< GitHub88>
bitcoin/master 21e45a0 Andrew C: Fix history deletion bug after font change...
< GitHub88>
bitcoin/master b24266c Jonas Schnelli: Merge #7668: Fix history deletion bug after font size change...
< GitHub180>
[bitcoin] jonasschnelli closed pull request #7668: Fix history deletion bug after font size change (master...rpcconsole-hist-fix) https://github.com/bitcoin/bitcoin/pull/7668
< jonasschnelli>
Hmm... getting a new warning:
< jonasschnelli>
torcontrol.cpp:365:24: warning: private field 'base' is not used [-Wunused-private-field]
< jonasschnelli>
But seems to be used in TorController::TorController constructor... strange..
< paveljanik>
there is a global base
< paveljanik>
it can be removed from TorController::TorController and commented out around l. 365
< paveljanik>
I do not know if the global should be prefer or the private one there.
< paveljanik>
let's wait for wumpus
< wumpus>
we should probably rename the global one
< wumpus>
(I'd like to go without it completely, but I don't think that's possible)
< wumpus>
it would work if our initialization sequence was structured differently, and StartTorControl would return an instance instead, but with the global StartTorControl/InterruptTorControl/StopTorControl flow, some state needs to be global
< wumpus>
it shouldn't be called the same as an instance variable though
< GitHub154>
[bitcoin] laanwj opened pull request #7676: torcontrol: rename global base, avoid overlap with field (master...2016_03_torcontrol_rename_base) https://github.com/bitcoin/bitcoin/pull/7676
< wumpus>
eventually we could want to move toward having one libevent base, instead of having separate ones for RPC and torcontrol, but best to do this after cfields_' net refactor
< wumpus>
(which will introduce a third one for the P2P)
< wumpus>
paveljanik: I don't like that warning - it's correct, after #7637 we don't need to keep `base` around at all in TorController right now because we only use it in the constructor. But I think it's good practice to do so anyway, in case it's needed again in later changes...
< GitHub155>
bitcoin/master 8a253b3 Pieter Wuille: Make the generate RPC call function for non-regtest
< GitHub155>
bitcoin/master c87f51e Wladimir J. van der Laan: Merge #7663: Make the generate RPC call function for non-regtest...
< GitHub3>
[bitcoin] laanwj closed pull request #7663: Make the generate RPC call function for non-regtest (master...generatenonreg) https://github.com/bitcoin/bitcoin/pull/7663
< paveljanik>
wumpus, 8) When it is needed in later changes, it can be resurrected from git history 8)
< wumpus>
paveljanik: of course, but I think the warning is over-eager
< paveljanik>
every warning is disturbing, hides probles etc. But I know you have a different opinion :-)
< paveljanik>
and I respect that.
< wumpus>
sure, they're annoying, but in some cases it makes more sense to disable the warning than change the code
< paveljanik>
I agree with that - in some cases. Like known compiler bug etc...
< wumpus>
it's not up to compiler authors to decide about our code style
< wumpus>
looks like they're making the compiler behave like a nanny with too much time on their hands
< paveljanik>
I agree with that completely ;-)
< wumpus>
all time spent on 'fixing' the code to be as acceptable as possible to the compiler is not spent on facing user-facing issues, ok and that concludes my rant on this
< MarcoFalke>
wumpus, what is the plan for 0.10.4 and 0.11.3?
< wumpus>
no plan at this moment
< wumpus>
another 0.10.x would require a critical reason, we could do a 0.11.x if there's good reason
< btcdrak>
the next maint release should be the CSV softfork imo
< wumpus>
sure, well, to be honest I don't really keep track of 0.11 backports, may make sense to appoint another release manager for backport releases
< wumpus>
(e.g. as linux has)
< MarcoFalke>
0.11 currently has my "[dbwrapper] Detect obfuscation" patch and the wallet issues
< MarcoFalke>
Though, it also includes the "paytxfee changed behavior without warning"
< wumpus>
possibly Luke-Jr, though maybe he is a bit too trigger-happy with backporting
< jonasschnelli>
paveljanik: I agree the var names should be different... but no strong opinion. The struct event_base *base; at L662 should not interfere the private struct event_base* base; of TorController (not even shadow).
< paveljanik>
jonasschnelli, and this is what your and wumpus 's change does :-)
< paveljanik>
but anyway, warning gone with your PR
< jonasschnelli>
+1
< jonasschnelli>
paveljanik: thanks for testing...
< paveljanik>
it was done almost without me, automagically 8)
< morcos>
btcdrak: ok, i submitted a PR to your repo for my new RPC test. it's a handful, but i think it now tests much much more of the consensus logic.
< btcdrak>
morcos: thank you very much! I'll take a look after dinner.
< GitHub73>
[bitcoin] sipa opened pull request #7688: List solvability in listunspent output and improve help (master...helpspendsolv) https://github.com/bitcoin/bitcoin/pull/7688
< GitHub87>
[bitcoin] sipa opened pull request #7689: Replace OpenSSL AES with our own constant-time version (edit of #5949) (master...const_aes) https://github.com/bitcoin/bitcoin/pull/7689