has had for a while, the openzfs issue is as hilarious as it is ten years old: https://github.com/openzfs/zfs/issues/405 (though, reading it back now, it seems that some madlad actually wrote it)
maybe im like too young or whatever, but i don't get the pokemon thing. why does everyone above the age of 27 seemingly have 300 sets of id/jpeg/6 moves/lore memorised. how, rather, maybe
one of my favourite openbsd features is how the documentation is false for no good reason
Replying to @nabijaczleweli
OpenBSD script(1) says
> script will exit with the status of 0 unless any of its child processes fail, in which case, script will return 1.
And, guess what, shock of shocks: no it doesn't. Unless setup fails (it won't) `script -c false` exits 0.
Replying to @nabijaczleweli
NetBSD script(1) is not plagued by this because it doesn't mention exit codes at all, because it just exits 0 always. OpenBSD attempts to shoehorn exit code handling into the 4.4BSD-Lite impl (which is kinda fucked in its own right), just straight-up fails, and lies it doesn't.
Replying to @nabijaczleweli
blergh, now i remember why i don't like writing in deep 4.4 derivatives
Replying to @nabijaczleweli
lmao. if you have a mailing list, please oh please make it so it won't be so :v (DMARC failures tagged in (1))
Replying to @nabijaczleweli
it's normal to get a few aggregate reports when pooosting (indeed, that's what they're for), but by god what the fuck is this. I've never seen this happen before. this really is the power of OpenBSD!!
Replying to @nabijaczleweli
hm, @dazabani is OpenBSD really always like this or do I keep holding it wrong because I'm not bought in?
Replying to @nabijaczleweli
the official archives link points to marc.info, which is very cool and not at all a piece of shit that explicitly sets charset=CP1252, doesn't transcode mails to it, /fails to decode/ GB2312 RFC2036-encoded headers, but does decode GB2312 bodies, and illegally breaks them at 80
Replying to @nabijaczleweli
fun fact: decoding RFC2036 and reinterpreting those bytes under the default encoding du jour (inv. CP1525) is also what outlook does. unlike the above, it requires a business relationship with and commercial licence from microsoft. who said you can't get enterprise quality free?
Replying to @nabijaczleweli
this is also where i post the patchset: https://www.mail-archive.com/misc@openbsd.org/msg180638.html
Replying to @nabijaczleweli
look! look! it's possible!!!!! you can decode mails that arent in 7-bit ascii!! the chinese one works too: https://www.mail-archive.com/misc@openbsd.org/msg180627.html!!!! it's truly a new era (1997)
Replying to @nabijaczleweli
at 39 DMARC reports just off these 4 posts now lol
if i were a tiny bit more unhinged id post shit like "bitches out here really pretending they're sulfite-free. please. #containssulfites" but this is a fake moral panic i don't support (unlike the good moral panics that exist, are my friends, and are good, such as birds)
Replying to @nabijaczleweli
(turns out its non-obvious to formulate a gag that works as a "birds are actually real and theyre our friends" moral panic in the context of a blatantly bird-free world when birds do actually exist, the audience for that is narrow, and without belabouring the point beyond belief)
Replying to @cketti
I did; you see how that's somehow even worse, though, right?
Replying to @dazabani and @FakeUnicode
fuckin A. next-level shit.
hyperbolic, nevertheless cold take: the more letters around the "sh" in a shell executable basename, the worse it is
Replying to @__eater__
topologically, on average, slightly more than 0, right?
Replying to @__eater__
of course, traditional wisdom says "between 1 and 1+the amount of the user's prince alberts", but then is it really a hole if it serves the business need of hooking your knob on your pant/does that hold in today's un-hooked pants era? the HMRC needs to opine on this imo
just walked like 3.5h consecutive, its hotly debated whether I still "have hips"
Replying to @nabijaczleweli
anyway, с новым годом or whatever
Replying to @nabijaczleweli
good boots will change your life, dude; the hip wear is as expected but my knees are in pristine state. schmoovement tech as comparatively incredible as it is unsurprising
Replying to @fucking_dildos
a guy with a fat ass crawls around (hence the Snake thing?) in military encampments taking out high-value targets while having a telenovella happen to him over shortwave radio
Replying to @fucking_dildos
like, i know on a logical level that there's "plot" and "deep deliberation on the concept of war" but AFAICT it's just a boss rush with funny sound effects
apparently, and this is very cool, docs@netbsd.org accepts your mail even if you're not subscribed; tech-userlevel@ doesn't, and, get this: straight-up devnulls it. Why? Who knows
Replying to @nabijaczleweli
I also don't know if I never got a reply for my docs@ mails because or if it archives but doesn't forward to the list. Love too majordomo (also, neither of them say they're members-only).
Replying to @nabijaczleweli
also also: tech-userlevel@ accepted 1-6/6 of my patchset but ate my cover letter twice. LOVE TOO MAJORDOMO
Replying to @nabijaczleweli
also, majordomo <2 (such as that used for the netbsd lists) doesn't let you configure mail delivery. why. i NEVER want to subscribe to a list im posting to.
Replying to @atax1a
> ezmlm is mailing list management software (MLM) by Daniel J. Bernstein. It is similar to GNU Mailman and Majordomo but only works with the qmail mail
ah, yes, djb, a mediocre gnu project, a shit-tier perl program, and fucking qmail of all things in one sentence. good god.
Replying to @hipsterelectron
:0 https://lfs.nabijaczleweli.xyz/0011-OneMix3-Win8-tablet-screen-recording/demo-output
Replying to @nabijaczleweli
ooohh budDYYYYY
i spent the best part of an hour re-structuring netbsd script(1) to use poll(2) like I saw the util-linux one use in the strace, but gave up after i had `script -c 'for i in 0-10; do echo a; sleep 0.1; done; read -r a; echo a$a' < bigfile` hang on the stdin->pty write(2)
Replying to @nabijaczleweli
(seemingly because even if POLLOUT is set, it'll block until it writes the whole buffer)
Replying to @nabijaczleweli
((do you fore-see what the kicker is))
Replying to @nabijaczleweli
yes, it's also fucking broken:
Replying to @nabijaczleweli
this is hilarious to me because /the 4.4BSD-Lite implementation is impervious to this/ and BOTH of them derive from it
Replying to @nabijaczleweli
"util-linux regression against 4.4BSD-Lite (1995)" is definitely a fucking Time https://bugs.debian.org/1003095
Replying to @sehetw
4.4BSD-Lite doesn't have this bug (i.e. doesnt deadlock, because there's two processes). all extant implementations (incl. util-linux, Net/Open/FreeBSD) have it as their common ancestor
Replying to @the6p4c
i just let'er rip tbh i dont think ive said a single cohesive sentence out loud in years
Replying to @nabijaczleweli
after off-list discussion I adopted someone else's SHA1 removal for another -3543 lines and total -37.2% lol
Replying to @the6p4c
she was wg14, she was glibc, need i make it any more obvious?
Replying to @nabijaczleweli
FreeBSD also uses a single-process model (with select(2)). it's also broken. very cool. love that.
Replying to @nabijaczleweli
my second FreeBSD bug in as many half-years, let's go for the same amount of comments!!!!! https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=260938
who'd've thought two-day-old remnants of a rotisserie chicken carcass would work so well as a seven-minute noodle
Replying to @nabijaczleweli
(well, I'd've thought, but had I spared another I'd've'd noticed the "No MSG" on the bag and tripled the salt. alas.)
truly an indictment of the rustup ecosystem that i managed to build dis.@yaxpeax.net for its native wasm target and could usefully apply the build errors to bump yx-superh up to 1.0.0 without getting void-called once
Replying to @nabijaczleweli
i cargo built and it just worked. took a bit longer than usual i guess but who cares. what's the damn point. i thought this was going to be "building for wasm just to feel something" like that one time i did c++ (those were good three hours, nothing got done) but i feel nothing
Replying to @the6p4c
hm.
what if we were both patterns 👉👈 and we autosuggested 🥺
good bug when third OP comment starts "Slept at 07:47 at home.", specs hotplug down to which holes two sticks entered, links lenovo model pages for EVERYthing, "Before leaving work yesterday afternoon, at 16:13", "At home this morning (early hours): … woke the computer at 02:16"
Replying to @nabijaczleweli
I mean, I'd be equally paranoid if my pools reported 16EiB free: https://github.com/openzfs/zfs/issues/12779, to be fair (also, that was AGONY to golf down to 280)
top: V5, bottom: V4. VERY cool and normal for a program that's supposed to insure your data lol
absolutely fucking kicking mesel, just spent the past god-damned two hours looking through my firefox history/linux-pre2.6/kernel.org-pub to find what fucking bdflush was called again after having looked at- and having been mesmerised by it like half a year ago
Replying to @nabijaczleweli
FINALLY. bdflushd-1.6.2 -> assembly -> lighter version of the cache flushing thing for laptops
Replying to @nabijaczleweli
yes, those are keywords, because yes i will forget again, and I'll, again, search through my tweets to find it, but this time I won't get all the way back to the 2015 cringe
Replying to @nabijaczleweli
all of this just to reference the updated/bdflushd in my sync(1) when noting 4.4BSD update(8) (which, now that I've found it: lol, the nomenclature here and the 30-second timer is 0% chance an accident)
Replying to @nabijaczleweli
> There's a bit confusion between two names "bdflush" and "update". Currently, executable is called update, but it is compiled from bdflush.c and authors refer to it as to bdflush.
Replying to @nabijaczleweli
return... 🥺
gcc __attribute__((section( doesn't let you specify flags and modes like gas .section does. society.
installed screen because i wanted to run a vm overnight. its telling me "screen caught signal 11. (core dumped)". i have ulimit -c 0. strace says its getting a SIGABRT. /run/screen is filling up. what the fuck.
Replying to @nabijaczleweli
GNU moment! https://bugs.debian.org/1003257
Replying to @atax1a
as I hope you understand, i've carefully avoided running any analysis because i scrolled through the first three pages of screen.c#main after getting "core dumped" in an undump{able,ed} scenario and decided id actually sooner die; sorry
western-european languages, or "latin distributions",
guy who speedslides on the freeway on his way to work
yeah haha im, like, so random lol!! *induces a catastrophic pool failure*
[exorcises yuor consistutional rights] whered u go ⌀щ⌀
Replying to @the6p4c
this plagued me for three solid hours as i was offline last night. this is less than nothing. its a poast vacuum
Replying to @astraleureka
out of those two: variable, obviously, but this is a false dichotomy imo. where my centre pointers at
Replying to @the6p4c and @astraleureka
caddr_t gang
a domain fit for a king (or, indeed, a big yellow)
Replying to @nabijaczleweli
bro what the fuck is the matter with the phone camera, honestly; also: pixma balls(?)
Replying to @Foone
cowgirls of the world, unite!! :/
we have peaked in video gameing when the yakuza 2 guy bonks into a door and he recoils a bit
Replying to @the6p4c
DEVELOPER 1 takes out a blowtorch, a small anvil, hammer, and TIG filler rods, and reforges his fork to have a bronze inlay, which doesn't appear to impact performance. This process takes approximately twenty-five minutes. He refuses to order before he's done.
Replying to @the6p4c
okay this is probably less funny if you havent seen the Super People crafting animation
hate it when i, as an author, am LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION. really binds my balls.
Replying to @nabijaczleweli
(also, didnt it get released on a different version? do javascripters not check their dependency updates?). thats my fill of Current Events for the month.
Replying to @nabijaczleweli
and its upstream! https://cgit.freebsd.org/src/commit/?id=c0ba4c2ee2c48ec9892d10c0aca797f3685c53ee
now, about that uname(1) being non-conformant with a two-line fix i outlined in the message thing,
Replying to @the6p4c
sus pdms
Replying to @the6p4c
sory sorry im tyring to deletei t
unbelievable!
never sleeping The Recommended Amount again; i had dreams, this shit sucks
Replying to @nabijaczleweli
such powerful click-bait and no-one even dared engage :/ truly an indictment of the tweeting class
Replying to @the6p4c
this is just greylisting
Replying to @the6p4c
okay, no, that depends on the quality on the originating MX; i have gotten registration mails with 15-minute-expiration links after two fucking hours before, so
Replying to @the6p4c
(bonus points if they use a different mx for re-sending so you need to snipe one of the 7 mx?.site.jpeg endpoints; i cant imagine that site sent+received more than a 200 mails a day; is this what SaaSing your mails ends up as?)
Replying to @nabijaczleweli
((saw the headline again (mistake). my mans didnt sabotage shit. he posted code in the usual place under the usual conditions with his legitimate credentials. shut the fuck up.))
yeah im btfo (bummed the fuck out)
Replying to @nabijaczleweli
googled what the canonical expansion of BTFO is after posting. this is middling.
Replying to @Ar1aAU
yeah i dont think ive ever seen it used in earnest, and even then i havent managed to discern any actual consistent meaning
Replying to @nabijaczleweli
when yuo see a poast.
Replying to @the6p4c
indeed, I just cropped out (1), which i picked for doing a dumb face; it seems that in my quest to make "shitty fake clickbait" i made "actual clickbait but just not very good"
Replying to @__eater__
just пии imo, no need to put too much sauce on it
liking ops is my biggest character flaw (← clueless)
Replying to @nabijaczleweli
hm. as i wrote this, to obtain the left arrow, i did `echo '\(<-' | nroff | clip` instead of googling it
Replying to @nabijaczleweli
in a brilliant move, i saved refuse from once-extracted chicken, ovenned it, and re-extracted, alongside cleaning out the older stuff from the fridge. this did /not/ take seven minutes, but did look like stomach contents prior to removing the tortured chicken remains
Replying to @nabijaczleweli
nevertheless i must've done /some/thing right because it doesn't need extra salt
Replying to @nabijaczleweli
grey remains unphased. im starting to think cats are anti-agriculture :/
Replying to @nabijaczleweli
lets fucking goooooooooooooooooooo https://github.com/openzfs/zfs/pull/12899#event-5897814864
Replying to @__eater__
sexual gone cinnamon challenge??!??!?
why the FUCK is mail(1) exploding with
mail: symbol lookup error: mail: undefined symbol: mu_stream_timed_getdelim
in 1:3.14-1 now. how does this happen. i can't imagine what the fucking reasoning is going to be (defaulting to "GNUwno" for now)
Replying to @nabijaczleweli
with each passing day im That much closer to switching over to s-nail even before v15.0
Replying to @nabijaczleweli
reporting grave on mailutils at 5am like a maniac https://bugs.debian.org/1003755
Replying to @nabijaczleweli
doesnt beat the casual slow-burn one-severity-per-mail bump from important all the way to critical in https://bugs.debian.org/998867 though
Replying to @nabijaczleweli
currently engaging in very delicate and highly erotic severity dance
this bread brought to you by the little dudes who be fartin
Replying to @nabijaczleweli
brought him a new box because the previous one expired and he likes it :)
Replying to @dazabani
the pre-2.6 git repository's a good read (mostly the release messages w/commentary), the shit that used to go in and out and then back in again is, comparatively, mind-blowing. the timeline's absolutely nuts half the time too, like 3 releases in 16h (cf. https://lwn.net/Articles/861431/)
Replying to @dazabani
did you install on btrfs? for some reason (maybe because i grew up with xp) the concept of xp-on-btrfs is the funniest fucking thing and iirc you can Just do that and its appx. as stable as on competing solutions
Replying to @the6p4c
in casual parlance this is known as "patches welcome :)" i think
Replying to @the6p4c
you should generally avoid posting those to a list, however; very easy mistake to make
we need to make doing dumbass shit like SCSU make sense again imo
unicode enthusiasts getting married, call that hand unification
Replying to @nabijaczleweli
:0 https://github.com/systemd/systemd/pull/21805#event-591154066
Replying to @__femb0t
narrow cleft for beer def. sus imo
Replying to @__femb0t
either shut it entirely and just compare still to barrel or somehow make it much wider but still avoid the scandies ig? bad chart
ive found the origin of sscjm and i feel a months-long void has been unvoided
Replying to @nabijaczleweli
Super Sucker Cum Jetting Machine.
Replying to @the6p4c
"∈ ∅" imo
Replying to @tyronerodriguez
i can't even derive what the Polish one is supposed to be. it should be like Czech – "w dupie mam"/"mam to w dupie" ("I've it in my ass"); for a more severe version (as in you actively don't care) you'd go with "mam wyjebane" which you'd be extremely hard-pressed to transliterate
"a + b + ratio" is the new "don't care, didn't ask, plus you're bald"; regress :/
Replying to @ErisMeatPuppet
don't care, didn't ask, plus you're ba^Wgay
Replying to @nabijaczleweli
https://github.com/openzfs/zfs/pull/12996
Replying to @nabijaczleweli
oh sweet jesus piss, nerd alert!!
Replying to @nabijaczleweli
but then i produced two separate 140- and 171-file PRs (+1718 -2863 sum!) today, so who's the loser here (it's still me)
Replying to @nabijaczleweli
just wait for the one where i replace the boolean_t (FUCK boolean_t) with actual god-forsaken _Bool. 7.5k in 418 files by git grep, /half/ of all *.[chS] (could run it automatically ig, but (a) what's the point, (b) == B_TRUE is pervasive, and (c) it's достопримечательности)
one day the macintosh will support standard PC104 keyboards :) too bad that's too difficult with today's limited computing power :/
will you compute with him 🥺
Replying to @evilgaywitch
the only actual noun that refers to the product in this is "roast" btw, everything else is saying how good it is (well, /that/ it's quality, which, thank.s). AFAICT more words are spent describing the 201{4,5,6} "world product" нижегород fair grand prix than the bag's contents
this shit (left) is in the official upstream. at *dell*/dkms. just.. how? how do you write this and think "yeah this is good." it's an optimal use of my and the users' time. dkms is written in bash btw. https://github.com/dell/dkms/pull/192
i come across maths academics/MATLAB complainers on my TL semi-regularly and, dunno, it seems pretty okay and vaguely irrelevant for today's day and age (then again, you could say the same thing about dc/bc, so)?
Replying to @sehetw
sometimes i worry i don't drive these gags hard enough, and im always happy to find out that there ain't no way to be too blunt
either juat caught a malaise or acquiesced low-key high-key smoke/CO poisoning. if i dont make it til the morrow kiss the big yellow from me
Replying to @nabijaczleweli
blacked out and hit my head at night in the bathroom. ive regained the ability to speak english over the last half an hour but reading back discord messages i was absolutely a scary like brain problem
Replying to @nabijaczleweli
definitely in the top 10 worst moments of my life. i spent literal minutes crafting each and it just.. wasnt working
Replying to @greenTetrah
real politicians moment huh
really worthwhile moments
holy shit bro just wait 'til american pundits find out you can speak english, russian, and your native language and its like not a big deal. just imagine the fucking day.
[induces psychosis in your warfarin] sahrry, sahry
thanks, doug :)
GONG(I) – evaluate process performance
The evaluation is performed by a set of three ‘‘panelist’’ routines. If any of these routines is not amused by the performance, the process is sent the gong(II) signal.
/dev/ding.dong is the standard gong.
Replying to @nabijaczleweli
I mean, they were really wilding in the seventies, as usual, but "On the Applicability of Gonging to the Performance and Merit Review Process, Journal of Irreproducible Results, vol. 263, issue 19, pp. 253-307.", like damn bro, could like narrow it down a bit maybe?
wow VERY cool, thanks, Mail Server <no-reply@outlook.com> that.. bounced my URGENT ACTION REQUIRED mail from aws off.. gmail?
there's a 5.3k HTML part too, but I like the plain "Test Mail", better. esp. since i got another (identical?) one dated 08:13:22 later, greylisted 39035s
Replying to @nabijaczleweli
left: the one from 23h, right: the one from 08h. the raw diff is really not that intriguing, beside I think this literally being composed in some insane microsoft WYSIWYG
Replying to @nabijaczleweli
okay no this is worse when viewed in a graphical browser what in the fuck.
Replying to @nabijaczleweli
the address is a link to #NOP(?)
23 link: https://firebasestorage.googleapis.com/v0/b/polastisca-malonei.appspot.com/o/indexx.html?alt=media&token=cb4b5cc1-4574-48e0-b801-e25336bb416e#nabijaczleweli@nabijaczleweli.xyz
08 link: https://firebasestorage.googleapis.com/v0/b/server-verification-portal34.appspot.com/o/indexx.html?alt=media&token=853e20be-9fdf-4de9-b2dd-75e1d13037f5#nabijaczleweli@nabijaczleweli.xyz
which are... (a) wha. what on earth, (b) POSTing my first three tries and then redirecting to http://nabijaczleweli.xyz. fucking ace.
(also twitter broke ZWNJs?)
wow thank.s :0
Replying to @nabijaczleweli
this is moments after it decided that the battery thats in it is not in it actually btw
Replying to @nabijaczleweli
???!??!? pog? https://lore.kernel.org/linux-input/20220126161832.3193805-1-benjamin.tissoires@redhat.com/T/
Replying to @nabijaczleweli
aaaaaaand it only applies to 5.17-rc1 and changes the ABI, can't just swap modules :v bindeb-pkg and strapping a clean sid at the end of my disk where I've, with a shocking amount of fore-thought, left 18 unallocated gigabytes, instead of trying to fast-track openzfs 2.1 for 5.17
Replying to @nabijaczleweli
thanks. love too log in and engage in an interactive session
Replying to @nabijaczleweli
okay, that seems to've been a classic kexec moment; actually booting from the EFI shell works :0
Replying to @nabijaczleweli
HOLY BRSCH
Replying to @nabijaczleweli
and the tests pass too, but, after a full year(! the root tweet here is Jan 29, 2021!), we've got functioning GXTP Win8 styli approaching mainline!!
Replying to @nabijaczleweli
:0 in what appears to be tradition in CVS users, the committer didn't notify me, but NetBSD script has -e now: http://cvsweb.netbsd.org/bsdweb.cgi/src/usr.bin/script/script.c#rev1.29!! (but also the bit where i replaced system(3) with a symmetric exec was Very Controversial, so -e with -c always exits 143, patch pending)
Replying to @nabijaczleweli
and my dd convtabs patch too !! :)
http://cvsweb.netbsd.org/bsdweb.cgi/src/bin/dd/conv_tab.c#rev1.10
http://cvsweb.netbsd.org/bsdweb.cgi/src/bin/dd/dd.1#rev1.38
http://cvsweb.netbsd.org/bsdweb.cgi/src/bin/dd/args.c#rev1.42
http://cvsweb.netbsd.org/bsdweb.cgi/src/bin/dd/extern.h#rev1.25 (but let it be known that /I/ didn't miss this header)
Replying to @nabijaczleweli
fucking CVS bro
Replying to @nabijaczleweli
hi :0
Replying to @nabijaczleweli
had a very cool moment where first they completely didn't find my registration, and then the shooter asked "did you get this season's flu shot?" – no – "you'll get it in the other arm, then", and, indeed, she gave me two quick ones
Replying to @nabijaczleweli
more hospital visits should be like this imo
Replying to @the6p4c
listing mesel on alibaba like