Fix issue 1780: Market fees of settle orders aren't shared to referral program #2132
Conversation
Fix: Market fees of settle orders aren't shared to referral program
Resolved conflicts: - libraries/chain/db_market.cpp - tests/tests/settle_tests.cpp
abitmore
force-pushed
the
way-old-hardfork-for-issue-1780
branch
from
fd408f0
to
23008cd
Compare
Pay market fee and share it with the referral program when settling an amount of a globally settled asset after the core-1780 hard fork, for issue #1780
* Add new test cases for instant settlement market fee sharing * Update old test cases to test more data
| // performance loss. Needs testing. | ||
| if( d.head_block_time() >= HARDFORK_CORE_1780_TIME ) | ||
| { | ||
| auto issuer_fees = d.pay_market_fees( fee_paying_account, settled_amount.asset_id(d), settled_amount ); |
There was a problem hiding this comment.
@abitmore This invocation of pay_market_fees from the evaluator is difficult to resolve and merge with another PR from simple maker-taker fees from BSIP81 because the market fee will be dependent on whether the order is considered to be a maker or taker.
I am attempting to find a reasonable manner to merge these two PRs. Because this line of logic applies during a global settlement where the settlement pool is waiting, is it reasonable to consider this settlement order as a taker?
There was a problem hiding this comment.
Yes, the settlement order is always taker.
There was a problem hiding this comment.
Actually, the settlement order can be maker sometimes if we implemented bsip73.
There was a problem hiding this comment.
A side note, since there is no virtual fill_order_operation generated for instant settlement, neither the price nor the volume would be in the HTLCV charts.
There was a problem hiding this comment.
If I merge this before the BSIP73 PR, can I ignore the problem? :)
There was a problem hiding this comment.
Sure. I believe 73 will be worked on only after this.
Follow-up PR for #1780.
Reviews were here: #1842.