9

I don't understand this post.

What is "ArtForz modified exploit"?

What does "Gavin's patch will not work for the exchanges" mean?

Any more important details I'm missing?

Murch
  • 71,155
  • 33
  • 180
  • 600
ripper234
  • 26,452
  • 30
  • 111
  • 246
  • 2
    [This](https://bitcointalk.org/index.php?topic=42417.msg517020#msg517020) is supposedly the code that is attacked. The timestamps on the blocks are manipulated to fool the difficulty algorithm into accepting a large number of blocks with rapidly decreasing difficulties. The basic idea is that you manipulate the timestamps on the first and last blocks in a set and you can get the difficulty to divide by 4 every time. – David Schwartz Sep 14 '11 at 08:34
  • 2
    Anyone know what "Gavin's patch" is? – David Schwartz Sep 14 '11 at 08:44
  • 1
    This question is becoming quickly dated. Is there any case to be made that this content is of lasting value? – eMansipater Sep 28 '11 at 19:20
  • @emansipater why do we need to make this case? The question served its propose, it has an accepted answer, just let it be. – ripper234 Sep 28 '11 at 20:01
  • 1
    On Stack Exchange sites, questions without lasting value are usually closed as [too localised](http://bitcoin.stackexchange.com/faq#close) in time. The idea is that the site exists primarily as a reference, and that quality is maintained by keeping it restricted to content that people will find useful. – eMansipater Sep 28 '11 at 20:16
  • @eMansipater: it still has an historic value: "Why did BitcoinEXpress plan a 51% attack on Namecoin in Sep 2011?" – o0'. Oct 09 '11 at 16:18
  • @Lohoris - changed to past tense. – ripper234 Oct 09 '11 at 16:24
  • 1
    Just adding that even in late '13, this question has some historical value as documentation for this historical attack. – user239558 Aug 20 '13 at 07:12

2 Answers2

9

This post by ArtForz on the forums explains the attack:

https://bitcointalk.org/index.php?topic=43692.msg521772#msg521772

Quote: "By exploiting the fact that retargeting ignores one block interval every period, it's possible for an attackers' fork chain to "jump backwards in time" and create lots of blocks at low difficulty without running nTime off into the far future."

This attack has been proven to work on GeistGeld (which uses 10 second blocks and 16 block retarget intervals). It was rather easy there, because of the low intervals and because the developers "forgot" to adjust other constants regarding timing and clocks. They since made it much better.

Now it seems it will be tried on namecoin, with the goal of making bitcoin more secure.

molecular
  • 1,904
  • 1
  • 10
  • 12
6

The answer to the title question "Why is BitcoinEXress planning a 51% attack on Namecoin?" has been given in the linked post itself:

Goals

  1. Developing protection for Bitcoin against exploits
  2. Developing a lethal alt chain killer.
  3. Making developers aware that when the "bad guys" make you aware of an exploit, FIX IT!

We know the 51% attack could be a danger to bitcoin (there are several threads about this IIRC), so they are trying that attack against a bitcoin weaker clone to see "live" if it really works, to raise public awareness of this problem.

o0'.
  • 5,240
  • 6
  • 38
  • 66
  • The said reason behind why someone is doing something and the actual reasoning behind it may be very different. Just because this man declares his intentions as genuine it may not be the case. By destroying a live chain will only put fear and concern to the unknowning and media about the safety of any technology built this way. – MaxSan Sep 14 '11 at 10:56
  • 4
    The reason stated by him is a fact, while what he might be actually thinking is pure speculation hence unfit for a SE site. – o0'. Sep 14 '11 at 12:16