< wumpus>
bsm117532: no, I don't think anyone is working on replay protection on our side of the chain, the risk is also that they might also try to thwart it, like w/ the service bits
< bsm117532>
wumpus: A proposal is far from complete, clearly.
< bitcoin-git>
bitcoin/master d0b1fc8 Vivek Ganesan: Docs: Change formatting for sequence of steps...
< bitcoin-git>
bitcoin/master f9b74ef MarcoFalke: Merge #11624: Docs: Change formatting for sequence of steps...
< bsm117532>
I think some mechanisms could not be thwarted, e.g. OP_CHECKBLOCKHASHATHEIGHT or committing to a chain's block hash with the signing algorithm.
< yurit>
Hello everyone. I am interested in learning about bitcoin one of the doubts that I have is technically. What makes the difficulty of mining increase?
< bsm117532>
yurit: try asking your question in #bitcoin
< bsm117532>
I'll answer there
< yurit>
bsm117532: ok
< cfields>
\o/ #10866
< gribble>
https://github.com/bitcoin/bitcoin/issues/10866 | Fix -Wthread-safety-analysis warnings. Compile with -Wthread-safety-analysis if available. by practicalswift · Pull Request #10866 · bitcoin/bitcoin · GitHub
< BlueMatt>
cfields: "Cannot post comment, commit is no longer a part of PR".......
< cfields>
eh?
< BlueMatt>
cfields: oh, ignore my comment
< cfields>
heh, comment at exactly the same time i force-pushed ?
< BlueMatt>
cfields: is SOCKET ever going to be a not-int?
< BlueMatt>
can we just pass-by-copy instead of pass-by-const-ref for an int?
< cfields>
BlueMatt: i think passing by const ref reinforces the idea that it's a handle. Otherwise I'm afraid someone may stash it somewhere and destroy it elsewhere
< BlueMatt>
hmm, alright fair
< cfields>
could wrap it up in a struct or something to force the semantics if you'd like?
< BlueMatt>
nah, all good
< BlueMatt>
can I get some comments on #11517? Specifically, it adds the ability for bench_bitcoin to generate html which has a <script ref="cdn.server"> - should we check in this plotly.js library (its apparently already mit) or would we prefer to keep the link in the generated output (its optional, you can still get the csv-style output on cli if you want)
< gmaxwell>
BlueMatt: is the plotting thing that you use with fibre that kills firefox?
< BlueMatt>
gmaxwell: heh, no, thats some google charts library thing
< BlueMatt>
gmaxwell: and, fwiw, I can now *only* plot it on firefox
< BlueMatt>
it blows chromium up by exceeding some per-tab memory limit they have
< BlueMatt>
but firefox does fine
< cfields>
BlueMatt: how much of plotly would we need?
< BlueMatt>
cfields: no idea, haven't looked, but the usage is rather trivial
< BlueMatt>
its just like making box plots of runtimes of benchmark runs
< cfields>
I'm confused as to why that wouldn't just be a server-side thing, though? Why not just parse a csv in realtime?
< BlueMatt>
server-side?
< BlueMatt>
there is no server?
< MarcoFalke>
The plotly printout is optional, right?
< cfields>
BlueMatt: I meant that outputting a reasonable csv seems like enough. If a nice plot/markup is desired, that seems beyond the scope of the bench tool
< BlueMatt>
MarcoFalke: yes, default is to print the csv-style output
< MarcoFalke>
I don't mind adding an option for common formats
< BlueMatt>
cfields: its nice to have a tool to do it....anyway, I was asking for comments, would like to get comments ont he pr
< BlueMatt>
on the
< cfields>
ok, will do
< MarcoFalke>
They would just sit around free for people to use and fix in case they break. Why not? Would at least encourage some to run the bench more often
< MarcoFalke>
Or we could host a nightly bench on github pages
< cfields>
if it can be setup so that you just supply your own plotly, that seems fine to me. But if it requires that we commit it, or fetch it, i'd be uneasy
< sipa>
cfields: yes, it's just a url
< phantomcircuit>
BlueMatt, 2.8 GB btw
< BlueMatt>
phantomcircuit: E_NO_CONTEXT
< phantomcircuit>
memory requirement under chromium for fibre stats page
< BlueMatt>
yea, well also that stats page hasnt been updated in months cause I changed the network topology and haven't updated the stats-parsing script to actually understand the new topology