428200+ entries in 0.309s
assbot: Logged on 14-07-2015 03:22:40; asciilifeform: a mistake
that is owned up
to, before it sinks any ships - can perhaps be accepted as a mistake, and forgiven
decimation: my point is
that it was made clear
that anyone who signed something
that fucked stuff up would suffer
the negrating consequences
assbot: Logged on 05-08-2015 04:19:08; decimation: plus, his original
theory was
that others should read and sign. now
this is good in
theory, but in practice has no upside and all downside
decimation: asciilifeform: are
the maps in question contained in your 'mempool zap patch'?
decimation: asciilifeform: yeah my browser eats
too (firefox)
mod6: it does
the same for me on multiple browsers
too
trinque: all I'm grumpy about is having something structured
to put
the patches in, really, with branching
ben_vulpes: patches and signatures in
the repository itself. i'd not be averse.
trinque: you can do
the same process better with a hierarchy of git repos
decimation: plus, his original
theory was
that others should read and sign. now
this is good in
theory, but in practice has no upside and all downside
☟︎☟︎ ben_vulpes: all i ever heard was "git=invitation for
tard collaborators"
☟︎ decimation: I
thought ascii was okay with repos as long as it was clear which patches 'caused'
the repo state
trinque: I watched a guy drink himself
to death because you know, unix wasn't everything-is-a-table
ben_vulpes: how much hell would i catch for sharing
this with others?
trinque: I bet it's in git
too, you filthy swine
ben_vulpes: i'm just
trying
to operate within
these crazy boundaries. i don't have any process improvements
that i'm sure wouldn't get shot down in a new york second from asciilifeform.
ben_vulpes: "what, you want
to automate
the nuclear power plant?"
trinque: the barrier can be "the devs are fucking vitriolic
towards laziness"
☟︎ ben_vulpes: "what, you want
to automate
this process
that keeps
the
tards at bay?"
ben_vulpes: "what's so hard about applying patches
to your private repository?"
ben_vulpes: "a real contributor would write all of
their own
tooling
to solve
this"
trinque: I had a hell of a
time figuring out where I should be re: patches, at least before giving up and using rotor instead.
trinque: will either need
tools built
to support it, or will need
to grab
tools
ben_vulpes: but
this is bitcoin, where since
time immemorial people have been shipping code
that can only ever be
tested in production and
then standing by in amazement claiming
that "nobody could have predicted!"
ben_vulpes: from experience i know
that
the more difficult it is
to
test
the software,
the less likely anyone is
to actually do so.
☟︎ ben_vulpes: this makes for quite
the
testing nightmare.
those interested in
testing must reverse applied patches
to your antecedent
tip and
then apply yours.
ben_vulpes: if you're up
to date on
the patching, submit a patch and note its antecedents.
ben_vulpes: last release, i'd prefer. asciilifeform
tends
to stack his patches atop one another.
BingoBoingo: Most intimidating part of current realbitcoin is where
the fuck do I submit a diff from?
☟︎ ben_vulpes: the fewer enforced constraints upon us,
the more discipline with which we must act.
decimation: I'm still confused about
this map
thing. does anyone know exactly which map ascii is
talking about?
ben_vulpes: "i don't know, let's
test it during code review"
ben_vulpes is struck by
the eerie resemblance
to $dayjaeorb here
ben_vulpes: i'm all for barriers
to entry here but
this is a goddamn development obstacle course
mircea_popescu: how about submiutting a "This doesn't work - don't use it"
thing on
the broken one
mircea_popescu: plox
tell me it's not one of
those where everyone is waiting for me
to do it first or something.
mircea_popescu: which is why
the fuck we evern have
the complexsignature scheme in place.
mircea_popescu: of course
the logical approach would be
to promote
to stable ml patches
that were signed
ben_vulpes: "all
testing" which is what we've seen
to date.
mircea_popescu: the more common cop out is "fuck you, it's all
testing".
mircea_popescu: i can appreciate
the heroism of "fuck you, it's all stable". but
then
that comes at a cost.
mircea_popescu: for one
thing, most everyone but us has a stable and a
testing list.
mircea_popescu: the ml process needs some reviewing. it's sort of grown organically
to a point where it's
troublematic.
ben_vulpes: not
that i'm not infinitely indebted
to you for unwedging my coracle but a hint of context'd be useful.
mircea_popescu: you know world is on fire
to
that degree pretty much weekly.
ben_vulpes: second patch contained a link
to mp's "dis wat i do, yo"
mircea_popescu: asciilifeform he still has a point. a) we're careening dangerously
towards -dev levels and b) people can't fucking follow wtf is on
that list.
☟︎☟︎ decimation: oh I
thought
this was all about
the stl map, not boost unordered map
ben_vulpes: plz for
to slightly more detail in patchdescs.
ben_vulpes: so i am
to expect
that patches arriving in
the ML aren't even
tested now?
ben_vulpes: asciilifeform: did you find a reason why maxint wouldn't work, or is
the second patch just using mircea_popescu's lock values for 'myzteeri0us reezuns'?
decimation: asciilifeform: so
the basic problem with
the map is
that it never goes out of scope?
ben_vulpes: however
the build i made on 7-15 with
the dumpblock patch boots from
the same datadir.
ben_vulpes: i'll
try
to nuke some state and come back with more useful information.