Well, the best word I've heard is servant or something like that, but what I read into your post, was that you suggested a server run by the university, which hasn't been an issue since napster. I don't even know if you can set up some kind of "hub" or "server" for the newer p2p programs, don't they just get a list of known hosts from some location and start from there? If so, then I don't see how you could really keep it local, except by making some sort of subnet/vpn/whatever.
I agree with all your points, except the one about the need for a programming language in the application.
Although the choice of language may be up for debate, I think on of the good points about Word, or Emacs which I personally prefer (It's been years since I've used Word on a regular basis), is the fact that they allow extensions to be made without changing the app itself. While I think that open source is a great thing, it helps a lot to be able to express the things you want to do with a program in terms of that program, rather than in terms of the implementation language/environment. It also should help in portability and maintainabillity, since the app may keep it's interface relatively stable (or just backwards compatible).
For those too lazy to follow the link, LFS is instructions on building a working (minimal/basic) linux system from sources. This can be quite a bit of work, but in return you get to have a lot of say in how your system works (depending on how adventurous you happen to be), although there are excact instructions (with the reasons for many of the choices).
Also there is the ALFS project (automated LFS) and the nALFS tool (which is great, I've built an entire system using it), which make it much easier to rebuild your system, if need be.
If you would like a system where you decide how things are done, LFS is certainlly worth a try.
Re:Preempt patch and 3rd party modules
on
Kernel 2.4.17 Out
·
· Score: 1
Have you tried to rebuild the modules, this used to fix similar problems for me. I know the modules are binary, but there is still some linking to be done before they load on a given kernel. I usually just make modules modules_install and rebuild any seperate modules (nvidia!) every time I mess with the kernel (not that often)...
Well, if the laser is set to protect your forces in a specific area, then no friendly rounds should be heading into this area. If there are, then either your defence zone is set too wide, or you are fireing close to your own troops, for which there should be special procedures. Or you've made a mistake, in which case it is good to shoot down the round. Basicly you just need to determine the approximate target of the trajectory, and match that against you troop deployment.
Could be a different model, but at my university (Aarhus, Denmark) we have airports all over the cs department (two different sites in fact). And I'm very sure it's directly on the ethernet (propably a different wlan or something, I'm not sure). Pretty sweet anyway... especially since hardly anybody uses it (I work as a student programmer, and we only have a few notebooks with 802.11b), so streaming video and stuff is not a problem.....
You are absolutely right, utilities have to support big files. I don't know how you made dd barf, but from my experience the one in the later Suse dists. have no problems with big files, or at least big device-files.
Once I used it to copy an entire 18 GB partition of a bad harddrive. The partition happened to be reiserfs, and the harddrive had gotten bad somewhere critical (no amount of tools/remounts/prayers/swear words could make it work), but dd just stopped when it reached the bad parts, so I restarted at the next good block, sticking the result together on a new partition. Since there were only a few errors in the early part of the partition, I think it was less than 1 GB I copied patchwork style. Great thing was that the result was a mountable partition, and hardly any data was lost:-).
So under the right conditions (some versions of) dd handles large amounts of data nicely.
And what's to prevent me from making a trojan out of it??? "This demonstrates an exploit in your system" could really mean "This fucks your system up beyond your wildest nightmares, no exploit needed, you executed it, fool!"
But still, the point is that this is being done by ME (or you or whoever) (ok, so I use software to do it, so what, I use software for lots of stuff, thats to annoying to do by hand), to my own copy, that I got from the server, just like everybody else, and I'm NOT distributing the copy. So unless you have better arguments, I'll view it as I see fit, I might even use lynx... hah...;-)
Well, there is a weakness in your argument. Since some of the tools you suggest look at the abstract syntax/compilation trees, they might catch more than what is wanted. Especially in this area, since, as mentioned elsewhere in this thread, the structures in question MUST be compatible with the hardware, and thus MUST have the same layout. So basically, if you abstract the comments, names and layout away, there can be NO difference between the two.h files, or one would be wrong.
I agree, that it can be a good model, for the consumers. But since it is the choice of the manufacturer, how to create a business model around some product, I don't go around feeling guilty if I should be so lucky to get cheap/free stuff because of some loss leader business model.
Basically I want to buy the things I think are a good deal, at the price they are sold at (or lower...;-) and let the people making the products worry about their end. Should they loose money on me, then they can regard it as a mistake or a statistical anomaly, but they (or anybody else) shouldn't start blaming me... they sold it, so it's their problem.
also, passwords/phrases are rarely shown, except as ****** or the like, and while the signals are there, I should think that Van Eck Phreaking gets a lot harder, when you are trying to get somethng like keyboard and mouse, than if it's a crt with a lot of wellknown frequencies (or ratios of same).....
When you say uncrackable, do you mean something like "would take enormous resources and much time" or do you mean flat out uncrackable? Just wondering... 'cause it seems like you know that there is no proof, that pgp is strong, just conjecture, and yet you use "uncrackable" which sound very much like something absolute???? While it would of cause be useless to bruteforce pgp under normal circumstances, it is possible, it is just that the universe might cease to exist before you got lucky...;-)
Well, this raises an interesting question... Would they have been able to find the passphrase within a reasonable timeframe? I think the tool used was PGP, and although the passphrase is not as strong, cryptographically, as the private key it unlocks, it might still be very large. Thus the feds might never have been able to crack the encryption without tapping the keyboard... Just something to think about...
Since when is DeCSS lame???
Surely you must be talking about the CSS encryption scheme, which is very lame indeed.
But to call DeCSS lame... OK, there are smarter, smaller, more efficient implementations, in more exotic languageges, but lame???
This makes me wonder... would you feel the same if I avoid ads by using lynx, or browsing with an old/uncommon (graphical) browser, thus rendering the pages in a way other than what the designer thinks it should look like in Internet Exploiter???
The way I see it, html was never designed to inforce any particullar way of formatting the pages, so if I'm not satisfied with the way a page looks, its ads or anything else, I will change it if I can.
Now if they were to charge up front for access, and use some sort of access control (beyond not publishing the link), then you could begin talking about theft (or whatever term you find best in this context). But as long as their server willingly sends the info my way, for a simple http-request on port 80, then I will view the result in whatever way I see fit... How about netcat and a nice print filter...;-)
I think the idea is, that the file is decrypted LOCALLY on the handheld. Thus there is NO cleartext document being communicated over GSM/Bluetooth for the feds to catch. Of cause they can still try to get at the handheld, but such is life...
Well, isn't it actually the trafic they pay for? Because that might very well increase, if you connect more boxes (assuming you are not already saturating your link;-)
Not thatI would even pretend to know.... But wouldn't it be possible then to take the signature of a "good" client and implant it into a hacked client???
Slashdot Mirror
Well, the best word I've heard is servant or something like that, but what I read into your post, was that you suggested a server run by the university, which hasn't been an issue since napster. I don't even know if you can set up some kind of "hub" or "server" for the newer p2p programs, don't they just get a list of known hosts from some location and start from there? If so, then I don't see how you could really keep it local, except by making some sort of subnet/vpn/whatever.
Isn't P2P server an contradiction in terms (oxymoron???)?
You're not worried that Bill might ask to have his monopoly back? Or are you just hoping he gets a clue too?
I agree with all your points, except the one about the need for a programming language in the application.
Although the choice of language may be up for debate, I think on of the good points about Word, or Emacs which I personally prefer (It's been years since I've used Word on a regular basis), is the fact that they allow extensions to be made without changing the app itself. While I think that open source is a great thing, it helps a lot to be able to express the things you want to do with a program in terms of that program, rather than in terms of the implementation language/environment. It also should help in portability and maintainabillity, since the app may keep it's interface relatively stable (or just backwards compatible).
For those too lazy to follow the link, LFS is instructions on building a working (minimal/basic) linux system from sources. This can be quite a bit of work, but in return you get to have a lot of say in how your system works (depending on how adventurous you happen to be), although there are excact instructions (with the reasons for many of the choices).
Also there is the ALFS project (automated LFS) and the nALFS tool (which is great, I've built an entire system using it), which make it much easier to rebuild your system, if need be.
If you would like a system where you decide how things are done, LFS is certainlly worth a try.
Have you tried to rebuild the modules, this used to fix similar problems for me. I know the modules are binary, but there is still some linking to be done before they load on a given kernel. I usually just make modules modules_install and rebuild any seperate modules (nvidia!) every time I mess with the kernel (not that often)...
Well, if the laser is set to protect your forces in a specific area, then no friendly rounds should be heading into this area. If there are, then either your defence zone is set too wide, or you are fireing close to your own troops, for which there should be special procedures. Or you've made a mistake, in which case it is good to shoot down the round. Basicly you just need to determine the approximate target of the trajectory, and match that against you troop deployment.
Could be a different model, but at my university (Aarhus, Denmark) we have airports all over the cs department (two different sites in fact). And I'm very sure it's directly on the ethernet (propably a different wlan or something, I'm not sure). Pretty sweet anyway... especially since hardly anybody uses it (I work as a student programmer, and we only have a few notebooks with 802.11b), so streaming video and stuff is not a problem.....
I'm think the Dell Inspiron 8100 looks even nicer... GeForce2 Go.... Any experiences??
You are absolutely right, utilities have to support big files. I don't know how you made dd barf, but from my experience the one in the later Suse dists. have no problems with big files, or at least big device-files. :-).
Once I used it to copy an entire 18 GB partition of a bad harddrive. The partition happened to be reiserfs, and the harddrive had gotten bad somewhere critical (no amount of tools/remounts/prayers/swear words could make it work), but dd just stopped when it reached the bad parts, so I restarted at the next good block, sticking the result together on a new partition. Since there were only a few errors in the early part of the partition, I think it was less than 1 GB I copied patchwork style. Great thing was that the result was a mountable partition, and hardly any data was lost
So under the right conditions (some versions of) dd handles large amounts of data nicely.
And what's to prevent me from making a trojan out of it??? "This demonstrates an exploit in your system" could really mean "This fucks your system up beyond your wildest nightmares, no exploit needed, you executed it, fool!"
But still, the point is that this is being done by ME (or you or whoever) (ok, so I use software to do it, so what, I use software for lots of stuff, thats to annoying to do by hand), to my own copy, that I got from the server, just like everybody else, and I'm NOT distributing the copy. So unless you have better arguments, I'll view it as I see fit, I might even use lynx... hah... ;-)
well, actually your two structs are not identical, the layout is diffenrent, the hardware expects to find the bits and pieces in the right order etc.
Well, there is a weakness in your argument. Since some of the tools you suggest look at the abstract syntax/compilation trees, they might catch more than what is wanted. Especially in this area, since, as mentioned elsewhere in this thread, the structures in question MUST be compatible with the hardware, and thus MUST have the same layout. So basically, if you abstract the comments, names and layout away, there can be NO difference between the two .h files, or one would be wrong.
I agree, that it can be a good model, for the consumers. But since it is the choice of the manufacturer, how to create a business model around some product, I don't go around feeling guilty if I should be so lucky to get cheap/free stuff because of some loss leader business model.
;-) and let the people making the products worry about their end. Should they loose money on me, then they can regard it as a mistake or a statistical anomaly, but they (or anybody else) shouldn't start blaming me... they sold it, so it's their problem.
Basically I want to buy the things I think are a good deal, at the price they are sold at (or lower...
also, passwords/phrases are rarely shown, except as ****** or the like, and while the signals are there, I should think that Van Eck Phreaking gets a lot harder, when you are trying to get somethng like keyboard and mouse, than if it's a crt with a lot of wellknown frequencies (or ratios of same).....
When you say uncrackable, do you mean something like "would take enormous resources and much time" or do you mean flat out uncrackable? Just wondering... 'cause it seems like you know that there is no proof, that pgp is strong, just conjecture, and yet you use "uncrackable" which sound very much like something absolute???? While it would of cause be useless to bruteforce pgp under normal circumstances, it is possible, it is just that the universe might cease to exist before you got lucky... ;-)
Well, this raises an interesting question... Would they have been able to find the passphrase within a reasonable timeframe? I think the tool used was PGP, and although the passphrase is not as strong, cryptographically, as the private key it unlocks, it might still be very large. Thus the feds might never have been able to crack the encryption without tapping the keyboard... Just something to think about...
Since when is DeCSS lame???
Surely you must be talking about the CSS encryption scheme, which is very lame indeed.
But to call DeCSS lame... OK, there are smarter, smaller, more efficient implementations, in more exotic languageges, but lame???
This makes me wonder... would you feel the same if I avoid ads by using lynx, or browsing with an old/uncommon (graphical) browser, thus rendering the pages in a way other than what the designer thinks it should look like in Internet Exploiter??? ;-)
The way I see it, html was never designed to inforce any particullar way of formatting the pages, so if I'm not satisfied with the way a page looks, its ads or anything else, I will change it if I can.
Now if they were to charge up front for access, and use some sort of access control (beyond not publishing the link), then you could begin talking about theft (or whatever term you find best in this context). But as long as their server willingly sends the info my way, for a simple http-request on port 80, then I will view the result in whatever way I see fit... How about netcat and a nice print filter...
I think the idea is, that the file is decrypted LOCALLY on the handheld. Thus there is NO cleartext document being communicated over GSM/Bluetooth for the feds to catch. Of cause they can still try to get at the handheld, but such is life...
Well, isn't it actually the trafic they pay for? Because that might very well increase, if you connect more boxes (assuming you are not already saturating your link ;-)
Just to nitpick:
It's "President Skroob"
Not thatI would even pretend to know.... But wouldn't it be possible then to take the signature of a "good" client and implant it into a hacked client???
Fierce Creatures.