29600+ entries in 0.017s
bvt: now, sse could
theoretically help, but
there is a question of whether sse operations are constant
time (in each generation of intel cpus)
bvt: asciilifeform: perhaps i could get a bit better performance after scrutinizing multiplication code a bit; however i don't
think it'll get much faster with current code structure
a111: Logged on 2018-04-11 15:29 asciilifeform: speaking of which, apparently AS OF
TODAY gentoo portage latest ver DEMANDS gpg2.
a111: Logged on 2019-06-23 21:34 asciilifeform:
http://btcbase.org/log/2019-06-23#1919680 << fwiw asciilifeform has not used 'portage' in year+ --
the heathen portage finally 100% broke
then. ( 'updated' on
the gentooist end, so
that no longer agrees
to build ~anything~ without 'new profile', and won't install 'new profile' because hard-contravenes asciilifeform's poetteringisms ban list... )
a111: Logged on 2019-06-23 18:00 mp_en_viaje: right. so portage might be
the pipe
that got you your galon of linux. but
that is what you wanted -- not
the faucet, but
the galon.
a111: Logged on 2019-06-23 09:41 mircea_popescu:
there's no rule you must publish
the signed
thing. you can publish 11 drafts before signing, what of it.
a111: Logged on 2019-06-23 06:44 mircea_popescu: here's
the
thing with
this reading business : most
things are read on
the second pass.
a111: Logged on 2019-06-22 22:34 mircea_popescu: whereas implicit v-ism would be "all code is made one pile,
then built once and linked, and
that's it".
trinque: we agree. I contend
that folks are going
to have
to build e.g. v, gnat, in order
to move forward, which is why I'm asking for ebuilds
mp_en_viaje: whole fucking interest for us here is
to do ~something~ (not entirely known what exactly yet) so
that we... don't fucking need portage ever again.
mp_en_viaje: right. so portage might be
the pipe
that got you your galon of linux. but
that is what you wanted -- not
the faucet, but
the galon.
☟︎ a111: Logged on 2019-06-22 21:34
trinque: my intent here has been
to stablize
the workbench and
then make a *hard* break with
the whole stack
mp_en_viaje: what it does is, give some measure of anti-friability
to
the everchange.
mp_en_viaje: the point was
to get it in a ~state~. something
to start actual decrufting work from. portage itself is mostly a raft for navigating
the phenomenology of fosstardation
trinque: that's not anyone's problem but mine, but pointing at it illustrates what I'd like
to avoid repeating
trinque: the cumulative work on gentoo was wasted cycles if portage is undesired, as most of
the work was around capturing portage in a given state
a111: Logged on 2019-05-17 23:17
trinque: republic is scant of profit centers
mp_en_viaje: obviously, on some level,
there's only so many hours in
the day, whether one works or doesn't, still
there's a finite
time.
mp_en_viaje: yes but look, is
the idea here
that for
the past whatever, 4-5 months you have been ~so busy~
that not only you had no
time
to say
things about your work on gentoo, but also somehow ended up isolated into yourself among sadnesses about how nobody reads and such ?
trinque: broadly, because I haven't yet extracted enough of
the inflationary fruits of
the former into coin.
a111: Logged on 2019-06-23 06:57 mircea_popescu: what worries me is
this approach whereby a)
things
tend
to be discussed in response
to some kind of prodding, and b) while
that feeling lasts, only
to c) fall back into an eerie silence within a few hours or maybe days, for WEEKS AND WEEKS AND MONTHS AND YEARS AT A
TIME!!!
mp_en_viaje: anyway, seems in
the ripe silence of lack of activty all sorta
things grew disjointed an' out of sync.
mp_en_viaje: diana_coman, but basically, it's running
the same
thing from 2018 ; and
the reason it doesnt move is
that a) we couldn't give less of a shit about ebuilds while b) we can't easily change
the pile of dependencies
to live with musl and so yet haven't.
diana_coman: so far other parts
took priority (smg comms and client) but at least my understanding was
that
there is work being done further on cuntoo, I had no idea
that it was stuck waiting on more intensive use or something
a111: Logged on 2019-03-09 22:45
trinque:
the remainder of work here is resolving
this issue (I have not had) with paths, after which we can start producing ebuilds for novel republican work atop
the genesis.
a111: Logged on 2018-11-27 18:11 diana_coman:
trinque,
that looks great; if I understand correctly,
the archive
there contains
the means
to a. install cuntoo b. make
the genesis of cuntoo/portage
that could in principle be used
to move an existing gentoo
to cuntoo; is
this correct?
a111: Logged on 2019-06-23 08:49 mircea_popescu: diana_coman, more practically, are we running cuntoo latest or what is
the situation
there ?
spyked: this makes perfect sense, I'ma do
that for my next code-report.
mircea_popescu: this "here i sit, my eyes intelligent, my expression absent" situation's for
the birds.
mircea_popescu: or alternatively,
there's all sorta greatness, phf has a perfectly workable code metadiscussion system on btcbase, jurov had a ver yworkable one for earlier
trb work and so fucking on.
mircea_popescu: because it occurs
to me
that ~if~ you publish your codebase you're working on, and if your blog works correctly in
the mp-wp sense,
then
therefore you get all
this for free, just make a metapost.
mircea_popescu: and so some other guy can go "that nfi part is just a discussion of so and so, like
this" and so on ?
mircea_popescu: can you do
this in some manner ? so people can follow your work if
they can be arsed
to import
the same ide / cms ? (here, it's a bit of js, and a browser,
that entire stack)
spyked: line-per-line annotation sounds like a neat idea, but I don't have anything like
that going
spyked: for
the moment all my notes are in a
text file
that's going
to grow into a blogpost as soon as it has a head and a
tail
mircea_popescu: how do you work, do you have
this codepile loaded up in some sort of management interface ? does it permit adnotation ?
spyked: but more specifically:
the code I've stolen works so far, I have it serving an instance of
thetarpit plus a somewhat-working experiment doing dynamic stuff PHP-style. but
there's a lotta complexity in
the work distribution code
that has me going "wtf". perhaps one good
task would be
to implement a worker-pool
thing, similarly
to what apache has.
mircea_popescu: there's no rule you must publish
the signed
thing. you can publish 11 drafts before signing, what of it.
☟︎ spyked: hitecture" is actually next blogpost in
the works.
a111: Logged on 2019-06-22 13:57 mircea_popescu:
the sort of "before publishing, I'm making sure" idiocy is exactly, but i do mean exactly, of
the same substance as
the fat girl's "this donut has no calories". it uses
the exact same parts of
the brain in
the exact same way.
spyked:
http://btcbase.org/log/2019-06-22#1919220 <-- well, I can't in all good faith sign and publish
the item before I understand it well enough
to maintain it myself -- for
the same money, I could just claim "use
the version on shithub" and be done with
that. I agree 100%
that I should do small regular updates, but in
this case breaking
the
thing into its constituent parts is
the challenge, and "hunchentoot arc
☝︎ mircea_popescu: diana_coman, more practically, are we running cuntoo latest or what is
the situation
there ?
☟︎ mircea_popescu: responsibility is a fine and great
thing indeed ; but it's not mechanizable, as part and parcel of why we have a wot as it is rather
than as
the lulzentreprise everyone else imagines it
to be.
mircea_popescu: there's
this lengthy list of examples of people coming
to grief
through
the process of "that mp so dumb, dun even know how
to horseride a shirt" or w/e such.
mircea_popescu: this pov is even voiced, but always in a chiding
tone in
training environments. it has its place
there, but i dunno how practicable it is in
the general.
a111: Logged on 2019-01-04 17:18 mircea_popescu: why
the fuck didn't i
think of
that.
mircea_popescu: diana_coman, i suppose in a very
theoretical, idealized sense. in practice however... what are we
to conclude on
the basis of
http://btcbase.org/log/2019-01-04#1884426 ?
that neither i (nor anyone else here for
that matter) had diffraction loaded in head ?
the wave model of energy ? what exactly ?
☝︎ a111: Logged on 2019-03-10 21:39
trinque: scripts/make_portage_tree.sh << line 14, I do string-munging on
the path
that's specific
to my own filesystem layout
diana_coman: I rather saw it as "can't do
this alone, need help but won't ask because it should come naturally"
diana_coman: uhm,
there seems
to be something out of sync in
there; perhaps you are right and
the problem is one of wrongly-attached self-worth
though I did not see it
that way
mircea_popescu: diana_coman,
this is like slavegirl saying "i don;'t understand why not ask detailedly for
tongue movements and assorted shenanigans". she gets punished if jher bj is subpar because i don't have
the fucking patience
to micromanage blowjobs. it'd be like writing
the book im about
to read, if i wrote it what sorta discovery is
this!
diana_coman: oh, I don't know if it's re self-worth; rather interpreting
the "not fully load-in-head" as "not read"
diana_coman: what I don't understand is why not ask specifically if one would like detailed feedback on some bit/part
they struggle with
mircea_popescu: in any case
the "i am of
the republic, i will now base my self worth on
the qty and amt of loading-in-head my works get" is, as of yet, a recipe for eternal sadness.
this is a new rather
than old republic.
mircea_popescu: but... what can you do. it's a crapshoot, you know
this going in.
mircea_popescu: i mean, i would also like people
to do more loading-into-head of whatever
their native language is, so i could get better movies, music lyrics and music videos, books, blog articles and everything else.
diana_coman: re reading of stuff I guess
there is also
the difference between one reading and full load-in-the-head; it struck me
that people would want others
to do more load-in-the-head of what
they produce so as
to get finer-detail feedback
mircea_popescu: and
this paroxistic-avoidant approach worries me universally, phf is stuck in
the exact same rut, for instance. i have never in my life seen
this approach
to
things work. not once, not FUCKING ONCE, and
this is all
the more indicative as it's one of
the most commonly deployed broken strategies.
☟︎ mircea_popescu: rather
than in
the normal, and correct manner in which
things are discussed, as
they are noticed,
to get
to
their bottom, which is
then gotten
to and marked as such and
things move on, with
the bottom in question an ever-reliable reference always at
the ready for having been correctly found, and always visible just under
the skin of
things, for having been correctly found.
mircea_popescu: what worries me is
this approach whereby a)
things
tend
to be discussed in response
to some kind of prodding, and b) while
that feeling lasts, only
to c) fall back into an eerie silence within a few hours or maybe days, for WEEKS AND WEEKS AND MONTHS AND YEARS AT A
TIME!!!
☟︎ a111: Logged on 2019-06-22 18:07
trinque: I noticed
that not a soul read
the scripts
that bootstrap cuntoo
mircea_popescu: whereas usg.blue reads not-at-all, never ever no matter what ; and usg.red reads always-aforehand. both of
these idiocies, equally idiotic as
they find
themselves, are
therefore always found in some uneasy
truce. but you can't properly notate "read on 2nd pass" as "didn't read" or "won't read".
mircea_popescu: because
the republic's a practice, not a "narrative" aka daydream. fundamentally, what distinguishes us from
the github hipster / foss moron / entire collected pile of avortons looking for
their abortion from "the bay area" & portland
to
eudemocracia and beyond is
that we don't read on
the first pass, BUT ON
THE SEC
a111: Logged on 2019-06-22 21:28
trinque: or is
this "in six years
time we'll have systems
that work"
mircea_popescu: this is exactly cosubstantial with
the
http://btcbase.org/log/2019-06-22#1919401 inquiry : "reading everything aforehand" as a strategy is just as dysfunctional as any other idealism aka premature optimization.
there's a fucking reason republican doctrine is "do it by hand first, automate later, and
those parts
that actually need it and benefit from it, rather
than randomly and abstractly like
the pantsuit do".
☝︎ mircea_popescu: as i sit here now,
there's
things around me i've never read. serial numbers on smoke detectors, handling and caring labels on rugs, brand names on plaster and wall insulation facing inside,
towards
the wall. i am not about
to
tear
the place apart
to read all possible strings around me. i might, ~if need be~, of course. but not before.
mircea_popescu: here's
the
thing with
this reading business : most
things are read on
the second pass.
☟︎