For a little background, http://coinsmack.com
is a news site similar to digg and reddit where users can vote up stories, but the catch is they vote up stories using micro-payments in bitcoins.
Users can post new stories to the site for free, and they receive 80% of the bitcoins used to vote up the story.
I made the first payout for coinsmack with this transactionhttp://blockexplorer.com/tx/69ae8ac2ecc66ef16a77be333b71a14d5f1d70c1cecf127e6ef47ef6b107b0e4#o8
It makes for 6.164 kilobytes with 2.0045 BTC sent in the transaction to multiple recipients. It is pretty heavy for the actually number of BTC being passed around, but really transaction size has nothing to do with the number of btc in the payload. I feel naughty that I wrote so much to the block for only 0.0005 bitcoins ($0.0025).
A) Remember the BTC 10K pizza ? Give it a couple of years
and you'll be sorry that you paid such a huge fee.
B) Supply and demand. It's up to the miners to decide if your fee is
worth the 6K you pushed in the chain. Looks like one of them did
think it was OK, which means you have nothing to feel guilty about.
C) But you raise an interesting point about BTC being a viable vehicle
for micro-payments in the long run, and how the whole fees vs.
mining profitability problem will evolve.
A) Yes, but that was just when it was something that people were futzing with, and variability has decreased by orders of magnitude. I doubt that 0.0005 bitcoins will ever be worth much in my lifetime.
B) True that, but miners I don't think are acting like true rational operators that are setting their own rules. They simply run the client. I think the cost of writing to the block should be very tied to actual number of bytes recorded and have nothing to do with btc value. Of course, that would probably kill micro-micro transactions like what I am using.
C) The thing is the miner does not really pay the long term cost of writing to the block. The actual cost is the total disk space taken up by that transaction on the entire network. The miner will always write to the block for nearly any value that is submitted since not doing so would leave money on the table for the next miner. Basically the total cost of block writing does not align with the actual cost to the miner. I sometimes consider tweaking the miner code and running it myself just to suck up all the near spam transactions that I know are floating around the network, that are being skipped by the hard coded rules. If you just save all those transactions over time, even when they are not being passed around, one day you might be able to write them all to a block, and make a full bitcoin from the spam you fill the chain with. I don't believe that transactions have a TTL. I suspect that there has been a lot of 0.00001 fee transactions that have been dropped.