< GitHub181>
bitcoin/master 889426d Andrés G. Aragoneses: autogen.sh: warn about needing autoconf if autoreconf is not found...
< GitHub181>
bitcoin/master 9b9bfce Wladimir J. van der Laan: Merge #7528: autogen.sh: warn about needing autoconf if autoreconf is not found...
< GitHub20>
[bitcoin] laanwj closed pull request #7528: autogen.sh: warn about needing autoconf if autoreconf is not found (master...warn-autoconf) https://github.com/bitcoin/bitcoin/pull/7528
< GitHub166>
bitcoin/master 1ecbb3b Luke Dashjr: depends: Use curl for fetching on Linux...
< GitHub166>
bitcoin/master 5d1148c Luke Dashjr: Travis: Use curl rather than wget for Mac SDK
< GitHub166>
bitcoin/master c01f08d Luke Dashjr: Bugfix: depends/Travis: Use --location (follow redirects) and --fail [on HTTP error response] with curl
< GitHub119>
[bitcoin] laanwj opened pull request #7536: test: test leading spaces for ParseHex (master...2016_02_test_parsehex_leading_space) https://github.com/bitcoin/bitcoin/pull/7536
< GitHub131>
[bitcoin] laanwj opened pull request #7537: wallet: Warn on unexpected EOF while salvaging wallet (master...2016_02_salvage_unexpected_eof) https://github.com/bitcoin/bitcoin/pull/7537
< cypherBlock>
for segwit transactions types that are nested in P2SH (p2sh-p2wsh & p2sh-p2wpkh), is the intent that those are there primarily for un upgraded wallets to send bitcoin to upgraded wallets? Any other usage?
< cypherBlock>
sipa: if you have a few secs, see above question about p2sh wrapped transactions. Thanx.
< morcos>
Any thoughts on what kinds of privacy protections I should think about building in?
< morcos>
Implementing this directly would make it trivial to correlate a public node and its onion counterpart.
< wumpus>
maybe quantize the result?
< morcos>
wumpus: yes thats not a bad idea, i thought about adding some variance to the sending of the message, but it seemed insufficient, combining that with quantization might help
< wumpus>
right, even randomizing just a bit will already make it much harder to correlate nodes
< wumpus>
you'd also have to take care to not send the feefilter message to all peers at once
< morcos>
it looks to me from some recent testing that in current conditions this will save a typical node from dowloading about 500MB of tx data a day
< wumpus>
awesome
< morcos>
of course it might be significantly less than that once 0.12 is rolled out everywhere anyway, not sure how much these small feerate txs will keep bouncing around, but its possible they will
< morcos>
oh yes, by randomizing the sending i meant randomizing the timing of the sending
< wumpus>
BTW: having a node listen both on .onion and public IPv4/6 is already not recommended for privacy-sensitive nodes, it's mostly useful to help .onion users, but there may be other ways to correlate as well
< morcos>
ok yeah i wasn't sure how much effort i should put into trying to avoid that correlation, but at least not making it trivial seems worthwhile
< wumpus>
well randomizing the value a bit for each node you send to makes sense too
< jarret>
I will have to relearn my screen metakeys too ;)
< cfields>
Luke-Jr: do you still have the list of ip ranges? that paste is gone
< Luke-Jr>
cfields: PM'd
< cfields>
thanks
< cfields>
Luke-Jr: done
< Luke-Jr>
k, I'll test it in a bit
< cfields>
btcdrak: file copied too
< btcdrak>
cfields: thanks
< michagogo>
Okay, I nuked my cache and rc5 sigs
< michagogo>
Going through with the long rebuild. Script *should* just work, as far as I know... but please don't push/merge to the sigs repo before my PR goes up
< michagogo>
(Either that, or pull from my fork first... When the builds finish, if git can't fast-forward (git pull --ff-only) it'll fail and not commit/push.)