Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
Remove coin age priority and free transactions - implementation #9602
This PR removes all coin age priority functionality from the codebase.
Some remaining open questions:
referenced this pull request
Jan 20, 2017
referenced this pull request
Jan 31, 2017
Nice! Code review ACK, will test.
Left a few non-blocking nits for you to consider; I think they would be fine to address/discuss in future PRs if you preferred.
- Changes make sense, though I think this might been have been easier to review if it consisted of a just few minimal commits that update the behavior, followed by a single larger commit that removes dead code without changing behavior.
- I didn't understand what you mean about potentially bypassing the mempool min fee in the event of mempool limiting. Could you say more about how this would work, what would be the advantages / disadvantages? Also, in the comment that "fLimitFree is now only set for transactions we are trying to reaccept to the mempool," assuming you meant "set false" not "set".
- Re: prioritisetransaction, one risk of reducing the number of arguments is that at some point in the future, somebody adds a new argument, and then old code calling the RPC with priority may appear to work but silently be broken. You could keep the priority_delta arg and require it to be null or 0 to prevent this. But I think getting rid of the argument is nice if you don't think it's likely someone will add a new prioritisetransaction argument in the future.
You may also wish to update the comment above paytxfee in contrib/debian/examples/bitcoin.conf
@jnewbery I didn't want to get into a whole big deal about what the current purpose of minrelaytxfee is so I left the comment minimally changed for now.
As for the backwards compatibility of prioritisetransaction, that was one of my open questions from the top of the PR. I'll defer to group consensus, but my viewpoint is if we never change it then we're stuck with a stupid dummy field for all eternity, and if we are ever going to change it, now is the time to do so. I think its a pretty safe change as it'll either work the old way or the new but they can't be mixed up.
utACK b421e6d (only reviewed tests and wallet, did not look closely at miner and mempool changes.) Some notes about doc fixes for later commits: * The change in check-doc.py seems no longer relevant * The rpc breakage should be mentioned in the release notes * The misleading comment in init.cpp about minrelaytxfee can be removed alltogether. Better no doc than wrong doc and inline comments are the wrong place for user documentation anyway.…
On Mon, Mar 6, 2017 at 5:17 PM, Matt Corallo ***@***.***> wrote: utACK b421e6d — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.