Feature Request for Bitcoin Core: Replace by Fee

[byt411]

Well on the flip side, banks also wouldn't carry your money from A to B if they didn't get payed for it. So there's that.

And BC.I isn't in theory preventing this transaction from happening; it's actually encouraging it by trying to get it to a miner, which it's failing at for obvious reasons. On the other hand, wallets are preventing you from double spending the transaction because BC.I appears to be persistent on relaying a doomed transaction, so in practice, it is preventing you from moving the funds - even if through another transaction, indirectly.

Rest assured all these 'irrelevant usability issues', as myself and many developers would call them, will be worked out before long.

Thanks for your thoughts on this matter. Makes perfect sense, all the way around.

And… great news!! My unconfirmed transaction finally "fell off" of Blockchain.info, and I got my coins returned to me!

Only took 4 days! 

Next time, try writing to blockchain.info support, and see if they can help. They're the retarded ones that still showed the tx, and everyone looks at blockchain.info, all clients do -.-

[BitcoinGirl325]

Well on the flip side, banks also wouldn't carry your money from A to B if they didn't get payed for it. So there's that.

And BC.I isn't in theory preventing this transaction from happening; it's actually encouraging it by trying to get it to a miner, which it's failing at for obvious reasons. On the other hand, wallets are preventing you from double spending the transaction because BC.I appears to be persistent on relaying a doomed transaction, so in practice, it is preventing you from moving the funds - even if through another transaction, indirectly.

Rest assured all these 'irrelevant usability issues', as myself and many developers would call them, will be worked out before long.

Thanks for your thoughts on this matter. Makes perfect sense, all the way around.

And… great news!! My unconfirmed transaction finally "fell off" of Blockchain.info, and I got my coins returned to me!

Only took 4 days! 

Next time, try writing to blockchain.info support, and see if they can help. They're the retarded ones that still showed the tx, and everyone looks at blockchain.info, all clients do -.-

I actually did write blockchain.info support… no response from them. 

Ugh. Annoying that everyone looks to them as their master source of information.

[amaclin]

And… great news!! My unconfirmed transaction finally "fell off" of Blockchain.info, and I got my coins returned to me!

Now you should spend your uxto. Just send your funds with 0.00001 fee (or 0.0001)
Because everybody is able to resubmit your stuck transaction

[BitcoinGirl325]

And… great news!! My unconfirmed transaction finally "fell off" of Blockchain.info, and I got my coins returned to me!

Now you should spend your uxto. Just send your funds with 0.00001 fee (or 0.0001)
Because everybody is able to resubmit your stuck transaction

Ha! Thanks!! I did go ahead and spend my uxto, so I think I'm okay now. Thanks to you and everyone else for your amazing help through these trying times!! All this for $1. But a great learning experience.

[PewDiePie]

It seems like there is nothing I can do to rebroadcast this transaction with a fee, and that I have lost this $1 forever. At least it was only $1, though.

You can. But let us start from the early beginning.
1) Can you show stuck transaction on the blockchain?
2) Do you have another utxo (sorry - this means some other funds on unspent tx output )?

1. The stuck transaction on the blockchain can be seen here: https://blockchain.info/tx/c238d939b330b682abe1c729300636600254c23c67e1eea4e8daa9abc26dd2c5

2. Sorry, I'm new to this. I don't know if I have other funds on the unspent output. How can I tell?

Thanks,
Scott

i think thats all about new code in running . or double spend ?
i never got problem like that 

[Newar]

Ugh. Annoying that everyone looks to them as their master source of information.

Not everyone. There's plenty of options.

[Peter Todd]

This thread is kinda depressing; this sillyness should have been fixed a year ago.

FWIW greenaddress.it was talking about implementing fee bumping to me; tell Lawrence you want to see it. I've also suggested it to the rest of the DarkWallet team, again, telling them you want to see this too helps. Also, everyone who runs my replace-by-fee patch helps out the network. Even if you aren't a miner, nodes that run it help get replacements to miners who will accept them for whatever reason.

[Razick]

If you have any change from your transaction, you can send the change to another address that you own with enough fee to cover both transactions and that may get your transaction confirmed.

I do agree that replace by fee should be added when the client is in expert mode. Also, miners really need to implement Child-pays-for-parent.

[BitcoinGirl325]

If you have any change from your transaction, you can send the change to another address that you own with enough fee to cover both transactions and that may get your transaction confirmed.

Can't be done, because the change transaction was also in an unconfirmed state.

[BitcoinGirl325]

This thread is kinda depressing; this sillyness should have been fixed a year ago.

It's a good wake-up call to how open source software almost never succeeds in the real world. And when open source does have some modicum of success, it is riddled with bugs and flaws and security holes (e.g. Android). It makes me realize that Bitcoin will probably never reach mainstream success.

It takes a for-profit corporation to take open source software and turn it into something proprietary that works for the mainstream (like how Apple took Gecko/KHTML and turned it into Safari).

[DannyElfman]

This thread is kinda depressing; this sillyness should have been fixed a year ago.

It's a good wake-up call to how open source software almost never succeeds in the real world.

You mean like linux, Firefox, open office, Wordpress, apache , MySQL, BSD, Magento, Filezilla, audacity, gimp, vlc, truecrypt, ubunto, keypass?

Just because there is a part that needs to be fixed ( I am not even entirely convinced it should be fixed on a protocol level!) , makes you think BTC will never reach mainstream success..

Well, I got bad news for you

[amaclin]

If you have any change from your transaction, you can send the change to another address that you own with enough fee to cover both transactions and that may get your transaction confirmed.

child-pays-for-parent is not implemented by miners

Can't be done, because the change transaction was also in an unconfirmed state.

No. There is no such restriction in protocol itself. Just use smart client software. Unfortunately, "smart" is not "friendly to novice"

[CJYP]

If you have any change from your transaction, you can send the change to another address that you own with enough fee to cover both transactions and that may get your transaction confirmed.

child-pays-for-parent is not implemented by miners

Hmm that's something I thought existed. It would profit miners to implement this; why don't they?

If Replace By Fee were standard, no transaction would ever be instant, even very small ones. Some see this as an improvement in security at the expense of convenience and choice, others see the improvement in security to be so minuscule, especially with small transactions, that the cost in convenience makes it not worthwhile (although I'm probably summarizing all this incorrectly).

Replace-by-fee, combined with child-pays-for-parent, seems to make 0/unconfirmed transactions much more secure for small transactions. Just read the thread where Peter Todd describes the proposal; he has a very well thought-out argument for why, which I'll summarize:
Current system:
- If a scamer manages to send out a double-spend transaction at the same time as you send the actual transaction, 50% chance yours is confirmed 50% theirs - theirs might even have a higher chance if it has higher fees and a miner (but not most clients) is implementing replace-by-fee.
- Expected value for potential scamer: 0.5*amount_paid
- Expected value for you: amount_paid - P(payee is not a scamer)*amount_paid
Replace-By-Fee system:
- If a scamer sends out a double-spend transaction at the same time as you send the actual transaction, you send another transaction paying the output of your transaction to a transaction fee. Scamer receives nothing, but neither do you.
- Expected value for potential scamer: 0
- Expected value for you: amount_paid
As you can see, there's no longer any motivation to scam. The only reason there would be motivation to scam is if the scamer has a vendetta against you, and if losing the money would cause significant damage you wait for a confirmation - which should be true now anyway.

[DannyElfman]

If you have any change from your transaction, you can send the change to another address that you own with enough fee to cover both transactions and that may get your transaction confirmed.

child-pays-for-parent is not implemented by miners

Hmm that's something I thought existed. It would profit miners to implement this; why don't they?

If Replace By Fee were standard, no transaction would ever be instant, even very small ones. Some see this as an improvement in security at the expense of convenience and choice, others see the improvement in security to be so minuscule, especially with small transactions, that the cost in convenience makes it not worthwhile (although I'm probably summarizing all this incorrectly).

Replace-by-fee, combined with child-pays-for-parent, seems to make 0/unconfirmed transactions much more secure for small transactions. Just read the thread where Peter Todd describes the proposal; he has a very well thought-out argument for why, which I'll summarize:
Current system:
- If a scamer manages to send out a double-spend transaction at the same time as you send the actual transaction, 50% chance yours is confirmed 50% theirs - theirs might even have a higher chance if it has higher fees and a miner (but not most clients) is implementing replace-by-fee.
- Expected value for potential scamer: 0.5*amount_paid
- Expected value for you: amount_paid - P(payee is not a scamer)*amount_paid
Replace-By-Fee system:
- If a scamer sends out a double-spend transaction at the same time as you send the actual transaction, you send another transaction paying the output of your transaction to a transaction fee. Scamer receives nothing, but neither do you.
- Expected value for potential scamer: 0
- Expected value for you: amount_paid
As you can see, there's no longer any motivation to scam. The only reason there would be motivation to scam is if the scamer has a vendetta against you, and if losing the money would cause significant damage you wait for a confirmation - which should be true now anyway.

Potential flaw, you double spend accidential yourself. I managed to do that too...

[amaclin]

> child-pays-for-parent is not implemented by miners
Hmm that's something I thought existed. It would profit miners to implement this; why don't they?

Most of current clients do not allow to spend unconfirmed transaction output. So, the profit is near to zero.
Right now the transaction fee is too small comparing with block reward.
It is not good reason to change existing code to increase profits to 0.00x btc per block. The error in code costs much more.

[TwinWinNerD]

> child-pays-for-parent is not implemented by miners
Hmm that's something I thought existed. It would profit miners to implement this; why don't they?

Most of current clients do not allow to spend unconfirmed transaction output. So, the profit is near to zero.
Right now the transaction fee is too small comparing with block reward.
It is not good reason to change existing code to increase profits to 0.00x btc per block. The error in code costs much more.

That is not true, most advanced clients allow it. Same with coincontrol.

Also you can always dump your private key and reimport it elsewhere to create a new transaction. This is a chicken-egg problem with your argumentation!

[BitcoinGirl325]

You mean like linux, Firefox, open office, Wordpress, apache , MySQL, BSD, Magento, Filezilla, audacity, gimp, vlc, truecrypt, ubunto, keypass?

I am totally pro-Bitcoin and a huge Bitcoin fan, but it's just not looking so great for mainstream adoption.

Here's a great article on how closed-source always beats open-source:
http://blog.erratasec.com/2014/07/cliche-open-source-is-secure.html#.U-Kj3Ui9KSN

The technologies you named are great for tech-savvy engineers, but not for the general public:

Linux - never made it to mainstream users. The general public uses OS X.
Firefox - surpassed in usability and security by Safari and Chrome. Apple's new web services don't even support Firefox anymore.
Wordpress and MySQL and Apache - arguably the 3 open source technologies that made it to mainstream users... BUT both are maintained & controlled by corporations (Wordpress and Oracle and Apache). Hence their success. Without these corporations funding the development of those products, they would likely not have taken off.
VLC - surpassed handily in functionality & features & ease of use by companies like LogMeIn. VLC stopped evolving years ago.
The other technologies you've named: not being used by any mainstream users, unless they were taken over by a corporation (e.g. how Apple used BSD for OS X).

I would LOVE for Bitcoin to go mainstream. I'm rooting for Bitcoin. But unless a for-profit corporation starts controlling & writing the code, it's looking increasingly unlikely that Bitcoin will go mainstream. The few developers working on the code can't even agree on a direction for its future! It's taken 5 years to get where we are today, which is... almost no progress. The Bitcoin Core client hasn't even reached version 1.0.

On the other hand, for example, look where Apple has come in 5 years with the iPhone: 5 major iOS releases, hundreds of millions of phones sold, rock solid stable OS with no security flaws, an entire worldwide ecosystem built around it.

[BitcoinGirl325]

Also you can always dump your private key and reimport it elsewhere to create a new transaction. This is a chicken-egg problem with your argumentation!

Not true. I reimported my private key to 4 different clients, and none of them let me create a new transaction while my unspent outputs were still unconfirmed. No client that I could find will allow for double spending.

[DannyElfman]

I think you confuse with "used by the huge majority" with mainstream use.

Sorry to say, but what you say is just not true.

Just a few pointers: Linux is the most used server OS, by far
Firefox still has over 20% market share
@others, again you confuse two concepts, mainstream and majority.

Even if "the majority" won't directly use Bitcoins, I am sure they will use it without knowing it.

[DannyElfman]

Also you can always dump your private key and reimport it elsewhere to create a new transaction. This is a chicken-egg problem with your argumentation!

Not true. I reimported my private key to 4 different clients, and none of them let me create a new transaction while my unspent outputs were still unconfirmed. No client that I could find will allow for double spending.

Blockchain infos custom send, does.