< gmaxwell>
I wonder why non-conservative estimator on my node here says 11s/b for 6 blocks, my mempool suggests thats enough to get in the current block ... and 21inc says 80 for 6 blocks and 280 for now.
< aj>
gmaxwell: some (1 in 6 by hashpower?) pools are mining empty blocks and ignoring transactions under 20s/b?
< gmaxwell>
ah, that would be not very helpful for the estimators.
< gmaxwell>
yea, bitcoin.com censored my transaction, indeed.
< gmaxwell>
looks like thats actually costing them a lot of fees.
< gmaxwell>
Kano's not a censor, hurray
< rodarmor>
How will a current bitcoin core node interpret a block with a transaction whose version is greater than CURRENT_VERSION? I can see in the code that it's used for relay policy, and for BIP68 enforcement, is that it?
< sipa>
correct
< rodarmor>
sipa: Sweet, thanks!
< dcousens>
BlueMatt: are you sure about the std::move breaking the CFeeBumper code?
< dcousens>
IIRC std::move on a const& simply _doesn't_ move, unless it can guarantee safety
< sipa>
it casts to a const&&
< sipa>
std::move never moves; it just casts it to an rvalue
< sipa>
which indicates to wherever it's passed that it does not maintain the state
< sipa>
const&& almost always means it's not going to be touched
< jnewbery>
Can I get #10711 added to https://github.com/bitcoin/bitcoin/projects/8 ? It's a test framework enhancement so not super urgent, but it moves quite a lot of code within test/functional/test_framework so I'd like to avoid too many rebases.
< bitcoin-git>
[bitcoin] laanwj closed pull request #10745: Be consistent in using "opt_into_rbf" parameter for Opt-In RBF (master...opt_in_rbf-param) https://github.com/bitcoin/bitcoin/pull/10745
< wumpus>
jnewbery: sure
< wumpus>
jnewbery: done
< BlueMatt>
promag: I believe I told morcos I'd take that over
< BlueMatt>
I havent done anythong on it except for rebase the first few commits, however
< gribble>
https://github.com/bitcoin/bitcoin/issues/0652 | Update contrib/debian/ for 0.5.0 release and fix copyright file. by TheBlueMatt · Pull Request #652 · bitcoin/bitcoin · GitHub
< promag>
ops, yeah
< promag>
BlueMatt: I have some nits for that PR, I can push to a branch and later you can rebase on that, wdyt?
< BlueMatt>
hmm? no, I havent yet taken it any further than that pr
< BlueMatt>
i plan to, but if you want to rewrite it, feel free
< promag>
kk
< BlueMatt>
promag: you almost certainly want to use 10652 as a base, fwiw, building on top of that should be easier, but its still a very thorny area
< promag>
BlueMatt: should we close 9447?
< BlueMatt>
no rush, but probably
< BlueMatt>
i mean open something to replace it first
< morcos>
promag: yeah, i'm fine if you or bluematt want to take it over, i just left it open b/c i feel we might still want it
< instagibbs>
wumpus, how annoyed would you be if I opened a PR that future-proofed rpc calls against failure to check isNull if additional arguments are added(as well as fixing named arg calls in a couple places)
< wumpus>
instagibbs: a lot of cleanup in regard to named argument handling needs to be done, all over the place, biggest issue is that intermediate nulls are not handled as default value by most calls
< wumpus>
instagibbs: in any case it seems too late for that for 0.15, but would be a nice goal for 0.16 to get that consistant
< instagibbs>
With only a few dozen easy to review line changes I think we can catch most of it for 0.15.
< rhavar>
I have the minizinc version if you're interested and keep it private :P
< rhavar>
murch: btw the dataset you're using is actually from bustabit not moneypot
< rhavar>
The project used to be called moneypot, but was renamed to bustabit
< Murch>
oh okay
< rhavar>
Then I created a totally unrelated project, and called it moneypot rofl
< rhavar>
(and then sold it)
< rhavar>
If you want, i can give you a lot more data for your dataset
< Murch>
rhavar: I would be very interested in that.
< Murch>
Especially amounts + fee levels would be exciting, as I could then start researching the effect of dynamic fee levels.
< Murch>
(fee levels for the outgoing transactions)
< Murch>
rhavar: So, I read your description, it seems to me that you only define the fitness function and then have some sort of optimizer solve it for you?
< Murch>
That's an interesting approach. :)
< rhavar>
Murch: correct yeah
< Murch>
Where does the "*10000" come from?
< Murch>
Is that a fee rate of 10k satoshi/Byte?
< rhavar>
No, that's just to break ties
< rhavar>
Its just a useless number
< Murch>
oh, I see
< rhavar>
It's so that things with more "optional outputs" are picked
< rhavar>
(if the fitness is the same)
< Murch>
Did you see achow101's push to get my Branch and Bound algorithm into Core?
< rhavar>
I heard of it, but I haven't looked at the code
< Murch>
Ah, thanks, I just realized that you don't work with a fixed set of outputs but optional number of outputs
< rhavar>
Murch: well that's for the future
< rhavar>
At the moment, i do
< rhavar>
But i'm using this strategy for queued withdrawals (in next release) i think it'll give a lot better results
< rhavar>
I don't have the feerates, or an easy way to query it. But I can give you the txid's if you sanitize the data before publishing
< Murch>
rhavar: wow, awesome :DDD
< Murch>
I'm interested, and can do that, but I have a meeting in 40 min, can I get back to you later? :)
< rhavar>
sure
< rhavar>
Murch: If you can, can you update your README.md with installation / run instructions?
< rhavar>
I'll add my minizinc one if you make it easy enough for me :P
< Murch>
Okay, will do.
< Murch>
Cliffnotes for the moment, you need to setup Scala, compile the Scala code and then run 'coinselection.Simulation'
< Chris_Stewart_5>
Murch: Do you use sbt?
< Murch>
Chris_Stewart_5: I cleaned up the repo at some point to use the sbt standard structure
< Murch>
I'm not sure if I ever tried it out
< sipa>
cliffnotes?
< Chris_Stewart_5>
rhavar: You should be able to download sbt (scala build tool) from this link and then run 'sbt test' in the root directory for Murch's project
< BlueMatt>
(used by american students to cheat by allowing them to "read the cliffnotes" instead of the book)
< rhavar>
(In English class they make you read books and then write about it. But only suckers actually read the book, everyone else reads the cliffnotes)
< Chris_Stewart_5>
I just tried it and it works (well one test fails for me) but whatever
< Murch>
Mh, I'll have to look into it, thanks Chris
< sipa>
Murch: ok, what does it mean in this context?
< rhavar>
"summary"
< sipa>
i'm still confused
< Murch>
rhavar wanted to know how to run my project, and I was giving a summary, because I don't have time right now to write a set of instructions in full
< rhavar>
sipa: It's just a way of saying: "The short version is..."
< Murch>
rhavar wanted to use my coinselectionsimulator
< sipa>
ah, ok
< * BlueMatt>
wants a fourth exlanation of "cliffnotes" both in general and in context. clearly we dont have enough yet :p
< sipa>
i missed that it was the answer to a question
< Chris_Stewart_5>
the fourth explanation is where you sick of studying a particular class's notes, so you find the nearest cliff and toss them off of it :P
< BlueMatt>
+1
< midnightmagic>
Origin is an actual series of book summaries *called* cliffnotes, as an academia short-circuit-to-study, and that company was named after an actual guy named cliff who started it.
< Murch>
oh, @rhavar and anyone else running Coin Selection Simulations. You will probably not want to run with all strategies at the same time as especially the Bitcoin Core based selection takes quite a lot of computation time.
< Murch>
If you run all strategies at the same time, my home desktop would take about four days.
< Murch>
rhavar: Regarding the full set of the transactions. Thank you for the offer. I'm thinking that I might get back later to you regarding that set, as I'd first have to change my simulator to work with changing fee levels for the first place. For the moment having a much bigger set to play with is already pretty awesome.
< bitcoin-git>
[bitcoin] TheBlueMatt opened pull request #10784: Do not allow users to get keys from keypool without reserving them (master...2017-07-keep-change) https://github.com/bitcoin/bitcoin/pull/10784