Topic: Some dummy necro-bumped this. Please ignore. (Read 5229 times)

I still see perfectly the same case - long chain of unconfirmed TXs and mine TX stuck too. Coinbase unable to fix it and cannot see what happened to MtGOX I guess.

There is plenty of "fault" to go around.  Just off the top of my head:

• blockchain.info could have done a better job of ensuring that users of their JSON RPC are aware of the fees they are including
• CoinAd.com could have written a better back end program that ensured that they used the proper fees
• Everyone in the chain of unconfirmed transactions could have made sure they were not using unconfirmed transactions as inputs
• Coinbase.com is one of those that could have made sure they were not using unconfirmed transactions as inputs
• Miners could use a forward looking algorithm to determine if a transaction should be included

I received my BTC. I would say major kudos to Coinbase for putting out a generous bounty to make things right. Major jeers to Coinbase for not having any customer service and for using unconfirmed transactions in their hot wallet. Apparently the second problem is fixed for the future.

I still don't know who's 'fault' this whole thing is, or if it's just a series of happy accidents. Thanks so much for all of your help and insight- esp DannyHamilton!

Unfortunately I don't see any mention of this on the CoinAd web page.  I suspect that very soon you will have customers complaining that they've been using the site, but aren't getting paid.

That's great news Carlos.   
I'm fairly certain that your market would prefer fewer larger payments instead of several smaller payments.  I like the solution of the website I cited.  A customer can have a balance build up and then request it to be sent at anytime.  After a request, payout is usually within 24 hours.  All the best Carlos.

I have emailed several pools when this problem occurred, and offered a bounty. The bounty was small, but it was something.

I said in a previous post that payments are suspended. It will remain like that until I find a way to prevent this.

I'm already working on it.

Whew, glad coinbase is making good, and improving their service.    

I should like Carlos L. to reconsider his business model and research suggested fees for the large(in bytes) transactions he does.  I would suggest that people use http://bitcoiner.net, there you can go back multiple days and not take a payout until it is larger.  This makes the fees on both sides more manageable.  

If you are going to try to run a business, you need to make sure you know what you are doing.  Your business is responsible for its actions.  Ignorance is not a mitigating excuse.

Coinbase did not intentionally send out transactions that were likely to never confirm, but they recognized that it was their responsibility to make sure that their customers got paid.  So they offered a bounty for the confirmation of your mistake, and they fixed their program so they wouldn't use unconfirmed inputs in transactions again.

I see that the coinad.com website is till up and running.  Have you fixed your process to ensure that you don't send large transactions with small fees yet?  Have you fixed your process to ensure that you don't double spend any inputs again?

Coinbase.com offered a 5 BTC bounty for a miner to confirm the original offending transactions.  It looks like your transaction has confirmed in block 221131.  Coinbase.com also claims that they have fixed their program to no longer use unconfirmed inputs in the transactions they send.

Note that CoinAd.com, which is who created the original 2 huge transactions with insufficient fee that have finally confirmed, had tried to double-spend the input from:
http://blockchain.info/tx/ccb75ea48acc40f249f445ab7b6913e6a805938c022fbc6c9dcbdc67e4cb8847

in the following transaction:
http://blockchain.info/tx/97825f2228c0087acf2b7edce81c79838931845c4b0595ed4088966ebdf9d633

Now that ccb75ea48acc40f249f445ab7b6913e6a805938c022fbc6c9dcbdc67e4cb8847 has confirmed, 97825f2228c0087acf2b7edce81c79838931845c4b0595ed4088966ebdf9d633 never will.

That sucks I will stop promoting them to my buddies until this is resolved.

So just to update: still no btc, still not a word of response from Coinbase to my case. I've sent multiple support tickets, I've emailed their support email address, and I have yet to have any of my efforts or my questions acknowledged at all. But I guess the good news is now that the price of btc has gone up, I'm actually out about $17 more than when I started!

The transfers in Coinbase still show 'pending' but the tx hash (9eb202978e1653a0d3b1c936c2a07228055ce434c6c065c727336cfc0a74d115) for my transfer has disappeared off of blockchain.info. This experience has taught me a bunch about bitcoin in general, but if Coinbase is going to be a big player in connecting technically inept people and the bitcoin economy they need to up their game. Or at least respond to a customer who has just lost $340 into thin air.

Looks like a blockchain bug.


18:13:53

getrawtransaction a752638db79143ab3cc23280125601193ee3e8b8fb272a52a6d6cffb82e2dc0e


18:13:53

0100000002a99dcc16468d4e370862d2b95f23a101a95aca9333341420870447132c2517c600000 0008a47304402205c2f0c685a005c98cc378f54419aacaa965708e51fe622bec96389dd3a4ca8d0 0220498aede1ed575a404717c3357eec80b249bfd65eb90ea3e22b5bcea12faa7fb0014104c87e8 2eeda25cd8546c636b7c2549a5ccfe4b3eb25290df989e13ae9192cf347a8e895ce452dca0cbb6b 3a4fa4aca593d0ebea800a1ac8443af317f3b797dcbeffffffff66e6a82dec1b7c344f000f9bc02 e4e3372cbee4207d2076d95a990f9780abd89010000008b4830450221008442bd2cd630b01cad72 6d22075afe047a4b805957d2bceb3d58b9dcfabec125022007df133b53b900c480da9de357a7d73 297a2dfcd1b6d19892f8f9d69997a250d0141045618440a6f417fb724cb7112b8213f5e30929e55 12045cebb2f7b8291e7901518374b31e3669fbeb514a309b72c9b2341d2cf1cd97174af1838209b b07b3b41affffffff0200c2eb0b000000001976a914451dd2dc37d7782760c728abbd1c42279fb4 bd1c88acc39d3200000000001976a914ab1fb59f5f3d346a0b82e39f16ec57afb311bbbd88ac000 00000


18:13:59

decoderawtransaction 0100000002a99dcc16468d4e370862d2b95f23a101a95aca9333341420870447132c2517c600000 0008a47304402205c2f0c685a005c98cc378f54419aacaa965708e51fe622bec96389dd3a4ca8d0 0220498aede1ed575a404717c3357eec80b249bfd65eb90ea3e22b5bcea12faa7fb0014104c87e8 2eeda25cd8546c636b7c2549a5ccfe4b3eb25290df989e13ae9192cf347a8e895ce452dca0cbb6b 3a4fa4aca593d0ebea800a1ac8443af317f3b797dcbeffffffff66e6a82dec1b7c344f000f9bc02 e4e3372cbee4207d2076d95a990f9780abd89010000008b4830450221008442bd2cd630b01cad72 6d22075afe047a4b805957d2bceb3d58b9dcfabec125022007df133b53b900c480da9de357a7d73 297a2dfcd1b6d19892f8f9d69997a250d0141045618440a6f417fb724cb7112b8213f5e30929e55 12045cebb2f7b8291e7901518374b31e3669fbeb514a309b72c9b2341d2cf1cd97174af1838209b b07b3b41affffffff0200c2eb0b000000001976a914451dd2dc37d7782760c728abbd1c42279fb4 bd1c88acc39d3200000000001976a914ab1fb59f5f3d346a0b82e39f16ec57afb311bbbd88ac000 00000


18:13:59

{
"txid" : "a752638db79143ab3cc23280125601193ee3e8b8fb272a52a6d6cffb82e2dc0e",
"version" : 1,
"locktime" : 0,
"vin" : [
{
"txid" : "c617252c134704872014343393ca5aa901a1235fb9d26208374e8d4616cc9da9",
"vout" : 0,
"scriptSig" : {
"asm" : "304402205c2f0c685a005c98cc378f54419aacaa965708e51fe622bec96389dd3a4ca8d00220498 aede1ed575a404717c3357eec80b249bfd65eb90ea3e22b5bcea12faa7fb001 04c87e82eeda25cd8546c636b7c2549a5ccfe4b3eb25290df989e13ae9192cf347a8e895ce452dc a0cbb6b3a4fa4aca593d0ebea800a1ac8443af317f3b797dcbe",
"hex" : "47304402205c2f0c685a005c98cc378f54419aacaa965708e51fe622bec96389dd3a4ca8d002204 98aede1ed575a404717c3357eec80b249bfd65eb90ea3e22b5bcea12faa7fb0014104c87e82eeda 25cd8546c636b7c2549a5ccfe4b3eb25290df989e13ae9192cf347a8e895ce452dca0cbb6b3a4fa 4aca593d0ebea800a1ac8443af317f3b797dcbe"
},
"sequence" : 4294967295
},
{
"txid" : "89bd0a78f990a9956d07d20742eecb72334e2ec09b0f004f347c1bec2da8e666",
"vout" : 1,
"scriptSig" : {
"asm" : "30450221008442bd2cd630b01cad726d22075afe047a4b805957d2bceb3d58b9dcfabec12502200 7df133b53b900c480da9de357a7d73297a2dfcd1b6d19892f8f9d69997a250d01 045618440a6f417fb724cb7112b8213f5e30929e5512045cebb2f7b8291e7901518374b31e3669f beb514a309b72c9b2341d2cf1cd97174af1838209bb07b3b41a",
"hex" : "4830450221008442bd2cd630b01cad726d22075afe047a4b805957d2bceb3d58b9dcfabec125022 007df133b53b900c480da9de357a7d73297a2dfcd1b6d19892f8f9d69997a250d0141045618440a 6f417fb724cb7112b8213f5e30929e5512045cebb2f7b8291e7901518374b31e3669fbeb514a309 b72c9b2341d2cf1cd97174af1838209bb07b3b41a"
},
"sequence" : 4294967295
}
],
"vout" : [
{
"value" : 2.00000000,
"n" : 0,
"scriptPubKey" : {
"asm" : "OP_DUP OP_HASH160 451dd2dc37d7782760c728abbd1c42279fb4bd1c OP_EQUALVERIFY OP_CHECKSIG",
"hex" : "76a914451dd2dc37d7782760c728abbd1c42279fb4bd1c88ac",
"reqSigs" : 1,
"type" : "pubkeyhash",
"addresses" : [
"17JTMRSdD2rWs1eSkFqZnXTKWKXLGQsVqY"
]
}
},
{
"value" : 0.03317187,
"n" : 1,
"scriptPubKey" : {
"asm" : "OP_DUP OP_HASH160 ab1fb59f5f3d346a0b82e39f16ec57afb311bbbd OP_EQUALVERIFY OP_CHECKSIG",
"hex" : "76a914ab1fb59f5f3d346a0b82e39f16ec57afb311bbbd88ac",
"reqSigs" : 1,
"type" : "pubkeyhash",
"addresses" : [
"1GbpY1ssJLPHmFYnNr7RhtYsYWNNM1ssvP"
]
}
}
]
}

I'm pretty sure that a client will not broadcast such a transaction.  So it's unlikely to confirm.  However I do not see a problem with a negative transaction fee as long as the block has only 25 BTC more out than in. 

Does the protocol allow miners to in theory include the transaction, taking a negative tx fee?

That just looks like a bad transaction. I'm a bit surprised that it would get relayed or stored in blockchain.info's memory pool. It shouldn't ever confirm.

http://blockchain.info/tx-index/52591276

Which I suppose means that none of these 655 transactions will ever confirm either since they can all trace their inputs back to that bad transaction:

http://pastebin.com/LYcMAi8X

EDIT: It looks like blockchain.info has now dropped the transaction. It no longer show up on the website.

http://blockchain.info/address/1NiamfxxTcGVDkAj5d5WcbxUiA41hR7GGM

It looks like that's a transaction that has more outputs than inputs.  I don't think the transaction will ever confirm.

Does a transaction removed from the memory pool result in a negative fee displayed on blockchain?

I suspect there are a lot of people hoping that it will confirm.  I believe coinbase.com has even paid a miner a bounty to confirm it:

If another nodes relays us the transaction again it will reappear on the site but for now it has been removed from the memory pool as it seems unlikely it will confirm.