Smarter CF tx change #27
Conversation
darosior
force-pushed
the
smarter_cfanout_chage
branch
from
2376f1f
to
cf741a0
Compare
darosior
force-pushed
the
smarter_cfanout_chage
branch
4 times, most recently
from
f00bed2
to
7659c04
Compare
|
Amended with a version that decides whether to create a new fb coin not based on Vb but based on a "minimum coin value". |
darosior
force-pushed
the
smarter_cfanout_chage
branch
from
7659c04
to
45f3d79
Compare
Model/statemachine.py
Outdated
| @@ -383,6 +379,14 @@ def grab_coins_2(self, block_height): | |||
| total_to_consume = total_unprocessed + total_unallocated + total_negligible | |||
| return total_to_consume, num_inputs | |||
|
|
|||
| def min_fbcoin_value(self, height): | |||
There was a problem hiding this comment.
This is very similar to the StateMachine._is_negligible() value. In that I put the coin should increase the cancel by at least 10 sats/vByte. Do you think 5 is sufficient? I don't know. They should be consistent I think. Also, I think if Vm is much smaller, that is also a valid min value for a feebumping coin. So I'd use
min(Vm, int(feerate * P2WPKH_INPUT_SIZE + self._feerate_to_fee(5, "cancel", 0)).
WDYT?
There was a problem hiding this comment.
This is very similar to the StateMachine._is_negligible() value. In that I put the coin should increase the cancel by at least 10 sats/vByte. Do you think 5 is sufficient? I don't know.
Wow, then is_negligible is way too high in my opinion. For a reserve feerate of say 600sat/vbyte (2017 levels), this is coin at 64240 + 12290 = 76530sats (~38$ at 50k$/BTC) that you consider negligible. This is unreasonable.
Same computation with a feerate bump of 5sat/vb is 64240 + 6145 = 70385sats (~35$) which IMO is clearly enough. Keep in mind that at a large but reasonable feerate of 100sat/vb this would be a bump of more than 48sat/vb!
They should be consistent I think.
Absolutely not i think.
To create them we should be conservative to avoid having to consolidate them and "bumps by at least 5sat/vb at the reserve feerate" is a reasonable value IMO (happy to bikeshed).
However, removing them from an allocated vault reserve is a dangerous operation that we ought to avoid unless it's absolutely necessary. So i'd say "bumps by at least 1sat/vb at the reserve feerate" but it's still a very high value and one based on recent fee estimates might be more reasonable.
Of course if it's not allocated then it's another matter: we can keep the 5sats/vb and consolidate it whenever the feerate is low.
Also, I think if Vm is much smaller
If it is then we have a bigger issue lol
|
Fixed the code, results are even worse now... |
darosior
force-pushed
the
smarter_cfanout_chage
branch
from
45f3d79
to
9caa5a0
Compare
darosior
force-pushed
the
smarter_cfanout_chage
branch
2 times, most recently
from
25a4202
to
a9c6403
Compare
|
Rebased on master, this is now a single small commit |
darosior
force-pushed
the
smarter_cfanout_chage
branch
from
a9c6403
to
ae38542
Compare
Signed-off-by: Antoine Poinsot <darosior@protonmail.com>
darosior
force-pushed
the
smarter_cfanout_chage
branch
from
ae38542
to
17b9fb1
Compare
|
So i fixed the logic and this (almost) doesn't make this worse anymore. |
No description provided.