Bitcoin Forum

EDITED - Changed link to the correct transaction (like I said, I was tired).

I'm in the middle of a dispute with a bitcoin supplier who have sent me a txn with what I believe to be a too low fee. I wanted to ask the community for their thoughts.

In short, this 1550 byte txn was sent with a 0.0002 BTC fee:
https://blockchain.info/tx/5b842d282c2086add3dc3eaa9f5d0f8d05a8056c16703b54607a3a635e43960e
it's been pending for 32 hours now when other txns with what I believe to be correct fees are being included in a couple of blocks.

Accoding to bitcoind's estimatefee, for a:
1 block confirm the fee is 0.00048871 BTC per kb
6 block confirm the fee is 0.00024609 BTC per kb

So, I suspect for a 1550 byte txn, the fee should be between 0.00038143 BTC and 0.00075750 BTC - so from 2x to 4x the actual fee used.

However, the exchange feels differently - here is there response in what is now a long email exchange:
 "As a bitcoin developer, you'll know that the minimal bitcoin fee is calculated based on transaction size, not cardinal number of inputs or outputs. Looking at the specs, you'll see that for a 1.55KB transaction, 1mBTC would be perfectly acceptable, and yet, we pay 2mBTC. That's why you'll notice that the blockchain.info page is showing this a "high priority"."

He says 2mBTC, but surely he means 0.2 mBTC?

Are my calculations wrong here? Sorry for the possibly dumb questions - I've not had much sleep the past 24 hours....

Use http://cointape.com/ for finding out the best fee for every KB according to the current number of uncomfirmed transaction.

It might be because of the ongoing blockchain attack

Yes, I think that backs up my calculations - but like I say I am really tired and would appreciate someone who is less so to either back my numbers up or show me where I am going wrong. :)

most companies probably just have a fixed fee and their customer support isnt usually doing more than repeating standard answers

in recent days (hours), the number of unconfirmed tx spiked, this allows the miners to cherry pick the highest paying tx and still fill a block, so they do.

Agree there's been a huge spike, but even with a lowish fee (say bitcoind's estimatefee 6) things are still going through in a couple of hours. This has been 32 hours!

They do use a fixed fee - it's 0.0002 BTC each time - which is fine when there's 1 input and 2 outputs, a 300byte txn - but when there are 10 inputs and 2 outputs and it's 1550 bytes that simply isn't enough (IMHO). Kraken on the other hand charge users a fixed fee, but on the actual transaction they apply the correct fee.

However, can anyone confirm or deny my maths on the fee for this txn?

.00062000 ~ .00077500 (40~50 satoshi per byte). or 1mbtc if you want to be sure. but 20~30 satoshi per byte shouldn't take more than 10 blocks.

Thank you --Entrypted--, glad my maths is not completely off.

by the way that part about blockchain showing the tx as high priority just because of the high fee are bull. the priority calculation doesn't have the fee on it.

15.2 satoshi per byte is the current fee level

so, 0.00023560 BTC

First, lets get the errors out of the way, shall we?

The TX pays a 79122 satoshi fee and has a size of 1555 byte. Thats a 50.88 satoshi fee per byte and a very good fee currently, according to your own (@OP) estimate of core as well. Im not sure where the 2mBTC (or 0.2mBTC) come from, but they have nothing to do with the TX linked.

Also, the linked transaction is now confirmed.

Sorry, sorry, sorry, it was the wrong transaction. I've updated the original post with the correct link and here it is again:
https://blockchain.info/tx/5b842d282c2086add3dc3eaa9f5d0f8d05a8056c16703b54607a3a635e43960e
(really, this is the correct one now - and still unconfirmed).