I personnally only learned recently of the awfull (and criminal)
turkey shooting on the highway of death that happend in gulf war I. I don't think we'll know any detail (that is outside the outcome of war) before very long.
Re:What if another coutry did the same ?
on
Strike on Iraq
·
· Score: 1
You're damn right. Expect Taiwan to be a threat for China anytime soon.;-(
Re:Support our troops. Bring them home NOW!
on
Strike on Iraq
·
· Score: 1
Same old slogan, since VietNam, still true. Don't put them in harm's way and then 'support' for them.
The memory subsystem is one leg of the northbridge (center of the chipset), (two channels allows the chipset to double the bandwidth, but not the latency)
The CPU(s) sit on another bus.
The PCI busses are interconnected through HUBs and specilised links. With this kind of architecture, you can reach 4 times 400 MB/s (1.6GB/s agregate) using the busses in PCI64bits/66MHz). Even better can be expected with PCI-X interfaces.
About the address tricks, you can do that kind of things, but in this case, expect to have to write many things ad-hoc, and forget the general-purpose side of your system. You usually want a real-time system, and I see no point in doing that for a simple web server. RADAR systems, avionics, and stuff like that can be expected to use that kind of trick and optimisations a lot (lots of processing done, and in a very systematic way). 3D rendering seems a nice application for that as well, but I don't know what the state of the art is for high quality (movies) rendering.
The hard part of Web sites is usually database access, which implies complex algorithm that don't fit well in specialised hardware. Compression is I think anacdotis in a web server.
You won't be gzipping faster that the bandwidth, which is the bottleneck (let's assume you double bandwidth with gziping). Usually, serving a lot of requests involves a load balancer plus many machines serving the actual request, because the serving is complex. The gzipping will be neglectable I think. It could also be a task devoted to the load balancer itself, if load permits (on the other hand, the load balancer is critical).
The general trend in the industry goes to non-intelligent interconnections (Gigabit card used to have a processor (Alteon), they don't anymore (see latest intels)). I2O never took off because you don't really need to relieve a computer from computation when your computation power is pletoric.
On a Xeon 2.8GHz, I just got 71 MB/s for gzip.
What's the use for such hardware then?
Plus it will eat the PCI bus because data has to go out of memory to processing card, back to memory, then to network card. You triple the PCI bus bandwidth. (Not true if the compression is embedded in the network card).
I used to run an engineering school's dorm network from 95-97. We enjoyed the school's connection, which was then dispatched to all rooms.
We were students ourselves, but acountable for what happend on the residence, and the school had one plug to pull to shut us out of internet (literally).
Our most improtant protection was simply about having people responsible. We had them sign a check of a big enough amount for a student that we would bring to the bank it they broke the loaned ethernet adapter of if they screwed up a lot.
So basically, pressure on us got translated to the students by ourselve.
By that time, security hole weren't that a big issue, and we just had to disconnect the visible warez. We also monitored MAC/IP matching, so we could pull the plug of the computer that fucked up easily, but that implied active monitoring (Ahhh the face of the guy to whom you say: "You were right changing your password today, but I found the previous one nicer...")
On the other hand, it's not so easy for the schools to pull the plug, because on the long run, we were saving them the money from upgrading the school's computers. (I guess now in the USA having a computer is a requirement for the students).
Surface pressure: 6.36 mb at mean radius (variable from 4.0 to 8.7 mb depending on season) [6.9 mb to 9 mb (Viking 1 Lander site)] Surface density: ~0.020 kg/m3 Scale height: 11.1 km Total mass of atmosphere: ~2.5 x 1016 kg Average temperature: ~210 K (-63 C) Diurnal temperature range: 184 K to 242 K (-89 to -31 C) (Viking 1 Lander site) Wind speeds: 2-7 m/s (summer), 5-10 m/s (fall), 17-30 m/s (dust storm) (Viking Lander sites) Mean molecular weight: 43.34 g/mole Atmospheric composition (by volume): Major : Carbon Dioxide (CO2) - 95.32% ; Nitrogen (N2) - 2.7% Argon (Ar) - 1.6%; Oxygen (O2) - 0.13%; Carbon Monoxide (CO) - 0.08% Minor (ppm): Water (H2O) - 210; Nitrogen Oxide (NO) - 100; Neon (Ne) - 2.5; Hydrogen-Deuterium-Oxygen (HDO) - 0.85; Krypton (Kr) - 0.3; Xenon (Xe) - 0.08
So we're talking carbon dioxide. Pressure is 7mb or 7hPa or 0.7kPa (earth pressure beeing around 1000hPa or 100kPa)
One thing I always wondered is why the hell rivers have to be water on mars.
Mars's surface temperature goes down pretty low at night to some -100 degree Celcius, at which nitrogen (roughly our air) is liquid as well (at earth ground pressures).
Can't all those riverbed come from other liquid that only flow at night time and vaporize during daytime. As we only observe the daytime mars, the "water" is always gone.
Using IMAP, with a decent email reader without preview allows you to download only the headers of the SPAMs, killing them at first sight, and proceeding through your legitimate emails.
If you're using a shitty and expensive internet connection in some remote place, you can make the effort not to download you whole mailbox.
Using fossil fuels in big plants/power stations is very likely to be cleaner and incredibly more efficient than having thousands of our little engines like today. Add to that that you can add expensive devices to clean up the fumes from your factory with no comparision to what is economically viable on the butt of a car.
This also paves the way to an easy shift to cleaner energies. Once every layman relies on hydrogen, it's easier to convert the plant.
In one word, you will concentrate your problem to a single point (the hydrogen generation industry) whereas having the problem with every of us. That's actually the hardest step.
No, no, you didn't get my point. Plentyfull, time to go, etc. is bull shit and no backed by figures. You say that we won't empty the bottle fast enough to feed our thist, I'm telling you the bottle will run DRY.
Figure 1: World Crude Oil and Natural Gas Reserves, January 1, 2002 (PDF) (Go directly to the bottom line of last page). World reserves, as of 1st january 2002: 1,018.7 billion barrels.
Figure 2: World Petroleum Consumption (idem) 77,125 thousand barrels per day.
So years to go at CURRENT RATE before there's no oil left on Earth:
1.0187E12/77.125E6/365=36.19 years.
So at current rate of consumption, all oil on Earth will be burnt in 36 year. As we extract only roughly 60% of the reserves due to technical reasons, it boils down to 21.7 years. You've got that long to buy a bike, if the consumption doen't skyrocket as China's life standards increases.
I reply to you again, on another point. You might want to add the oil gives chemical industry, which produces plastic and medecines, so in 20 years (se my other post about the deadline), we might have a bid health problem if the chemical industry gets broken.
Now you know why Bush goes for oil. (As an oil man, he knows that).
On the other hand, we can grow crop that yield hydrocarbons, but farming requires fuel, etc...
Kind of a vicious circle we're in, as you point out.
We've got 30 year of oil to go maximum. I made the calculation 2 week ago. I think this figure is broadly overestimated (yes, over).
It's basically all world proven reserves of oil (1000 Bilion barels) divided by the world daily consumption (75 milion barrel per day).
That brings us to 36,5 years (2040).
BUT the reserves can't be fully extracted (usually only 60%) and the consumption is likely to go up, so 20-25 years seems more likely, or even less. (ouch, isn't it?)
Yes my experience is the same in many cases. In one defense company, the only internet-connected machine of a 1000 people sized site was a few machines in the library.
And anyway in a major computer manufacturer's network, you didn't see much of internet except through the web proxy and soxyfied telnets. That's of course the way to go.
If you want real security, you are likely not to want a machine connected to the main power lines as well (tempest protection). I guess an off line UPS does the job.
The map posted in this comment
http://www.state.nv.us/nucwaste/yucca/seismo01.h tm
reminds me of an area I already read about...
Did anyone spot that Yucca Mountain is conveniently located in the highly secured Nellis Air Force Range, just north of Armagosa Valley. Funny this area is also the world famous Area 51! (home of the stealth planes^H^H^H^H^H^H^H^H^Haliens)
Compare with this map :
http://www.ufomind.com/area51/orgs/nellis/articl es/1996/sun_map.jpg
You don't need it. The demanding IO is on the PCI bus (North). I don't think the south bridge can feed more than 100MB/s. See http://www.amd.com/products/cpg/server/athlon/chip set.html. It's only IDE (how much can the HDD MECANICS give you) plus USB. No big deal.
"the LVM included with HP-UX is actually a licensed Veritas VxFS."
This is false. You mix up filesystem and LVM. LVM is a software layer between physical devices (physical volumes) and block device (logical volumes). The logical volume is growable and shrinkable. But wether you can grow a mounted partition or not depends on you FILESYSTEM. So it just depends on you ext2/ext3 filesystem.
On HPUX VxFS (also called JFS for jourmaling is growable, but I don't remember if it's shrinkable).
Re:This it intended to stop dual booting Linux/BSD
on
Copyrant
·
· Score: 1
You're not forced to repartition your drive in order to have a runing linux. See Loopback Root FS Mini Howto. I don't pretend it's easy, but it's even technically feasible to have a distro installing that way.
Slashdot Mirror
I personnally only learned recently of the awfull (and criminal) turkey shooting on the highway of death that happend in gulf war I. I don't think we'll know any detail (that is outside the outcome of war) before very long.
You're damn right. Expect Taiwan to be a threat for China anytime soon. ;-(
Same old slogan, since VietNam, still true. Don't put them in harm's way and then 'support' for them.
On current PCI architectures, you already have that implemented.
Here is the description of the Serverworks chipset (Scroll down to the drawings) Intel's (e7500/7501) is very similar, in architecture at least.
The memory subsystem is one leg of the northbridge (center of the chipset), (two channels allows the chipset to double the bandwidth, but not the latency)
The CPU(s) sit on another bus.
The PCI busses are interconnected through HUBs and specilised links. With this kind of architecture, you can reach 4 times 400 MB/s (1.6GB/s agregate) using the busses in PCI64bits/66MHz). Even better can be expected with PCI-X interfaces.
About the address tricks, you can do that kind of things, but in this case, expect to have to write many things ad-hoc, and forget the general-purpose side of your system. You usually want a real-time system, and I see no point in doing that for a simple web server. RADAR systems, avionics, and stuff like that can be expected to use that kind of trick and optimisations a lot (lots of processing done, and in a very systematic way). 3D rendering seems a nice application for that as well, but I don't know what the state of the art is for high quality (movies) rendering.
The hard part of Web sites is usually database access, which implies complex algorithm that don't fit well in specialised hardware. Compression is I think anacdotis in a web server.
You won't be gzipping faster that the bandwidth, which is the bottleneck (let's assume you double bandwidth with gziping). Usually, serving a lot of requests involves a load balancer plus many machines serving the actual request, because the serving is complex. The gzipping will be neglectable I think. It could also be a task devoted to the load balancer itself, if load permits (on the other hand, the load balancer is critical).
Could be worse...
I think you should expect Sun's US branch to shrink and Indian branch to grow accordingly if Sun loses the case.
Indian enginners are said to be quite good and damn cheep if I recall well.
As the booming years are away, we can expect tech industry to go down on its costs. Like an other industry would.
The general trend in the industry goes to non-intelligent interconnections (Gigabit card used to have a processor (Alteon), they don't anymore (see latest intels)). I2O never took off because you don't really need to relieve a computer from computation when your computation power is pletoric.
On a Xeon 2.8GHz, I just got 71 MB/s for gzip.
What's the use for such hardware then?
Plus it will eat the PCI bus because data has to go out of memory to processing card, back to memory, then to network card. You triple the PCI bus bandwidth. (Not true if the compression is embedded in the network card).
I used to run an engineering school's dorm network from 95-97. We enjoyed the school's connection, which was then dispatched to all rooms.
We were students ourselves, but acountable for what happend on the residence, and the school had one plug to pull to shut us out of internet (literally).
Our most improtant protection was simply about having people responsible. We had them sign a check of a big enough amount for a student that we would bring to the bank it they broke the loaned ethernet adapter of if they screwed up a lot.
So basically, pressure on us got translated to the students by ourselve.
By that time, security hole weren't that a big issue, and we just had to disconnect the visible warez. We also monitored MAC/IP matching, so we could pull the plug of the computer that fucked up easily, but that implied active monitoring (Ahhh the face of the guy to whom you say: "You were right changing your password today, but I found the previous one nicer...")
On the other hand, it's not so easy for the schools to pull the plug, because on the long run, we were saving them the money from upgrading the school's computers. (I guess now in the USA having a computer is a requirement for the students).
OK one step further: Martian Atmosphere
So we're talking carbon dioxide. Pressure is 7mb or 7hPa or 0.7kPa (earth pressure beeing around 1000hPa or 100kPa)
Here's a phase diagram of CO2
So at such low pressures, CO2 is vapor at diurnal temperature ranges. My theory seems not to hold. Please go back to sleep.
One thing I always wondered is why the hell rivers have to be water on mars.
Mars's surface temperature goes down pretty low at night to some -100 degree Celcius, at which nitrogen (roughly our air) is liquid as well (at earth ground pressures).
Can't all those riverbed come from other liquid that only flow at night time and vaporize during daytime. As we only observe the daytime mars, the "water" is always gone.
Anybody have an idea about that?
Using IMAP, with a decent email reader without preview allows you to download only the headers of the SPAMs, killing them at first sight, and proceeding through your legitimate emails.
If you're using a shitty and expensive internet connection in some remote place, you can make the effort not to download you whole mailbox.
Agreed, but with a caveat(tm) ;-)
Using fossil fuels in big plants/power stations is very likely to be cleaner and incredibly more efficient than having thousands of our little engines like today. Add to that that you can add expensive devices to clean up the fumes from your factory with no comparision to what is economically viable on the butt of a car.
This also paves the way to an easy shift to cleaner energies. Once every layman relies on hydrogen, it's easier to convert the plant.
In one word, you will concentrate your problem to a single point (the hydrogen generation industry) whereas having the problem with every of us. That's actually the hardest step.
You say that we won't empty the bottle fast enough to feed our thist, I'm telling you the bottle will run DRY.
Figure 1:
World Crude Oil and Natural Gas Reserves, January 1, 2002 (PDF) (Go directly to the bottom line of last page). World reserves, as of 1st january 2002: 1,018.7 billion barrels.
Figure 2:
World Petroleum Consumption (idem) 77,125 thousand barrels per day.
So years to go at CURRENT RATE before there's no oil left on Earth:
1.0187E12/77.125E6/365=36.19 years.
So at current rate of consumption, all oil on Earth will be burnt in 36 year. As we extract only roughly 60% of the reserves due to technical reasons, it boils down to 21.7 years. You've got that long to buy a bike, if the consumption doen't skyrocket as China's life standards increases.
I reply to you again, on another point. You might want to add the oil gives chemical industry, which produces plastic and medecines, so in 20 years (se my other post about the deadline), we might have a bid health problem if the chemical industry gets broken.
Now you know why Bush goes for oil. (As an oil man, he knows that).
On the other hand, we can grow crop that yield hydrocarbons, but farming requires fuel, etc...
Kind of a vicious circle we're in, as you point out.
We've got 30 year of oil to go maximum. I made the calculation 2 week ago. I think this figure is broadly overestimated (yes, over).
It's basically all world proven reserves of oil (1000 Bilion barels) divided by the world daily consumption (75 milion barrel per day).
That brings us to 36,5 years (2040).
BUT the reserves can't be fully extracted (usually only 60%) and the consumption is likely to go up, so 20-25 years seems more likely, or even less. (ouch, isn't it?)
The figures were comming from the Official Energy Statistics from the U.S. Government
Yes my experience is the same in many cases. In one defense company, the only internet-connected machine of a 1000 people sized site was a few machines in the library.
And anyway in a major computer manufacturer's network, you didn't see much of internet except through the web proxy and soxyfied telnets. That's of course the way to go.
If you want real security, you are likely not to want a machine connected to the main power lines as well (tempest protection). I guess an off line UPS does the job.
Because christians *ARE* morons.
(applies to any religion to be honest)
The map posted in this commenth tm
l es /1996/sun_map.jpg
http://www.state.nv.us/nucwaste/yucca/seismo01.
reminds me of an area I already read about...
Did anyone spot that Yucca Mountain is conveniently located in the highly secured Nellis Air Force Range, just north of Armagosa Valley. Funny this area is also the world famous Area 51! (home of the stealth planes^H^H^H^H^H^H^H^H^Haliens)
Compare with this map :
http://www.ufomind.com/area51/orgs/nellis/artic
Opps posted in the wrong place, sorry, this goes to the VISA story...
Yes, they were cracked. See: http://parodie.com/monetique/ (In French, but with the full explaination.) This site covers many possible VISA frauds.
Yes, they were cracked. See: http://parodie.com/monetique/ (In French, but with the full explaination.) This site covers many possible VISA frauds.
You don't need it. The demanding IO is on the PCI bus (North). I don't think the south bridge can feed more than 100MB/s. See http://www.amd.com/products/cpg/server/athlon/chip set.html. It's only IDE (how much can the HDD MECANICS give you) plus USB. No big deal.
"the LVM included with HP-UX is actually a licensed Veritas VxFS."
This is false. You mix up filesystem and LVM. LVM is a software layer between physical devices (physical volumes) and block device (logical volumes). The logical volume is growable and shrinkable. But wether you can grow a mounted partition or not depends on you FILESYSTEM. So it just depends on you ext2/ext3 filesystem.
On HPUX VxFS (also called JFS for jourmaling is growable, but I don't remember if it's shrinkable).
You're not forced to repartition your drive in order to have a runing linux. See Loopback Root FS Mini Howto. I don't pretend it's easy, but it's even technically feasible to have a distro installing that way.