IIRC, mine took about 20 minutes or so, on a Dell Precision 486DX2/66, with 256MB cache RAM, 20MB of RAM, Quantum 170MB IDE. That was 0.99something kernel, Slackware.
The rotational latency of new 5400 RPM disk drives is going to be the same today and tomorrow as it was in 1998, and the major portion of random access time. Luckily nonserver systems don't have as much of a problem with this, access being more sequential, so your point is still a good one.
There isn't any particular reason that fast drives cannot run cooler, though, with improvements. Except when spinning up, which is not much of the duty cycle. I have a pair of 10K WD Raptors that don't get very hot with a moderate airflow.
We know for certain, for instance, that for some reason, for some time in the beginning, there were hot lumps. Cold and lonely, they whirled noiselessly through the black holes of space [reverb effect here]. These insignificant lumps came together to form the first union, our sun, the heating system. And about this glowing gas bag, rotated the earth, a catseye among aggies, [reverb begins to really build] blinking in astonishment across the face... of time. [reverb overwhelmes everything, sounds disappear]
Why is it surprising that tall mountain ridges are found on small (relative to planets) moons, where there may be little weather and low gravity to cause their erosion?
It's the ban on pet sugar gliders that really gets me. What's do they expect to happen if a population of sugar gliders somehgow did established in the wild anyway? Their main food is the sap from incisions in the the dang eucalyptis trees from Australia spreading all over the place here, that extremely few native CA creatures benefit from, plus invertebrates found under dead tree bark, nectar, and pollen.
Come to think of it, maybe the florescent zebra danios never really had a chance...
I thought it was quite interesting to read the quote. Rather frightening also to realize that a Game and Fish Commission is actually willing to admit they don't care about science, and feel free to act as martinets imposing prejudices they can't support with facts on the largest state population in the United States. This is on a par with Italy's banning pet spiders for no better reason than the current head of state there having a phobia about them.
Luckily most of the rest of the United States, at least, is less tolerant of this type of crap, and people with similar views to the commissioner quoted can exercise them by not buying the fish involved - or, if they choose, not buying anything from the pet shops that carry them, picketing such shops, and so on, as is their right under the set of liberties recognized outside the bottom part of the Left Coast.
In this case pet shops can still apparently display these beautiful genengineered Zebrafish, bought from outside the state. Along with signs noting the reasons (or lack of reason) that they cannot be sold, and petition forms for recalling commissioners, for laws requiring scientific evaluations in such cases, and so on. Hint, hint.
I gather that the theory's difference it that it insists that, on the contrary, there is not a single instant that is the "beginning" or "end" of an event - they have a duration, as well as the event itself.
IANAL, but as for the "if you are in California" part, I do recall reading of a California labor regulation that requires giving a person at least one day off a week, and specifically a day off after each six days working (the last presumably to prevent an employer on Sunday telling a worker his day off has been moved from Monday to Saturday). The law states the employer is otherwise guilty of a misdemeanor.
And it goes on for a while after that, to eventually state the rewatd for being a registered geologist:
(b) Each specialty geologist certified under this chapter may, upon certification, obtain a seal of the design authorized by the board bearing the registrant's name, number of his certificate and the legend "certified specialty geologist. "
So the question would seem to be, is this company fraudulently displaying a certificate with the legend "certified specialty geologist"? How exactly does this forbid practicing geology without a license?
I am by no means a fan of RedHat (Slack for me), but I think that they have a legit and smart model here. You're essentially buying support when you buy their (reasonably priced) distribution. If you don't like it, then don't pay them.
Yeah, I've bought a lot of boxed sets of Red Hat. Actually, what you were buying changed over time. When I started, it came with some nice extras that made it worth it, 3rd party X server with an ability do a multiheaded display which XFree86 has only recently matched for instance, BRU backup software, and so on. While the price of the succeeding boxed sets rose, the amount of added value Red Hat contributed dropped. The licensed programs disappeared and were replaced with trial versions disabled after an evaluation period. The manuals shrank and lost detail. The number of CDROMS grew, but probably not as fast as the cost of producing them has dropped.
Still, there is that support, and indeed eventually that's all buying the sets really got you. Or at least, you thought it did. With that support, what you paid for was the implied assurance that gotten past that 8.0 install, you could reasonably expect little trouble from your Red Hat systems, for the next few years, provided you were concientious with updates.
Well, at least that's what you *thought* you paid for. Now, just to keep your systems minimally safe and under your own control (assiming internet connections), you are going to have to pay for it again and again, every year.
Kind of messes up your users when they need to access a URL that specifies a port number (other than the default 80), doesn't it?
Such cases are rare and you can adjust your firewall to open holes for those must-access services (e.g., www.im-running-a-warez-server-on-my-cable-modem.co m:8181).
Ooo, you *really* don't like the idea of non-root users being able to run server programs, without kissing your butt, do you? I'd be sure to consider you if I needed someone for the job of Preventer of Information Services, except that I'd have questions about why you seem to have no problems with warez servers on port 80. You're aiming at the wrong target, the port and not what the service is.
Don't try to tell me I have an obligation to shove it down the throats of mine.
You have an obligation to see that your network is not used to propagate worms and participate in DDOS attacks. If you can do that without a firewall, good for you. But if your users piss on my networks, I'll do everything I can to get your upstream provider to shut you down until you resolve the problem.
Shrug. I've worked in the role of resolving threats like that for a backbone provider, everything from legitimate problems like those caused by this worm, to idiots like those livid because they installed their new wonderful protective software and noticed someone maliciously pinging them four whole times, to the guy who was going to sue us for not cutting off someone offering a non-http service on port 80 he didn't want to let his users access.
No way in hell are you going to get that "upstream provider" of Internet services to impose a "you gotta, in general, firewall outgoing ports from your users' network" condition on their customers. Even if you happened to be a larger customer of the same provider it would probably violate the service contract. You're blowing steam here.
Kind of messes up your users when they need to access a URL that specifies a port number (other than the default 80), doesn't it? I don't recall any RFC that *requires* TCP/UDP services in general to use particular ports, so this cuts users off from a potentially good part of the Internet.
Fine for your own users.
Don't try to tell me I have an obligation to shove it down the throats of mine.
I care. It makes it a royal pain in the ass to change nameservers' IP addresses.
The problem arises when the name server is not necessarily in the same domain as the DNS name you are looking up. I.E., nameserver for "fie.com" may be "poorbastard.foo.com". The root nameserver, asked for "host.fie.com", returns the IN A for poorbastard from their "fie.com" records. This means that in order to change either the IP address or the DNS name of "poorbastard.foo.com", you must go through a laborous process, fie.com and foo.com must have their DNS records changed multiple times:
1. administrator of fie.com must ask their domain registrar to remove poorbastard.foo.com as the nameserver of fie.com - otherwise the domain registrar will reject the change requested by foo.com because "another domain uses the DNS server". Note that this requires fie.com to replace poorbastard with another name server to maintain the minimum number of nameservers for a domain record. Otherwise the registrar will reject this request from fie.com.
After this change has propagated to the other root name servers, then:
2. Admin of foo.com must ask their registrar to change poorbastard's IP address in both the root name server's record and themselves change foo.com's own DNS servers' record. After this has propagated:
3. Admin of fie.com must ask their registrar to change fie.com's root DNS records back to poorbastard, using the new IP address.
Consider the case where you are greatbigisp.net and "ns1.greatbigisp.net" is the server - even just a secondary name server - that handled dozens, even hundreds, of customer domains for their customer. How do you go about changing the IP of ns1?
Answer: Effectively you can't. You're totally fucked.
Of course this could be easily fixed a couple of different ways - first, by only keeping one IP address definition for a nameserver listed as a primary/secondary for a domain on a root server. instead of keeping multiple definitions in every domain's records - looking up the address of poorbastard in foo.com's record when fie.com is being resolved - or alternately, when foo.com asks to change poorbastard's ip address, the registrar could change it in fie.com's record too, and all other places it occurs.
Why is this not done? According to the explanation I got from Network Solutions (when it was still "Network Solutions") the first method would require - gasp - another database lookup to get the IP - unacceptable! The second, I was told brusquely, is not allowed because it would allow foo.com to change fie.com's record, evidently considered a terrible thing - even though fie.com has no business whatsoever determining the IP address of foo.com's nameserver.
Frankly my suspician is that they were just being lazy as hell, and don't want to bother changing anything if they can come up with an excuse, given the obnoxious and absurd reasons used to arbitrarily reject DNS changes. Such as attempting to reply to their confirmation email when you have a "Reply-to:" address, for chrissake.
Perhaps you are mistaking how the root name servers are *supposed* to work, for how they *do* work.
For example, given that the domain "domain.tld" owns name servers x, y, and z:
If you submit a request for "x.domain.tld" to the root nameserver for the "tld" top level domain, you should, according to all the RFCs I have seen, get the IP address of the nameservers defined for domain.tld from the root server. Using this, your resolver should then query the nameservers for that domain to find out what the IP address of your "x" host is.
However, what actually *does* happen is that the root nameserver short-circuits the process and simply returns the IP address of x.domain.tld itself; the dnsserver for domain.tld is never even touched, nor can it cache.
If a terrorist organization chose to deliberately release the plague or smallpox in an urban area, millions would die even with the tremendous quality of care available in the United States. Hundreds of millions would die if such a release occurred in Asia, in the crowded cities of India or China.
No, what would happen if smallpox were released would be hundreds of deaths in places with the quality of care available in the United States, mostly caused by vaccinia side-effects. Probably followed by millions of deaths in the poorer third world countries if, as is likely, it spread there before being recognized.
Bubonic plague is endemic to most of the world and is both curable and easily suppressed in a variety of ways ranging from halfway decent sanitation to flea powder.
The big surprise is likely to be a perfectly natural influenza virus pandemic.
Unfortunately, once you have taken the opposite stand you're on the other slippery slope towards banning birth control, penalizing people for not having enough children, et cetra, aren't you? Or even falling over the mental cliff of declaring morality trumps logic and the math of compound-interest population growth, in the worst case.
Slashdot Mirror
Ah, so evolution == Survivors survive!
IIRC, mine took about 20 minutes or so, on a Dell Precision 486DX2/66, with 256MB cache RAM, 20MB of RAM, Quantum 170MB IDE. That was 0.99something kernel, Slackware.
The rotational latency of new 5400 RPM disk drives is going to be the same today and tomorrow as it was in 1998, and the major portion of random access time. Luckily nonserver systems don't have as much of a problem with this, access being more sequential, so your point is still a good one.
There isn't any particular reason that fast drives cannot run cooler, though, with improvements. Except when spinning up, which is not much of the duty cycle. I have a pair of 10K WD Raptors that don't get very hot with a moderate airflow.
I think it is the remains of the ancient circum-Iapetus particle accelerator.
Why is it surprising that tall mountain ridges are found on small (relative to planets) moons, where there may be little weather and low gravity to cause their erosion?
it doesn't seem to affect Lynx.
I'm interested in the possibility of a competition writing programs that would do the hunting for you.
Think of it - who can do the best open source cybernetic sniper program? Remember those neat antipersonnel guns in Aliens?
Well, yeah. The infection, worm, virus, SQL script, whatever is itself third-party software.
All a matter of semantics - just a nonanswer.
As for the "our own fault" part, that wasn't actually in the article.
Damnit, I want to see a film version of Bored of the Rings.
It's the ban on pet sugar gliders that really gets me. What's do they expect to happen if a population of sugar gliders somehgow did established in the wild anyway? Their main food is the sap from incisions in the the dang eucalyptis trees from
Australia spreading all over the place here, that extremely few native CA creatures benefit from, plus invertebrates found under dead tree bark, nectar, and pollen.
Come to think of it, maybe the florescent zebra danios never really had a chance...
I thought it was quite interesting to read the quote. Rather frightening also to realize that a Game and Fish Commission is actually willing to admit they don't care about science, and feel free to act as martinets imposing prejudices they can't support with facts on the largest state population in the United States. This is on a par with Italy's banning pet spiders for no better reason than the current head of state there having a phobia about them.
Luckily most of the rest of the United States, at least, is less tolerant of this type of crap, and people with similar views to the commissioner quoted can exercise them by not buying the fish involved - or, if they choose, not buying anything from the pet shops that carry them, picketing such shops, and so on, as is their right under the set of liberties recognized outside the bottom part of the Left Coast.
In this case pet shops can still apparently display these beautiful genengineered Zebrafish, bought from outside the state. Along with signs noting the reasons (or lack of reason) that they cannot be sold, and petition forms for recalling commissioners, for laws requiring scientific evaluations in such cases, and so on. Hint, hint.
I gather that the theory's difference it that it insists that, on the contrary, there is not a single instant that is the "beginning" or "end" of an event - they have a duration, as well as the event itself.
IANAL, but as for the "if you are in California" part, I do recall reading of a California labor regulation that requires giving a person at least one day off a week, and specifically a day off after each six days working (the last presumably to prevent an employer on Sunday telling a worker his day off has been moved from Monday to Saturday). The law states the employer is otherwise guilty of a misdemeanor.
I was thinking, "LA has so many already, seems like a lot of concern over this fault threatening to rip it a new one."
And it goes on for a while after that, to eventually state the rewatd for being a registered geologist:
(b) Each specialty geologist certified under this chapter may, upon certification, obtain a seal of the design authorized by the board bearing the registrant's name, number of his certificate and the legend "certified specialty geologist. "
So the question would seem to be, is this company fraudulently displaying a certificate with the legend "certified specialty geologist"? How exactly does this forbid practicing geology without a license?
IIRC in the Bay area here, there was a scandal a couple years back about an unqualified person serving corporations as a feng shui consultant.
Yeah, I've bought a lot of boxed sets of Red Hat. Actually, what you were buying changed over time. When I started, it came with some nice extras that made it worth it, 3rd party X server with an ability do a multiheaded display which XFree86 has only recently matched for instance, BRU backup software, and so on. While the price of the succeeding boxed sets rose, the amount of added value Red Hat contributed dropped. The licensed programs disappeared and were replaced with trial versions disabled after an evaluation period. The manuals shrank and lost detail. The number of CDROMS grew, but probably not as fast as the cost of producing them has dropped.
Still, there is that support, and indeed eventually that's all buying the sets really got you. Or at least, you thought it did. With that support, what you paid for was the implied assurance that gotten past that 8.0 install, you could reasonably expect little trouble from your Red Hat systems, for the next few years, provided you were concientious with updates.
Well, at least that's what you *thought* you paid for. Now, just to keep your systems minimally safe and under your own control (assiming internet connections), you are going to have to pay for it again and again, every year.
Ooo, you *really* don't like the idea of non-root users being able to run server programs, without kissing your butt, do you? I'd be sure to consider you if I needed someone for the job of Preventer of Information Services, except that I'd have questions about why you seem to have no problems with warez servers on port 80. You're aiming at the wrong target, the port and not what the service is.
You have an obligation to see that your network is not used to propagate worms and participate in DDOS attacks. If you can do that without a firewall, good for you. But if your users piss on my networks, I'll do everything I can to get your upstream provider to shut you down until you resolve the problem.
Shrug. I've worked in the role of resolving threats like that for a backbone provider, everything from legitimate problems like those caused by this worm, to idiots like those livid because they installed their new wonderful protective software and noticed someone maliciously pinging them four whole times, to the guy who was going to sue us for not cutting off someone offering a non-http service on port 80 he didn't want to let his users access.
No way in hell are you going to get that "upstream provider" of Internet services to impose a "you gotta, in general, firewall outgoing ports from your users' network" condition on their customers. Even if you happened to be a larger customer of the same provider it would probably violate the service contract. You're blowing steam here.
Kind of messes up your users when they need to access a URL that specifies a port number (other than the default 80), doesn't it? I don't recall any RFC that *requires* TCP/UDP services in general to use particular ports, so this cuts users off from a potentially good part of the Internet.
Fine for your own users.
Don't try to tell me I have an obligation to shove it down the throats of mine.
Yes, that describes the problem. :)
Where does it say they are allowed/supposed to do that?
I care. It makes it a royal pain in the ass to change nameservers' IP addresses.
The problem arises when the name server is not necessarily in the same domain as the DNS name you are looking up. I.E., nameserver for "fie.com" may be "poorbastard.foo.com". The root nameserver, asked for "host.fie.com", returns the IN A for poorbastard from their "fie.com" records. This means that in order to change either the IP address or the DNS name of "poorbastard.foo.com", you must go through a laborous process, fie.com and foo.com must have their DNS records changed multiple times:
1. administrator of fie.com must ask their domain registrar to remove poorbastard.foo.com as the nameserver of fie.com - otherwise the domain registrar will reject the change requested by foo.com because "another domain uses the DNS server". Note that this requires fie.com to replace poorbastard with another name server to maintain the minimum number of nameservers for a domain record. Otherwise the registrar will reject this request from fie.com.
After this change has propagated to the other root name servers, then:
2. Admin of foo.com must ask their registrar to change poorbastard's IP address in both the root name server's record and themselves change foo.com's own DNS servers' record. After this has propagated:
3. Admin of fie.com must ask their registrar to change fie.com's root DNS records back to poorbastard, using the new IP address.
Consider the case where you are greatbigisp.net and "ns1.greatbigisp.net" is the server - even just a secondary name server - that handled dozens, even hundreds, of customer domains for their customer. How do you go about changing the IP of ns1?
Answer: Effectively you can't. You're totally fucked.
Of course this could be easily fixed a couple of different ways - first, by only keeping one IP address definition for a nameserver listed as a primary/secondary for a domain on a root server. instead of keeping multiple definitions in every domain's records - looking up the address of poorbastard in foo.com's record when fie.com is being resolved - or alternately, when foo.com asks to change poorbastard's ip address, the registrar could change it in fie.com's record too, and all other places it occurs.
Why is this not done? According to the explanation I got from Network Solutions (when it was still "Network Solutions") the first method would require - gasp - another database lookup to get the IP - unacceptable! The second, I was told brusquely, is not allowed because it would allow foo.com to change fie.com's record, evidently considered a terrible thing - even though fie.com has no business whatsoever determining the IP address of foo.com's nameserver.
Frankly my suspician is that they were just being lazy as hell, and don't want to bother changing anything if they can come up with an excuse, given the obnoxious and absurd reasons used to arbitrarily reject DNS changes. Such as attempting to reply to their confirmation email when you have a "Reply-to:" address, for chrissake.
Perhaps you are mistaking how the root name servers are *supposed* to work, for how they *do* work.
For example, given that the domain "domain.tld" owns name servers x, y, and z:
If you submit a request for "x.domain.tld" to the root nameserver for the "tld" top level domain, you should, according to all the RFCs I have seen, get the IP address of the nameservers defined for domain.tld from the root server. Using this, your resolver should then query the nameservers for that domain to find out what the IP address of your "x" host is.
However, what actually *does* happen is that the root nameserver short-circuits the process and simply returns the IP address of x.domain.tld itself; the dnsserver for domain.tld is never even touched, nor can it cache.
Bubonic plague is endemic to most of the world and is both curable and easily suppressed in a variety of ways ranging from halfway decent sanitation to flea powder.
The big surprise is likely to be a perfectly natural influenza virus pandemic.
Unfortunately, once you have taken the opposite stand you're on the other slippery slope towards banning birth control, penalizing people for not having enough children, et cetra, aren't you? Or even falling over the mental cliff of declaring morality trumps logic and the math of compound-interest population growth, in the worst case.