Show Idle (>14 d.) Chans

← 2021-05-16 | 2021-05-18 →
08:03 jurov slackware?
~ 1 hours 24 minutes ~
09:27 asciilifeform wb jurov !
09:28 * asciilifeform not tried slackware.
09:29 asciilifeform did try a purportedly systemd-free 'modern' linux -- 'artix'. needed to serve a commercial client who traditionally used 'arch'.
09:30 asciilifeform 'artix' indeed is systemd-free. but many of the packages (e.g. 'cups') do not work; and device hotplugging is mysteriously broken (mass storage -- worx; usb-to-serial -- not; and this despite a known-good kernel. it's whatever the thing has instead of udev that's a dud)
09:33 asciilifeform jurov: iirc BingoBoingo also experimented with a 'modern' and purportedly depoetteringized linux. but can't recall which
~ 50 minutes ~
10:23 trinque devuan's the no-systemd debian fork
~ 27 minutes ~
10:51 trinque none of these are particularly satisfying these days, which is why I tend to use gentoo or my own thing.
~ 2 hours 1 minutes ~
12:53 asciilifeform trinque: would luvv to find 1 of these that 100% worx, even in the sense of 'all proggys which ran on asciilifeform's gentoo, run identically' but so far no dice
12:54 asciilifeform in 'artix', for instance, traditional 'ifconfig' util is vanished and no ready way to restore. instead of it there's some heathen thing 'ip' with entirely different commands, for no detectable reason
12:55 asciilifeform and, as mentioned earlier, given as it sucks packages from 'arch' repos, many proggys refuse to run after installation, as they expect systemd. 'cups' printer util is one example.
12:56 asciilifeform in that same linux, classical 'alsa' worx, but only for pci soundcard, and not for usb. (the mobo in question had built-in audio, which turned out to sit on usb internally..)
12:57 asciilifeform and other weird dysfunctions. scrollback in shell, for instance, leaves visible artifacts/garbage. found 0 info on www re why.
12:57 asciilifeform perhaps this has been a thing for years on heathen linux, and asciilifeform only now found out 'new normal'(tm)
13:07 trinque you know, for all the scoffing at openbsd, every time I pick it up, works.
13:12 asciilifeform trinque: i had a number of issues with it.
13:12 snsabot (trilema) 2017-07-22 asciilifeform: mod6: i'm using the i386, on old lappy, and it dies similarly
13:12 snsabot (trilema) 2017-06-08 asciilifeform: in other noose, nobody on planet3 seems to know how to disable serialport flowcontrol on openbsd.
13:12 snsabot (trilema) 2017-03-28 asciilifeform: and i thought my openbsd 'libretto' that loses its x11 frame buffer was rotten
13:13 asciilifeform ( complete chronology for the curious.)
13:13 asciilifeform most direly, there is no equivalent of gentoo's USE flags, so there's no reliable way to depoetteringize the ports, other than by not using the ports tree at all.
13:15 asciilifeform i suppose, like all other sadware, there's some circumscribed universe where it 'works' (e.g. irons from bush II era, no softs but apache, etc) -- but for workstation, can't imagine who and how lives with it
13:17 asciilifeform afaik to this day there is not a reliable port of trb to openbsd. (and not because no one tried)
13:20 trinque phf did, just didn't publish, irrc
13:20 trinque *iirc
13:20 trinque at any rate, "every man for himself". I published an item tiny enough it doesn't need use flags to cull further.
13:21 asciilifeform apparently he did mention it once and no moar
13:21 trinque last movement on that was dpb couldn't be brought to give a shit, since not spoonfed
13:21 asciilifeform trinque: am i mistaken that your recipe required a debian box to bootstrap from?
13:22 trinque the use of debootstrap to bootstrap is an option. I'm not tuning it to everyone's pet gentoo.
13:23 trinque trivially, some set of binaries is required to begin.
13:24 asciilifeform trinque: indeed
13:24 trinque also, the idea that folks bootstrap from wherever they sit, and see if they get bin-identical output was perhaps lost on all but asciilifeform
13:24 asciilifeform trinque: i dun expect you to tune it to other people's gentoo, lol
13:24 asciilifeform whole point is that, aha, erryone must find path from his to working build, a la rotor
13:25 * asciilifeform not, sadly, had time to go through this process yet
13:25 trinque sure, and how interesting if not bin-identical
13:25 asciilifeform aha!
13:25 * trinque does not blame asciilifeform, who has actual projects.
13:25 trinque others may consider picking up an oar.
13:26 * trinque recently has been filling in gaps in the kademlia guy's online-codes paper. much was unspecified other than the core algo
13:26 asciilifeform trinque: the fountain-code algo ? (i recall we did several of these, which one was this?)
13:29 trinque yep
13:29 asciilifeform a.
13:29 trinque huh, guess doesn't eat pdf well
13:29 trinque guess one downloads the zip for that
13:29 * asciilifeform pulls the hardcopy from file cabinet.
13:30 * trinque has backups also, figures this webpage will eventually rot, or maymounkov found to have once fucked someone, or etc
13:42 asciilifeform trinque: don't hesitate to post your 'filling in', if you've the time
~ 29 minutes ~
14:12 trinque yeah, in time. no surprises though. author assumes "friendly wire", i.e. one where udp checksum can be trusted. certain essential metadata is external to the encoding.
14:12 trinque lots of places for enemy to impose unbounded cost
14:14 asciilifeform trinque: i'm not aware of any fountain schemes where the fragging can be used 'naked' i.e. w/out signatures
14:14 asciilifeform it aint a substitute for rsa.
14:19 * trinque will have to load the thing back into head later to speak more specifically, but thought that the enemy not knowing what xor paths to take across the check-block graph combined with the fact that xor does not lose randomness might yield something useful
14:19 trinque but 100% agree that RSA comes in.
14:19 trinque also interesting that xor paths don't have to transit just "your" blocks. can transit any number of blocks you may have seen in the past.
14:20 trinque anyhow, orig. author didn't even specify a binary protocol, just fountain algo, that I've seen
14:21 asciilifeform trinque: both of the above were applications we spoke of for fountains (in one case, as a pill against known-plaintext cryptoanalysis and possib. substitute for traditional 'padding' schemes; in the other -- as a proof-of-having-blocks-on-disk scheme) but were separate things
14:22 asciilifeform the application i was thinking of in this thread was the third one -- where civilized replacement for tcp.
14:29 trinque yep, the frame I've been thinking in is p2p chat, as I don't intend to remain on freenode forever.
14:30 trinque that frame easily generalizes to other graphs of nodes hucking wads of ??? at each other.
14:35 asciilifeform makes sense.
14:36 asciilifeform tho if narrowly specific to chat, rather than general-purpose parcels a la tcp, not clear why needs fountains -- a signed 'line' of reasonable length could fit in 1 packet
~ 4 hours 54 minutes ~
19:30 asciilifeform $ticker btc usd
19:30 btcinfobot Current BTC price in USD: $43168.79
19:31 asciilifeform !w poll
19:31 watchglass Polling 16 nodes...
19:31 watchglass : Could not connect!
19:31 watchglass : Could not connect!
19:31 watchglass : ( Alive: (0.082s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Return Addr= Blocks=684027
19:31 watchglass : ( Alive: (0.034s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027 (Operator: asciilifeform)
19:31 watchglass : Alive: (0.082s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Return Addr= Blocks=683874
19:31 watchglass : ( Alive: (0.084s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027
19:31 watchglass : ( Alive: (0.107s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027
19:31 watchglass : ( Alive: (0.172s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=674732 (Operator: trinque)
19:31 watchglass : Alive: (0.204s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027
19:31 watchglass : Alive: (0.145s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Return Addr= Blocks=684027 (Operator: whaack)
19:31 watchglass : Alive: (0.221s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027
19:31 watchglass : ( Alive: (0.256s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=412049 (Operator: jurov)
19:31 watchglass : Alive: (0.283s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=684027
~ 18 minutes ~
19:49 verisimilitude I'm also slowly working on something tangentially related.
19:50 verisimilitude I've yet to review other papers to understand online codes, but any reading guidance would be appreciated.
19:51 verisimilitude OpenBSD only works well enough for me on my Lemote Yeeloog; it often froze on other machines, although I ran WWW browsers on those.
19:54 verisimilitude That is my Lemote Yeeloong.
~ 2 hours 39 minutes ~
22:34 asciilifeform verisimilitude: didja read any of the linked threads?
22:35 asciilifeform just about anyffin can be said to 'it worx' if you limit to particular irons, small set of ultra-popular proggys, etc
22:38 asciilifeform verisimilitude: this aside -- 'yeelong' is an interesting box, and i tried for many yrs to get hold of one. but no longer, even if found one today, would be 'museum queen', would not want to leave the house with it -- say it gets stolen or rained into etc., where wouldja get another ?
22:39 asciilifeform (recall rms and his nervous breakdown when he lost just such a box in argentina)
~ 48 minutes ~
23:27 verisimilitude I've not read them yet, no.
23:30 verisimilitude I mostly just use the Yeeloong for IRC, and more recently what APL programming I do, and most recently managing my rented server and so website and whatnot; it's much nicer with the trackball and BAT.
23:30 verisimilitude I prefer to leave my home without any electronics.
← 2021-05-16 | 2021-05-18 →