But complaining at MtGox will make it get fixed faster. Of course.
As long as MtGox was the only one affected by this, people spent their time complaining about MtGox. Now when they finally see what he issue is, they finally start eating their words and fixing their s**t. And I hear a lot less complaining about MtGox now that all major exchanges have halted BTC withdrawals. Even those who previously claimed they were immune (e.g. BTC-E).
The official client is fixed already.
https://github.com/bitcoin/bitcoin/commit/1bbca24This is only a workaround for a part of the problem, it doesn't fix it. Even with this option on the client will spend one-confirmation outputs, which will break again in case of a reorg of one block. A fix would e.g. use another unique ID, independent of the txid, for internal accounting, notice when a txid has changed for a transaction, and resend transactions which depend on the changed txid.
This prevents the current DDOS which is currently affecting Bitstamp
No, it doesn't. It is just some tape to hold it better together and make it happen less frequently.
- which is a DIFFERENT issue. MtGox's problem is that they get transactions rejected because they don't format them properly:
Mallable transactions is not their biggest problem, sending broken transactions is.
Nope. Wouldn't make much difference. Their problem is treating the txid as if the confirmed txid for a transaction will be identical to the txid of the transaction they sent. And this, incidentally, is the cause of Bitcoin-Qt/bitcoind failure as well. A broken transaction (in that the signature won't get accepted by clients version 0.8 and above) just make the exploit easier because the attacker has more time.
You are basically spreading FUD. Your story is that everyone else has mtgoxproblems.gif because someone decided to do a ddos on the bitcoin network today - except that's simply not true. There is a difference between a bug which displays your balance incorrectly until you rescan your wallet and sending someone withdraws twice.
The problem is the same (change of txid confuses clients), the consequences differs between implementations.