With Ion engines, the weight is neither in the engine nor in the reaction mass; it's in the solar panels. Ion engines require lots of power to operate, since the xenon is only a reaction mass. The ISS has a pretty tight power budget, and there is no way enough power can be diverted to Ion engines.
So, installing Ion stationkeeping engines on the ISS would also require installation of large new solar panels. The current system with Progress ships boosting the station is actually quite nice because the Progress space-robots are going to visit the station anyway - even if you had stationkeeping engines, you'd still have Progress ships visiting the station, laden with spare fuel and engines.
As for lowering the orbit of the station, well... It would degrade much more rapidly, requiring more power to keep it up - and power requirements increase exponentially. Also, the thin atmosphere would start ablating at the station, and as you pointed out, building the station aerodynamically might start making sense at some point. However, this would make the station incredibly more difficult to build.
I feel that the current solution with Progress space-robots boosting the station is a pretty good one.
Not so insane. The current device is pretty bulky; if someone was looking for a portable audio recorder, they might prefer a small, less expensive device.
I have reviewed all information available at this time, including discussion threads on many sites more specialised than Slashdot. This is bad. Very bad. Right now, there are millions of Internet users with partial connectivity.
But the action of Level3 is not merely an inconvenience to end users; it is hurting a great many small businesses, badly. There are thousands of small businesses that depend on single-homed Internet connectivity and that cannot afford dual-homing. There are dozens of low-cost datacenters that provide single-homed bandwith to tens of thousands of servers.
As we speak, the livelyhoods of thousands of entrepreneurs are being threatened. Many people depend on being able to offer internet services to any peer on the net. But today, Level3 has changed the rules of the game, and have split the Internet into two somewhat isolated internets.
This is happening on a very large scale. Sure, most of the affected people and businesses are going to get through it just fine. But given the sheer scale of the Internet, a small percentage of those depending on full connectivity will not escape this ordeal unscathed.
You can be sure that a few small businesses will close because of this, the reputations of a few persons will be damaged, and there will be a few bankruptcies - all because of Level3's evil actions. You won't hear about it in the media - nobody cares about such small-scale damage. But the damage is already done, and it is getting worse with every passing hour.
I urge you to join me in a five-minute hate against Level3 and all that their evil discriminative ways stand for. While Cogent is widely recognized for its shitty cut-rate network, they are the good guys here. In the past few years, Cogent has been a major driving force for lowering bandwith costs. Level3 is fighting back, and they long for the days where they charged 5000$/mbps. I say: down with Level3 !
Read carefully. Limewire users will still be able to download anything they want off the Gnutella network. However, Limewire will not SHARE copyrighted files, transforming it as the uber-leech client. Use Limewire to download copyrighted files, but never share them back. This protects the user and makes him safer from prosection, hence encouraging him to pay for the Limewire Pro registration.
Now, this will work because Gnutella is an open network, with many clients. A commercial leech client like the new Limewire will drain some ressources off the network, and in time, other clients may adapt to detect the newer Limewire versions, at which point Limewire will fight back...
Electrical transmission and distribution losses in the USA were estimated at 7.2% in 1995. See http://en.wikipedia.org/wiki/Electric_power_transm ission for details. Electrical power generation is very efficient, and overall pollutes far less than most small-scale energy production operations.
Secondly, "nothing is created, nothing is lost". When you're trying to heat up a room with electricity, waste heat is a good thing. This hydrogen fire device has multiple conversion stages, all of them inefficient - in that they release waste heat. In the end, all of the energy that goes into the system is converted into heat.
In fact, most of the heat of the device probably comes from the electrolysis rather than from the burning. But in the end, it's meant to be a room heater, and is doing a fine job as that. It is as efficient a furnace as a normal heater, or as a beowulf cluster. That's right, a beowulf cluster is a very good way to heat your room, and it's just as efficient as a purpose-made heater.
Do you know about heat pumps ? Those devices are basically air conditioners acting in reverse, taking heat from the outdoors, during the winter, and pumping it inside. At first glance, it doesn't make much sense: pumps and compressors are very inefficient devices, aren't they ? Plus, there's not much heat outside... But then you realise that the waste heat of the whole heat pump is a good thing - it's kept inside the house and used to heat it up. So all the heat pump has to do is extract a little bit of energy from the outside and spit out lots of waste heat, hence making it a tad more efficient than a device which merely spits waste heat.
Any electrical devices that doesn't move outside air around is an efficient heater. Your toaster, your computer and your electrical chainsaw are just as efficient as your room heater, when it comes to producing heat.
Anyway, your post is a travesty of science and logic. You were inspired by a hampster and your reasoning smells of elderberries.
You are wrong about the resolution. He's sampling every 3 seconds and making a min/max/mean database entry out of 20 samples, every minute.
You are wrong about the importance of the power factor. It is not even measured by the power companiesl; they ignore it in their billing. He cares about the bill, not some unmeasurable factor.
This is a common mistake to make for young scientists who fancy themselves engineers, electrical or otherwise.
Once upon a time, a young scientist learned of an home power monitoring system. It worked well, produced numbers that could be renconciled with the bills, was automated and not too expensive.
So the young scientist said: "Heresy ! This system does not meet the holy five-nines precision! Operating is is akin to spitting on the tomb of Nyquist while claiming that phase information is more of a theory than evolution ! The readouts are meaningless and design negligent !"
But he had a solution: "You will buy 15 of these meters and spread them around your house !" As for the automation, he figured you'd put a couple grad students on shifts for readout collection duty.
Open Source must be free for all or else it isn't Free. Should the GPL include a clause that prevents military contractors, neo-nazis, child molesters and Bill Gates from using the software ?
Never. We cannot exclude a single group from using Free software. This would be a desecration of the Freedom that the software stands for. Also, every OSS author would use the license as a political platform to condemn people at random: "This software cannot be used by Southern Baptists, Wahhabists and the followers of Ayn Rand."
--- By having read this post, you have already agreed to the Mostly Open Posting License. You are hereby granted the generous authorisation to read this post, but only if you are not a Witch, the daughter of Bill Gates or a fat-cat billionnaire. If you have read this post illegally, in defiance of this read-through license, you must paypal ten thousand american dollars to the author of this post, before Big Ben has sounded the upcoming call of midnight.
Okay. Some of you think that this means giving spammers a list of addresses confirmed of belonging to children.
If the list is well implemented, that won't happen. The published list will only contain hashes of the addresses.
Let's say I have a list of 1 million do-not-email addresses. From it, I'll compile a list of 1 million hashes.
Spammers will run their own address list through the hash, and check the do-not-email hash list against theirs.
Now, there is a problem with that system: spammers could generate a list of addresses by using common domains and usernames. And they'd probably find a few confirmed-valid addresses in the list.
So, there really isn't a good way to fix this, is it ?
Well, I've got one idea. Mass-sender registration. It requires the weight of law to work, which makes it an imperfect solution, it's not a universal fix to spam by any measure. But the fight against spam can only be won by the death of a thousand cuts.
To defeat spam, we need to wage a many-pronged attack against it and all that it stands for. The need for technological solutions is evident, and antispam software is critical.
However, legal attacks can help us fight spam. Mass-sender registration is simple. Everyone who sends over x amount of emails that will reach people within a specific jurisdiction need to first register themselves and their mailing in a purely electronic procedure.
A common registration protocol, simple to interface with, must be adopted by these governments, but registrars must remain separate.
For example, let's say I want to send 1 million spams. Let's say the Utah, Michigan and Wyoming have laws that require me to register my mailing if I mail more than 15 citizens of each state. If I don't register my mailing, I must include valid contact information in the mailing, and be prepared to provide all registration information on demand. Failure to comply with the law gives the right for individual citizens on the do-not-email list, who find the spam inacceptable, to sue individually for 250$ or run a class action suit for the same amount.
Registration is very simple: my mailing software automatically sends a copy of the mailing to the registrar - through SMTP. Seconds later, I receive a confirmation email that lets me send my mailing.
I do not need to register private mailings, nor mailing list messages for lists with double opt-in confirmation, list admin contact information and unsubscribe address or link.
This solution, while flawed, lets us enlist vast armies of greedy lawyers, who sue spammers for a living. They'll keep most of the class action fees for themselves, but they'll bleed spammers dry.
We need laws and technologies that makes it hard for spammers to make a profit. The public sector cannot stop spammers, but the private sector can. If there is a bounty for stopping spammers, around 500$ per illegal spam... It will be claimed by greedy lawyers and their spammed clients.
The key to the spam problem to unleash ONE THOUSAND ATTACK LAWYERS upon evil spammers.
How expensive is data storage, really ? I'll design a ten petabyte (10PB) storage system. You'll see how much it costs. To build this monster machine, I'll be using commercial off-the-shelf hardware organised as a massive Linux cluster.
You may ask "why do you want to build the most powerful Beowulf cluster on Earth when storage companies have all these amazing storage systems ?" Well, this system needs to be an open solution. The system will need to grow and evolve as the needs change. Vendor lock-in is simply unacceptable.
"Current storage software for Linux is useless for this massive an archive", you retort. Let me make it clear that there is no software anywhere that can fulfill the needs of the NARA. No matter what solution they choose, they will have to commission the creation of custom software. This software must be open-source, for the sake of the future growth and survival of the system.
Our basic storage unit will be a 300GB SATA drive. It's inexpensive, fairly reliable, and availaible in bulk quantities. We're going to need a lot of these little drives.
I'm going to create a very reliable system. First, it'll be completely redundant, with two identical systems, kept in sync, at two different geographical locations. I'll first consider the cost of a single location.
So, I need to store 10PB in a single building. I want reliability, and I'll get that by using two different levels of ECC: RAID-5 and distributed data. All disk drives will be part of 4-drive RAID-5 arrays. All these arrays will be part of larger 4-array RAID-5 arrays that are distributed across storage nodes.
Thus, each drive is regrouped with 16 others to form a 2-redundancy-layer 16 drive array, with 7 drives dedicated to parity. This leaves us 9/16 usable drives, or 2.7TB.
We therefore have 4-node storage subunits, with 5.4TB usable storage space. Inter-node communications go through gigabit ethernet with jumbo frames. The second gigabit ethernet interface of each node is also connected to a secondary network, for outside access to the node. Here, storage groups of 4-subunits are channeled through a single GBe port. That's 21.6TB which can be accessed through GBe.
Now, we have 512 such storage groups, each addressable at a speed of 1GBps. At this speed, the complete data store of each group can be transferred in or out in less than 72 hours.
We're using 2U cases for each system. This means that each subgroup uses 32U for servers, and we're using 42U racks. We have free space in every rack to allow for our high power density and communications and service equipment.
Let us consider the cost of a single such 21.6TB storage unit. We have 16 servers and 128 drives. Each server costs with storage costs around 2.5K. Each storage unit will be contained in a single rack with independent power conditioning, cooling and communications, at a cost of 15K$/rack, if we use high-end COTS equipement. So, we have a cost of 55K$/unit.
We need to make a cluster out of all this. We need extra computational power. We need a network backbone. We need tape drives to put data in the system. Front-end systems. A way to communicate with the outside world. A network operations center. That's going to add 5M$ to the system.
This means that the total equipment cost for a single location, including 5% of spare units, is around 35M$, including power and cooling equipment.
We need around 35000 sq.ft of floor space to host all this equipment and an 5000 sq.ft operations center. At a cost of 30$/sq.ft, we're talking about 1M$/year for the building.
Let's consider power costs, and do so very conservatively, with and a generous overhead on all figures. Each drive uses 15W of power. Each server uses a total of 250W of power. Each storage group uses 4kW of power. Each location thereforce uses 2MW of power. Over the course of a year, we're talking about 12.7GWh of power.
Now, we still need more power for cooling, which comes in at about 60% of total power costs. T
Apple has always craved publicity. This looks like what could be their smartest move in a long time.
Okay. You've got a keynote coming up on monday, you want people to listen. What do you do ? Leak a very, very controversial story to the mainstream technological press. This story is sure to get every Mac fan riled up.
And they'll be thinking about it all weekend, and they'll be listening to the keynote. Apple has generated massive interest in their keynote speech.
So here's my prediction for Monday's announcement. Apple announces Dual G5s and confirms that the PPC architecture is moving forward - also mentionning the amazing new PPC chips made for the Xbox360 and the PS3.
Apple also announces a new chip supplier: Intel. Those chips will not be used in Macs; they will be used in a new, low-end computer line that Apple will sell. This new line, which won't be called Macs, will not compete directly with Macs. Here's my guess:
For the sake of argument, let's call the new Apple computer line the e-Ntels. Low end machines, powered by Intel chips. They're computer appliances. They come pre-loaded with OS X x86-64 and Apple apps.
You can't install conventional software on them - every piece of software comes pre-installed by Apple, and gets automatically updated. The only software that you can run on it is approved by Apple, and is automatically deployed on all the e-Ntels.
The form factor might be a tablet, or it may be an HTPC, or both ! It's not going to be a conventional computer form factor. It's not going to be a general purpose computer. It's going to be an extremely inexpensive computer appliance.
You can pick one up like you pick up an iPod or a console. Costs at most 300$. You don't need to know anything about computers to use one - they run the incredibly user friendly OS X, which has been stripped down to be even easier.
So, what about the Macs, then ? Well, if you ever overgrow your i-Ntel, if you ever become a power user.. If you ever need software past that which is preinstalled on the i-Ntel (and it'll include everything most home users need), you get a real Mac.
And those Macs will be powerful. They will have fast dual-core PPC chips, and everything you expect from a Mac. They'll be able to network with your i-Ntel appliance, without any configuration, thanks to Rendezvous.
You're a current Mac user ? You'll want an inexpensive i-Ntel tablet and HTPC. It runs OS X and communicates with your real Mac.
Not a Mac user right now ? Tired of your shitty old PC ? Apple has a very inexpensive solution to your problem. Pick up the tiny i-Ntel box availaible at all good computer stores, plug it in, and you're good to go.
I can't wait to know what they're going to announce on Monday. But trust me: they're not dropping PPC - and the use of Intel chips is going to enhance their offerings. Everybody is going to be pleased, including hardcore Mac fans.
Mr. Illserve, you imply that problems best solved by genetic algorithms may well be full of potholes and bugs. Your comment is overrated FUD; you are trying to scare unwary slashdotters away from genetic algorithms !
First of all, the possible parameters that the genetic scheduler can affect can all be changed safely. The genetic scheduler does not try to invent new scheduling algorithms; rather, it tests different parameter sets and existing algorithms to find which works best with the current workload.
You can think of it as acting on variables that control how the scheduling is performed. It does not do the scheduling itself, nor does it try to evolve new algorithms; it merely tweaks scheduler operation in an attempt to better configure it.
Genetic algorithms are not inherently more prone to bugs than other kinds of algorithms. And there is a problem space in which they perform admirably. It appears that the linux kernel scheduler may well be on such problem space.
The giant robots of the future will probably not be bipedal... In a combat situation, they will surely be able to adopt a low profile. I can imagine a spider-like giant robot being somewhat practical.. Multiple redundant legs, and perhaps the ability to stand taller on part of them if a high profile or climbing is required.
We're talking about XGA projectors, not TVs. Those are availaible for 1400$ on the open market. And the Alienwares may be cheaper, too. As for the software, you're overestimating the cost.
Anyway, just wanted to say that your numbers don't make sense; your analysis is flawed and useless. You're just throwing numbers up in the air.
A major cost is power and cooling requirements. According to Apple, a single drive, dual processor 2.0GHz XServe will use about 250W peak. Virginia Tech has 2.3GHz machines and Infiniband PCI-X cards. There's also networking gear and other support equipment to consider. So, we'll use a high figure of 350W per node, giving us a 40% overhead. As for heat production, same overhead, 1200BTU/h per node.
We're going to consider the worst-case scenario, under which we have a 100% load, year round, on all 1100 nodes. That gives us a power consumption of 385kW and 1320kBTU/h of heat generation.
Now, we need to get rid of that heat, and that's going to require a lot of power. My research indicates up to 300kW may be required, but that's a high number and actual requirements may be lower.
So, here we are, with 685kW required for power and cooling. That means a 6000MW/h a year.
Now, the cost of power is high, since you need to amortize and maintain the UPS equipment and the generators. We'll use a figure of 0.15$/kW/h, or 150$/MW/h. Very generous.
So here we are. The absolute worst case for power and cooling. Full load, year round, expensive cooling, overpriced power and amortized UPS and generators.
900 000$/yr. Below a million. It's not that bad, is it ? The real cost is likely below a half-million.
As for the rest, well, how much pizza is really required to entice graduate students and professors to work on that machine ?
http://www.monolithic.com/ has already constructed many inhabitable earthly structures using this technology.
They inflate a large plastic dome, and then they start building - from the inside. They first spray 3 feet of polyurethane, which provides excellent insulation. They use the polyurethane as a support for an inner shell of steel reinforcement bars. Once the bars are in place, they spray 3 feet of concrete.
So, by building from the outside in, they can keep building no matter what the outside temperature is. The plastic dome that gives shape to the structure remains as the outer skin. The building has the strength and thermal capacity of concrete, the incredible insulation of polyurethane, and has the strongest shape known to man.
This is better than geodesic domes. A bit more expensive, though.
"I'd like to have a car for under $1000, and since this is not offered, I should be able to steal one, right?"
"I'd like to have a car for under 1000 $, and as this isn't offered, I should be able to take my remote replicator (which doesn't hurt the replicated object.. in fact, the replicated object doesn't even notice) and copy yours."
Ah.. that can be fixed with great ease, you know ? It wouldn't be hard to produce stimulation tapes. Play them 16 hours a day. Have them developed by "experts"
With some luck, the produced babies will be even more intelligent than "normal" babies.
And then, you could have "special" tapes which would help the baby develop certain reactions. Such as agression. That would be useful if you were trying to develop killer ninja babies.
Grow them in vats, and create a lot of automatons that will teach them to fight, as well as other required lifeskills. Such as learning a 'newspeak' type language which will form their view of the world.. And enable you to cheaply produce an army of drone-babies ! In only about 18 years after they're born !
You say "teach children morals and a sense of right and wrong and let them make their own decisions".
But, whose morals are you talking about ?
Yours ? Jack Valenti's ? Rufus Shinra ? Osama Bin Laden's ? Ghandi ? Mother Teresa (I should hope not !) ? The RIAA's ?
I don't think we should be teaching them more than the most basic morals. To a point, they all have to be subjective. I say, give them knowledge, and let them decide for themselves.
Have you ever seen an IME ? The program a Japanese person would use to enter their 10,000 characters ?
You spell out the word phonetically, and press space as you complete each word - the computer will show possible kanji, and you can cycle through them with the space key.
It actually works pretty well. Their keyboards pretty much look just like ours.
You're right ! Killing an industry for the sake of progress is totally unacceptable !
We should reinstate the hundred thousand of telephone operators and dismantles the evil computers that have replaced them ! Why destroy an industry for the sake of cheaper, higher quality product ?
We should abolish potato peeling machines at once, providing jobs to thousands of people that'll be able to lead a fulfilling life as potato peelers !
I believe that your thinking has a few flaws. You do overlook very important facts.
In the current economic and political situation, it is quite difficult to start a new business and try to compete with the dinosaurs. Regulations make it much harder for an enterprising young fellow to start his own business and make a profit. He'll have to follow countless rules, file tons of documents, pay fees, and waste a lot of time doing accounting and management to satisfy the lawmakers. As an entrepreneur myself, I can testify to the high barrier of entry into the club of business creation that has been instated by the governement.
Those regulations are there and make it much harder for anyone to become a threat to the larger corporations. You could invest your money in smaller businesses, but the thought of having the governement seize most of your profits make your teeth grind. So, why not take a tax-sheltered account and invest in Blue Chips ? That's easy..
Corporations abuse workers, you say ? Why w
You say that cars weren't as safe before government regulations, and you're partly right. Cars were 1/2 the price back then. And if you thought safety was important, you could buy a Volvo. But people didn't buy Volvos in droves, because they didn't want to pay a 100% premium for safety. Now, Big Government feels that you shouldn't be able to opt-out of having 5000$ worth of airbags and safety systems in every car. Effectively increasing the price of cars in a significant way. Instead of the less fortunate being able to afford recent yet less safe cars, they have to purchase old piles of mobile rust.
Governement makes it harder for anybody to make money, and to spend such that money. It makes it easier on Big Corporations that claim 100% tax deductions through creative accounting, and hurts the smaller companies through laws that make running a small business much harder.
Imagine if you had a no tax to pay on anything. Going in business would become second nature to anybody. Start selling stuff on the side, start consulting on the side. There is no penalty. There is no 500$ incorporation fee, no 50 page tax document to file.
Actually, producing hydrogen is very expensive, especially when using eletrolysis. The steam reforming method might be another story, though. But let's look at the cost of getting a megaJoule worth of hydrogen.
Electrolysis is approximately 15% efficient, and you're going to pay about 0.04$ for a kWh of power. so.04/.15/3.6
That means a megaJoule worth of hydrogen is going to cost you 0.075$.
Now, gasoline costs 0.5$ per liter, and a liter of gasoline is good for about 31.5 megaJoules. That means gasoline costs 0.016$ per megaJoule.
Conclusion ? You're going to need to produce lots of power to make your oxygen. Power isn't free, or even cheap. Worse, you're using a 15% efficient process. It's going to cost you a lot of money to make your hydrogen on a commercial scale.
Now, if you're using the latest hydrogen making processes, you might be able to achieve 80% efficiency. Over 5 times as efficient ! It drops the price of the power needed to make your megaJoule worth of hydrogen to less than 0.015$. That doesn't include all the equipement, personnel and taxes, methods of transportation and everything that you'll need to add to the cost of your hydrogen..
Slashdot Mirror
With Ion engines, the weight is neither in the engine nor in the reaction mass; it's in the solar panels. Ion engines require lots of power to operate, since the xenon is only a reaction mass. The ISS has a pretty tight power budget, and there is no way enough power can be diverted to Ion engines.
So, installing Ion stationkeeping engines on the ISS would also require installation of large new solar panels. The current system with Progress ships boosting the station is actually quite nice because the Progress space-robots are going to visit the station anyway - even if you had stationkeeping engines, you'd still have Progress ships visiting the station, laden with spare fuel and engines.
As for lowering the orbit of the station, well... It would degrade much more rapidly, requiring more power to keep it up - and power requirements increase exponentially. Also, the thin atmosphere would start ablating at the station, and as you pointed out, building the station aerodynamically might start making sense at some point. However, this would make the station incredibly more difficult to build.
I feel that the current solution with Progress space-robots boosting the station is a pretty good one.
Not so insane. The current device is pretty bulky; if someone was looking for a portable audio recorder, they might prefer a small, less expensive device.
I have reviewed all information available at this time, including discussion threads on many sites more specialised than Slashdot. This is bad. Very bad. Right now, there are millions of Internet users with partial connectivity.
But the action of Level3 is not merely an inconvenience to end users; it is hurting a great many small businesses, badly. There are thousands of small businesses that depend on single-homed Internet connectivity and that cannot afford dual-homing. There are dozens of low-cost datacenters that provide single-homed bandwith to tens of thousands of servers.
As we speak, the livelyhoods of thousands of entrepreneurs are being threatened. Many people depend on being able to offer internet services to any peer on the net. But today, Level3 has changed the rules of the game, and have split the Internet into two somewhat isolated internets.
This is happening on a very large scale. Sure, most of the affected people and businesses are going to get through it just fine. But given the sheer scale of the Internet, a small percentage of those depending on full connectivity will not escape this ordeal unscathed.
You can be sure that a few small businesses will close because of this, the reputations of a few persons will be damaged, and there will be a few bankruptcies - all because of Level3's evil actions. You won't hear about it in the media - nobody cares about such small-scale damage. But the damage is already done, and it is getting worse with every passing hour.
I urge you to join me in a five-minute hate against Level3 and all that their evil discriminative ways stand for. While Cogent is widely recognized for its shitty cut-rate network, they are the good guys here. In the past few years, Cogent has been a major driving force for lowering bandwith costs. Level3 is fighting back, and they long for the days where they charged 5000$/mbps. I say: down with Level3 !
Read carefully. Limewire users will still be able to download anything they want off the Gnutella network. However, Limewire will not SHARE copyrighted files, transforming it as the uber-leech client. Use Limewire to download copyrighted files, but never share them back. This protects the user and makes him safer from prosection, hence encouraging him to pay for the Limewire Pro registration.
Now, this will work because Gnutella is an open network, with many clients. A commercial leech client like the new Limewire will drain some ressources off the network, and in time, other clients may adapt to detect the newer Limewire versions, at which point Limewire will fight back...
But remember: Limewire may well be evil.
You, sir "Doc" Ruby, are spreading lies.
m ission for details. Electrical power generation is very efficient, and overall pollutes far less than most small-scale energy production operations.
Electrical transmission and distribution losses in the USA were estimated at 7.2% in 1995. See http://en.wikipedia.org/wiki/Electric_power_trans
Secondly, "nothing is created, nothing is lost". When you're trying to heat up a room with electricity, waste heat is a good thing. This hydrogen fire device has multiple conversion stages, all of them inefficient - in that they release waste heat. In the end, all of the energy that goes into the system is converted into heat.
In fact, most of the heat of the device probably comes from the electrolysis rather than from the burning. But in the end, it's meant to be a room heater, and is doing a fine job as that. It is as efficient a furnace as a normal heater, or as a beowulf cluster. That's right, a beowulf cluster is a very good way to heat your room, and it's just as efficient as a purpose-made heater.
Do you know about heat pumps ? Those devices are basically air conditioners acting in reverse, taking heat from the outdoors, during the winter, and pumping it inside. At first glance, it doesn't make much sense: pumps and compressors are very inefficient devices, aren't they ? Plus, there's not much heat outside... But then you realise that the waste heat of the whole heat pump is a good thing - it's kept inside the house and used to heat it up. So all the heat pump has to do is extract a little bit of energy from the outside and spit out lots of waste heat, hence making it a tad more efficient than a device which merely spits waste heat.
Any electrical devices that doesn't move outside air around is an efficient heater. Your toaster, your computer and your electrical chainsaw are just as efficient as your room heater, when it comes to producing heat.
Anyway, your post is a travesty of science and logic. You were inspired by a hampster and your reasoning smells of elderberries.
You are wrong about the resolution. He's sampling every 3 seconds and making a min/max/mean database entry out of 20 samples, every minute.
You are wrong about the importance of the power factor. It is not even measured by the power companiesl; they ignore it in their billing. He cares about the bill, not some unmeasurable factor.
This is a common mistake to make for young scientists who fancy themselves engineers, electrical or otherwise.
Once upon a time, a young scientist learned of an home power monitoring system. It worked well, produced numbers that could be renconciled with the bills, was automated and not too expensive.
So the young scientist said: "Heresy ! This system does not meet the holy five-nines precision! Operating is is akin to spitting on the tomb of Nyquist while claiming that phase information is more of a theory than evolution ! The readouts are meaningless and design negligent !"
But he had a solution: "You will buy 15 of these meters and spread them around your house !" As for the automation, he figured you'd put a couple grad students on shifts for readout collection duty.
Open Source must be free for all or else it isn't Free. Should the GPL include a clause that prevents military contractors, neo-nazis, child molesters and Bill Gates from using the software ?
Never. We cannot exclude a single group from using Free software. This would be a desecration of the Freedom that the software stands for. Also, every OSS author would use the license as a political platform to condemn people at random: "This software cannot be used by Southern Baptists, Wahhabists and the followers of Ayn Rand."
--- By having read this post, you have already agreed to the Mostly Open Posting License. You are hereby granted the generous authorisation to read this post, but only if you are not a Witch, the daughter of Bill Gates or a fat-cat billionnaire. If you have read this post illegally, in defiance of this read-through license, you must paypal ten thousand american dollars to the author of this post, before Big Ben has sounded the upcoming call of midnight.
Build a massive network of solar-powered laser beams pointed at the sky. Or maybe we could just cover the land with mirrors.
Okay. Some of you think that this means giving spammers a list of addresses confirmed of belonging to children.
If the list is well implemented, that won't happen. The published list will only contain hashes of the addresses.
Let's say I have a list of 1 million do-not-email addresses. From it, I'll compile a list of 1 million hashes.
Spammers will run their own address list through the hash, and check the do-not-email hash list against theirs.
Now, there is a problem with that system: spammers could generate a list of addresses by using common domains and usernames. And they'd probably find a few confirmed-valid addresses in the list.
So, there really isn't a good way to fix this, is it ?
Well, I've got one idea. Mass-sender registration. It requires the weight of law to work, which makes it an imperfect solution, it's not a universal fix to spam by any measure. But the fight against spam can only be won by the death of a thousand cuts.
To defeat spam, we need to wage a many-pronged attack against it and all that it stands for. The need for technological solutions is evident, and antispam software is critical.
However, legal attacks can help us fight spam. Mass-sender registration is simple. Everyone who sends over x amount of emails that will reach people within a specific jurisdiction need to first register themselves and their mailing in a purely electronic procedure.
A common registration protocol, simple to interface with, must be adopted by these governments, but registrars must remain separate.
For example, let's say I want to send 1 million spams. Let's say the Utah, Michigan and Wyoming have laws that require me to register my mailing if I mail more than 15 citizens of each state. If I don't register my mailing, I must include valid contact information in the mailing, and be prepared to provide all registration information on demand. Failure to comply with the law gives the right for individual citizens on the do-not-email list, who find the spam inacceptable, to sue individually for 250$ or run a class action suit for the same amount.
Registration is very simple: my mailing software automatically sends a copy of the mailing to the registrar - through SMTP. Seconds later, I receive a confirmation email that lets me send my mailing.
I do not need to register private mailings, nor mailing list messages for lists with double opt-in confirmation, list admin contact information and unsubscribe address or link.
This solution, while flawed, lets us enlist vast armies of greedy lawyers, who sue spammers for a living. They'll keep most of the class action fees for themselves, but they'll bleed spammers dry.
We need laws and technologies that makes it hard for spammers to make a profit. The public sector cannot stop spammers, but the private sector can. If there is a bounty for stopping spammers, around 500$ per illegal spam... It will be claimed by greedy lawyers and their spammed clients.
The key to the spam problem to unleash ONE THOUSAND ATTACK LAWYERS upon evil spammers.
How expensive is data storage, really ? I'll design a ten petabyte (10PB) storage system. You'll see how much it costs. To build this monster machine, I'll be using commercial off-the-shelf hardware organised as a massive Linux cluster.
You may ask "why do you want to build the most powerful Beowulf cluster on Earth when storage companies have all these amazing storage systems ?" Well, this system needs to be an open solution. The system will need to grow and evolve as the needs change. Vendor lock-in is simply unacceptable.
"Current storage software for Linux is useless for this massive an archive", you retort. Let me make it clear that there is no software anywhere that can fulfill the needs of the NARA. No matter what solution they choose, they will have to commission the creation of custom software. This software must be open-source, for the sake of the future growth and survival of the system.
Our basic storage unit will be a 300GB SATA drive. It's inexpensive, fairly reliable, and availaible in bulk quantities. We're going to need a lot of these little drives.
I'm going to create a very reliable system. First, it'll be completely redundant, with two identical systems, kept in sync, at two different geographical locations. I'll first consider the cost of a single location.
So, I need to store 10PB in a single building. I want reliability, and I'll get that by using two different levels of ECC: RAID-5 and distributed data. All disk drives will be part of 4-drive RAID-5 arrays. All these arrays will be part of larger 4-array RAID-5 arrays that are distributed across storage nodes.
Thus, each drive is regrouped with 16 others to form a 2-redundancy-layer 16 drive array, with 7 drives dedicated to parity. This leaves us 9/16 usable drives, or 2.7TB.
We therefore have 4-node storage subunits, with 5.4TB usable storage space. Inter-node communications go through gigabit ethernet with jumbo frames. The second gigabit ethernet interface of each node is also connected to a secondary network, for outside access to the node. Here, storage groups of 4-subunits are channeled through a single GBe port. That's 21.6TB which can be accessed through GBe.
Now, we have 512 such storage groups, each addressable at a speed of 1GBps. At this speed, the complete data store of each group can be transferred in or out in less than 72 hours.
We're using 2U cases for each system. This means that each subgroup uses 32U for servers, and we're using 42U racks. We have free space in every rack to allow for our high power density and communications and service equipment.
Let us consider the cost of a single such 21.6TB storage unit. We have 16 servers and 128 drives. Each server costs with storage costs around 2.5K. Each storage unit will be contained in a single rack with independent power conditioning, cooling and communications, at a cost of 15K$/rack, if we use high-end COTS equipement. So, we have a cost of 55K$/unit.
We need to make a cluster out of all this. We need extra computational power. We need a network backbone. We need tape drives to put data in the system. Front-end systems. A way to communicate with the outside world. A network operations center. That's going to add 5M$ to the system.
This means that the total equipment cost for a single location, including 5% of spare units, is around 35M$, including power and cooling equipment.
We need around 35000 sq.ft of floor space to host all this equipment and an 5000 sq.ft operations center. At a cost of 30$/sq.ft, we're talking about 1M$/year for the building.
Let's consider power costs, and do so very conservatively, with and a generous overhead on all figures. Each drive uses 15W of power. Each server uses a total of 250W of power. Each storage group uses 4kW of power. Each location thereforce uses 2MW of power. Over the course of a year, we're talking about 12.7GWh of power.
Now, we still need more power for cooling, which comes in at about 60% of total power costs. T
Apple has always craved publicity. This looks like what could be their smartest move in a long time.
s p - it's about Dual-Core G5s.
Okay. You've got a keynote coming up on monday, you want people to listen. What do you do ? Leak a very, very controversial story to the mainstream technological press. This story is sure to get every Mac fan riled up.
And they'll be thinking about it all weekend, and they'll be listening to the keynote. Apple has generated massive interest in their keynote speech.
So, what will they actually announce on Monday ? Well, there's another leaked story today, at http://www.eweek.com/article2/0,1759,1823273,00.a
So here's my prediction for Monday's announcement. Apple announces Dual G5s and confirms that the PPC architecture is moving forward - also mentionning the amazing new PPC chips made for the Xbox360 and the PS3.
Apple also announces a new chip supplier: Intel. Those chips will not be used in Macs; they will be used in a new, low-end computer line that Apple will sell. This new line, which won't be called Macs, will not compete directly with Macs. Here's my guess:
For the sake of argument, let's call the new Apple computer line the e-Ntels. Low end machines, powered by Intel chips. They're computer appliances. They come pre-loaded with OS X x86-64 and Apple apps.
You can't install conventional software on them - every piece of software comes pre-installed by Apple, and gets automatically updated. The only software that you can run on it is approved by Apple, and is automatically deployed on all the e-Ntels.
The form factor might be a tablet, or it may be an HTPC, or both ! It's not going to be a conventional computer form factor. It's not going to be a general purpose computer. It's going to be an extremely inexpensive computer appliance.
You can pick one up like you pick up an iPod or a console. Costs at most 300$. You don't need to know anything about computers to use one - they run the incredibly user friendly OS X, which has been stripped down to be even easier.
So, what about the Macs, then ? Well, if you ever overgrow your i-Ntel, if you ever become a power user.. If you ever need software past that which is preinstalled on the i-Ntel (and it'll include everything most home users need), you get a real Mac.
And those Macs will be powerful. They will have fast dual-core PPC chips, and everything you expect from a Mac. They'll be able to network with your i-Ntel appliance, without any configuration, thanks to Rendezvous.
You're a current Mac user ? You'll want an inexpensive i-Ntel tablet and HTPC. It runs OS X and communicates with your real Mac.
Not a Mac user right now ? Tired of your shitty old PC ? Apple has a very inexpensive solution to your problem. Pick up the tiny i-Ntel box availaible at all good computer stores, plug it in, and you're good to go.
I can't wait to know what they're going to announce on Monday. But trust me: they're not dropping PPC - and the use of Intel chips is going to enhance their offerings. Everybody is going to be pleased, including hardcore Mac fans.
Mr. Illserve, you imply that problems best solved by genetic algorithms may well be full of potholes and bugs. Your comment is overrated FUD; you are trying to scare unwary slashdotters away from genetic algorithms !
First of all, the possible parameters that the genetic scheduler can affect can all be changed safely. The genetic scheduler does not try to invent new scheduling algorithms; rather, it tests different parameter sets and existing algorithms to find which works best with the current workload.
You can think of it as acting on variables that control how the scheduling is performed. It does not do the scheduling itself, nor does it try to evolve new algorithms; it merely tweaks scheduler operation in an attempt to better configure it.
Genetic algorithms are not inherently more prone to bugs than other kinds of algorithms. And there is a problem space in which they perform admirably. It appears that the linux kernel scheduler may well be on such problem space.
The giant robots of the future will probably not be bipedal... In a combat situation, they will surely be able to adopt a low profile. I can imagine a spider-like giant robot being somewhat practical.. Multiple redundant legs, and perhaps the ability to stand taller on part of them if a high profile or climbing is required.
We're talking about XGA projectors, not TVs. Those are availaible for 1400$ on the open market. And the Alienwares may be cheaper, too. As for the software, you're overestimating the cost.
Anyway, just wanted to say that your numbers don't make sense; your analysis is flawed and useless. You're just throwing numbers up in the air.
A major cost is power and cooling requirements. According to Apple, a single drive, dual processor 2.0GHz XServe will use about 250W peak. Virginia Tech has 2.3GHz machines and Infiniband PCI-X cards. There's also networking gear and other support equipment to consider. So, we'll use a high figure of 350W per node, giving us a 40% overhead. As for heat production, same overhead, 1200BTU/h per node.
We're going to consider the worst-case scenario, under which we have a 100% load, year round, on all 1100 nodes. That gives us a power consumption of 385kW and 1320kBTU/h of heat generation.
Now, we need to get rid of that heat, and that's going to require a lot of power. My research indicates up to 300kW may be required, but that's a high number and actual requirements may be lower.
So, here we are, with 685kW required for power and cooling. That means a 6000MW/h a year.
Now, the cost of power is high, since you need to amortize and maintain the UPS equipment and the generators. We'll use a figure of 0.15$/kW/h, or 150$/MW/h. Very generous.
So here we are. The absolute worst case for power and cooling. Full load, year round, expensive cooling, overpriced power and amortized UPS and generators.
900 000$/yr. Below a million. It's not that bad, is it ? The real cost is likely below a half-million.
As for the rest, well, how much pizza is really required to entice graduate students and professors to work on that machine ?
http://www.monolithic.com/ has already constructed many inhabitable earthly structures using this technology.
They inflate a large plastic dome, and then they start building - from the inside. They first spray 3 feet of polyurethane, which provides excellent insulation. They use the polyurethane as a support for an inner shell of steel reinforcement bars. Once the bars are in place, they spray 3 feet of concrete.
So, by building from the outside in, they can keep building no matter what the outside temperature is. The plastic dome that gives shape to the structure remains as the outer skin. The building has the strength and thermal capacity of concrete, the incredible insulation of polyurethane, and has the strongest shape known to man.
This is better than geodesic domes. A bit more expensive, though.
"I'd like to have a car for under $1000, and since this is not offered, I should be able to steal one, right?"
"I'd like to have a car for under 1000 $, and as this isn't offered, I should be able to take my remote replicator (which doesn't hurt the replicated object.. in fact, the replicated object doesn't even notice) and copy yours."
Ah.. that can be fixed with great ease, you know ? It wouldn't be hard to produce stimulation tapes. Play them 16 hours a day. Have them developed by "experts"
With some luck, the produced babies will be even more intelligent than "normal" babies.
And then, you could have "special" tapes which would help the baby develop certain reactions. Such as agression. That would be useful if you were trying to develop killer ninja babies.
Grow them in vats, and create a lot of automatons that will teach them to fight, as well as other required lifeskills. Such as learning a 'newspeak' type language which will form their view of the world.. And enable you to cheaply produce an army of drone-babies ! In only about 18 years after they're born !
You say "teach children morals and a sense of right and wrong and let them make their own decisions".
But, whose morals are you talking about ?
Yours ? Jack Valenti's ? Rufus Shinra ? Osama Bin Laden's ? Ghandi ? Mother Teresa (I should hope not !) ? The RIAA's ?
I don't think we should be teaching them more than the most basic morals. To a point, they all have to be subjective. I say, give them knowledge, and let them decide for themselves.
Have you ever seen an IME ? The program a Japanese person would use to enter their 10,000 characters ?
You spell out the word phonetically, and press space as you complete each word - the computer will show possible kanji, and you can cycle through them with the space key.
It actually works pretty well. Their keyboards pretty much look just like ours.
You're right ! Killing an industry for the sake of progress is totally unacceptable !
We should reinstate the hundred thousand of telephone operators and dismantles the evil computers that have replaced them ! Why destroy an industry for the sake of cheaper, higher quality product ?
We should abolish potato peeling machines at once, providing jobs to thousands of people that'll be able to lead a fulfilling life as potato peelers !
I believe that your thinking has a few flaws. You do overlook very important facts.
In the current economic and political situation, it is quite difficult to start a new business and try to compete with the dinosaurs. Regulations make it much harder for an enterprising young fellow to start his own business and make a profit. He'll have to follow countless rules, file tons of documents, pay fees, and waste a lot of time doing accounting and management to satisfy the lawmakers. As an entrepreneur myself, I can testify to the high barrier of entry into the club of business creation that has been instated by the governement.
Those regulations are there and make it much harder for anyone to become a threat to the larger corporations. You could invest your money in smaller businesses, but the thought of having the governement seize most of your profits make your teeth grind. So, why not take a tax-sheltered account and invest in Blue Chips ? That's easy..
Corporations abuse workers, you say ? Why w
You say that cars weren't as safe before government regulations, and you're partly right. Cars were 1/2 the price back then. And if you thought safety was important, you could buy a Volvo. But people didn't buy Volvos in droves, because they didn't want to pay a 100% premium for safety. Now, Big Government feels that you shouldn't be able to opt-out of having 5000$ worth of airbags and safety systems in every car. Effectively increasing the price of cars in a significant way. Instead of the less fortunate being able to afford recent yet less safe cars, they have to purchase old piles of mobile rust.
Governement makes it harder for anybody to make money, and to spend such that money. It makes it easier on Big Corporations that claim 100% tax deductions through creative accounting, and hurts the smaller companies through laws that make running a small business much harder.
Imagine if you had a no tax to pay on anything. Going in business would become second nature to anybody. Start selling stuff on the side, start consulting on the side. There is no penalty. There is no 500$ incorporation fee, no 50 page tax document to file.
Ah ! You haven't read the article right.
It takes 15 1-megaton (read: SMALL) nukes to create an explosion equivalent to 15 H-bombs (read: BIIIIIIG).
Actually, producing hydrogen is very expensive, especially when using eletrolysis. The steam reforming method might be another story, though. But let's look at the cost of getting a megaJoule worth of hydrogen.
.04/.15/3.6
Electrolysis is approximately 15% efficient, and you're going to pay about 0.04$ for a kWh of power. so
That means a megaJoule worth of hydrogen is going to cost you 0.075$.
Now, gasoline costs 0.5$ per liter, and a liter of gasoline is good for about 31.5 megaJoules. That means gasoline costs 0.016$ per megaJoule.
So:
0.075$/MJ Hydrogen (electrolised)
0.016$/MJ Gasoline
Conclusion ? You're going to need to produce lots of power to make your oxygen. Power isn't free, or even cheap. Worse, you're using a 15% efficient process. It's going to cost you a lot of money to make your hydrogen on a commercial scale.
Now, if you're using the latest hydrogen making processes, you might be able to achieve 80% efficiency. Over 5 times as efficient ! It drops the price of the power needed to make your megaJoule worth of hydrogen to less than 0.015$. That doesn't include all the equipement, personnel and taxes, methods of transportation and everything that you'll need to add to the cost of your hydrogen..
What about TurboLinuxCare ? It fits, too..
I really hate any name with "Turbo" in it, though.
Turbo button on a 486.. augh.. Turbonium.. Turbolinux.. Don't they have any idea what a turbo is ?