Quote from: mat5x on May 14, 2013, 08:08:56 PM

Quote from: WindMaster on May 14, 2013, 06:03:00 PM

It seems Milkshake did manipulate the blockchain with these checkpoints. Now the question is why. An analysis of the first ~5000 blocks should prove insightful.

see the irc logs. i can paste the whole set if you want to read through about 500 newbs trying to learn linux in a coin channel.

there might be some logic to it, but there was _A LOT_ of worry about bogey-man "AWS GUY" (please choose a ragecomic for this mythical character) was fucking up the blockchain.

early on, around 1000 blocks in or so, maybe 2000, i saw a REORG of _120 blocks_ in one shot. damn i dont have the debug.log from then Sad

((( blew away .elacoin with a recompile to connect to more hosts at one point. my debug shows biggest reorg was 18 blocks now only. which is pretty big, but typical of early low diff launches.

he was paranoid so he checkpointed it in, fearing a total takeover of 1000s of blocks by 'aws guy'.

lack of experience ruled, methinks (i certainly dont have it, but there's a rumour going around that we all learn somewhere).

So basically, he intentionally deleted a significant amount of legitimately mined blocks out of paranoia? Not once, but twice. A significant number of miners were hard forked and their earnings erased by this choice by Milkshake.

Judging by the name "AWS GUY", how would someone have legitimate concerns that someone would rent enough servers to CPU mine and 51% a GPU coin? That would be rather expensive to accomplish, with basically no benefit for the person renting them.

When it comes to "fucking up the blockchain", isn't that what Milkshake did, twice? All of my blocks mined below block 2500, some with >2200 confirmations, were wiped out by this choice.

Quote from: mat5x on May 14, 2013, 08:08:56 PM

early on, around 1000 blocks in or so, maybe 2000, i saw a REORG of _120 blocks_ in one shot.

Oh, I certainly noticed the REORG's, since elacoind segfaulted on each one. I was running one of the two servers that I believe a significant number of people here used as a seed node (I had up to 750 inbound connections to my elacoind at one point). The person running the other server people were using as a seed node experienced the same crashes. On just about every REORG.