1 2013-01-19 00:04:55 <Akiraa> Currently, transactions involving less than 0.01 BTC will incur transaction fees?
 2 2013-01-19 00:05:38 <gmaxwell> Akiraa: payments with outputs less than 0.01 BTC, yes.
 3 2013-01-19 00:05:56 <Akiraa> the fee in that case is what, perchance
 4 2013-01-19 00:06:27 <Akiraa> presumably, a value hardcoded in the semiofficial client
 5 2013-01-19 00:06:50 <gavinandresen> yeah, un-harcoding that is on my TODO list.  It's half-done.
 6 2013-01-19 00:07:12 <sipa> the fee, if used, is base_fee * size_in_KB
 7 2013-01-19 00:07:27 <sipa> the required minimum base_fee is 0.00005 BTC or so iirc
 8 2013-01-19 00:07:54 <gavinandresen> one too many zeros, I think sipa
 9 2013-01-19 00:08:00 <gavinandresen> 0.0005 iirc
10 2013-01-19 00:08:02 <sipa> right
11 2013-01-19 00:08:17 <sipa> i wanted to say 00.0005, but misplaced the dot!
12 2013-01-19 00:08:26 <sipa> ... and if you don't believe that
13 2013-01-19 00:09:15 <gavinandresen> sipa: speaking of un-hardcoding the fee??? I've been thinking that redesigning the memory pool could help with the estimate-fee-needed code
14 2013-01-19 00:09:29 <sipa> how so?
15 2013-01-19 00:09:51 <gavinandresen> the idea being:  make the memory pool like an oversized block, and fill it as if you were a miner choosing transactions
16 2013-01-19 00:10:21 <gavinandresen> When sending transactions, you can then ask yourself:  If I was a miner, would I put this new transaction in my block?  If not, add some fees and ask again.
17 2013-01-19 00:10:43 <sipa> that sounds fragi;e
18 2013-01-19 00:10:46 <sipa> fragile
19 2013-01-19 00:11:25 <sipa> others may have a very different view of the current memory pool
20 2013-01-19 00:11:36 <gavinandresen> fragile how?  if it was on the raggedy-edge of not making it into the block then it'd be 'add fees'
21 2013-01-19 00:12:16 <gavinandresen> sipa: mmm, true, although I think the defaults are pretty widely used right now
22 2013-01-19 00:12:29 <sipa> i think it's not reasonable to expect that clients need to maintain a memory pool to be good at guessing fees
23 2013-01-19 00:12:52 <sipa> if your fee rules are so complex you need that, you're instantly killing SPV clients from every making a good guess
24 2013-01-19 00:13:50 <gavinandresen> mmm.  We need to give SPV clients information about transaction fees so they can estimate for themselves
25 2013-01-19 00:14:05 <gavinandresen> ??? but I see that as a separate issue
26 2013-01-19 00:14:30 <Scrat> is in * 180 + out * 34 + 10 a good guess for transaction size?
27 2013-01-19 00:15:17 <gavinandresen> Scrat: until multisig transactions get popular that's probably fairly close
28 2013-01-19 00:15:50 <gavinandresen> Scrat: why are you guessing transaction sizes?
29 2013-01-19 00:16:24 <sipa> Scrat: 151.5 bytes for compressed-pubkeyhash prevout inputs, 183.5 for non compressed ones
30 2013-01-19 00:16:36 <Scrat> I want to know what the fee will be for highly fragmented wallet
31 2013-01-19 00:17:16 <sipa> wait, i'm wrong
32 2013-01-19 00:18:51 <sipa> 148 / 180