Quote from: Jollyburner on March 07, 2014, 06:38:35 AM

Quote from: BitcoinEXpress on March 07, 2014, 06:13:25 AM

Quote from: efx on March 07, 2014, 06:04:58 AM

Now now, you probably didn't deduce the time warp aspect, BCX gave that away for free Wink

I would also be interested in further detail.

A blast from the blast from that genius kid ArtForz.

I wonder what ever happened to him ? ? ?

~BCX~

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

thanks, its almost as complicated as actual time-travel Cheesy

I think I get it tho.

but I don't understand this part:

Quote from: ArtForz on September 12, 2011, 09:39:28 PM

Now here's the weird part, we retarget after blocks 3, 7, 11, 15, and for block 3 we use 0 as first and 3 as last, for 7 we use 4 as first and 7 as last, ...
so what happens if an attackers chain has blk timestamps like this:

Code:

blk# 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
time 0 1 2 30 4 5 6 70 8 9 10 110 12 13 14 150

?
first period (#3 - #0) is 30s as before
2nd period is (#7 - #4) ... 66s
3rd period is (#11 - #8) ... 104s

what does minusing #4 from #7 have to do with it? why does it use only the timestamps from particular blocks to calc the diff adjustment?
how can u do less work but do more work? its because its not counting all the work its skipping some and only counting others towards the adjustment/ ?

plz excuse my ultra nub

Isn't this fixed, if the code walks throught the timescale and takes oldest and newest block (not lowest and highest) to calculate the time diff?