110800+ entries in 0.062s
a111: Logged on 2018-04-09 15:58 asciilifeform: funnily enuff ~nobody alive knows how
to actually build a ruby interpreter from src
mircea_popescu: "technology we've been working at for five years" "because you're retarded, and your
tooling sucks"
mircea_popescu: all
these fucktards, you know. oh, "smashing success". dude, it's an off
the cuff item gets done if it even needs any doing over a weekend.
phf: fwiw i have it, i'll put mine online in
the next day or
two
mircea_popescu: phf ben_vulpes Framedragger any one of you want
to add a gopher service
to
teh logotron ?
mircea_popescu: the funny
thing is
the cocksuckers aren't even PAID for
the fluff pieces.
mircea_popescu: ("oh mp,
that's racis, it just went unreported before" "your mother went unreported.")
mircea_popescu: (fun fact : installing security cameras increases
the chances of sex act occuring in
the space like 30x)
mircea_popescu: so
they came up with
the emo spew of
the next decade : github contributions!
mircea_popescu: it's a means of expression, you udnerstand
this ? emo mostly died because it was evidently pointless,
to
the poor misunderstood darlings
themselves, "all
this '''poetry'''
that does nothing, not even rhyme."
mircea_popescu: (this is whart you get if you permit chelsea barabas and other knitters into "tech" : inept articles about how "my ass is really really fat and
the couch is really really comfortable".)
mircea_popescu: asciilifeform i dunno, something about how "my delicious cake is in
the fridge"
mircea_popescu: it's all an exercise in paper-over, and
their only incentive is
to claim surface.
mircea_popescu: the motivations are
transparently unprincipled. i mean --
they published "support" at
the bottom, which 100% consists of cheap linkbait by assholes a la "pump.io" and so forth. support my foot.
mircea_popescu: no because germany is not a handpuppet. it has like, courts and shit. respek
the sovereign cumrag!
a111: Logged on 2018-04-06 20:58 asciilifeform: mircea_popescu: it is not even clear
to me
that eben moglen (note, not rms) gpl experiment , i.e. 'use copyrasty against itself', worked for a short while (
to naked eye it would seem
that it did, bsd routines got subsumed into microshit, gnu -- not , but
there ~are~ alternative plausible explanations
that don't involve 'fear of gpl violation' )
mircea_popescu: asciilifeform
there was also
that hellvig lawsuit (which afaik went exactly
the same place all other "rule of law" challenges
to notes-from-stalin ever went)
☟︎ mircea_popescu: incidentally, anyone looked into
the guts of vmware's esx (now "esxi") see how exactly it does
the hardware driving (ie which bsd it lifted from) ?
douchebag: Honestly,
Texas seems like a pretty nice place
to live
trinque: guy makes good stuff, so I never questioned his
technique
trinque: girlfriend's pop is from east
texas, where
the coon-ass leaks over from louisiana. any
time he cooks something
tasty, it's "I put my foot in
that!"
mircea_popescu: and in
today's funny spanish lesson : pie, in spanish, means foot.
this, of course, does not prevent
the uppity coffee house i favour from being all international and offering pie de manzana.
trinque: dunno what
to grasp foar here
mircea_popescu: the good news for you mod6 being
that you don't really have
to do anything ; v.pl is fine as it is, for now, and once phf has a prototype walker you can just move on
to
that and it'll be cheaper. continue work on your ada guy in
the meanwhile at your leisure, seems we're converging
that way.
douchebag: That's what I
told
them, but you know how women are
trinque: typically I run
them in 24hrs or fewer.
trinque: get
to
them both in a bit. doing some surgery.
douchebag: trinque: One of
the girls is wondering when
the
transactions from last night will go
through
mircea_popescu: ben_vulpes : is
this a contract you folks want, incidentally ?
mircea_popescu is commissioning a new server with a 1gbps unmetered port
to act as a
torrent client, because i've finally decided i wish
to keep copies of all
the films i review.
trinque: wont be
the vtronic portage just yet, but yeah, I can install an entirely usable gentoo for ya.
mircea_popescu: soo
trinque, can i hire you
to install your blessed gentoo in
there ?
☟︎ mircea_popescu: (the great win being 注：爱比特.聂意译，翻译的不一定准确，也可能理解的完全相反，欢迎高手指教。~aka attention: my
translation is not necessarily accurate,
trilema may very well mean
the exact opposite. such is
the world of experts.)
mircea_popescu: anyway,
this PARTICULAR item is a chinese forum
that actually
tracks us / includes (not
terrible)
translations of
trilema pieces for commentary etc.
mircea_popescu: asciilifeform
there's lots of mfa using my name / stuff. apparently i'm a valuable google adwords item.
mircea_popescu: but
the devil is in
the details. consider how
this flows : "so mp, do you want a gui ?" "uh ok" "should it have any buttons on it ?" "kinda what a gui is" "every button will have
to be specified, as
to location, as
to..." "just get on with it" "6mb, will
that be cash or check on other people's
time ?" "motherfucker..."
mircea_popescu: 165
thousand lines! hundreds upon hundreds of files. IT DOES PRETTY MUCH NOTHING!
a111: Logged on 2018-04-07 01:25 mircea_popescu: 6mb
to get a fucking blog cms going. oh epic world of bits...
mircea_popescu: need-driven rewrite, best engineering known
to man. it's certainly how i produce
trilema.
phf: i suppose
that's one way
to release btcbase source: slowly rewrite it all in ada
mircea_popescu: the easiest press walk may in principle become very VERY expensive on large enough projects (in
terms of
the bitcoin blockchain large, 6 digit patches with 1-3 digit files included each it'd be a
true unassailable nightmare). but
then again it also bakes in some incentive
to NOT DO
THAT.
mircea_popescu: phf i guess next on your plate is make a prototype proper grapher for all
time ?
then mod6 can put
that in v and avoid further such wtfs.
☟︎ mircea_popescu: yeah. so on one hand indeed
the problem was as suspected ; but on
the other hand phf is correct and yes
the graph IS constrained.
phf: obviously
that works because btcbase produces
the correct dependency graph
phf: fwiw
there's no ambiguity in
the patchset,
the dependency graph _is constrained_. but you can't make
that constraint rule based, you can e.g. discover
that constraint by walking
the graph while keeping
the
track of
the checksums of
the press (i suspect
there's a smarter graph
theoretical way of doing it, but a "press walk" is
the easiest).
hanbot: right, correct patchset on
the left
then presses + flows
to vtools_vpatch_newline
mircea_popescu: which is a useful warning, because in principle if $1 and $2 could both delivered wanted $file, but
the author intended for $1
to be used, $2 might produce different $otherfile
than author needed, and
then you can't press because your hashes no longer match downstream"
mod6: mircea_popescu: solid plan. i can add something like
that perhaps when alternate paths are detected.
mod6: asciilifeform: ahhhh interesting,
thanks for reporting.
mircea_popescu: no, what it should do is yell "alternate paths
to press $vtools_fixes_static_tohex.vpatch requirement $x.alloc : $vtools+patch and $vdiff_sha_fixes_newline_gcc."
mod6: mircea_popescu: i can
try
the leftmost
trunck on phf's viewer, but not right
this minute. maybe in a few hours. i'm pinned down in
the mud. gotta kill
the guy in
the pillbox first.
mod6: yeah,
this is
true
too.
mod6: in
this case, it
throws out: vdiff_sha_static.vpatch and vtools_fixes_static_tohex.vpatch
mircea_popescu: wouldn't have gained us anything here, it'd just have pointed
to vtools_fixes_static_tohex, but we pretty much knew
that one is where it dies anyway.
mod6: mircea_popescu: well, for instance, it would be nice if I'd have put a statement in
the release version
that yells when it
throws a vpatch out for rule breakage.
mircea_popescu: it's not directly clear just ~how~ it could be more verbose here, seeing how it doesn't magically know all
the "meanings" we associate
to
the patches
themselves.
mod6: So indeed, better failure messages are needed. And if we're going
to allow
this
type of strange project structures,
that'll have
to get built into
the rules.
mod6: alf already instructed me a number of months ago
to be more verbose about failures. I didn't get
that done as much on
the last release as I was
trying
to limit
the scope
to
the last change for
the press-paths.
mircea_popescu: (which is not even wrong, as
the serial application of patches from different branches can result in a messed up codebase ; but it could be more verbose about what happened)
mircea_popescu: mod6 i suspect we ran into a novel situation : if
there's ambiguous file usage in multi-head bundles, v panics.
mircea_popescu: a right. hanbot do me a favour : download ~only~
those patches which are in
the leftmost
trunk seen on phf's viewer (so exclude vtools_vdiff_sha, and its dependents) and
try
to flow again ?
☟︎☟︎☟︎ phf: mircea_popescu:
that's one of
the 98BA's
mircea_popescu: phf what is
the vdiff_sha_fixes_newline_gcc link
to vtools_fixes_static_tohex ? as now apparently
the
two patches are entirely disjunct, did you update or something ?
BingoBoingo: <mircea_popescu>
textile engineering !? << You know... Applied HAIR
THEORY!
mod6: If you understand my meaning. I.e. I would need
to add an exception
to
the rule for
that.
mod6: So made routines
that do just
that.
There almost certainly is not a built in case where
two antecedents have
the same output hashes and can use a single in
that csae.
mircea_popescu: lemme reload all
this pile of stuff here, i closed it all last night.