8 ms·
XT was intentionally designed to split the Bitcoin transaction history into two rapidly-diverging versions - the XT and official chains - each with enough suppo
by makomk 11y ago
XT was intentionally designed to split the Bitcoin transaction history into two rapidly-diverging versions - the XT and official chains - each with enough support to continue to exist for some time. This would have killed Bitcoin - and I mean actually killed it, complete with plummeting exchange rate and termination of trading and acceptance all over the place, not just caused a disgruntled ex-developer to write blog posts about how it was dead. Even supporters of larger blocks were largely unwilling to support it for this reason.
This was a deliberate design decision. Mike Hearn didn't think he was going to get widespread support, so he set the threshold for it to fork Bitcoin at 75% of the last 1000 blocks indicating support, which actually means it would activate with less than 75% support from miners due to variance and the fact that he was retesting the condition every block until it succeeded. Potentially quite a lot less, if I remember correctly, maybe even a minority.
- teacup50 11y agoAnd yet, miners still had the choice whether or not to run XT. If enough did so, then isn't that a vote? If Bitcoin can't survive without centralized control over miner software, isn't that a sign of broken decentralization?
- makomk 11y agoEnough miners did not in fact do so, and the "centralized control over miner software" was anything but - bitcoin.org pulled links to the one or two services that used XT because there was a huge amount of demand from Bitcoin users to do so, and they were making the same demands of everyone else. Also, part of the problem was that the threshold for enough miners was intentionally set dangerously low. If it had succeeded then I'd say it would be a sign that Bitcoin was fundamentally broken and couldn't survive without centralized control, but it didn't.