Topic: QT strange DOUBLE SPEND (malleable) - page 3. (Read 13046 times)

So any conclusion on what this means (in terms of security, usability of bitcoin)? Any changes to the conclusion that its all mt gox' incompetence and bitcoin is not flawed?

The guy is having fun. Probably also looking for cheap coins.

It happened to me around 10 minutes ago sending from the android wallet. I now have a red cross for the double spend and 4 confirms on the real tx. My balance has since corrected it's self.

this is what i really need to know. Will the blockchain eventually force a 'reject' response or do i need to 'delete' the transaction record out of the wallet.dat? I need to make sure there isnt 0.22BTC in the bitcoin-qt wallet that isnt recorded or will be lost to the ether rather than moved to a change address

At worst he is spreading some FUD among novices, what is clear is that regular users using the reference implementation are not affected. Minor inconveniences, but their wallets are safe.

On the contrary incompetent exchanges who base their accounting on transaction IDs and manage poorly their customer services might be affected.

Sounds like someone nefarious is creating double spends of tons of transactions for nefarious purposes.

I had a friend send me a BTC from multibit yesterday that showed as a doublespend sent 1 second apart.

today i sent from bitcoin-qt and it seems to have created a double-spend (both txids show in the bitcoin-qt and blockchain.info)

did my wallet deduct the send value twice, and/or will the bad transaction (which isnt being confirmed of course) be deleted? I wasnt watching my wallet balance and am not sure if it dropped by my transfer amount (0.2BTC) or by 2x that (0.4btc)

and why is this happening right after/around mt.Gox crying wolf about the problem? some tinfoil-hats among us might be inclined to point fingers

That explains a lot... so basically we have a dark individual or a group of ppl who just wants cheap coins.

That makes sense  

What happens here is that the transactions are malleable.

Roughly speaking, when you send a transaction, it has some data in it which is not really relevant to the transaction. This data can be changed, and the same transaction looks slightly different.

This would not really be a problem, but the effect is that the changed transaction gets a new txid. The effect is what you observed: two transactions with different txid's which really have the same inputs and outputs.

Obviously someone is changing the transactions and broadcasting them again (for kicks, or whatever).

It happened a lot yesterday, some node seems to be relaying transactions twice or something.

I know very well how it is resolved, the question is why the double spend appear in first place.

When you double spend a race condition starts.
Each spend has to reference coins that already belong to you and transfer the ownership to someone else.
Whichever transaction wins the race gets the money, the other is rejected because the coin is already recorded as spent.

How Bitcoin Works Video...
http://www.youtube.com/watch?v=Lx9zgZCMqXE

I deleted the transactions from the wallet and let it to the network. Today one of them is confirmed other is gone.

The question is... Why is this happening ?!

I run a full node and I've been seeing a number of double spends in the log.  So it is not just blockchain.info that is detecting the double spends.  For the ones that I checked, they have the same inputs and outputs and were created at almost the same time.  Sometimes they are a double spend, although I found a couple were included in blocks (which could just be the result of a race condition as to which transaction is seen first)

Here are a couple of them:

  6e21cf5c9dc0320d09084a6426ec360f53bfa0fa9f981d4b26e76fd77a19b93c
  20acb23194be4af89538e720af2507c2ad4b0d179fce9c39e3cae0518d6e7593
  15b94449d35dac6c31b181d6384f3120f8caf700e628b6241cce76a0727567ff
  4837a47929d05e723abc8a5b4c50193ed559c4a73537582584db2f8a677554a9
  43ccf90a7b445acc8265d08c9d48ef99ec17337b56effc850977872d567ddada
  66e507b9f079434ff223fdb54f7f1f88695179b38efac5185809387375ad0e6b
  478143a7acd3325ca0218a7d10b43ce16694f5836a7948f13ca90cb024426042

And many more ...

I should add that the node that is reporting the double spends identifies itself as /Satoshi:0.8.99/ and protocol version 70002.  So this might be something in the new client release.  It sends 'reject' messages whenever it is unhappy (which I think is a good thing because it lets you know why a peer is unhappy with you).  It just reported a couple more duplicates and I checked and they are all in the block chain, so I'm not sure what is going on.

I got the exact same problem earlier today...
I was using qt 0.8.6, and shocked to find the double-spend on blockchain.info.

BTW, I also checked blockr.io at that moment, and found only one tx.

IMHO, it seems to be a blockchain.info problem, rather than bitcoin-qt...

One time, i double spent when i switching up my wallets. I changed my wallet when a transaction was pending, and opened a new wallet.dat file.

When i went back to the 1st wallet.dat (the one the transaction was being sent from, but never confirmed) It showed a balance of my BTC and that the other transaction i sent.

So i tried sending that balance again, hence trying to double spend.

Never got confirmed

Hi,
I'm using version v0.8.5-beta, made a regular send and it appears to be a double spend:

That's the tx I see in the QT:
https://blockchain.info/tx/41f1cfe70cbbb45ed1efeafed8213cbdee14fe772b8643e0f37259791ce6039f

as it seems the wallet generates a second tx with the same input/outputs
https://blockchain.info/tx/22b00c6516676d401e48cc4617ce5cad2b708e0780b72669da9a67deb52d773b

This is very odd... 
maybe a bug