In December 2014, I was in NYC with my wife for the Blockchain(dot)info Christmas party.
We played laser tag, had a great staff BBQ dinner, and went back to our Airbnb to get some sleep.
I woke up to a nightmare.
The creator of the site pushed an update to the wallet, which resulted in an entropy bug. Any user who created a wallet during a specific timeframe was generating insecure addresses. This made the funds very easy to steal.
~1000 BTC were stolen, but some were returned. How?!
White hat legend, Johoe, is how.
He posted about it on BitcoinTalk:
He would go on to "steal" funds before the hacker(s) could take all of them, and he returned them to Blockchain. We obviously paid him a bounty.
We were able to make up the difference and pay everyone back who was affected by this. Who ended up having to figure all that out?
Me.
I had to go through each individual case, use a complicated method to make sure I was refunding funds to the actual person who lost them, and then send the funds.
Yes, I was given hundreds of Bitcoin and spent the next few weeks distributing them to the people impacted by this entropy issue.
I still have nightmares. The burden of responsibility weighed heavily on me at the time. I was so stressed about screwing something up, triple and quadruple checking everything before moving on to the next step.
This Coldcard situation is many magnitudes worse.
I feel for those immediately impacted by this - the ones who lost their funds.
The hardest part of working customer service at an early bitcoin hot wallet was explaining how Bitcoin transactions couldn't be reversed, and any lost funds were gone forever.
I also feel for everyone at Coinkite.
It's easy to hate them right now, but those guys are true Bitcoiners. A lot of them didn't need to keep working, and may very well vanish from the space after this. It sucks.
Why risk your reputation and everything you've built if one little mistake in 5 year old code can ruin you?
Why keep building when you can retire and enjoy life on a beach somewhere?
I suspect many are thinking this very thought right now.
Can't say I blame them.
Reused R values again
Reused R values again


