MtGox in trouble
-
Time to make some quick cash.
P&D ;D
Price is slowly creeping back up.
-
[quote name=“pauldodd” post=“57725” timestamp=“1391770996”]
This is a good article from someone that went to visit them.http://www.reddit.com/r/Bitcoin/comments/1x9gue/my_protest_at_mtgox_offices_5_to_7th_february/
[/quote]That’s a very interesting read. I feel sorry for him.
-
Press Release:
-
So basically MtGox is blaming Bitcoin for their withdrawal problems.
-
I thought people were also having trouble getting $'s outta the system too :?
-
[quote name=“Ruthie” post=“58179” timestamp=“1392031022”]
I thought people were also having trouble getting $'s outta the system too :?
[/quote]Yea seemingly that’s a separate problem.
-
A good article on technical aspects of the issues:
http://blog.oleganza.com/post/76213549017/mtgox-and-malleable-transactions -
Lol… I just caught 40 BTC at $300 USD each on BTCe.
The bot went berserk, and it tries hard to clear off as many BTC as it can on a huge sell-off. ( I assume its MT4, the sell off volume triggered some sort of ‘bankrupt’ signal.)
HFT used for the stock market doesn’t always work for Bitcoin ;D because the stock market does sort of cap the loss at max 10% daily.Lowest point today: 102 USD
Volume: ~20,000 BTC at that minute[img]http://imageshack.com/a/img30/8907/6nm8.jpg[/img]
Effectively doubled my holdings :P thank you! MTGox.
-
[quote name=“eaxvac” post=“58192” timestamp=“1392038096”]
Lol… I just caught 40 BTC at $300 USD each on BTCe.
The bot went berserk, and it tries hard to clear off as many BTC as it can on a huge sell-off. ( I assume its MT4, the sell off volume triggered some sort of ‘bankrupt’ signal.)
FYI: HFT used for the stock market doesn’t always work for Bitcoin ;D because the stock market does sort of cap the loss at max 10% daily.Lowest point today: 102 USD
Volume: ~20,000 BTC at that minute[img]http://imageshack.com/a/img30/8907/6nm8.jpg[/img]
Effectively doubled my holdings :P thank you! MTGox.
[/quote]holy fuck I missed this! :o
-
How long did it take to get back up from $102 ?
-
Sat there watching but was not logged in :/ so didn’t quite see the drop so slow.
-
[quote name=“futureproof” post=“58201” timestamp=“1392040187”]
How long did it take to get back up from $102 ?
[/quote]I think it went straight back up I was watching but wasn’t logged in so it didn’t refresh as quick as when you are logged in. (thats just a thought).
-
[quote name=“futureproof” post=“58201” timestamp=“1392040187”]
How long did it take to get back up from $102 ?
[/quote]It just kept flashing 100, 200, 400 for around 10 seconds until the BTC bag is fully sold. Back to 620 instantly.
Prove one thing for sure, no whale can ever dump to profit. They can only pump to attract greedy bustards into dump.
-
Check out realsolid’s comments …
[url=http://tuckfheman.com/post/76211621530/mtgox-bitcoin-exchange-halts-all-bitcoin-withdrawals]http://tuckfheman.com/post/76211621530/mtgox-bitcoin-exchange-halts-all-bitcoin-withdrawals[/url] -
I have been live tweeting the Gox drama directly from their hidden IRC channel https://twitter.com/Feathercoin
-
Makes me wonder how one can run a multimillion exchange and get fooled by a trick 3 years old. If a transaction doesn’t come through, must check inputs and outputs.
-
[quote name=“ghostlander” post=“58215” timestamp=“1392045005”]
Makes me wonder how one can run a multimillion exchange and get fooled by a trick 3 years old. If a transaction doesn’t come through, must check inputs and outputs.
[/quote]It’s actually a very hard problem.
You have an index, because god help you if you have to walk the blockchain every time you want to send from a wallet. So you index the ins and outs by wallet, and you use the most obvious thing in the world to id them: The transaction id.
And then suddenly, a new item appears in your index. Except really, it’s the old item, and the old item has actually been removed from the index, but no one told you, but what really happened is the new item just showed up with a completely different identifier. An identifier that is the result of a deterministic process which states that if the same data is used to form the id, the id will be the same.
Now, how on earth are you supposed to determine that this new transaction which has a new id is really exactly the same as the old transaction even though the id is different? This way lies madness.
Except it turns out that this is exactly the wrong way to approach wallets. Wallets are not a series of transaction ids. They’re a ledger of account changing ownership between entities, and if you’re following the bitcoin protocol, this means that each input references an output. If your index was correct, it would model that linkage of coins transfering between addresses, not the txid series, and such bugs couldn’t possibly happen: You’d correctly overwrite your index when the new modified transaction got included [u][i][b]because there can only be one spend of an output[/b][/i][/u].
[b]MtGox screwed up the protocol implementation.[/b] And it’s easy to see why: [i]It’s hard to do right.[/i]
-
It’s their fault, no doubts about it ( wrong implementation of the protocol ), but we must keep in mind that someone is exploiting that bug and that’s sort of attack. Everyone needs a time to recover from tech. issue specially when they transmitting virtual goods. According me the biggest error was when they mention bitcoin protocol as “broken” and got strong reactions from foundation… After that “To D Moon” guys came with the torches :D
-
[quote name=“Kevlar” post=“58429” timestamp=“1392106140”]
[quote author=ghostlander link=topic=7477.msg58215#msg58215 date=1392045005]
Makes me wonder how one can run a multimillion exchange and get fooled by a trick 3 years old. If a transaction doesn’t come through, must check inputs and outputs.
[/quote]It’s actually a very hard problem.
You have an index, because god help you if you have to walk the blockchain every time you want to send from a wallet. So you index the ins and outs by wallet, and you use the most obvious thing in the world to id them: The transaction id.
And then suddenly, a new item appears in your index. Except really, it’s the old item, and the old item has actually been removed from the index, but no one told you, but what really happened is the new item just showed up with a completely different identifier. An identifier that is the result of a deterministic process which states that if the same data is used to form the id, the id will be the same.
Now, how on earth are you supposed to determine that this new transaction which has a new id is really exactly the same as the old transaction even though the id is different? This way lies madness.
Except it turns out that this is exactly the wrong way to approach wallets. Wallets are not a series of transaction ids. They’re a ledger of account changing ownership between entities, and if you’re following the bitcoin protocol, this means that each input references an output. If your index was correct, it would model that linkage of coins transfering between addresses, not the txid series, and such bugs couldn’t possibly happen: You’d correctly overwrite your index when the new modified transaction got included [u][i][b]because there can only be one spend of an output[/b][/i][/u].
[b]MtGox screwed up the protocol implementation.[/b] And it’s easy to see why: [i]It’s hard to do right.[/i]
[/quote]Can something like this happen with ACP in place?
-
[quote name=“netnerd” post=“58431” timestamp=“1392107567”]
[quote author=Kevlar link=topic=7477.msg58429#msg58429 date=1392106140]
[quote author=ghostlander link=topic=7477.msg58215#msg58215 date=1392045005]
Makes me wonder how one can run a multimillion exchange and get fooled by a trick 3 years old. If a transaction doesn’t come through, must check inputs and outputs.
[/quote]It’s actually a very hard problem.
You have an index, because god help you if you have to walk the blockchain every time you want to send from a wallet. So you index the ins and outs by wallet, and you use the most obvious thing in the world to id them: The transaction id.
And then suddenly, a new item appears in your index. Except really, it’s the old item, and the old item has actually been removed from the index, but no one told you, but what really happened is the new item just showed up with a completely different identifier. An identifier that is the result of a deterministic process which states that if the same data is used to form the id, the id will be the same.
Now, how on earth are you supposed to determine that this new transaction which has a new id is really exactly the same as the old transaction even though the id is different? This way lies madness.
Except it turns out that this is exactly the wrong way to approach wallets. Wallets are not a series of transaction ids. They’re a ledger of account changing ownership between entities, and if you’re following the bitcoin protocol, this means that each input references an output. If your index was correct, it would model that linkage of coins transfering between addresses, not the txid series, and such bugs couldn’t possibly happen: You’d correctly overwrite your index when the new modified transaction got included [u][i][b]because there can only be one spend of an output[/b][/i][/u].
[b]MtGox screwed up the protocol implementation.[/b] And it’s easy to see why: [i]It’s hard to do right.[/i]
[/quote]Can something like this happen with ACP in place?
[/quote]Yes
