WOW ... ein sehr guter Abriss der IRC Ereignisse der Nacht. Liest sich wie ein Krimi. Bemerkenswert: Gavin war für 0.8. ..... 'Going backwards is the wrong thing to do, in my opinion'
Die anderen Entwickler und vor allem BTCGuild waren jedoch für 0.7 und so wurde es dann gemacht.
mtgox hat auch in Minutenfrist reagiert und Überweisungen eingefroren.
-->>
QuelleDie längste Zeit die vergangen ist, ist die während man auf das überholen der 0.8er Version durch die 0.7er Version wartete.
11:30pm problem starts to come to light with one user struggling with their version 0.7 bitcoin client.
00:00 The problem didn’t really come to light however until about midnight, when a couple of devs starting noticing something was up.
00:05 It took a mere 5 minutes for “Luke-Jr” to correctly identify the hardfork with the message “Luke-Jr: so??? Yay accidental hardfork? ”
00:08 Just three minutes after Luke-Jr first identified the issue, user “sipa” tracks the problem to a bug (as yet unidentified) in the version 0.7 software. Also within three minutes The MtGox Bitcoin Exchange confirmed it was on Version 0.7 “old” software.
At this point Luke-Jr raises a concern “my immediate concern is, which fork are the majority of miners on?” which is echoed for the next few hours.
00:11 The Mt.Gox exchange stops accepting bitcoin transactions (concerned they could be double spent)
00:20 a few people start to freak out, and messages such as “omg what is going on” and “oh fuck” begin to appear in the chat.
But – in the same minutes (less than 15 minutes since the hardfork was identified) Luke-Jr (aka “hero for the evening”) suggests what will eventually become the short term fix, saying to a fellow dev “sipa: seems to me the best short-term fix is to get miners off 0.8”
00:22 merely two minutes after this fix was suggested a discussion begins between the devs on whether or not they can reach a consensus on this.
This is when lead developed Gavin Andresen joins the conversation and is initially very resistant to the proposal.
00:31 while a consensus has not yet been reached, major mining pool BTC Guild begins to take action on rolling back their software to version 0.7
00:39 Gavin Andresen expresses strong resistance to the proposed solution “Going backwards is the wrong thing to do, in my opinion”
00:43 the bitcoin price on MtGox begins to suffer dropping $2 to $46/BTC
00:45 Despite Gavin Andresens initial protest, the devs eventually reach consensus when BTC Guild announces they have the balance of hashing power and all agree to request pools downgrade to version 0.7 as a matter of urgency.
00:46 A bitcoin gambling website which has been at the brunt of a lot of blame, Satoshi Dice, comes to the party and pauses all transaction processing, lightening the load on the network.
00:50 the bitparking mining pool begins downgrading their software to version 0.7
01:15 the Mt Red mining pool shuts down its services as a quick solution to reduce hashing power on the version 0.8 blockchain.
01:19 Major bitcoin payment processer BitPay intentially freezes transactions to reduce the impact of a potential (but unlikely) double spend attack
01:23 Although struggling through a number of issues surrounding moving large files between servers, BTC Guild mining pool finally downgrades the first of its servers.
02:36 the BTC Guild mining pool downgrade is completed across all of its servers.
03:00 A temporary bandaid patch for software version 0.8 is release which should allow it to rejoin the 0.7 blockchain and ditch the problematic 0.8 blockchain
06:21 Everyones hard work is reworded as the fix proposed at 00:20 dominates the blockchain and the version 0.8 blockchain merges once more with the version 0.7 blockchain
06:23 The Mt Gox exchange announced they would wait until a further 6 block confirmations came through to ensure there was a clear winning/leading chain. This wasn’t necessary, but a cautious safeguard.
Conclusion:
While I don’t doubt at some point another accidental hardfork event might take place (although this was the first in over 225,000 blocks and 4 years), If such an event were to take place I would sleep easy knowing the bitcoin developers and community have things in hand.
My only other takeaway from this situation is a slight loss of faith in the lead developer of bitcoin, Gavin Andresen, who in my mind could have acted quicker and more decisively, and could in theory have lessened the impact by a matter of hours.
My own view is that Luke-Jr was the star of the evening (although he assumed the problem was with 0.8 software, not with 0.7 software, which was an easy assumption to come to in the heat of the moment) – Sipa was also a stand out contributor of the evening, taking an equally strong role in the recovery and working extremely effectively with Luke-Jr. In my mind these two deserve the praise.
I do wonder if there is scope for the creation of a centralised (although voluntary) notification system for the key players and network at large to close ranks more quickly in any future event such as this one, but I also think this would need to be managed with a very level head. In my mind last nights even should have been an amber warning event, not a panic event, as the solution was clear, simple and quick.
One honest bitcoin user has reported he was able to double spend $10,000 worth of bitcoin, but I believe this is more down to the payment processors flaws than it is the network and bitcoin protocol.