First order bias isn't a problem, most systems use something similar to von Neumann's bias elimination: pair the inputs up, and discard 00's and 11's. If the input is 01, output 0, if the input is 10, output 1.
the system I use redhat packages on started off as a redhat 5.1 install. Over the years I've upgraded bits and pieces but never done a complete upgrade. The system is a bit of a shambles, but most of that isn't due to the redhat parts. Upgrading rpm with rpm might be a bit of a pain in the neck, but it's quite doable.
Depends on what you want to use it for. It's easy to generate two pieces of data that hash to the same value, but so far I haven't seen any techniques for generating collisions for a given hash ("inverting the hash").
MD4 is very fast though, especially on x86 hardware. All in all I'd say the edonkey people made a good, informed decision.
Check the entire message, it's probably multipart/alternative with junk in the text/plain part to reduce it's score with spam filters, and the actual spam in the text/html part (which the most likely to respond to spam will be reading anyway)
This is the place in this thread where the maths jokes come to hang out? Okay, here goes...
What's the contour integral around Western Europe? Zero, because all the Poles are in Eastern Europe! (Actually, there ARE some Poles in Western Europe, but they are removable)
Why did the math professor name his dog Cauchy? Because he left a residue at every pole!
Quite doable, nearly all dotmatrix printers are chainfeed and support running the paper back the other way. So you just roll the paper back a few pages and print all black (or dense random dots) over it a few times. Good luck getting your logs out from under that.
It doesn't matter if it's albert fucking einstein, there is no functional difference. Both methods have the same result: a blob of disk from which you can recover a few partitions of data, and cannot prove there aren't any more.
It's still no different than filling a number of partitions with random binary garbage (or if you're paranoid about the randomness of the ciphertext of your chosen cipher, you encrypt 0's with a random key which you discard after initialisation), and using the rest.
A harddrive can only store as much data as it can store, so there's always an upper limit to the amount of data you can store. Using rubberhose you may be able to store even less than that (I don't know rubberhose), but that doesn't add any security. Once they know you're using stego (at which point it really isn't stego anymore), the two are equivalent.
Nice, but considering it'll be obvious from the start THAT you're using "stego" software, you might as well just make a number of encrypted partitions using linux cryptoAPI loopback, and only use a few of them.
Re:Better have a high-quality surge suppressor...
on
Aquarium Modcase
·
· Score: 1
How exactly are the fish supposed to get electrocuted? They're swimming around in water with all kinds of stuff dissolved in it, which has a much lower resistance than the fishies.
Re:Some things to point out.
on
Perl 1.0?
·
· Score: 4, Informative
There still isn't a switch statement you know... Well, not in perl 5 anyway. There'll be one in perl 6.
The text in the FAQ is mine, I don't think Ian is claiming freenet will get better throughput/latency for browsing random websites, however freenet can be faster for downloads from websites, websites with flash animations or big applets, those kinds of things. (At current the anonymity filter (a piece of code that filters potentially anonymity-compromising parts from freenet websites) will remove all plugins and applets, but we're talking middle to long-term future here).
Basicly, freenet latency is bad, freenet throughput is good. (and freenet reliability is different:-))
Possibly, yes. But any country in which the courts rule forwarding requests to illegal content to be illegal will also have made public HTTP proxies (like anonymizer.com) illegal. At that point a different system will be necessary, something where it would be impossible to distinguish publisher/retriever from 'innocent' bystander. There are possibilities for such systems, but they're going to be even less efficient/fast/simple, so let's hope they aren't necessary.
You don't need to have room in your datastore for all the files you're inserting, your machine won't be the only machine that ends up holding the data. You can insert 100G of mp3s with only a 100M datastore. As for large numbers of files not being handled properly, we've never had reports of it. If you have a reproducible bug, please report it.
Bollocks. If you want to share many files on freenet it's a small shellscript away. It is however going to take you a significant amount of time to insert your 100GB mp3 collection into freenet. This isn't because the developer cabal doesn't want you to share many files, this is because of the way freenet is designed. If you were to leave the files on your computer only until someone requested them it wouldn't be very anonymous.
(as for the first gripe, granted. Though the new NIO release is a significant improvement, though it doesn't cure the main problem that it's written in java.
You have to distribute the preferred form for editing, at least for the GPL. So encrypted source wouldn't be allowed. If you wanted to design a new language, convert the GPLed code you want to use to that language and write a compiler for that language which you don't distribute, you could do that however.
How about freenet? Got your anonymity straight out of the box, and if you're happy with just passing patches to the maintainers you've already got messaging via frost or FMB over freenet. There is some thought going on about a RCS over freenet, but that's not there yet.
Slashdot Mirror
First order bias isn't a problem, most systems use something similar to von Neumann's bias elimination: pair the inputs up, and discard 00's and 11's. If the input is 01, output 0, if the input is 10, output 1.
the system I use redhat packages on started off as a redhat 5.1 install. Over the years I've upgraded bits and pieces but never done a complete upgrade. The system is a bit of a shambles, but most of that isn't due to the redhat parts. Upgrading rpm with rpm might be a bit of a pain in the neck, but it's quite doable.
Or you could just download the RPM's listed in the RHSA and install those.
Who's to say? It hasn't been released yet.
Depends on what you want to use it for. It's easy to generate two pieces of data that hash to the same value, but so far I haven't seen any techniques for generating collisions for a given hash ("inverting the hash").
MD4 is very fast though, especially on x86 hardware. All in all I'd say the edonkey people made a good, informed decision.
Check the entire message, it's probably multipart/alternative with junk in the text/plain part to reduce it's score with spam filters, and the actual spam in the text/html part (which the most likely to respond to spam will be reading anyway)
This is the place in this thread where the maths jokes come to hang out? Okay, here goes...
What's the contour integral around Western Europe?
Zero, because all the Poles are in Eastern Europe!
(Actually, there ARE some Poles in Western Europe, but they are removable)
Why did the math professor name his dog Cauchy?
Because he left a residue at every pole!
Quite doable, nearly all dotmatrix printers are chainfeed and support running the paper back the other way. So you just roll the paper back a few pages and print all black (or dense random dots) over it a few times. Good luck getting your logs out from under that.
Perhaps you'd like to read the text you quoted again.
It doesn't matter if it's albert fucking einstein, there is no functional difference. Both methods have the same result: a blob of disk from which you can recover a few partitions of data, and cannot prove there aren't any more.
It's still no different than filling a number of partitions with random binary garbage (or if you're paranoid about the randomness of the ciphertext of your chosen cipher, you encrypt 0's with a random key which you discard after initialisation), and using the rest.
A harddrive can only store as much data as it can store, so there's always an upper limit to the amount of data you can store. Using rubberhose you may be able to store even less than that (I don't know rubberhose), but that doesn't add any security. Once they know you're using stego (at which point it really isn't stego anymore), the two are equivalent.
Nice, but considering it'll be obvious from the start THAT you're using "stego" software, you might as well just make a number of encrypted partitions using linux cryptoAPI loopback, and only use a few of them.
How exactly are the fish supposed to get electrocuted? They're swimming around in water with all kinds of stuff dissolved in it, which has a much lower resistance than the fishies.
There still isn't a switch statement you know... Well, not in perl 5 anyway. There'll be one in perl 6.
(oh, 5.8 has "use Switch;", but that's cheating)
Wouldn't that be 'reachable'? Not that my german's any less rusty than the grandparent post's poster.
The text in the FAQ is mine, I don't think Ian is claiming freenet will get better throughput/latency for browsing random websites, however freenet can be faster for downloads from websites, websites with flash animations or big applets, those kinds of things. (At current the anonymity filter (a piece of code that filters potentially anonymity-compromising parts from freenet websites) will remove all plugins and applets, but we're talking middle to long-term future here).
:-))
Basicly, freenet latency is bad, freenet throughput is good. (and freenet reliability is different
Possibly, yes. But any country in which the courts rule forwarding requests to illegal content to be illegal will also have made public HTTP proxies (like anonymizer.com) illegal. At that point a different system will be necessary, something where it would be impossible to distinguish publisher/retriever from 'innocent' bystander. There are possibilities for such systems, but they're going to be even less efficient/fast/simple, so let's hope they aren't necessary.
You don't need to have room in your datastore for all the files you're inserting, your machine won't be the only machine that ends up holding the data. You can insert 100G of mp3s with only a 100M datastore. As for large numbers of files not being handled properly, we've never had reports of it. If you have a reproducible bug, please report it.
Bollocks. If you want to share many files on freenet it's a small shellscript away. It is however going to take you a significant amount of time to insert your 100GB mp3 collection into freenet. This isn't because the developer cabal doesn't want you to share many files, this is because of the way freenet is designed. If you were to leave the files on your computer only until someone requested them it wouldn't be very anonymous.
(as for the first gripe, granted. Though the new NIO release is a significant improvement, though it doesn't cure the main problem that it's written in java.
You have to distribute the preferred form for editing, at least for the GPL. So encrypted source wouldn't be allowed. If you wanted to design a new language, convert the GPLed code you want to use to that language and write a compiler for that language which you don't distribute, you could do that however.
Not sure about BSD, but linux has ltrace which monitors all dynamic library calls. A real lifesaver in many situations.
Almost there, wrong level. The 802.11 frames contain the mac addresses, the (tcp/)ip stack sits a level higher. Still, end result's the same.
How about freenet? Got your anonymity straight out of the box, and if you're happy with just passing patches to the maintainers you've already got messaging via frost or FMB over freenet. There is some thought going on about a RCS over freenet, but that's not there yet.
Ehm, no, he passed to the dark side too. Sorry.