Actually, it wouldn't. I believe that Palladium/etc are designed to encrypt anything not on the CPU die. They assume that an attacker will be capable of capturing data off the system bus. Considering that this is how the XBox was opened it isn't surprising that they'd be this paranoid. They'd encrypt stuff within the CPU as well if they could accomplish this, but obviously at some point some part of the system needs access to the key. However, intercepting signals inside the CPU at several GHz on nm distance scales is not really practical currently.
Firefox took off on the PC because they marketed it to consumers - not megacorps.
Mozilla being upset about cell phone companies not deploying firefox is about as silly as them being upset about Microsoft bundling it with windows. What did you expect?
What they need is a PC-based installer that will bluetooth sync with a phone, identify the appropriate version to install, hack into it (with the owner's permission), and install the browser. Then it will take off. That's exactly how it took off on the PC - it still doesn't get preinstalled by most PC OEMs as far as I know.
Bypass the megacorp and their controls, and market direct to consumer. And make it easy to install...
I think the big problem is that depression is a very generic diagnosis.
I was chatting with a psychiatrist about this a few months ago. He shared his frustration with the difficulty of diagnosing psychological problems - he used an excellent analogy.
200 years ago if you went to a doctor with a heavy cough and difficulty breathing you'd be diagnosed with "cough". If they had the best modern drugs available and gave you one suitable for "cough" chances are they wouldn't be very effective in their treatment. Somebody with "cough" could have so many different actual ailments that there is a good chance that they would be getting incorrect treatment. Likewise, doing bypass surgery on anybody with pain in their chests would be only somewhat effective and might kill more people than it helps.
Drugs don't cure "depression". They treat people with high or low levels of particular neurotransmitters, often in a crude and untargetted manner. Sometimes this works, and sometimes it doesn't. The problem is that we don't understand depression well enough to actually define it in terms of an actual physical cause that has an actual physical solution.
This is also why I've always maintained that "me too" drugs are GOOD for the market. Besides keeping costs down through competition they give doctors more choices. Even if a given drug is provably a "better" drug it doesn't mean that it will be more effective for EVERYBODY who takes it. I know somebody who has trouble tolerating certain drugs that everybody would agree are "better" than some of the alternatives, but the alternatives work fine.
Unfortunately, our knowledge of medicine is not progressing as fast as our CPU designs are. This is why even the best medical studies sometimes contradict each other. There are just too many variables to perfectly control them all...
Yes, but why would the ISP want to support port-forwarding at all?
If they don't forward ports, email still works fine. The web (outgoing) works fine as well. That's all the internet they really care to sell in the first place.
I'd think NAT would be a good excuse to kill off P2P and everything else once and for all.
And I can't imagine that all this port-forwarding would be easier to implement than IPv6...
That's exactly what it means. Thanks for completely conceding the precept. The map shown in link #1 of GGP is a map of those hosts who ping or have open ports. Its pretty darn sparse.
Uh, could it be possible that a host not respond to any packets sent to it by a spider, and yet it accepts packets that are incoming from other hosts, and consequently needs a publicly-routable address?
Also - wouldn't it be nice that when two mega-corporations merge that their networks could just be joined with no migration needed at all? That would be because their address spaces didn't collide - unlike what you get with non-routable addresses?
NAT is a hack. It really isn't the way things ought to be.
You're also neglecting that the big iron routers still suffer capacity and performance problems given the insane address length of IPv6. This has been and remains one of the principal barriers to deployment. IPv6 simply costs a heck of a lot more per route and per packet routed.
I doubt it is the address length on its own that causes the problems. Most likely it is a failure to have optimized hardware to handle it - due to a lack of demand for this capability. The presentation linked in the original article states this. There is no reason that doubling the address length increases the routing complexity on its own. Now, if you hand out routes in an insane manner that would mess things up. However, having more breathing space should in theory make it EASIER to make nice consolidated routes. ISPs could request a single huge block of addresses and then split them up in a tidy way with plenty of room to grow. They wouldn't need to request them in blocks of 10-100k at a time and have a bazillion fragmented networks.
There is no doubt that routes will only get more complex in the future, and that routing hardware will become more capable.
I'm not suggesting that problems don't exist. However, turning the internet into a "network for networks" isn't really the solution (regardless of how it started). Maybe the internet would be better off without P2P, small servers, e-commerce, and games - but if that is how you feel why don't you just start your own internet? You won't have to worry about complex route tables if it is so featureless that nobody wants to bother to connect to it...:)
This is wrong. You clearly didn't take the trouble to look at the links I posted in the GP. While the instances of 'class a' abuse are somewhat constrained, there are plenty of poorly utilized allocations. Take a look, its shockingly sparse.
How would anybody know how that space is allocated. It is all behind firewalls. If you pinged every address in my company's class A you'd get no response unless you were on the network.
Ironically the whole class A connects to the internet via a NAT. I'm not sure how many hosts could even be pinged on it. That doesn't really mean the space is "wasted".
In any case, your proposal was to take back space from the class-A owners that aren't using most of it. There are only a handful of those - claiming that space won't help much at all. The problem is that EVERYBODY is wasting space. And current routing technologies basically make that inevitable. Unless routing tables grow to a point where/24s or smaller are routable at the top level you're going to end up making much larger allocations than are needed.
Re:Micro-Transactions and game balance
on
The Future of MMOs
·
· Score: 1
Just about every MMO out there officially allows people to buy levels/items/etc. The only thing that varies is the currency.
In most MMOs the only allowed currency is time spent playing the game.
It can be very frustrating to play an MMO that you'd only like to invest an hour in here or there without commitment, and find that anybody you make friends with ends up playing 40x more than you and is quickly onto challenges that your character just can't handle.
You make friends with a noob and they're like "wow - look how powerful a char you are". A month later they're looking up to you still. A month later they're your equal. A month later you're holding them back a little on quests. A month after that you're just a noob to them.
The grind is the real killer in these games. People who spend money are trying to avoid the least enjoyable aspect of these games in favor of the more enjoyable ones. You can't stop it - if the game trades time for levels then you just need to find somebody whose time is worth less than yours is...
Well, it is all inter-related, but color absorption is all about electrons and energy levels. Photon hits electron, electron pops to higher level, electron falls to lower level, electron emits photon.
The energies of those orbitals have everything to do with the sizes/masses/etc of the atoms they're bound to, and the number of electrons around them.
The electrons that matter generally aren't bound to a single atom - they move in molecular orbitals around groups of atoms or larger. If you take a piece of steel the electrons in the conductivity bands basically form one massive orbital the size of the piece of steel and just move around freely from one side to the other (well, to the extent that electrons actually move as opposed to existing in a given state that has some charge distribution). Hence metal conducts really well - electrons go in on one side and electrons pop out the other.
When you get into stuff like crystal field theory and molecular orbitals in chemistry you can start to get a feel for the responsible phenomena. The fact that we generally can't model orbitals perfectly for objects of any macroscopic size complicates things. However, for small molecules you can often predict the orbitals that will exist and which ones are populated and empty, and as a result get an idea for the spectral absorption of the molecule.
Disclaimer - I'm a chemist but certainly not an expert in molecular orbitals/etc...
Sure, I work for one of those companies - my laptop right now is connected to the VPN and is on one of those class-A networks. It is fairly well segmented across the company although obviously not all the address space is strictly necessary.
However, as others have pointed out if you actually got all those companies to give up all their address space it would buy you 6-12 months max. There aren't really that many of them. The problem is that address space demand is increasing exponentially.
And in some sense those companies helped get the internet started. There are always perks to being an early adopter. By the time you'd be able to take that space back in an orderly way it would be a sizzle in the pan.
NAT to ISP customers is EXACTLY what people are concerned about. ISPs would almost encourage it since it helps them to reduce the internet to email + large-scale websites, which is easier to support and extract ad revenue from. Stuff like games, bittorrent, etc is just a pain to them and the idea of customers not being herded to preferred sites paying ad revenue is just abhorrent...
Uh, I don't think you understand what is being proposed:
1) Periodically, software in CRON determines my current IP address, then it logs into an account via SSH and uploads it to a config file on some hosted server somewhere, and
Ok, your home computer (or router - whatever connects to your ISP directly) has the IP 192.168.234.123. You upload that.
2) The client software for my server (say, a Java client I'm using on a laptop, on vacation at a hotel) logs into the same account via SSH and checks the hosted server to see the current address of my home server, then uploads ITS IP address back to the hosted server, and
Ok, your laptop in the hotel knows your home is at 192.168.234.123. Your laptop has the address 192.168.123.234. It uploads that.
3) While the client software is setting up its connection (it retries until connected), the home server has already logged in and checked the hosted server to see which IP address the client is using, opening up the firewall for that IP address.
Ok, your home firewall is opened up to 192.168.123.234. Too bad the incoming connections would be coming from 192.168.234.001 - the NAT router at its ISP. EVERY incoming connection would be coming from there. But, that isn't the problem since you won't be getting any incoming connections at all. That's because when your laptop tries to open a connection to 192.168.123.234 the hotel filters it because that isn't a routable internet address.
The only exposure most people have to NAT is when their home firewall gets a routable IP and then issues non-routable IPs internally. That works fine because you own the firewall with the routable IP, and you can forward any ports that you'd like.
Now, imagine that I set up a router in your home. It is a black box to you - it hands out 192.168.x.y addresses to anything in your house, and doesn't forward anything to you. You can't even find out what IP it has from upstream because it won't tell you. Now you can't run anything - the only way you can find out your public IP is to connect to something else that does have a public IP and ask it where you seem to be coming from. However, that won't do you good since the NAT router will drop any incoming connections to that IP.
The only way you could do what you want to do is to set up some kind of tunnel on a routable host. Both sides of the link would connect out to the proxy, and then the proxy would forward packets between the connections. It would work, but only with a lot of hacking. It would be difficult to use standard web tools and your aunt edna definitely couldn't connect to your home web server...
I must then be imagining the public web server that I run over my NAT'd DSL connection.
You probably are if you are really behind an ISP-run NAT. We're not talking about the Linksys router that you can tell to forward port 80. We're talking about the ISP handing you a non-routable 192.168.x.x address and not forwarding anything to it. Outward-ONLY connections...
The irony of this of course is that the universe could really be only 6000 years old or whatever under this model. I mean, when you play an RPG and you read some book in the game referring to a battle that took place 500 years prior, that doesn't mean that the game is 500 years old...
For all you know the universe was created this morning and you just think you remember everything prior.
Check out this program. It calculates the impulse of a bottle rocket.
Using the defaults you get a specific impulse of about 33Ns/kg. That isn't much impulse. Most fuels are on the order of 4000Ns/kg.
Fuel requirements essentially go up exponentially as impulse decreases (I'm not a rocket scientist though). So having 100x less specific impulse is just going to kill your rocket almost entirely, as e^100 is a HUGE number. That's why ion engines are so attractive - if you can make your impulse 10-100x higher then your fuel requirements almost vanish. This is like an ion engine in reverse. You can get around this a little with multiple stages, but only to a degree.
Just think about this - a two stage bottle rocket can hit a few hundred meters altitude. You can do that with a chemical engine and some balsa wood with no trouble at all. A chemical rocket with two stages is going to be able to shoot down spacecraft. You only need big multi-stage chemical rockets to haul cargo into true orbit, and we're talking about multiple stage bottle rockets just to get about as high as a model airplane...
For typical RFIDs this is a very valid concern. On the other hand, if you had a powered RFID that contained a processor you could make it secure. The RFID could contain an SSL cert and the drive could issue a challenge against it.
Better still if you had enough bandwidth you could establish an SSL connection between drive and tag, and the decryption could take place inside the tag - so that the key never leaves the tag.
A less bandwidth-intensive solution might be to encrypt individual blocks on the drive with separate keys, and then the tag could authenticate access to each block based on some set of rules and deliver keys for only those blocks as they are needed. I'm sure some PK-based algorithm could be designed that would let the hard drive ask the fob to allocate a new key to encrypt a block of data, and then the fob could return an encrypt-only key, and sufficient metadata to store with the block so that the fob could later determine what the decrypt key is without needing too much internal storage.
There are lots of possible solutions out there. It just requires people to insist that their vendors not be lazy...
Well, if he wants to actually create space junk he'll need a lot more velocity to enter orbit... If you just launch it straight up it will only be space junk for about 30 seconds...
That is only a factor of about two. Sure, that is significant, but how much space will all that fancy conversion and sequestration technology take up? Why not just make the LH2 tank that much bigger?
I can't see this being cost-effective. How do you obtain H2 from hydrocarbons without producing CO2? It has to be cheap and take little energy or there is no point in it.
Actually, the Sun's gravity doesn't really help. You are correct that it is more difficult to insert a probe into orbit without an atmosphere to aerobrake into.
Let's suppose you have a probe in solar orbit at the radius of Mercury and want to get to Earth. Let's suppose you have another probe in solar orbit at the radius of Earth and want to get to Mercury. Which takes more effort?
The effort is actually the same in both cases - you need to enter a transfer orbit which is an ellipse that is tangental to both orbits. To enter this orbit you need to change your velocity relative to the Sun. If you're in orbit near the Earth you need to slow down quite a bit, and if you're in orbit near Mercury you need to speed up quite a bit. I believe the amount of delta-v is about the same (if not the same then the difference is pretty small compared to the total effort required), and speeding up and slowing down both require about the same amount of impulse - aka the same amount of fuel.
Read up on orbital mechanics - objects in orbit don't seem to act like objects on the Earth. Sure, when you drop a marble in space it falls towards the sun, but if it is in solar orbit it also is flying tangental to the sun with tremendous velocity so falling toward the sun means going in an ellipse around it instead of flying out into space. That's how black holes like the one in the center of the Milky Way can lie dormant for millions of years - tons of matter nearby but it is all in orbit so nothing falls in...
I think that the whole system needs a major revamp, but other than getting rid of traditionally-edited journals I'm not sure how. $3k/yr sounds like a lot of money - but that is for an institutional subscription, and maybe they have 1000 of those (are there any stats on this?). That is only $3M/yr to operate a very lengthy journal (ever see an issue of JACS?). Maybe the figures are larger, but even without profit it probably isn't a lot of money for what is being published. More exotic journals are even worse.
The issue is that the amount of labor going into each article is high compared to the number of people who will ever read them.
Either the amount of labor needs to go down, or the payment model needs to change. I don't have easy access to journals and I'd love to be able to browse them. However, the fact is that somebody needs to pay for this stuff - or decide that PDFs on websites is good enough.
In fact, I would say screw those patents and what-nots and let others produce the cure too. But then again, I'm not such a big fan of patents in the first place...
That's ok - most world governments tend to agree with you. Then again, maybe that's why most private pharma companies don't spend much money working on a cure for AIDS.
I see it the other way. Where do you want companies focusing their efforts - on erectile dysfunction medication or cures for plagues? If you want them to focus on the latter, then don't reward them for the former while punishing them for the latter! Life-saving medicines are exactly the kinds of things that companies should make lots of money on. Governments just need to stop spending so much on bombs and make it a little easier for people who can't afford the drugs to obtain them - not try to make them free...
Actually, the parent was correct. Patents apply to ideas - not implementations (like copyright).
This is probably appropriate, but in many cases patent lifetimes should probably be adjusted.
Tivo did come up with a truly innovative idea - nobody was doing anything like that at the time, and it wasn't just "the next step" or an obvious idea at the time. They should at least get license fees for a few years from the date of the invention.
The problem with patents isn't that they're hard to avoid - but that they last too long. Software patents should really only last a year or two unless there was some truly expensive ground-breaking research involved. Many highly profitable drug patents should probably bee a little shorter than they are now (although people get carried away with this - the fact is that drugs do cost serious money to develop - regardless of whether they originated in industrial R&D or a government/university lab).
Basically there needs to be a reasonable profit expectation for R&D or we discourage it. On the other hand, consumers shouldn't be saddled for decades by a single invention. Patents that hold up progress rather than promoting it are always bad.
Somebody needs to make a simple DNS server that just resolves requests. No need for BIND/etc - just something that is one-click on install and then you point your resolv.conf or windows networking setup to localhost for DNS. A standard DNS query that doesn't use the ISP would bypass all of this nonsense.
When you find an item for sale in a flea market at a really good price, do you shout that out to the whole world so that others might offer more for it and drive up the price? Or do you just quietly pay for it? Does Warren Buffet announce his next big buy the day before he does it so that everybody else can take advantage of some undervalued stock? An auction is a competition - I follow the rules and I always honor my bids. I don't help my competitors out.
Slashdot Mirror
Actually, it wouldn't. I believe that Palladium/etc are designed to encrypt anything not on the CPU die. They assume that an attacker will be capable of capturing data off the system bus. Considering that this is how the XBox was opened it isn't surprising that they'd be this paranoid. They'd encrypt stuff within the CPU as well if they could accomplish this, but obviously at some point some part of the system needs access to the key. However, intercepting signals inside the CPU at several GHz on nm distance scales is not really practical currently.
Firefox took off on the PC because they marketed it to consumers - not megacorps.
Mozilla being upset about cell phone companies not deploying firefox is about as silly as them being upset about Microsoft bundling it with windows. What did you expect?
What they need is a PC-based installer that will bluetooth sync with a phone, identify the appropriate version to install, hack into it (with the owner's permission), and install the browser. Then it will take off. That's exactly how it took off on the PC - it still doesn't get preinstalled by most PC OEMs as far as I know.
Bypass the megacorp and their controls, and market direct to consumer. And make it easy to install...
Hmm - just a smaller version of the OJ Simpson trial.
For whatever reason in the US you can basically tie up courts as long as you'd like. First person to run out of money loses.
Just about any kind of litigation costs a fortune here - mainly because stuff gets dragged out forever...
At the risk of beating this to death...:
Cue Horatio Caine.. *tilts head*
'Looks like someone's been left in the dark.. '
*removes sunglasses*
'permanently.'
I think the big problem is that depression is a very generic diagnosis.
I was chatting with a psychiatrist about this a few months ago. He shared his frustration with the difficulty of diagnosing psychological problems - he used an excellent analogy.
200 years ago if you went to a doctor with a heavy cough and difficulty breathing you'd be diagnosed with "cough". If they had the best modern drugs available and gave you one suitable for "cough" chances are they wouldn't be very effective in their treatment. Somebody with "cough" could have so many different actual ailments that there is a good chance that they would be getting incorrect treatment. Likewise, doing bypass surgery on anybody with pain in their chests would be only somewhat effective and might kill more people than it helps.
Drugs don't cure "depression". They treat people with high or low levels of particular neurotransmitters, often in a crude and untargetted manner. Sometimes this works, and sometimes it doesn't. The problem is that we don't understand depression well enough to actually define it in terms of an actual physical cause that has an actual physical solution.
This is also why I've always maintained that "me too" drugs are GOOD for the market. Besides keeping costs down through competition they give doctors more choices. Even if a given drug is provably a "better" drug it doesn't mean that it will be more effective for EVERYBODY who takes it. I know somebody who has trouble tolerating certain drugs that everybody would agree are "better" than some of the alternatives, but the alternatives work fine.
Unfortunately, our knowledge of medicine is not progressing as fast as our CPU designs are. This is why even the best medical studies sometimes contradict each other. There are just too many variables to perfectly control them all...
Yes, but why would the ISP want to support port-forwarding at all?
If they don't forward ports, email still works fine. The web (outgoing) works fine as well. That's all the internet they really care to sell in the first place.
I'd think NAT would be a good excuse to kill off P2P and everything else once and for all.
And I can't imagine that all this port-forwarding would be easier to implement than IPv6...
That's exactly what it means. Thanks for completely conceding the precept. The map shown in link #1 of GGP is a map of those hosts who ping or have open ports. Its pretty darn sparse.
:)
Uh, could it be possible that a host not respond to any packets sent to it by a spider, and yet it accepts packets that are incoming from other hosts, and consequently needs a publicly-routable address?
Also - wouldn't it be nice that when two mega-corporations merge that their networks could just be joined with no migration needed at all? That would be because their address spaces didn't collide - unlike what you get with non-routable addresses?
NAT is a hack. It really isn't the way things ought to be.
You're also neglecting that the big iron routers still suffer capacity and performance problems given the insane address length of IPv6. This has been and remains one of the principal barriers to deployment. IPv6 simply costs a heck of a lot more per route and per packet routed.
I doubt it is the address length on its own that causes the problems. Most likely it is a failure to have optimized hardware to handle it - due to a lack of demand for this capability. The presentation linked in the original article states this. There is no reason that doubling the address length increases the routing complexity on its own. Now, if you hand out routes in an insane manner that would mess things up. However, having more breathing space should in theory make it EASIER to make nice consolidated routes. ISPs could request a single huge block of addresses and then split them up in a tidy way with plenty of room to grow. They wouldn't need to request them in blocks of 10-100k at a time and have a bazillion fragmented networks.
There is no doubt that routes will only get more complex in the future, and that routing hardware will become more capable.
I'm not suggesting that problems don't exist. However, turning the internet into a "network for networks" isn't really the solution (regardless of how it started). Maybe the internet would be better off without P2P, small servers, e-commerce, and games - but if that is how you feel why don't you just start your own internet? You won't have to worry about complex route tables if it is so featureless that nobody wants to bother to connect to it...
This is wrong. You clearly didn't take the trouble to look at the links I posted in the GP. While the instances of 'class a' abuse are somewhat constrained, there are plenty of poorly utilized allocations. Take a look, its shockingly sparse.
/24s or smaller are routable at the top level you're going to end up making much larger allocations than are needed.
How would anybody know how that space is allocated. It is all behind firewalls. If you pinged every address in my company's class A you'd get no response unless you were on the network.
Ironically the whole class A connects to the internet via a NAT. I'm not sure how many hosts could even be pinged on it. That doesn't really mean the space is "wasted".
In any case, your proposal was to take back space from the class-A owners that aren't using most of it. There are only a handful of those - claiming that space won't help much at all. The problem is that EVERYBODY is wasting space. And current routing technologies basically make that inevitable. Unless routing tables grow to a point where
Just about every MMO out there officially allows people to buy levels/items/etc. The only thing that varies is the currency.
In most MMOs the only allowed currency is time spent playing the game.
It can be very frustrating to play an MMO that you'd only like to invest an hour in here or there without commitment, and find that anybody you make friends with ends up playing 40x more than you and is quickly onto challenges that your character just can't handle.
You make friends with a noob and they're like "wow - look how powerful a char you are". A month later they're looking up to you still. A month later they're your equal. A month later you're holding them back a little on quests. A month after that you're just a noob to them.
The grind is the real killer in these games. People who spend money are trying to avoid the least enjoyable aspect of these games in favor of the more enjoyable ones. You can't stop it - if the game trades time for levels then you just need to find somebody whose time is worth less than yours is...
Well, it is all inter-related, but color absorption is all about electrons and energy levels. Photon hits electron, electron pops to higher level, electron falls to lower level, electron emits photon.
The energies of those orbitals have everything to do with the sizes/masses/etc of the atoms they're bound to, and the number of electrons around them.
The electrons that matter generally aren't bound to a single atom - they move in molecular orbitals around groups of atoms or larger. If you take a piece of steel the electrons in the conductivity bands basically form one massive orbital the size of the piece of steel and just move around freely from one side to the other (well, to the extent that electrons actually move as opposed to existing in a given state that has some charge distribution). Hence metal conducts really well - electrons go in on one side and electrons pop out the other.
When you get into stuff like crystal field theory and molecular orbitals in chemistry you can start to get a feel for the responsible phenomena. The fact that we generally can't model orbitals perfectly for objects of any macroscopic size complicates things. However, for small molecules you can often predict the orbitals that will exist and which ones are populated and empty, and as a result get an idea for the spectral absorption of the molecule.
Disclaimer - I'm a chemist but certainly not an expert in molecular orbitals/etc...
Sure, I work for one of those companies - my laptop right now is connected to the VPN and is on one of those class-A networks. It is fairly well segmented across the company although obviously not all the address space is strictly necessary.
However, as others have pointed out if you actually got all those companies to give up all their address space it would buy you 6-12 months max. There aren't really that many of them. The problem is that address space demand is increasing exponentially.
And in some sense those companies helped get the internet started. There are always perks to being an early adopter. By the time you'd be able to take that space back in an orderly way it would be a sizzle in the pan.
NAT to ISP customers is EXACTLY what people are concerned about. ISPs would almost encourage it since it helps them to reduce the internet to email + large-scale websites, which is easier to support and extract ad revenue from. Stuff like games, bittorrent, etc is just a pain to them and the idea of customers not being herded to preferred sites paying ad revenue is just abhorrent...
Uh, I don't think you understand what is being proposed:
1) Periodically, software in CRON determines my current IP address, then it logs into an account via SSH and uploads it to a config file on some hosted server somewhere, and
Ok, your home computer (or router - whatever connects to your ISP directly) has the IP 192.168.234.123. You upload that.
2) The client software for my server (say, a Java client I'm using on a laptop, on vacation at a hotel) logs into the same account via SSH and checks the hosted server to see the current address of my home server, then uploads ITS IP address back to the hosted server, and
Ok, your laptop in the hotel knows your home is at 192.168.234.123. Your laptop has the address 192.168.123.234. It uploads that.
3) While the client software is setting up its connection (it retries until connected), the home server has already logged in and checked the hosted server to see which IP address the client is using, opening up the firewall for that IP address.
Ok, your home firewall is opened up to 192.168.123.234. Too bad the incoming connections would be coming from 192.168.234.001 - the NAT router at its ISP. EVERY incoming connection would be coming from there. But, that isn't the problem since you won't be getting any incoming connections at all. That's because when your laptop tries to open a connection to 192.168.123.234 the hotel filters it because that isn't a routable internet address.
The only exposure most people have to NAT is when their home firewall gets a routable IP and then issues non-routable IPs internally. That works fine because you own the firewall with the routable IP, and you can forward any ports that you'd like.
Now, imagine that I set up a router in your home. It is a black box to you - it hands out 192.168.x.y addresses to anything in your house, and doesn't forward anything to you. You can't even find out what IP it has from upstream because it won't tell you. Now you can't run anything - the only way you can find out your public IP is to connect to something else that does have a public IP and ask it where you seem to be coming from. However, that won't do you good since the NAT router will drop any incoming connections to that IP.
The only way you could do what you want to do is to set up some kind of tunnel on a routable host. Both sides of the link would connect out to the proxy, and then the proxy would forward packets between the connections. It would work, but only with a lot of hacking. It would be difficult to use standard web tools and your aunt edna definitely couldn't connect to your home web server...
I must then be imagining the public web server that I run over my NAT'd DSL connection.
You probably are if you are really behind an ISP-run NAT. We're not talking about the Linksys router that you can tell to forward port 80. We're talking about the ISP handing you a non-routable 192.168.x.x address and not forwarding anything to it. Outward-ONLY connections...
The irony of this of course is that the universe could really be only 6000 years old or whatever under this model. I mean, when you play an RPG and you read some book in the game referring to a battle that took place 500 years prior, that doesn't mean that the game is 500 years old...
For all you know the universe was created this morning and you just think you remember everything prior.
Check out this program. It calculates the impulse of a bottle rocket.
Using the defaults you get a specific impulse of about 33Ns/kg. That isn't much impulse. Most fuels are on the order of 4000Ns/kg.
Fuel requirements essentially go up exponentially as impulse decreases (I'm not a rocket scientist though). So having 100x less specific impulse is just going to kill your rocket almost entirely, as e^100 is a HUGE number. That's why ion engines are so attractive - if you can make your impulse 10-100x higher then your fuel requirements almost vanish. This is like an ion engine in reverse. You can get around this a little with multiple stages, but only to a degree.
Just think about this - a two stage bottle rocket can hit a few hundred meters altitude. You can do that with a chemical engine and some balsa wood with no trouble at all. A chemical rocket with two stages is going to be able to shoot down spacecraft. You only need big multi-stage chemical rockets to haul cargo into true orbit, and we're talking about multiple stage bottle rockets just to get about as high as a model airplane...
For typical RFIDs this is a very valid concern. On the other hand, if you had a powered RFID that contained a processor you could make it secure. The RFID could contain an SSL cert and the drive could issue a challenge against it.
Better still if you had enough bandwidth you could establish an SSL connection between drive and tag, and the decryption could take place inside the tag - so that the key never leaves the tag.
A less bandwidth-intensive solution might be to encrypt individual blocks on the drive with separate keys, and then the tag could authenticate access to each block based on some set of rules and deliver keys for only those blocks as they are needed. I'm sure some PK-based algorithm could be designed that would let the hard drive ask the fob to allocate a new key to encrypt a block of data, and then the fob could return an encrypt-only key, and sufficient metadata to store with the block so that the fob could later determine what the decrypt key is without needing too much internal storage.
There are lots of possible solutions out there. It just requires people to insist that their vendors not be lazy...
Well, if he wants to actually create space junk he'll need a lot more velocity to enter orbit... If you just launch it straight up it will only be space junk for about 30 seconds...
That is only a factor of about two. Sure, that is significant, but how much space will all that fancy conversion and sequestration technology take up? Why not just make the LH2 tank that much bigger?
I can't see this being cost-effective. How do you obtain H2 from hydrocarbons without producing CO2? It has to be cheap and take little energy or there is no point in it.
Actually, the Sun's gravity doesn't really help. You are correct that it is more difficult to insert a probe into orbit without an atmosphere to aerobrake into.
Let's suppose you have a probe in solar orbit at the radius of Mercury and want to get to Earth. Let's suppose you have another probe in solar orbit at the radius of Earth and want to get to Mercury. Which takes more effort?
The effort is actually the same in both cases - you need to enter a transfer orbit which is an ellipse that is tangental to both orbits. To enter this orbit you need to change your velocity relative to the Sun. If you're in orbit near the Earth you need to slow down quite a bit, and if you're in orbit near Mercury you need to speed up quite a bit. I believe the amount of delta-v is about the same (if not the same then the difference is pretty small compared to the total effort required), and speeding up and slowing down both require about the same amount of impulse - aka the same amount of fuel.
Read up on orbital mechanics - objects in orbit don't seem to act like objects on the Earth. Sure, when you drop a marble in space it falls towards the sun, but if it is in solar orbit it also is flying tangental to the sun with tremendous velocity so falling toward the sun means going in an ellipse around it instead of flying out into space. That's how black holes like the one in the center of the Milky Way can lie dormant for millions of years - tons of matter nearby but it is all in orbit so nothing falls in...
I think that the whole system needs a major revamp, but other than getting rid of traditionally-edited journals I'm not sure how. $3k/yr sounds like a lot of money - but that is for an institutional subscription, and maybe they have 1000 of those (are there any stats on this?). That is only $3M/yr to operate a very lengthy journal (ever see an issue of JACS?). Maybe the figures are larger, but even without profit it probably isn't a lot of money for what is being published. More exotic journals are even worse.
The issue is that the amount of labor going into each article is high compared to the number of people who will ever read them.
Either the amount of labor needs to go down, or the payment model needs to change. I don't have easy access to journals and I'd love to be able to browse them. However, the fact is that somebody needs to pay for this stuff - or decide that PDFs on websites is good enough.
If you were truly paranoid you wouldn't be sleeping in a bed under several layers of heavy metal plating in a building not designed to handle it...
In fact, I would say screw those patents and what-nots and let others produce the cure too. But then again, I'm not such a big fan of patents in the first place...
That's ok - most world governments tend to agree with you. Then again, maybe that's why most private pharma companies don't spend much money working on a cure for AIDS.
I see it the other way. Where do you want companies focusing their efforts - on erectile dysfunction medication or cures for plagues? If you want them to focus on the latter, then don't reward them for the former while punishing them for the latter! Life-saving medicines are exactly the kinds of things that companies should make lots of money on. Governments just need to stop spending so much on bombs and make it a little easier for people who can't afford the drugs to obtain them - not try to make them free...
Actually, the parent was correct. Patents apply to ideas - not implementations (like copyright).
This is probably appropriate, but in many cases patent lifetimes should probably be adjusted.
Tivo did come up with a truly innovative idea - nobody was doing anything like that at the time, and it wasn't just "the next step" or an obvious idea at the time. They should at least get license fees for a few years from the date of the invention.
The problem with patents isn't that they're hard to avoid - but that they last too long. Software patents should really only last a year or two unless there was some truly expensive ground-breaking research involved. Many highly profitable drug patents should probably bee a little shorter than they are now (although people get carried away with this - the fact is that drugs do cost serious money to develop - regardless of whether they originated in industrial R&D or a government/university lab).
Basically there needs to be a reasonable profit expectation for R&D or we discourage it. On the other hand, consumers shouldn't be saddled for decades by a single invention. Patents that hold up progress rather than promoting it are always bad.
Somebody needs to make a simple DNS server that just resolves requests. No need for BIND/etc - just something that is one-click on install and then you point your resolv.conf or windows networking setup to localhost for DNS. A standard DNS query that doesn't use the ISP would bypass all of this nonsense.
Nope - I just want the best price.
:)
When you find an item for sale in a flea market at a really good price, do you shout that out to the whole world so that others might offer more for it and drive up the price? Or do you just quietly pay for it? Does Warren Buffet announce his next big buy the day before he does it so that everybody else can take advantage of some undervalued stock? An auction is a competition - I follow the rules and I always honor my bids. I don't help my competitors out.
Ebay tells me to "shop victoriously" - so I do.