< Naphex> "In the problematic situation, packets queued in a network buffer are only dropped if the buffer is full. Having a big and constantly full buffer which causes increased transmission delays and reduced interactivity, especially when looking at two or more simultaneous transmissions over the same channel, is called bufferbloat. Available channel bandwidth can also end up being unused, as some fast
< Naphex> destinations may not be reached due to buffers clogged with data awaiting delivery to slow destinations, what is caused by contention between simultaneous transmissions competing for some space in an already full buffer." https://en.wikipedia.org/wiki/CoDel
< Naphex> a fair queue / codel-like scheduling algorithm for the mempool is starting to be needed
< gmaxwell> this is unrelated to bitcoin---- an entirely different and largely inapplicable area of science which refers deals with adaptive rate control against invisible remote buffers of unknown capacity.
< Naphex> yes
< Naphex> but i am looking for some of its properties
< Naphex> we have a known capacity limit (1MB) and invisible remote buffers of unknown capacity(mempools?)
< gmaxwell> I haven not seen anything in this space which is applicable (and I was participating in the IETF AQM working group since the first BOF session on it.)
< gmaxwell> Naphex: mempools are fixed capacity.
< gmaxwell> (and priority ordered)
< Naphex> yep but the capacity is a user preference
< Naphex> although defaults are presented
< Naphex> yep - and at the moment it displays increased delays and low interactivity
< gmaxwell> Naphex: what you're doing there is pattern matching words; this doesn't mean that the things are at all related.
< Naphex> CoDel distinguishes between two "types" of a queue (or rather, the effects produced by a queue): (mempool beeing the queue / confirmed getting pased)
< Naphex> gmaxwell: i'm trying to explain it better; i am only trying to think of unconfirmed transactions as a queue
< gmaxwell> You might want to take your brainstorming elsewhere, this channel is for concrete activity around bitcoin core today.
< Naphex> i'll take your hint
< gmaxwell> Naphex: you can try #bitcoin-wizards; but without a more coherent view of the applicability I don't know if it will be productive, but someone there might want to discuss it with you.
< AaronvanW_> is CPFP in use today?
< aj> eligius uses it i think
< zibbo> Does anyone have statistics on how many miners implement opt-in rbf logic?
< GitHub126> [bitcoin] instagibbs closed pull request #7602: [WIP] [RPC] Add call zaptransaction to delete transaction from wallet (master...zaponetx) https://github.com/bitcoin/bitcoin/pull/7602
< GitHub188> [bitcoin] laanwj closed pull request #7461: Make internal (core) errors show up in the Qt client. (master...propagateAlert) https://github.com/bitcoin/bitcoin/pull/7461