechanism:
Host A contacts host B and sends its FQDN (fully qualified domain name) and a request for a mail transfer connection
Host B performs a DNS lookup on the FQDN sent from host A and connects back to the host identified by the resolved FQDN. Hostnames that don't resolve, or which aren't in the FQDN form, are ignored.
Once the connection back to the originating site is established, the rest of the existing SMTP protocol transaction occurs. The sequence of validated hostnames would be processed into the 'Path:' mailheader, or another mailheader as determined when the protocol was updated.
This would establish a traceable chain of resolved hosts from the point at which the email entered the SMTP routing to its destination. Putting an email message into a mail transfer agent would still be vulnerable to the use of hacked or temporary accounts, but the upload would still require a trackable username and password for an account on the MTA. From that point, getting an MTA to accept an SMTP connection from a bogus host would require hacking the DNS server chain so that, when the receiving MTA host received the request, the IP address the passed hostname resolved to pointed back at the spammer's machine -- otherwise, you'd get a mail transaction sequence that looked like this:
Spam.com: Hello, [mta.com], [realhost.com] has mail to send. Mta.com: (resolves 'realhost.com') Mta.com: Hello, [realhost.com]; you have mail to send me. Realhost.com: [Mta.com], I don't have any mail to send you.
Not a panacea, but it would make the mail hop path trustable until you start seeing hacked mail daemons that would mangle the mail hop path of any mail going through it -- but that would still leave the host with the hacked daemon having to identify itself, from which it could be blocked.
The use of 'en' as a plural form actually is English usage, derived from the Germanic input into Old English, although its use is highly archaic now and survives only in a few words -- 'oxen', 'children', 'brethren' (which is mostly supplanted by 'brothers'), and 'kine' (plural of 'cow'). Formerly, hanging on into early modern English, you might also see eyen, shoon, hosen, and treen. The N suffix, had it survived in all the words it applied to in Old English, would also have given us namen, sunnen, moonen, starn, timen, churchen, hearten, tonguen, and ladyen.
The plurals for 'brother' and 'cow' have the vowel shift called 'umlaut', arising from a following vowel I or consonant J (Y) in the early Old English period, which disappeared after altering the previous vowel; the effect also appears in the plurals for 'man', 'tooth', 'mouse', and 'goose', among others.
If the '-en' plurals are annoying, then perhaps it's good that one of the classes of plurals that left no survivors added -U: we longer have 'shipu', 'weapnu', 'devlu', 'headu', or 'wondru'. Other classes of plurals also disappeared entirely because of the loss of final vowels. In Middle English the S ending absorbed almost all the others, resulting in the usages we have now, with its few holdouts.
Makes me wonder if I should append a "property of owner XYZ, please do not copy or link this article without permission"... at least to cover my ass in some form if such an article got in the wild.
If your site is on someone else's hostserver, you're pretty much SOL for doing anything about it, but if you are running on your own webserver, you should be able to go into the server configuration and configure your site so that articles are restricted by the HTTP-REFERER header field (admittedly, it can be spoofed, but it's better than nothing), so that your webserver denies direct linking to the articles unless you've specifically configured your server to allow linking to an article from a particular host, which you wouldn't do until you got a request to link.
Doing this can still get you slashdotted, because denying a connection still takes system resources, but you don't face the bandwidth costs of serving N+1 the article to an arbitrarily large number of requests. They would still be able to link to the root of your website and traverse the site to the article, but that's cumbersome enough that it's not practical to do for a link to the article such as those in most/. articles.
What's in a marketing campaign? That which we call a cowflop By any other name wouldst smell as vile.
(with apologies to Wm. Shakespeare, or whomever actually wrote it)
The MPAA is bemoaning the fact that modern technology is letting moviegoers find out that a new movie is crap many times faster than they were able to in the past. So what they are complaining about is that they're unable to hoover the money from as many people who will be unhappy at the amount of money they had to cough up to watch a bad movie before the public wises up and stays away in droves. But is this going to motivate the production houses to produce fewer big-budget flops and more good movies? It is to laugh.
"There has grown up in the minds of certain groups in this country the notion that because a man or corporation has made a profit out of the public for a number of years, the government and the courts are charged with the duty of guaranteeing such profit in the future, even in the face of changing circumstances and contrary public interest." -- Robert A. Heinlein
What you see, way too often, is a horse that looks pretty, but is completely screwed up in the head. And that's with traditional breeding (and I'm including the straw o'semen in the "traditional" category). I can only imagine the neurotic, unpredictable horses that will come from cloning the "best" show horses. They'll be useless for any actual work, probably won't be able to reproduce without assistance (already a problem today), and will be a danger to their rider and anyone nearby.
Some years ago, Scientific American had an article on the genetics of race horses. The thoroughbred racehorse has the best kept pedigree records of any form of livestock. The original founding stock was a mere 3 stallions and 40 mares; all racehorses alive today are descendants of those 43 horses, and every racehorse can have its pedigree traced back to the origin of the breed. Every racehorse is distantly or closely related to every other racehorse in the world. Racehorses must pass a "survival of the fittest" test, The Racecourse Test, and only the best horses are used for producing the next generation. If an expensive stallion proves to be a producer of poor stock, then he is banished from further breeding activity, despite his initial value; he is removed permanently from the breeding stock. Stallions of poor fertility are not persevered with, since they produce few offspring, and mare owners prefer to breed foals. Male thoroughbreds have a strong sex-drive and because of this, at least 90% of males are permanently excluded from procreating by castration. There are very few "genetic ailments" within the thoroughbred population, and weaklings are ruthlessly culled. One of the first use of cloning in racehorses is going to be its use to recover the genetic heritage of a gelding that proves to be a champion in its own right -- but only until that use demonstrates whether it's an effective way to recover the genetic line.
In order to seize your property, the federal government only requires probable cause that there is a nexus, or connection, between your property and a drug crime. Once the agents have probable cause they can seize your property. You have 10 days in which to post a bond equal in value to 10% of the property. If you fail to post the bond you lose your property, and you never get your day in court. Assuming you can afford to post the bond, the burden of proof is on you to prove that you didn't know about the connection between your property and the crime. The federal government uses the forfeited property to help fund the "drug war". Can you say 'conflict of interest'?
At this point, any further divergence onto the subject of asset forfeiture would be off-topic; if you're interested in more information, go to the F.E.A.R. website
The Microsoft link for the download is for the web-install version of 9.0b; you have to be connected to the Net to install it. I did some searching, and I found a download site for the redistributable full download of DirectX 9.0b here; it's a 32.6 Mb download, but you can put it on a CD or a fileserver, and you won't have everybody at your site pounding your Net connection to hit Microsoft for the pieces they need.
Interesting that in 3DMARK, the FX 5900 ran away with it. Hmmmm..
With all of the flap recently (referenced here, here, here, here, and here) regarding nVidia writing custom benchmark- and application-specific code into their drivers for the purpose of getting higher ratings, the value of benchmark ratings for evaluating video card performance is diminishing for benchmark software as it currently stands.
Perhaps what is needed is some kind of "drunkard's walk" scene traversal, where a scene is set up for rendering and, based on the sequence produced by a pseudo-random number generator, a traversal and view direction route is defined through the scene. Since the output from a pseudo-random number generator is deterministic given the initial seed, the same seed will produce the same route for testing different video cards, but a sufficiently large seed would produce so many different possible traversals of the scene that writing application-specific speedups would not be practical -- and a rigorous stochastic testing with different initial seeds would be able to average out any individual traversal variations. It would make the benchmarking process take longer, but it would produce results less vulnerable to special-case tweaking by the manufacturers.
And yes, it's what it sounds like. Ever used a belt sander and had the sander get away from you because you weren't holding it tightly enough? Well, get a bunch of people sitting around in a bar drunkenly arguing about whose belt sander can get away from them the fastest, and belt-sander racing is born.
We pay $20-$45 as it is for cable/satellite T.V. in order to watch a bunch of ads mixed in with the content.
We pay $20-$45 per month for access to the channels -- for the quality signal compared to a roof antenna -- not for the channel conent. The channels that we have to pay _extra_ for provide content that isn't 'edited for broadcast' or 'edited to fit in the time slot' or interrupted every twelve minute to tell us we smell bad.
Hopefully, this engine will be mod-able, as the original HL engine was
The previews explicitly describe this, and from the description, there's going to be a lot more automatic support in the engine for object properties than in the original:
Similarly, while the Half-Life 2 engine might not have floored us from a purely visual standpoint, we were definitely impressed with Valve's intentions to release the engine and the "Hammer" toolset to the mod community even before the release of the game itself. Valve explained that the engine is designed in such a way that objects can automatically acquire the properties of their textures. For example, a mod maker can take a rectangular object and apply a bricklike texture to it, and that object will then automatically function like a brick of that approximate size. It will sink if dropped in water, shatter if shot, make an appropriate noise if stepped on, and so on. Basically, the engine will allow mod makers to quickly and intuitively construct environments using a broad range of realistic surfaces and textures.
So instead of creating an object, texturing it, and then having to define all of its properties, it appears that you will create a texture and define the properties that objects with that texture possess, so that you only have to define properties once unless you want to special-case certain objects. More work for the setup process, but once you've got the textures laid out, much faster on the downside end of things.
As much as I will argue against the notion of the Internet as a lawless environment, the bottom line is that it is without borders, and spammers will easily be able to find an offshore haven from which to send their sexual enhancement ads. To assume that a US or UK law charging a per e-mail tax will somehow eliminate spam is unrealistic and unworkable.
The answer to that is to charge the $.01 per email back to the originating ISP if the email is received from a country that does not assess the email charge. Ultimately, though, there will have to be a fundamental alteration in the mechanics of processing SMTP mail in order to verify that the host from which mail is originating is the host it claims to be, thereby establishing a chaint of host authentication back to the originating host. This would still be able to be spoofed, but it would require more extensive work with the nameserver tree to do so.
It will also significantly reduce the incentive to use e-mail for appropriate means, such as operating an e-mail discussion list.
Which problem also isn't addressed by most of the other proposed solutions to the spam problem.
Slashdot Mirror
It would make the universe a poorer place to think that we had discovered all of the fundamental ways to manipulate it, and that all we can look forward to is ever more expensive ways of fiddling further and further to the right of the decimal point. The problem with truly revolutionary discoveries is that they are, by their very nature, impossible to search for directly -- you can't know what it is that you do not know. All we can do is keep pushing out the limits of what we do know until someone looks at a set of results for a hitherto-predictable test and says, "That shouldn't be happening", and then follows it out to find out why. Over and over again.
You mean besides winning a Locus award for Best Critic? Besides being book reviewer for Galaxy, Analog and New Destinies magazines for nearly a decade, and continuing to write occasional book reviews and a regular Op-Ed column, "Future Tense," for The Globe and Mail, Canada's national newspaper.? Nothing, I guess...
The people who voted to award him three Hugo awards (science fiction's top honor), a Nebula award, the John W. Campbell Award for Best New Writer, the E.E. "Doc" Smith Memorial Award (Skylark), the Pat Terry Memorial Award for Humorous Science Fiction, and a second Locus award for Best Novella would appear to disagree with you. But you can always define 'speculative fiction' to be whatever you want, and set up your definition to exclude what he writes.
You might want to look at some of Vernor Vinge's work. Much of his work has centered around the concept of 'the Singularity' -- a point at which we create, through technology, an entity of greater than human intelligence (whether machine intelligence, technologically-augmented humans, or bioengineered humans, or something else entirely), which results in an exponential runaway in the rate of technological advance, such that we can no longer even imagine what advances the future will bring. Dr. Vinge has written a paper on his view of the Singularity, and there is a critical discussion of the concept.
The point of science fiction, with the rate of technological advance going beyond where a writer can hope to have the background to predict in even the broadest terms over more than a short time, isn't about the technology itself, the way much of the space-opera of the '40s and '50s focussed on the 'gosh-wow' hardware, but on the way people deal with living in a world where it has become impossible to understand how things work -- and you begin to approach the embodiment of Clarke's Law: "Any sufficiently advanced technology is indistinguishable from magic." When everybody has a device you can pour dirt into and have it produce anything that can be made by recombining the atoms in the dirt, it changes the focus of people's drives -- but not the existence of those drives.
Recently out in paperback by Brian Herbert and Kevin J. Andersen is The Butlerian Jihad, fourth of the 'Dune prequels'. In the book (presented as background, so I'm not giving away plot), the advances in robotics created worlds where humans fell into idleness with no drive. One individual tried to turn humanity aside from this, and eventually collected a small group of people who renamed themselves after mythological figures, setting out to conquer the worlds to direct them outward again. Eventually this group chose to have their brains placed in life-support systems, allowing them to control robotic bodies, becoming 'cymeks' with no life-span limits. One of the cymeks, through lack of foresight, turned over too much of the operation of his domain to his AI servitors, which had been programmed with a desire for expansion. When he returned from a conquest, he found that the AI had taken over his capitol world and spread to other worlds; the cymeks were subjugated to the great AI Omnius by sheer force of numbers, and the AI conquered more and more worlds (sending out periodic memory updates to the AIs on other planets to keep them from diverging too far). It's at this point that The Butlerian Jihad starts, with the free humans defending themselves against the forces of the Synchronized Worlds, with the war continuing in The Machine Crusade (to be released in hardcover in about a week).
It's not that they believed that technology would advance much more quickly than it did, but that they misjudged the shape of technological advances, resulting in gross mismatches between the predicted technology and what we actually have.
For example, look at Heinlein's novel Starman Jones. In it, ships jump between stars at 'congruencies', where applying the right amount of thrust at the right time along the right vector causes you to jump 'around' folded threespace to another location lightyears away. But the navigation crew who guide the ship take manual readings from instruments, set up calculations using these readings by hand, use large 'secret' books that are essentially decimal-to-binary conversion tables to translate them into the format the computer understands, enter them into the computer, then translate the binary results from the computer back into decimal. We don't have the torch drive, or any of the sources of power described in the book -- but modern computer technology would be hooked directly into the navigation instruments, have the congruency locations stored in non-volatile memory, and all the 'navigator' would have to do is pick a congruency and specify what thrust to use getting there; the computer would handle all of the navigation functions itself.
Unfortunately, 'fench' (a portmanteau word merging 'fen', the fannish plural of 'fan', and 'stench') has been with us for decades; it's not a recent change.
"Ahh, the heady air of a con." "Euuw. What's that stench?" "Smells like... gamers."
-- Frank Cho, "Liberty Meadows"
Also, the good geek shows actually expect the viewer to think about the plot and the characters, rather than just opening the top of your skull and pouring in the predigested 'entertainment experience'. The vast majority of viewers don't want to have to think about their entertainment, and are turned off by the show because of it. Also, shows like that are also more 'difficult' (read 'expensive') to write, because the production house's stable of writers just churn out the same pap with a new coat of paint, so they either pay for better writers, or use the schlock writers and watch the viewer rating tank because the scripts are dreck. And paying for better writers reduces their profits, even when they don't hogtie them by having them make changes to the plots and characters to turn the series into another schlocky clone of all the other programs...
Firefly had the potential to be a good show. Unfortunately, like other shows that have relied on character development and plot, rather than flashy action or T&A, it was going to live or die on the strength of its scriptwriting... and Fox didn't let it tell a coherent story, then killed it even before lame scriptwriting could ruin it -- it was still trying to deepen the characterization, and hadn't been far enough along to slip into lame 'Old West in outer space' recycled plots.
Laugh not; a recent kidnapping attempt (quick Google search for news story about the incident) had the teenager being accosted pull out his cellphone, take pictures of the attempted kidnapper and the license plate of the car he was driving, and ran off; the pictures were used by the police to identify the kidnapper.
Actually, the anti-forgery algorithms that have been developed can tell when you're doing that; the better algorithms don't look for the outline of the letters you write (sign your name five times, and there will be differences between them), but instead look for the patterns of movement and speed while you're writing your name; if you're tracing a signature, your pen movements are going to be very different than the original signer's was.
The host that is trying to send mail is the host that gets verified; if there is a private mailserver trying to send mail out, it would need to have a resolvable hostname for the receiving mail server to use to connect back to it to fetch the mail it has.
While a spammer could set up a complete system as a mailhost, it would still need to have a registered FQDN in order to be a valid mail forwarding host. What I proposed eliminates the problem of J. Random Spamhost connecting to a mailhost, claiming to be, say, mail.yahoo.com, and dumping several hundred thousand mail messages onto the Net; there still will have to be some verification process to keep the backtrail through prior systems from being forged.
It's going to be functionally impossible to fix the problem of spammers opening an account and pumping email through it until it gets closed, but the transmission of email could be hardened by changing the SMTP protocol from 'call-up' to 'call-back'.
The SMTP protocol is set up to allow a host to contact another host and dump mail to it; there's no validation that the originating host is who it claims to be in the SMTP transaction. If you change the setup for the mail transfer connection to use the following mechanism:
This would establish a traceable chain of resolved hosts from the point at which the email entered the SMTP routing to its destination. Putting an email message into a mail transfer agent would still be vulnerable to the use of hacked or temporary accounts, but the upload would still require a trackable username and password for an account on the MTA. From that point, getting an MTA to accept an SMTP connection from a bogus host would require hacking the DNS server chain so that, when the receiving MTA host received the request, the IP address the passed hostname resolved to pointed back at the spammer's machine -- otherwise, you'd get a mail transaction sequence that looked like this:
Not a panacea, but it would make the mail hop path trustable until you start seeing hacked mail daemons that would mangle the mail hop path of any mail going through it -- but that would still leave the host with the hacked daemon having to identify itself, from which it could be blocked.
The use of 'en' as a plural form actually is English usage, derived from the Germanic input into Old English, although its use is highly archaic now and survives only in a few words -- 'oxen', 'children', 'brethren' (which is mostly supplanted by 'brothers'), and 'kine' (plural of 'cow'). Formerly, hanging on into early modern English, you might also see eyen, shoon, hosen, and treen. The N suffix, had it survived in all the words it applied to in Old English, would also have given us namen, sunnen, moonen, starn, timen, churchen, hearten, tonguen, and ladyen.
The plurals for 'brother' and 'cow' have the vowel shift called 'umlaut', arising from a following vowel I or consonant J (Y) in the early Old English period, which disappeared after altering the previous vowel; the effect also appears in the plurals for 'man', 'tooth', 'mouse', and 'goose', among others.
If the '-en' plurals are annoying, then perhaps it's good that one of the classes of plurals that left no survivors added -U: we longer have 'shipu', 'weapnu', 'devlu', 'headu', or 'wondru'. Other classes of plurals also disappeared entirely because of the loss of final vowels. In Middle English the S ending absorbed almost all the others, resulting in the usages we have now, with its few holdouts.
If your site is on someone else's hostserver, you're pretty much SOL for doing anything about it, but if you are running on your own webserver, you should be able to go into the server configuration and configure your site so that articles are restricted by the HTTP-REFERER header field (admittedly, it can be spoofed, but it's better than nothing), so that your webserver denies direct linking to the articles unless you've specifically configured your server to allow linking to an article from a particular host, which you wouldn't do until you got a request to link.
Doing this can still get you slashdotted, because denying a connection still takes system resources, but you don't face the bandwidth costs of serving N+1 the article to an arbitrarily large number of requests. They would still be able to link to the root of your website and traverse the site to the article, but that's cumbersome enough that it's not practical to do for a link to the article such as those in most
By any other name wouldst smell as vile.
(with apologies to Wm. Shakespeare, or whomever actually wrote it)
The MPAA is bemoaning the fact that modern technology is letting moviegoers find out that a new movie is crap many times faster than they were able to in the past. So what they are complaining about is that they're unable to hoover the money from as many people who will be unhappy at the amount of money they had to cough up to watch a bad movie before the public wises up and stays away in droves. But is this going to motivate the production houses to produce fewer big-budget flops and more good movies? It is to laugh.
Some years ago, Scientific American had an article on the genetics of race horses. The thoroughbred racehorse has the best kept pedigree records of any form of livestock. The original founding stock was a mere 3 stallions and 40 mares; all racehorses alive today are descendants of those 43 horses, and every racehorse can have its pedigree traced back to the origin of the breed. Every racehorse is distantly or closely related to every other racehorse in the world. Racehorses must pass a "survival of the fittest" test, The Racecourse Test, and only the best horses are used for producing the next generation. If an expensive stallion proves to be a producer of poor stock, then he is banished from further breeding activity, despite his initial value; he is removed permanently from the breeding stock. Stallions of poor fertility are not persevered with, since they produce few offspring, and mare owners prefer to breed foals. Male thoroughbreds have a strong sex-drive and because of this, at least 90% of males are permanently excluded from procreating by castration. There are very few "genetic ailments" within the thoroughbred population, and weaklings are ruthlessly culled. One of the first use of cloning in racehorses is going to be its use to recover the genetic heritage of a gelding that proves to be a champion in its own right -- but only until that use demonstrates whether it's an effective way to recover the genetic line.
Unless, of course, you fall afoul of the 'asset forfeiture' laws, which seize your property and charge it (as in "UNITED STATES of America, Plaintiff-Appellee, v. LAND, WINSTON COUNTY, Certain Real Property Located near Highway 195, Winston County, Alabama, together with all improvements, fixtures and appurtenances thereon"), requiring you to prove that the property was not obtained either illegally or with the profits of illegal activity.
In order to seize your property, the federal government only requires probable cause that there is a nexus, or connection, between your property and a drug crime. Once the agents have probable cause they can seize your property. You have 10 days in which to post a bond equal in value to 10% of the property. If you fail to post the bond you lose your property, and you never get your day in court. Assuming you can afford to post the bond, the burden of proof is on you to prove that you didn't know about the connection between your property and the crime. The federal government uses the forfeited property to help fund the "drug war". Can you say 'conflict of interest'?
At this point, any further divergence onto the subject of asset forfeiture would be off-topic; if you're interested in more information, go to the F.E.A.R. website
The Microsoft link for the download is for the web-install version of 9.0b; you have to be connected to the Net to install it. I did some searching, and I found a download site for the redistributable full download of DirectX 9.0b here; it's a 32.6 Mb download, but you can put it on a CD or a fileserver, and you won't have everybody at your site pounding your Net connection to hit Microsoft for the pieces they need.
With all of the flap recently (referenced here, here, here, here, and here) regarding nVidia writing custom benchmark- and application-specific code into their drivers for the purpose of getting higher ratings, the value of benchmark ratings for evaluating video card performance is diminishing for benchmark software as it currently stands.
Perhaps what is needed is some kind of "drunkard's walk" scene traversal, where a scene is set up for rendering and, based on the sequence produced by a pseudo-random number generator, a traversal and view direction route is defined through the scene. Since the output from a pseudo-random number generator is deterministic given the initial seed, the same seed will produce the same route for testing different video cards, but a sufficiently large seed would produce so many different possible traversals of the scene that writing application-specific speedups would not be practical -- and a rigorous stochastic testing with different initial seeds would be able to average out any individual traversal variations. It would make the benchmarking process take longer, but it would produce results less vulnerable to special-case tweaking by the manufacturers.
Damn, screwed up the links. Here they are again:
National Belt Sander Racing Association
International Belt Sander Drag Race Association
Tyrol Basin Belt Sander Races
North Liberty Belt Sander Races
You should go look at the places where they really race power tools:
National Belt Sander Racing Association International Belt Sander Drag Race Association Tyrol Basin Belt Sander Races North Liberty Belt Sander Races
And yes, it's what it sounds like. Ever used a belt sander and had the sander get away from you because you weren't holding it tightly enough? Well, get a bunch of people sitting around in a bar drunkenly arguing about whose belt sander can get away from them the fastest, and belt-sander racing is born.
The enemy of my enemy is my fall-guy.
We pay $20-$45 per month for access to the channels -- for the quality signal compared to a roof antenna -- not for the channel conent. The channels that we have to pay _extra_ for provide content that isn't 'edited for broadcast' or 'edited to fit in the time slot' or interrupted every twelve minute to tell us we smell bad.
The previews explicitly describe this, and from the description, there's going to be a lot more automatic support in the engine for object properties than in the original:
So instead of creating an object, texturing it, and then having to define all of its properties, it appears that you will create a texture and define the properties that objects with that texture possess, so that you only have to define properties once unless you want to special-case certain objects. More work for the setup process, but once you've got the textures laid out, much faster on the downside end of things.
The answer to that is to charge the $.01 per email back to the originating ISP if the email is received from a country that does not assess the email charge. Ultimately, though, there will have to be a fundamental alteration in the mechanics of processing SMTP mail in order to verify that the host from which mail is originating is the host it claims to be, thereby establishing a chaint of host authentication back to the originating host. This would still be able to be spoofed, but it would require more extensive work with the nameserver tree to do so.
Which problem also isn't addressed by most of the other proposed solutions to the spam problem.