IRC meeting summary for 2015-11-12
- transaction priority for 0.12
- Opt-in replace-by-fee
- Chain limits
transaction priority for 0.12
Each transaction is assigned a priority, determined by the age, size, and number of inputs. Which currently makes some transactions free.
This currently has a large amount of code, which makes it harder to maintain, and is not that optimal since you can’t expect miners to include 0-fee transactions.
Most people seem fine with removing priority in the mempool, but people should be notified ahead of time this is coming.
sdaftuar proposed a staggered approach, setting the default value for priority to 0, and remove it entirely in the next release.
petertodd notes there will be a natural staggered process since not everyone will upgrade to 0.12 instantly and some implementations might not remove priority at all.
Most wallet-software outside of bitcoin-core don’t implement priority calculations.
As fee estimation becomes more important and many wallet providers use the bitcoin-core fee estimation, improvements on that are welcome.
Luke-Jr doesn’t agree with removing priority, particularly with changing the mining code to use the priority a transaction has when it enters the mempool.
Sipa has the idea to add a small fraction of bitcoin days destroyed divided by the average UTXO age to the fee, so that non-spam-attack transactions are viewed as if they have a larger fee.
While most agree with the proposal to remove the current priority, there’s still much debate on whether it needs to be replaced for 0.13, and if so, how.
Currently when a node sees a transaction that spends the same output it ignores it. With replace-by-fee it replaces the current transaction in the mempool if it has a higher fee.
This allows for things like spending “stuck” transactions, adding more recipients to a transaction in order to prevent chaining, etc.
Since there are people that accept 0-confirmation transactions and this would make it extremely easy to double spend them, this is made opt-in.
The sender can choose to opt-in to replace-by-fee by changing an input in the nSequence field.
Peter Todd wrote some tools to use replace-by-fee. link
It would be nice to have opt-in per output instead of the whole transaction, however that would be very hard to implement and would have some privacy concerns.
Luke-Jr would like to see an option to toggle between first-seen-safe/full RBF and never/opt-in/always. Since there are possibly some objections with the “always” toggle it should be a separate pull-request.
review and merge nSequence-based Full-RBF opt-in
Peter Todd will write a mail to the list to explain how it works and how people can not accept opt-in transactions.
Currently softforks have been done by the isSuperMajority mechanism, meaning when 95% of the last X blocks has a version number higher than Y the fork is deployed.
A new way of doing this is currently being worked on and that uses all bits of the version number, appropriately being called versionbits. So instead of a fork happening when the version is larger than (for example) 00000000011 (3), a fork happens when (for example) the 3rd bit is up (so 00100000011).
This way softforks can be deployed simultaneous and independent of each other.
There are 2 different implementations. One from Codeshark and one from Rusty
jtimon thinks both implementations are more complicated than they need to be.
There needs to be a minor revision namely a starting time for proposals.
In general we’d like to get this in soon, but existing softforks need to complete first.
CodeShark adds a starting time to versionbits.
Chain in this context means connected transactions. When you send a transaction that depends on another transaction that has yet to be confirmed we talk about a chain of transactions.
Miners ideally take the whole chain into account instead of just every single transaction (although that’s not widely implemented afaik). So while a single transaction might not have a sufficient fee, a depending transaction could have a high enough fee to make it worthwhile to mine both.
This is commonly known as child-pays-for-parent.
Since you can make these chains very big it’s possible to clog up the mempool this way.
With the recent malleability attacks, anyone who made transactions going multiple layers deep would’ve already encountered huge problems doing this (beautifully explained in let’s talk bitcoin #258 from 13:50 onwards)
Proposal and github link.
Wumpus doesn’t feel comfortable with merging it because there’s some controversy from companies who exceed the limits (or could be/want to).
jgarzik does feel comfortable with it, and many think it should be merged as it’s easy to revert if needed.
There’s little choice as it’s not safe from attacks without limits.
We should communicate the replace-by-fee sendmany alternative to long chains (adding new recipients on existing non-confirmed transactions), although it won’t show up in users wallet yet and block-explorers probably aren’t ready to display it correctly.
Emphasis on the fact it’s a change in default values, not a consensus change, however default values have a lot of power.
The final limits are 25 transactions and 101kb total size for both ancestor and descendant packages.
jgarzik will merge the pull-request.
Morcos will mail the list once it’s merged.
BlueMatt Matt Corallo petertodd Peter Todd morcos Alex Morcos jgarzik Jeff Garzik gmaxwell Gregory Maxwell wumpus Wladimir J. van der Laan Luke-Jr Luke Dashjr jtimon Jorge Timón btcdrak btcdrak phantomcircuit Patrick Strateman sipa Pieter Wuille CodeShark Eric Lombrozo sdaftuar Suhas Daftuar jg_taxi jg_taxi gavinandresen Gavin Andresen cfields Cory Fields bsm1175321 Bob McElrath
19:53 sipa new topic? 19:53 wumpus any other topics? 19:53 petertodd <crickets> 19:53 jgarzik did we cover jonas while I was in the taxi? 19:54 sdaftuar ? 19:54 jtimon ? 19:54 CodeShark not sure I want to know 19:54 jgarzik proposal for new GUI maintainer 19:54 CodeShark sounds kinky, though 19:54 petertodd CodeShark: GUI's are pretty kinky 19:56 BlueMatt ok, end meeting? 19:56 btcdrak if we can remember the command this week :-) 19:56 wumpus #meetingend 19:56 gmaxwell #destroymeeting 19:56 wumpus #endmeeting 19:56 Luke-Jr #endmeeting 19:56 lightningbot Meeting ended Thu Nov 12 19:56:42 2015 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) 19:56 BlueMatt #magicmeetbotincantation 19:57 petertodd #DoWhatIMean
This summary was originally compiled by Stefan Gilis aka “G1lius” and posted to the bitcoin-discuss mailing list with the disclaimer, “Please bear in mind I’m not a developer so some things might be incorrect or plain wrong.” and placed copyright in the Public Domain.