It's just a matter of time, as well, it's not such a big deal to have your AI or actual render threads separated out. It's amazing what getting rid of a few context switches can do for performance.
Not to bash your point too much, but my brain works that way:) I remember in school we had to write an assembly program in gcc and I used jmp instructions to dance around a block of code so much (I was trying for the smallest program) that I got my A paper back with my favorite teacher comment ever... "too complex."
I understand that NAT is considered a hack, but isn't the fact that a device's real address is hidden a security feature for the user? Wouldn't it be that much harder for malicious users to track my internet usage? This would be especially true if I had a mobile device, since moving from one NAT system to another would make following my movements remotely more difficult. So I'd think NAT would be considered a privacy boon. The article doesn't really address this effectively. Also, since most mobile devices have limited bandwidth, I'd think that having a constantly changing IP address, or hiding behind a NAT would mean that DOS attacks against them would be more difficult. If most big mobile device ISPs like the blackberry and sidekick folks offered NAT based access in the future, I'd think that we'd be relatively safe from IPv4 address exhaustion. So stating the main reason for IPv6 being address exhaustion I think is crap. It IS very useful for other reasons though, and I think those reasons warrant it being switched to.
Actually, what's really needed is a really smart AI that you can just talk to and say "Yo, computer, mess with this stuff and make it cool". But seriously, I think the trend towards more conversational:) programming is leading to a place where we will eventually just use voice commands to get the computer to do most common programming tasks. Obviously the real science related heavy math alg. stuff will still need to be programmed by hand until one of the sentient AIs decides it's time to take over.
From what I've been reading asia is going IP6 much faster then we (USA) are. In China & other places I think it's because they're still building a lot of infrastructure so they can start with the latest.
Maybe you already know this, but you shouldn't use the framebuffer kernel driver with the proprietary nvidia drivers. It makes it much more finicky. Since I do almost everything in X I don't really need 1280x1024 VGA console:)
I don't understand all this excitement over fuel cells being used in mobile devices. No electricity isn't everywhere, but then again, neither are hydrogen stations to buy refills at. If you think electricity is spotty in the 3rd world, just wait till you try to get hydrogen. This whole idea is dumb. Electricity is the most commonly available energy source. I can understand making cars and other bigger machines run on hydrogen, since they already have refueling infrastructure, but the energy problem could be better served by using higher efficiency transmission of electricity with something like the SuperGrid would be a much more achievable goal and easier to phase in then a switch to hydrogen.
With so many places in the world don't have enough water already, or their water is poor quality, for them getting hydrogen for water would not be a good thing.
This reminds "Prey" by Michael Crichton where they use bacteria to create nanobots. I know it makes sense from a production standpoint, but man I really don't like the possibilities here.
This reminds me of the line from a good book "Scientists were so caught up with what they could build, that they never stopped to think if they should" Maybe I don't WANT artificial life forms roaming around inside my body:) Thing is, once you release (or something escapes) like this that can live at the single cell level what's to stop it from spreading? At least with nanobots they can be made to not reproduce.
I don't claim to understand the whole Jabber protocol, but it seems to me that having decentralized servers might be a security problem, and be a disadvantage when everyone and their brother can throw together a jabber server.
Once Google allows the other servers to be used with it's own service, it's giving them legitimacy. That's something that people should be careful with. It would be cool if there was some kind of verification system with the server operators, say something similar to personal SSL key verification that people like Thawte were doing before being aquired by Verisign.
There's also the issue of everyone using google's server's as the nexus of the entire jabber network if google started working with other jabber servers.
I just hope they design these things to be secure (however you do that) against hijacking. Imagine what a supersonic plane or two could do to a city... ouch.
Well, hopefully we'll get flying cars and floating highways soon that allow people to spread out more and not be so concentrated in cities.
I'm sorry, but anyone that thinks a 1 button mouse is the height of user interface design needs to be shot. Context menus make so much more sense being available with a right click. That aside, OS X does have some nice design principles, and like all developers I think OSS should seek to stea^F^F^F^Fcopy those features.
I think one of the better net admin jokes on this date was using the swedish chef text filter on all webpages in certain sections of the my college's site:)
Tethers can be used in space to do everything from adjusting orbits to generating electricity from orbiting planets that have magnetic fields.
http://www.strangehorizons.com/2003/20030414/rop e. shtml http://www.vectorsite.net/tarokt5.html
Tethers of (only) 20km long or less are much less likely to have problems then some ground to space tether that has to deal with wind conditions and other weather on the ground.
First off the author is looking at the wrong thing. For Linux to become mainstream means it needs to be commercialized and accepted in corporate arenas. When you ask your friends what versions of Linux they use you're mostly talking about personal preference work hobbiests/enthusiasts. What he really should have done is looked at the situation from big business' perspective.
Yes, Linux on the desktop isn't quite there. But it doesn't have to be right now either. The important thing that author has missed is that it's going in the right direction. It's not getting more fragmented. The LSB project has all the major players commited who could kill Linux by forking. All the major Linux distros are fully aware of the danger of them forking. So the LSB is going in the right direction and I think the kernel development process is as well (though that isn't the real issue here).
As far as ready-for-the-desktop goes, the big development there is the current maturing of the plug-and-play model in Linux. Eventually all distros will make it possible to never have to manualy mount your CD Rom and USB Flash drives manually. Right now for many self-roled kernels (and for me Slackware) out of the box automounting doesn't happen. That's an issue. But overall the picture is getting better not worse, and Linux' momentum is not going away quite yet.
The truth is though, that in the big picture the hobby distros don't count in the race to global domination. It's the big 3 that do. So in forgetting that fact, the author of the article missed the boat.
Well what I'm saying is, if people did know it was illegal and didn't care (which is what I'm assuming is the case here) then we're talkinga bout avoiding prosecution. I would think that on a small island with no/few international legal agreements it would be possible to maintain anonymity and thus avoid lawsuits, and since you're paying the dough, your site wouldn't be taken down so you could maintain web presence.
My question is, why do people host possibly infringing sites in territories that outlaw it? I mean why don't all these bittorrent sites host on little jungle islands in the pacific or other places that have no/fewer/looser international law agreements?
Ok, enough conjecture here's some numbers.
Assuming that a particular key would take 1billion years to crack by today's computers. That same key would be crackable in ~1 month using computers available 50 years from now. (Assuming 18 months to processor speed doubling, and that my math is right).
You can call that irrelevent, or besides the point if you want. But that's a much smaller number to work with then 1 billion. I'm not saying that isn't a formidable number, just that it's not the "uncrackable" 1 billion people bandy about all the time. If we're talking 4 billion that would take only 3 years longer in order to crack it in 1 month. And all of this is assuming that people weren't working on the keyspace during that time period, which would cut the time required dramatically.
However, I still agree with you that for most, people they don't have to worry about anyone carying about their encrypted data in 50 or 100 years:) I'm just saying, people get all excited about these huge numbers when their data isn't quite that safe.
Best Linux Article Ever
on
Linux, Inc.
·
· Score: 1
I have to say this is the best example of journalism regarding Linux that I have seen yet. It is comprehensive and doesn't just regurgitate all the usual stuff for corporate drones. It actually has somewhat rare/new content about how actual decisions were made etc. I wish there was an email/feedback link to the author, cause I would tell him this myself.
I agree that it's still ok for near term stuff. But if you actually want something to NEVER be unencrypted you can't get all cozy with times like 1billion years to crack your messages.
Use of such numbers is a PR stunt that doesn't really reflect reality.
... is that people don't factor in that computers increase in speed by almost a factor of 2 each 18 months or so. So basically, that's like removing 1 bit from your encryption key every 18 months. Factor that in, and you begin to see cracking time numbers orders of magnitude smaller then before.
So basically, public key encryption should still only be used for things that are time sensitive that no one will care about in 5 years. If that's not the case, then you should think about using extremely large keys, like 4000bit, or use quantum encryption.
There is not such thing as absolute security/secrecy.
Am I the only one that thinks that plane is the ugliest I've ever seen? I understand function coming before form, but man, that thing is ugly.
Being american and somewhat nationalistic I'm rooting for Boeing but I still am glad to see a good competition with Airbus forming. That's good for everyone.
You're right, that is what PGP has been all about. And that is also why it never took off. It's too much of a hassle for people to distribute/verify keys and also to revoke them. In order for something like PGP to be used it has to be ubiquitous. With P2P key sharing this will never happen.
And just how do you make it so the program does things better? If you default all keys to die in two years then you have to get all the signatures by other people again for your new key, and all the people using your old key have to reaquire you key again. Those two things are a huge hassle.
Doing key verification and revokationg via a server is the only sane way to do this. So having keys that never expire or expire after say, 10-20 years and are checked regularly against the server is the best way.
iF you'd even read the paragraph summary at the top of this story you'd see that it's easy remove keys from this central registry if you have control of the email address associated with them. But I agree that in the past this has been a real problem.
Slashdot Mirror
It's just a matter of time, as well, it's not such a big deal to have your AI or actual render threads separated out. It's amazing what getting rid of a few context switches can do for performance.
Not to bash your point too much, but my brain works that way :) I remember in school we had to write an assembly program in gcc and I used jmp instructions to dance around a block of code so much (I was trying for the smallest program) that I got my A paper back with my favorite teacher comment ever... "too complex."
I understand that NAT is considered a hack, but isn't the fact that a device's real address is hidden a security feature for the user? Wouldn't it be that much harder for malicious users to track my internet usage? This would be especially true if I had a mobile device, since moving from one NAT system to another would make following my movements remotely more difficult. So I'd think NAT would be considered a privacy boon. The article doesn't really address this effectively. Also, since most mobile devices have limited bandwidth, I'd think that having a constantly changing IP address, or hiding behind a NAT would mean that DOS attacks against them would be more difficult. If most big mobile device ISPs like the blackberry and sidekick folks offered NAT based access in the future, I'd think that we'd be relatively safe from IPv4 address exhaustion. So stating the main reason for IPv6 being address exhaustion I think is crap. It IS very useful for other reasons though, and I think those reasons warrant it being switched to.
Actually, what's really needed is a really smart AI that you can just talk to and say "Yo, computer, mess with this stuff and make it cool". But seriously, I think the trend towards more conversational :) programming is leading to a place where we will eventually just use voice commands to get the computer to do most common programming tasks. Obviously the real science related heavy math alg. stuff will still need to be programmed by hand until one of the sentient AIs decides it's time to take over.
From what I've been reading asia is going IP6 much faster then we (USA) are. In China & other places I think it's because they're still building a lot of infrastructure so they can start with the latest.
Maybe you already know this, but you shouldn't use the framebuffer kernel driver with the proprietary nvidia drivers. It makes it much more finicky. Since I do almost everything in X I don't really need 1280x1024 VGA console :)
I don't understand all this excitement over fuel cells being used in mobile devices. No electricity isn't everywhere, but then again, neither are hydrogen stations to buy refills at. If you think electricity is spotty in the 3rd world, just wait till you try to get hydrogen. This whole idea is dumb. Electricity is the most commonly available energy source. I can understand making cars and other bigger machines run on hydrogen, since they already have refueling infrastructure, but the energy problem could be better served by using higher efficiency transmission of electricity with something like the SuperGrid would be a much more achievable goal and easier to phase in then a switch to hydrogen.
With so many places in the world don't have enough water already, or their water is poor quality, for them getting hydrogen for water would not be a good thing.
This reminds "Prey" by Michael Crichton where they use bacteria to create nanobots. I know it makes sense from a production standpoint, but man I really don't like the possibilities here.
This reminds me of the line from a good book "Scientists were so caught up with what they could build, that they never stopped to think if they should" Maybe I don't WANT artificial life forms roaming around inside my body :) Thing is, once you release (or something escapes) like this that can live at the single cell level what's to stop it from spreading? At least with nanobots they can be made to not reproduce.
I don't claim to understand the whole Jabber protocol, but it seems to me that having decentralized servers might be a security problem, and be a disadvantage when everyone and their brother can throw together a jabber server.
Once Google allows the other servers to be used with it's own service, it's giving them legitimacy. That's something that people should be careful with. It would be cool if there was some kind of verification system with the server operators, say something similar to personal SSL key verification that people like Thawte were doing before being aquired by Verisign.
There's also the issue of everyone using google's server's as the nexus of the entire jabber network if google started working with other jabber servers.
I just hope they design these things to be secure (however you do that) against hijacking. Imagine what a supersonic plane or two could do to a city... ouch. Well, hopefully we'll get flying cars and floating highways soon that allow people to spread out more and not be so concentrated in cities.
I'm sorry, but anyone that thinks a 1 button mouse is the height of user interface design needs to be shot. Context menus make so much more sense being available with a right click. That aside, OS X does have some nice design principles, and like all developers I think OSS should seek to stea^F^F^F^Fcopy those features.
I think one of the better net admin jokes on this date was using the swedish chef text filter on all webpages in certain sections of the my college's site :)
http://www.cs.utexas.edu/users/jbc/home/chef.html
Tethers can be used in space to do everything from adjusting orbits to generating electricity from orbiting planets that have magnetic fields.
p e. shtml
http://www.strangehorizons.com/2003/20030414/ro
http://www.vectorsite.net/tarokt5.html
Tethers of (only) 20km long or less are much less likely to have problems then some ground to space tether that has to deal with wind conditions and other weather on the ground.
First off the author is looking at the wrong thing. For Linux to become mainstream means it needs to be commercialized and accepted in corporate arenas. When you ask your friends what versions of Linux they use you're mostly talking about personal preference work hobbiests/enthusiasts. What he really should have done is looked at the situation from big business' perspective.
Yes, Linux on the desktop isn't quite there. But it doesn't have to be right now either. The important thing that author has missed is that it's going in the right direction. It's not getting more fragmented. The LSB project has all the major players commited who could kill Linux by forking. All the major Linux distros are fully aware of the danger of them forking. So the LSB is going in the right direction and I think the kernel development process is as well (though that isn't the real issue here).
As far as ready-for-the-desktop goes, the big development there is the current maturing of the plug-and-play model in Linux. Eventually all distros will make it possible to never have to manualy mount your CD Rom and USB Flash drives manually. Right now for many self-roled kernels (and for me Slackware) out of the box automounting doesn't happen. That's an issue. But overall the picture is getting better not worse, and Linux' momentum is not going away quite yet.
The truth is though, that in the big picture the hobby distros don't count in the race to global domination. It's the big 3 that do. So in forgetting that fact, the author of the article missed the boat.
Well what I'm saying is, if people did know it was illegal and didn't care (which is what I'm assuming is the case here) then we're talkinga bout avoiding prosecution. I would think that on a small island with no/few international legal agreements it would be possible to maintain anonymity and thus avoid lawsuits, and since you're paying the dough, your site wouldn't be taken down so you could maintain web presence.
My question is, why do people host possibly infringing sites in territories that outlaw it? I mean why don't all these bittorrent sites host on little jungle islands in the pacific or other places that have no/fewer/looser international law agreements?
Maybe it's just ignorance.
Ok, enough conjecture here's some numbers. Assuming that a particular key would take 1billion years to crack by today's computers. That same key would be crackable in ~1 month using computers available 50 years from now. (Assuming 18 months to processor speed doubling, and that my math is right). You can call that irrelevent, or besides the point if you want. But that's a much smaller number to work with then 1 billion. I'm not saying that isn't a formidable number, just that it's not the "uncrackable" 1 billion people bandy about all the time. If we're talking 4 billion that would take only 3 years longer in order to crack it in 1 month. And all of this is assuming that people weren't working on the keyspace during that time period, which would cut the time required dramatically. However, I still agree with you that for most, people they don't have to worry about anyone carying about their encrypted data in 50 or 100 years :) I'm just saying, people get all excited about these huge numbers when their data isn't quite that safe.
I have to say this is the best example of journalism regarding Linux that I have seen yet. It is comprehensive and doesn't just regurgitate all the usual stuff for corporate drones. It actually has somewhat rare/new content about how actual decisions were made etc. I wish there was an email/feedback link to the author, cause I would tell him this myself.
I agree that it's still ok for near term stuff. But if you actually want something to NEVER be unencrypted you can't get all cozy with times like 1billion years to crack your messages.
Use of such numbers is a PR stunt that doesn't really reflect reality.
... is that people don't factor in that computers increase in speed by almost a factor of 2 each 18 months or so. So basically, that's like removing 1 bit from your encryption key every 18 months. Factor that in, and you begin to see cracking time numbers orders of magnitude smaller then before.
So basically, public key encryption should still only be used for things that are time sensitive that no one will care about in 5 years. If that's not the case, then you should think about using extremely large keys, like 4000bit, or use quantum encryption.
There is not such thing as absolute security/secrecy.
Am I the only one that thinks that plane is the ugliest I've ever seen? I understand function coming before form, but man, that thing is ugly.
Being american and somewhat nationalistic I'm rooting for Boeing but I still am glad to see a good competition with Airbus forming. That's good for everyone.
You're right, that is what PGP has been all about. And that is also why it never took off. It's too much of a hassle for people to distribute/verify keys and also to revoke them. In order for something like PGP to be used it has to be ubiquitous. With P2P key sharing this will never happen.
And just how do you make it so the program does things better? If you default all keys to die in two years then you have to get all the signatures by other people again for your new key, and all the people using your old key have to reaquire you key again. Those two things are a huge hassle.
Doing key verification and revokationg via a server is the only sane way to do this. So having keys that never expire or expire after say, 10-20 years and are checked regularly against the server is the best way.
iF you'd even read the paragraph summary at the top of this story you'd see that it's easy remove keys from this central registry if you have control of the email address associated with them. But I agree that in the past this has been a real problem.