177200+ entries in 0.109s
a111: Logged on 2017-03-20 17:00 asciilifeform:
http://btcbase.org/log/2017-03-20#1629256 << even reiser is almost certainly waste of
time, general-purpose fs is very sharply
the opposite of what we want,
they are all optimized for mutability (can delete/rename/resize/etc) and fast reads at
the expense of slow entity creation, as well as carrying out silent rebalances/defrags/etc.
a111: Logged on 2017-03-15 23:46 asciilifeform: in unrelated noose, 'nqb' reads & parses a full 1MB block, with 2218
tx, and recreates it from fast-form, again
to disk, in 0.123 sec. on a 3GHz opteron cum ssd.
a111: Logged on 2017-04-08 17:13 Framedragger: do you recall when you described
the storage of nqb?
a111: Logged on 2017-07-08 02:31 mircea_popescu:
this locks into an older discvussion re bitcoin fs, which was iirc still stalled at perf-ing
the various available fs
thoroughly for massive directory/file usage.
ben_vulpes: i have other exciting glyph monstrosities occupying my stack
these days like unicode
mircea_popescu: mod6 problem with
the very specific words is
that often enough conversations are carried in wholly metaphorical
terms.
ben_vulpes: a reasonable daring, my patience for aramaic went
the way of my patience for kanji etc
mircea_popescu: asciilifeform hey,
the history of roman empire was attempted 500x, why should one fellow stop
the rest.
ben_vulpes: relatedly, i started rereading
the old
testament on "kindle"
this year, have been missing proper
talmud for whole sojurn
ben_vulpes: asciilifeform:
talmud likely five hundred years
to early of a comparison
mod6: and as far as pete_dushenski or whomevers quest
to write a log digest or whatever, it never happened.
the foundation briefly considered a role for gathering up
trb related parts, but
that was set aside in exchange for
tb0t
ben_vulpes: basically unsafe
to use unix
tools without reading manpages exhaustively, and probably source
to boot
ben_vulpes: asciilifeform: yeah, found
the flag at about
the same
time i discovered
the idiocy
mod6: sometimes its a
tmsr-ism like 'printolade'
mod6: oh yeah,
those are
the hardest
to
track down. i find myself having
to search for very specific words and hope i find it.
mod6: ah, yeah,
the "long walk"
mod6: yeah, i
try never
to get more
than 1 day behind. and sometimes, i
take long walks
through
the l0gs on various
threads.
a111: Logged on 2017-07-08 02:07 mircea_popescu: mod6 i used
to enjoy
the luxury of multipass reads
through
the log. but it's under
threat
these days
a111: Logged on 2017-07-08 02:01 mircea_popescu: certainly above qs could have been asked just as well by a noob, and vague "large
threads" reference dun help him any.
a111: Logged on 2017-07-08 02:00 phf: ah, so it's a subtle "what
tenets would
that be? handy if you made a LIST of
them here, eh!"
BingoBoingo: Well, alf did recommend
that film
to rest of republic
ben_vulpes: am i
to
take from
this line in
the od manpage
that it doesn't actually output faithfully without extra flags?
a111: Logged on 2017-07-08 01:24 phf: no, it's more like alfs rhetoric approach at some point was encouraged, where's now it seems
to be recognized, and often
treated with "oh it's just alf doing his
thing".
ben_vulpes: > do not use *
to mark line suppression
a111: Logged on 2017-07-08 01:13 phf: pretty sure asciilifeform is actively ignoring
the whole initiative, with only periodic pounces. as opposed
to his older strategy of annihilating
the whole
thing upfront. must be getting old
BingoBoingo: <asciilifeform> sina:
trolling asciilifeform is actually pretty hard <<
Then why are
the neighborhood's small mammals so successful with
their verminating?
mod6: not sure it'll ever see
the light of day
tho.
mod6: as a way
to
teach it
to myself, i've been poking around with a v impl.
mod6: im ~starting~
to get
the hang of
this ada stuff.
mod6: looks like was
taken on a proper ny roof
mod6: sina: sure,
think on it for a bit. no rush.
mircea_popescu: sina does sqlite have unicode support ? if it does,
then it will necessarily be less lines of code.
mircea_popescu: sina simple example and i don't pretend like
this is useful : /assignments/0f0a.txt contains or does not contain "hurrdurr". by checking if it does you know hurrdurr is assigned
to 0f0a.
sina: and you contend
the actual final implementation of such a
thing will actually be less lines of a code
than
the existing
thing
☟︎ mircea_popescu: well, for you in
that you get
to say you did it (ie, have
the experience), and for alf
that he's not stuck importing ALL
THE REST of sqlite.
mod6: i gotta look at
this code again here
sina: mircea_popescu: I guess
the gist of
the question is really, how is implementing my own btree different at all
to using
the pretty much identical
thing in sqlite
sina: although I don't get
that either now
that I
think about it, how does
the "assignments" file know which key is which?
sina: how does
that related at all
to
the "make file with keys"
mircea_popescu: this locks into an older discvussion re bitcoin fs, which was iirc still stalled at perf-ing
the various available fs
thoroughly for massive directory/file usage.
☟︎ mircea_popescu: sina you can implement your whatever as a binary
tree, leverage
the directory structure, and simply check if
there's a file / write it.
mircea_popescu: sina or just you know, look into how
to do btree and have a hash something ?
sina: mircea_popescu: but still even in
that case I need
to "walk"
the list of assignments, looking for ^available, so I need
to use grep or write a iterating-finder-thingo myself and
then something like sed
to change
the line or write a
text-changer-at-a-line myself
sina: ok
true, I didn't
think of
that
mod6: mircea_popescu: re, logs, inline-pr0n helps
too :D
sina: imagine flatfile example of "assigning" generated key from "available" => "user" state, or "user" => "bogus" state,
that's moving a keyfile from gossipd/keys/available
to gossipd/keys/users/foo or similar action, now my program has
to either invoke "mv" or write mv-like functionality into my app
mircea_popescu: i don't know it has
to be shitty, unless
there's some other constraint, like "make it overnight"
sina: and now if I want
to do it in flatfile, gossipd code becomes gossipd + "sinas shitty db-less attempt"
a111: Logged on 2017-07-08 00:38 asciilifeform: ( it adds a screamingly unwarranted runtime and nonfitsinhead complexity
to just about any proggy )
sina:
http://btcbase.org/log/2017-07-08#1680437 << why I used a db: because
the spec said use flatfile, I first
tried
to implement flatfile one and after realising I would need
to either shell out
to utilities like "touch" and/or "find"/"ls" etc, or implement some of
their functionality myself, I decided
to import a library
that does
that stuff not
terribly, called sqlite
☝︎ a111: Logged on 2017-07-05 15:39 phf: sina: is
that ^ a correct method?
sina: I return,
temporarily
mircea_popescu: mod6 i used
to enjoy
the luxury of multipass reads
through
the log. but it's under
threat
these days
☟︎ mod6: i
think it's highly interesting, but indeed,
there's a lot of depth. sometimes
takes a while
to grok
the
threads.
that's my own personal
take. getting old, not for me anyway.
mircea_popescu: feels like headlifting
three
times
the weight
these days
mircea_popescu: and in other news,
the logs stay about
the same 500 lines
they always are, but holy shit it seems somehow
they get denser or what
the fuck, it's unseemly.
mircea_popescu: certainly above qs could have been asked just as well by a noob, and vague "large
threads" reference dun help him any.
☟︎