For address 1GtTPepFqS8MWZzFKsE62bNowY2LZos8Lq, in transaction 36f191348998adb993a4d2e4041a0aeb2cc108521d1a48a04c484baa9a1d3216 this is not a rounding error - from my understanding the coins bought value is 0.01088890 rather than the 0.01088888 you guys have.
Coins bought is based on bitcoin payment even if over the originally accepted amount, as long as those extra coins are available (and they were in this case). This bitcoin payment was for 0.0021777 BTC, multiplied by the unit price of 0.2 gives us 0.01088890.
Hope that makes sense
Zathras,
It seems J.R. agreed with me that we should disable this feature of over payment.
Repeating J.R. on the mailing list:
The over-payment logic has been there from the beginning (it's not new),
and is another annoyance which only affects BTC/MSC exchange (can't overpay
when buying GoldCoin with MSC, for instance). I figured if the buyer
overpaid and there were still coins for sale, why not give them to the
buyer? But, now I see why not.
Grazcoin, you make a very convincing set of arguments that we should not
count overpayments, especially in regards to complexity, which we all hate.
Bitoy and Zathras, can you guys confirm that "All implementations currently
work with the assumption that over payment is simply a buyer's mistake and
gives him no benefits."?
Assuming that is true, I'll just go update the spec to disable
over-payments entirely. It's obviously not core to what we are doing, and
we might as well make things simpler.
Simplicity is our biggest advantage over our alt-chain competitors, and
this is a small opportunity to increase our lead in that area.
If you guys ARE handling over-payments currently, I hope it is not too
difficult to simply disable that code?
Thanks!
-J.R.
On Fri, Feb 28, 2014 at 11:57 PM, <
[email protected]> wrote:
> I suggest to change the spec that over payments do not increase the amount
> purchased for the following reasons:
>
> If a user sees that the updated (more expensive) offer got already
> accepted, he could snap the offer by overpaying.
>
> The reserved amount kept on the address after an update must include then
> the whole previous sell offer at least until all accepts to original offer
> expire.
>
> The complexity is one magnitude higher, and the gain is some doubted
> feature/behaviour that users would not expect, and suspect that it is a
> bug, as it exists no where in the finance world.
>
> The sell offer update with 2 versions feature introduced enough complexity
> which still prevents us from consensus, so please do not add any more. We
> may have to postpone the 15.3.
>
> All implementations currently work with the assumption that over payment
> is simply a buyer's mistake and gives him no benefits.
>
> Grazcoin
But:
For the tx you mention, I don't think it got over paid.
The accept:
https://masterchain.info/sellaccept.html?tx=36f191348998adb993a4d2e4041a0aeb2cc108521d1a48a04c484baa9a1d3216¤cy=TMSCThe sell offer:
https://masterchain.info/selloffer.html?tx=effcb1a90f15b39ef2295d8b1d5b3fa5c4a156936b5cfcb770ca68ac1e7ace0e¤cy=TMSCThe payment:
https://masterchain.info/btcpayment.html?tx=0b77033abe7d9c2c06120d5fbd3708bba4a8da39989c9431704c9fd95412d97a¤cy=TMSCor on blockchain:
https://blockchain.info/tx/0b77033abe7d9c2c06120d5fbd3708bba4a8da39989c9431704c9fd95412d97aAmount accepted 0.01088888
Price per coin 0.2 ฿/TMSC
payment expected = 0.01088888*0.2=0.002177776 rounding to 8 decimal digits 0.00217778
actual payment is exactly 0.00217778 to 1BKpa19m5Xy9SvSzC5djPWtCfbuynSDwmb