211100+ entries in 0.156s
a111: Logged on 2015-08-02 20:02 mircea_popescu: siemens is "german" in
the sense
tsipras is "greek" and so on. absolutely nothing
to do.
trinque: it may sadden
those
that don't yet know
that open firmware was a dead simple forth nugget. i.e. in
the sense
that apple for a brief
time meant dead simple.
Framedragger: i see - i'm behind on
the internet of shit it seems
Framedragger: i
thought huawei only made consumer shit? (not
that
that's reassuring...)
BingoBoingo: ty asciilifeform, figured I'd
take big picture leaving opportunity for experts
to fill in mined lulz with further submissions
davout: (this unconfirmed
tx
thing)
davout: i really don't feel strongly about
this, just
thinking out loud
trinque: and if I wanted
that I'd want
to do it as a "mempool explorer"
tool
davout: asciilifeform: mebbe operator wants
to check his
tx propagated?
mircea_popescu: i concur. "unconfirmed
tx" is improperly
thought of as a
tx.
trinque: why does
that help anything?
davout: asciilifeform: knowing about incoming
txes before
they've confirmed?
mircea_popescu: asciilifeform incidentally your ssh
thing may well have been a major step forward for
this split.
trinque: why would it send either? crap a signed
tx and operator feeds it
to network node
mircea_popescu: if net-trb decides
to engage in a reorg, it will eventually induce one in all
the wallet-trbs connected
to it.
mircea_popescu: meh. wallet doesn't NOT store
the chain. it's just stock
trb.
davout: for example, how does
the wallet perform a reindex, without storing
the chain itself
trinque: wallet node could be munching with
that hole
mircea_popescu: asciilifeform just about. and large in
the sense of alcatel not in
the sense of alphabet, at
that.
trinque: davout:
there is also eatblock
davout: see, i don't see how it could be done without designing another way for nodes
to communicate in another way
than
the currently existing protocol
mircea_popescu: i suspect it's
the simplest-possible, which still doesn't promise it's simpler
than any arbitrary level.
davout: this split really doesn't sound
that simple
to me
mircea_popescu: you understand ? backfeed fucking computing, "if i do
this
this happens so you press
thios button" sorta excel-dev.
a111: Logged on 2017-03-07 17:22 davout: which allowed any existing node
to
trivially serve any arbitrary wallet, at
the cost of a 2gb memory scan for unspent outs
mircea_popescu: well no, let
them
talk
to each other as
they do now exactly.
mircea_popescu: then also conceivably
the changes will be backported into
trb main once well
tested
davout: design a way for
them
to
talk
together
davout: yeah,
take a few
trbs and cut different parts from each basically
mircea_popescu: conceivably can use current
trb as your genesis for
this.
mircea_popescu: then you can
take stock
trb, operate on it, and so people interested can pick both version, have internal-external couple.
mircea_popescu: you can just
take stock
trb, operate on it, and place it so it only connects
to operator-specified
trb nodes.
davout: well, as i understand it
the end result is a different set of binaries
that
talk
together
through a $to-be-defined queue mechanism
davout: so now, re your description, i'm not sure what a good start would be, because cutting it
the way you described sounds like a mega-project
that encompasses more
than
the mere wallet
davout: it would have allowed some level of split, but anyway,
this particular idea is dead since
trb doesn't have
this in-memory data structure
davout: which allowed any existing node
to
trivially serve any arbitrary wallet, at
the cost of a 2gb memory scan for unspent outs
☟︎ davout: i
thought
trb was keeping
them in ram already
mircea_popescu: keeping a pile of ALL utxos, as
the alternative, is not very good, because you end up storing a bunch of (potentially
toxic) crap
davout: in essence,
that's already what
trb does
mircea_popescu: i don't see
the user has a leg
to stand on, "i'm not sure which my addresses are". if you add some -
there's a (minor) penalty.
davout: as far as
the wallet is concerned it's pretty much "when new address is added, rescan blockchain, have relevant
txes in specific collection"
davout: mircea_popescu: i agree
that
the description made in
the convo you linked is a very desirable
target architecture
a111: Logged on 2017-01-03 20:51 davout: hrm, i'll have
to re-read
mircea_popescu: lol ok. anyway,
the issue is
that you seemed diverged, ie conversation is
there, you're doing something else with no real discussion of why / what's wrong with it / etc.
mircea_popescu: well, if you're not willing
the read
the jan 3rd discussion, i guess.
davout: so basically i
thought it would be quite easy
to cleanly decouple
the wallet from
the node, but i'm now stuck with
the realization
that for now
the most workable option will simply be
to make small improvements here and
there
davout: has
the index, should ask user
to select outputs, not automagically generate
transaction
mircea_popescu: i also dun get 2. how's
the fact you already have
the index you need related
to
the fact prb fucks up output selection ?
davout: 2.
the way
the existing
trb wallet currently works, mainly
the automatic utxo selection
mircea_popescu: i
thought
that was adequately resolved
through layering.
davout: 1.
the principle of
the
thing where
the wallet is coupled
to
the node and chain storage
davout: the existing wallet already does
that