POSIX who? https://bugs.debian.org/1016456
Replying to @nabijaczleweli
htnaks
Replying to @__phantomderp
floating-point is an adjective, floating point is a noun phrase, so 98% former in your domain; werent you like an anglo?
Replying to @__phantomderp
l + ratio + floating-point amount of bitches + you have a doctorate
Replying to @__phantomderp
that said, in modern floats i cannot imagine when you would actually have a floating point. the primary driving idea behind IEEE754 (and contrary to some heirloom floats) you always know where the point is floating
sourdough is a blockchain of sorts
Replying to @nabijaczleweli
okay im p sure coreutils who is just terminally broken when presented with IPv6 addresses
Replying to @nabijaczleweli
yeah, 100%
Replying to @nabijaczleweli
well, worse than being broken (cf. cutting ut_host off at the first colon), it uses the first byte of ut_addr_v6 as an IPv4 address and presents that as gospel. love that.
Replying to @XenonNSMB
i mean in this case this is fine, the XSI portions of POSIX who are horseshit and not fit for implementation, but the fact that they just straight-up copied this very obvious lie into the documentation is what tickles me the most
Replying to @nabijaczleweli
i love debbugs https://bugs.debian.org/1016492
Replying to @eatijr
ULTRAKILL (2020)
Replying to @eatijr
okay clown moment i misremembered because i used the shotgun exclusively with the punch-boost, theres no actual rocker launcher at present
Replying to @eatijr
:0 i played through in like february, this is a pog gamer moment /for/ sure
Replying to @nabijaczleweli
oh no! our white balance! it's broken! &c., &c.
Replying to @Hoenn_Otaku
..tree?
Replying to @nabijaczleweli
one got broken by wind. this shit a bean
definite бухарка and potential pokchoi moment
they call me post-nasal the way i be dripping
Replying to @nabijaczleweli
okay i searched, this is the sixth tweet that says this on the platform. -2 me
Replying to @nabijaczleweli
as a wise man once said: And i hope all your donuts come out looking like Fanny's.
Replying to @nabijaczleweli
hehe 666/69 https://bugs.debian.org/1016669
Replying to @nabijaczleweli
lmfao, spot the filtering bug (4.3BSD-Reno, same in 4.4)
Replying to @nabijaczleweli
kernel modesetting
if your car alarm goes off for longer than a minute (incl. leap second) you should be fined i think
Replying to @nabijaczleweli
IME no-one has ever cared about their car alarm going off, so i also dont see the point really
Replying to @nabijaczleweli
like 2 months since i ate commercial bread 🧐
Replying to @nabijaczleweli
AGAIN
Replying to @nabijaczleweli
wtf bublik inflation?????
Replying to @dazabani
piss room*.fextralife.com when
Replying to @dazabani
p.a. to http://darksouls.wiki.fextralife.com in this case, i forgot the wiki. part because im stupid; no relation to fetlife
Replying to @nabijaczleweli
poo print
Replying to @nabijaczleweli
new winner :0
they call them "chief /executive/ officers" because the equivalent position at disnep oversees staff executions
for a nation where proper nutritive health is oft taboo, americans are really making strides with GI Joe
Replying to @nabijaczleweli
[tapping my earpiece] producers are telling me that gastrointestinal joseph is a jigoist
got a new disk (36mo warranty!) to replace the dying one in my desktop (well, I think it's dying, but hearing the spindle resonate is probably bad); manufacturing date: June of 2013. note how the paper around the spindle high-spot is worn down
Replying to @nabijaczleweli
smell the freshness!
Replying to @nabijaczleweli
compare: the clapped-out one that reads 34732 Power_On_Hours (48mo continuous)
Replying to @nabijaczleweli
and they make fans with rat tails on them so i can stop using the shitty splitter i soldered????? 21st century is fast approaching!
Replying to @nabijaczleweli
162M/s on LBA0 to 74 on the innermost ring :0
Replying to @twinkmodin
Dead Cells/Gungeon?
holy shit have i found the fabled freshavocadoo??
Replying to @nabijaczleweli
lmao its so much worse. it shakes in every axis
Replying to @nabijaczleweli
this is a cp /dev/zero /dev/disk/by-id/ata-TOSHIBA_DT01ACA200_635G59PGS moment
Replying to @iximeow
Replying to @nabijaczleweli
thank god its compatible with Windows 10 *and* 8.1!!!!!!
Replying to @nabijaczleweli
its all zeros. what the fuck is the compat thing for then lol
Replying to @nabijaczleweli
:0 with an updated output format for time-elapsed
Replying to @nabijaczleweli
holy shit the write speed is so variable im seeing like 300/70/180 on consecutive seconds. your brain on nand
Replying to @nabijaczleweli
holy shit, this was folded in with the receipt (sic EN in alt)
Replying to @nabijaczleweli
(5-second samples, min 79.6, max 275)
Replying to @nabijaczleweli
im in general not super-convinced about the panaceumness of SSD but checksumming at 390±5MiB/s (first shelf at 237, tail one at 333) with a newly-splicing head -c (https://git.sr.ht/~nabijaczleweli/voreutils/commit/4d1b3a286e37f090e27032fcd6a1ace5ab957255) is p nice ngl
Replying to @nabijaczleweli
i had these splitters in here for like 2? 3? (so probably 4) years. insane to me anyone would make 2-pin fans. not pictured: a piece of tissue from the disk cubby folded and compressed in the same way as i folded an A4 sheet 2wk ago as antivibration, i think original to the build
Replying to @nabijaczleweli
fools – big or small, i was like 13? 14? – clearly rarely differ
Replying to @nabijaczleweli
as expected it got dry after the season, but artificial insemination was successful :0
Replying to @nabijaczleweli
getting the jpegs off my camera was Effort while I was off the oil rig, so: 2022.08.09; its surprising how a roll of dough Just rolls into a Classic Pretzel Shape if you twist it far enough
Replying to @nabijaczleweli
this post is approved by Wheat Gang
Replying to @nabijaczleweli
all бухарки flowered :0
Replying to @nabijaczleweli
:/
Replying to @__phantomderp
okay, stupid question, but you'd know probably: since when targeting windows wchar_t is sixteen bits......... how does windows represent characters larger than 2^16 like 1FB00 BLOCK SEXTANT-1?
Replying to @__phantomderp and @rpjohnst
my question could be more precisely formulated as: for Windows, in a Unicode locale, given that sizeof(wchar_t)==2, how does one represent code-points above U+FFFF? in this case I'm reading this as: "fuck you lmfao eat shit idiot"
Replying to @__phantomderp and @rpjohnst
the proto-unicode pre-print that called it a "16-bit encoding" and its consequences have been a disaster for the human race
RT @nabijaczleweli: @__phantomderp @rpjohnst the proto-unicode pre-print that called it a "16-bit encoding" and its consequences have been…
Replying to @nabijaczleweli
uClibc-ng 1.0.42 out with patch :0 https://cgit.uclibc-ng.org/cgi/cgit/uclibc-ng.git/commit/?h=v1.0.42&id=6d45c0fe89e6de7d35a991051ec7f9feaaaf3729
Replying to @nabijaczleweli
bugzilla moment https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106604
guy who thinks ambulation is driving around in a big truck
its actually wild how easy it is to fuck up your halogens so badly you're pouring your water intake directly into the sewer just slightly later and not notice and if you eat a handful of salt its basically fine
Replying to @nabijaczleweli
update: it was (and still is) the PSU fan
Replying to @nabijaczleweli
as validated by: the PSU vibrates to the touch
debashified all dracut modules where possible :/
Replying to @nabijaczleweli
wild how bash has all those insane extensions but this is net -121
Replying to @nabijaczleweli
gamer moment. 10 straight hours. https://github.com/dracutdevs/dracut/pull/1899
Replying to @nabijaczleweli
lol
the dracut thing has led me to the printf zone https://lore.kernel.org/dash/ba1440bf05e22d3cae38521a25cb3ca90b30680d.1660436383.git.nabijaczleweli@nabijaczleweli.xyz/T/#u
Replying to @nabijaczleweli
https://bugs.debian.org/1017110
why am i getting Žižek in my youtube recommendations. am i about to get balkanpilled
Replying to @nabijaczleweli
i am no more balkanpilled than prior but thats a relatively high bar as-is
Replying to @eatijr
if i didnt want people to call my computer i would simply not connect my computer to the telephone
Replying to @eatijr
all acl is downhill from here
Replying to @nabijaczleweli
and a sublime unicode bug(?) https://github.com/sublimehq/sublime_text/issues/5530
Replying to @nabijaczleweli
lmfao
Replying to @nabijaczleweli
cringe-ass nae-nae baby
Replying to @nabijaczleweli
i would turn it around so the intake was from the inside but the cables are like 20-40 too short. this is awful. too me.
Replying to @nabijaczleweli
NEVER put a moving part in a computer. especially one that takes more than 20 seconds to remove
Replying to @__phantomderp
which iconv :)
Replying to @nabijaczleweli
elisabeth holmes voice
introducing: the nürburgcockring
update on this: sendfile() is literal magic https://twitter.com/nabijaczleweli/status/1521167227478810624
get absolutely bright-yellowed
Replying to @nabijaczleweli
drip update
she cult on my lamb 'til I sermon
of course im anti-wax, just get to the point!
not every day that you find a grave POSIX conformance dash bug
Replying to @nabijaczleweli
https://bugs.debian.org/1017531
its only a sport if aomeone's corrupt, otherwise it's just sparkly running around
Replying to @nabijaczleweli
> I just checked my reference semantics shell, and it fails this test as well. Notably, [it] passes the POSIX test suite---which means that the POSIX folks are not properly testing for this behavior.
as worrying as it is unsurprising tbh. https://lore.kernel.org/dash/m2sfltskbi.fsf@greenberg.science/T/#mb0e7b96dedde115bd924c45a7701469b870a087b
Replying to @nabijaczleweli
ol' lizzy ol' pal
who would win: sigma male or sigma balls?
Replying to @AssembledGhost
this is just the part where the balls fused, which has different porosity so it bakes differently (well, hardly bakes at all) I think
Replying to @nabijaczleweli
lmfao, this account sent me a friend request yesterday(!) (no servers or friends in common), as @{Something}Harris, hence my kamala gag; now it's System Message and the same scam
how is this possible
they call me trans the way i fucking send it
Replying to @nabijaczleweli
what the FUCK are they cooking. this is /3.5x/ the instructions to do the same thing.
Replying to @__the_sylph__
coreutils (bullseye) vs https://git.sr.ht/~nabijaczleweli/voreutils, yeag
Replying to @NireBryce
ngl this is a certified GNU undocumented-by-design-fuck-you moment
Replying to @NireBryce
curious what you think of this (no ls(1) at present, but pick your similarly-awful-in-coreutils poison) in light of your FreeBSD enjoyment: https://srhtcdn.githack.com/~nabijaczleweli/voreutils/blob/man/man0/index.0.html
Replying to @NireBryce
the coreutils manuals are Like That because they're just preprocessed usage strings, and the Canonical manual is the info page. of course, no one has ever or will ever read the info page, so you get neither a manual nor a usage string
Replying to @NireBryce
(EXAMPLES being super fucked up on the left is a mandoc -Thtml bug – https://inbox.vuxu.org/mandoc-discuss/20220606153911.w4bvcz6zqsjde7so@tarta.nabijaczleweli.xyz/T/#t, cf. (1) for the correct rendering; but besides that i think it stacks up relatively okay?)
Replying to @linuxtardis
coreutils uniq doesn't handle locales at all (save for messages, but none were issued; cf. http://bugs.debian.org/1017482, http://bugs.debian.org/1017643), but this mode is entirely byte-oriented anyway
hm. glibc bug? (2) is a diff of Issue 7 to issue 7 TC2, cf. (3) Issue 8 Draft 2.1
Replying to @nabijaczleweli
well, okay, it's Wrong: that's obvious. but whether it's a glibc bug depends on if it targets Issue 7 (2008) or latest Issue 7 (TC2, 2017); it'll definitely be a bug once 202x drops
Replying to @nabijaczleweli
debbugs, my saviour! "C locale is 7-bit (127 characters), must be 8-bit (256 characters) since POSIX Issue 7 TC2/Issue 8": https://bugs.debian.org/1017852
Replying to @nabijaczleweli
i did make like 2 poops by accident but
Replying to @nabijaczleweli
i hate the begging for an account bull shit
Replying to @nabijaczleweli
:v https://sourceware.org/bugzilla/show_bug.cgi?id=29511
Replying to @EmilTheBonobo
oh yeah thats quite pogged, even if it suffers from unsuppressable colour vomit, thanks
she welt on my schmerz till i zeitgeist
Replying to @nabijaczleweli
FreeBSD moment https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266001
Replying to @nabijaczleweli
big yellow chopping??
they say i ligma, the way im fuckin ballin
Replying to @__phantomderp
does your preferred document scripting language not have automatic footnotes lol
Replying to @__phantomderp
wait weren't you using tex or whatever. how is it possible after 30+ years
Replying to @__phantomderp
is this too difficult for knuth's brain child
to think this is what people had to do before comm. V3 was a dark time
Replying to @nabijaczleweli
a − b = dedup(duplicates(dedup(a) ∪ unique(dedup(a) ∪ dedup(b))))?
Replying to @nabijaczleweli
assuming no duplicates, since V4 you could spell this as sort a a1; sort b b1; comm -23 a1 b1, otherwise inject sort a a1; sort b b1; uniq a1 a2; uniq b1 b2; comm -23 a2 b2
Replying to @hipsterelectron
pbuh
Replying to @hipsterelectron
the most fundamental operation here is analysis (dismantling) of lhs rhs into lines to analyse (inspect, but i guess also dismantle into characters v/v the collation sequence) their relationship so idk i think it makes sense
Replying to @hipsterelectron
but the five-word Dd and the first two verbs of the DESCRIPTION are always hard. too me. like, what does uniq.. do? "merge or filter adjacent identical lines" is both way too long /and/ overly specific because that's not what happens in -G, and not necessarily the case with slici
Replying to @hipsterelectron
pat^Wopinions welcome, &c., &c.
Replying to @pixelbeat_
hm. these two don't need join and can be better served with comm:
Intersection of sorted is comm -12 f1 f2
Difference of sorted is comm -13 f1 f2
the other two would need a sed to trim off the separator (or to have comm --o-d='' be empty, not NUL) as plain comm and comm -3
big score at the schweinefleisch factory today (jpeg of raw turkey neck in plastic bag elided for obvious reasons)
Replying to @nabijaczleweli
cope | seethe
Replying to @nabijaczleweli
we b breaded fr fr
Replying to @nabijaczleweli
what da hell
Replying to @nabijaczleweli
:0
Replying to @nabijaczleweli
top 10 scariest bright yellow moments
Replying to @__phantomderp
mfer really out here shaving off each −-marked sexion by hand
Replying to @StellaFoxxie
i actually tried, and it's never been buildable :) https://github.com/adobe/brackets-shell/wiki/Building-Brackets-Shell says to run "grunt setup or simply make"; the linux-flavoured build script does (1); there has /never/ been /any/ makefile in the repository. adobe distributions were likely heavily patched anyway, but
Replying to @StellaFoxxie
as a bonus, it starts by trying to download a 2016 binary CEF which 404s, and doesn't build against neither the nearest version available from upstream or latest. i continue to think that it was just never actually buildable
Replying to @nabijaczleweli
pog
Replying to @xsphi
this is infinitely funnier given that the observation that "u32 as an identifier is too small" is an observation that followed the introduction of gethostid(2) in 4.1cBSD by less than six months in 4.2BSD in /1983/
Replying to @xsphi
im older than the that note was when i was born by a good few years. how is this happening again
Replying to @xsphi
(this isnt even "dude just trust me": cf. https://www.tuhs.org/cgi-bin/utree.pl?file=4.1cBSD/usr/man/man2/gethostid.2 vs https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/man/man2/gethostid.2; ive unfortunately written about this at length at https://srhtcdn.githack.com/~nabijaczleweli/voreutils/blob/man/man1/hostid.1.html#HISTORY)
Replying to @nabijaczleweli
{gag here}print
Replying to @greenTetra_
mf on his lbj king shit. truly transcendent
guy who's anti-MSG but only because they stopped making it from crude
Replying to @nabijaczleweli
huh how https://builds.sr.ht/~nabijaczleweli/job/832746#task-build-gcc-304
Replying to @nabijaczleweli
what the hell is mandoc cooking that 2>&1ing fucks up the output so bad but stdbuf -oL -eL fixes it
OpenBSD moment lol
Replying to @dazabani
you're an OpenBSD enthusiast, right? why does /tmp 1777 0:0 cause subdirectories to be created owner:0 instead of owner:owner? I don't see any documentation for this anywhere across chmod([12]), sticky(8) :v
Replying to @nabijaczleweli
nevermind im fucking stupid thats just BSD group behaviour not an OpenBSD thing. the only place where this is mentioned is mkdir(2) apparently
Replying to @nabijaczleweli
https://www.mail-archive.com/tech@openbsd.org/msg71180.html
Replying to @atomicthumbs
zstd?
just fell victim to american cultural hegemony (didnt check the recipe sizes until too late when i poured in 1080 kcal of oil)
Replying to @nabijaczleweli
also 800 of sugar, 250 of sour cream, 1600 of chocolate, 950 of flour, 2 eggs and 40g of cocoa
Replying to @nabijaczleweli
4680 kcal + eggs + cocoa. thats literally three days' worth of food. how is this the default recipe scaling
Replying to @nabijaczleweli
behold: 3500 kcal
Replying to @nabijaczleweli
im gonna be like that australian what ate just beans for a month except its these and for three days
Replying to @autotheaphilia
okay but thats already a good chunk of the way to some solid drip
Replying to @mycoliza, @Wxcafe and @boring_cactus
uhoh does this make me a Damn Weapons Manufacturer or just a Fucking Collaborator
Replying to @nabijaczleweli
kms https://sourceware.org/bugzilla/show_bug.cgi?id=29511#c1
Replying to @eatijr
f for eater. americans are just not ready to have their cum carbonated
Replying to @nabijaczleweli
i replied, i was very nice, i proposed a shockingly sensible solution, i expect to get a "fuck you" again
Replying to @nabijaczleweli
pog pog he uninvalided it
Replying to @nabijaczleweli
lol musl does exactly what i suggested here and uses the first PUA (although starting at DF80 instead of E080)
Replying to @nabijaczleweli
nevermind i cant fucking read, that's surrogates, not a PUA, but that's even better, since
> Isolated surrogate code points have no interpretation; consequently, no character code charts or names lists are provided forthis range.
Replying to @eatijr
the against-all-odds answer to "what's a musl distro that isnt a pain in the pisspipe"
Replying to @nabijaczleweli
my proverbial soul is eternally tainted now https://sourceware.org/pipermail/libc-alpha/2022-August/141785.html
Replying to @nabijaczleweli
poglik
just two classic C parsing moments today, one of which @pixelbeat_ has already fixed upstream: http://bugs.debian.org/1018790, http://bugs.debian.org/1018803 :)
Replying to @nabijaczleweli
okay this is a psycho bug though https://bugs.debian.org/1018806
Replying to @nabijaczleweli
i am a golden god
Replying to @GNUxeava
constexpr is part of the function declaration, not the type; in the right conditions you will be allowed to run this function as part of a constant expression context like static_assert
Replying to @eatijr
centrifungal separation
Replying to @nabijaczleweli
maggots in your stty? more likely than you'd think! https://bugs.debian.org/1018844