Gpcode

Because you haven't read it properly. What i said is that if they crack this they could crack anything. Why? I highly doubt there is any crypto glitch this time and if they manage to crack RSA itself... do i have to tell more?

 

Yes, as a matter of fact you do, because I don't see for what reason successfully cracking a 1024 RSA would imply that any other algorithm with any other key size can also be cracked. It just doesn't make any sense.

 

Oh dear. Nevermind...

 

Very good argument, I'm convinced now. If they can crack 1024 RSA some way, they can obviously also crack anything else with any other key size.

And why not?

Never mind that it can be a flaw in the implementation, never mind that it can be a flaw in the algorithm that the NSA knows but that that is not present in other algorithms, never mind that (as unlikely as it is that they have it) the computing power to brute force a 1024 bit key is nothing when it comes to say, a 4096 bit key, never mind that that are completely uncrackable encryption schemes such as OTP. I suppose I must believe your "oh dear nevermind" and think "yeah, if they can crack this, they can crack ANYTHING"...

 

RejZoR,

Like I said in post 22, it depends on what they find is broken. If it is just the implementation in Gpcode, then the effect will be minimal. If there is a problem is the actual cipher, then there will be problems.

 

saw this for extra info:
http://people.csail.mit.edu/tromer/gpcode/

This looks very bad

 

Well a "partial" recovery of files is now possible but not by decrypting the files:

http://google.com/translate?u=http:...php?showtopic=72853&hl=en&ie=UTF8&sl=ru&tl=en

 

I doubt it.

The order the files are encrypted will most likely prevent recovery. You have file A being encrypted, that is copied to encrypted file CA with file A then deleted. Then you have file B encrypted by being copied to CB and so forth. The problem is when file B is copied to encrypted file CB, it will most likely overlay the space used previously by file A and then some (encrypted files are usually longer). This cycle repeats with every encryption so that the constant creation of encrypted files is enough disc activity to overlay most of the free space from the original un-encrypted files that were deleted.

 

This is not a mistake in the implementation of the crypto but an implementation issue nonetheless. The deleted files weren't wiped so it may be possible to recover the files using undelete tools.

Compare the virus to AxCrypt (not a virus but a well implemented file encryption program). With this program, when files are encrypted, the actual file is actually wiped from the drive so it can't be recovered.

 

Source: Kaspersky Weblog

 

IMO RejZoR's "Oh dear Nevermind" was intended as a graceful way of backing off from earlier statements. Hey TNT old bean -- no use shooting the wounded, wot?

Meanwhile, back at the thread -- am I missing something here? If GPcode or such got into my computer's knickers, then (once I recognized the problema) I would merely restore an uninfected image and Poof! -- no more GP code. Uh... is there anything erroneous with this concept?

 

Nope but how many of us keep backups?

 

don't count on me

 

If by "us" you mean computer users as a whole, then the answer is probably "very few."

However, if "us" refers to the denizens of Wilders, IMO the answer would be "many."

I not only do periodic full back-ups, I do so on external drives that are turned off when not in actual use. I learned to do this from the many helpful & knowlegeable posters here at Wilders. They taught me WHAT I should do. Threads like this present GPcode thread teach me WHY I should do it.

These are just a few of the reasons why I hang out here a lot.

 

Yes by 'us' I mean the general population. Which is very little. That is why viruses like Gpcode work. Because people just dont think to backup.

I do full backups to an external harddrive and back up documents online via mozy.