Show Idle (>14 d.) Chans
← 2022-01-25 | 2022-01-27 → ↓
shinohai[asciilifeform]: testing pingback
shinohai[billymg|asciilifeform]: now to get bot reworked
billymg: ok, if literally all i do is unpack the 2016 gnat bin (no ./doinstall or anything) and run ldd on the gnatmake that's included, i get missing libraries
billymg: so i only have one question, should those binaries in that linked gnat tarball run on a musl system? i have musl 1.2.2 (perhaps that's the issue, i assume others here are on older versions)
billymg: also my system doesn't have a /lib64 directory, only /lib
u0_a185[billymg|asciilifeform]: I forgot how I did that billymg lol
billymg: is that you, shinohai?
billymg: i read through all of ave1's old articles on this and i didn't see any mention of this
shinohai[asciilifeform|billymg]: oh hey yeah that's me, im experimenting with blatta from android lol
billymg: nice
billymg: shinohai: but you followed ave1's instructions on a musl-based system, correct?
u0_a185[billymg|asciilifeform]: Funny in this station I see you as heresay of shinohai billymg, but as self see you as heresay of awt
shinohai[billymg]: billymg: yup I was using musl profile from eselect, ran `ldconfig` and built fine iirc
billymg: ah, i did neither of those things
billymg: i mean i'm guessing my profile is probably set to musl, but i didn't do the `ldconfig` thing
billymg: `ldconfig`: Updating links is not implemented
billymg: alright, will google this path and see what i can find
billymg: my profile is set to "default/linux/amd64/17.0/musl/hardened (exp)
shinohai[asciilifeform|billymg]: yeah that should be correct
shinohai[billymg|asciilifeform] is terrible about documenting precise recipes for things like this ......
billymg: i mean it sounds like i'm just in a weird state, i'm sure if you or others had run into the 2016 gnat bootstrap bins simply *not working* it would've been mentioned somewhere
billymg: probably cause i started from a bleeding edge 2022 stage3 tarball
billymg[asciilifeform]: probably cause i started from a bleeding edge 2022 stage3 tarball
billymg: and somewhere something broke along the way
shinohai[asciilifeform|billymg] hangs head because using glibc on daily driver atm ....
u0_a185[billymg|asciilifeform]: >testing greentext in weechat
asciilifeform: shinohai: fwiw asciilifeform currently only has traditional glibc boxen
asciilifeform: (it's what 'dulap gentoo')
asciilifeform so far only has muslism in individual proggies, e.g. trb, ffa builds
shinohai[asciilifeform]: I have a chroot with signpost 's cuntoo from times past still
shinohai[asciilifeform]: and an experimental profile of own, but nothing to write about
billymg: i probably should've made this a glibc system
awt[asciilifeform]: Anyone know how to remap right alt to ctrl in X?
asciilifeform: xbindkeys
u0_a185[asciilifeform]: The 24-Hour VWAP for BTC is $ 37917.30 USD
billymg: awt: i use xmodmap, with this in .Xmodmap: http://paste.deedbot.org/?id=m6OW
billymg: and then `[[ -f ~/.Xmodmap ]] && xmodmap ~/.Xmodmap` inside .xinitrc to load it when xorg starts
awt[asciilifeform]: asciilifeform: xbindkeys only seems to map keys to commands
awt[asciilifeform]: asciilifeform: not sure how to handle versioning of blatta. I've been releasing incomplete implementations so far, which seems to have been useful.
awt[billymg|asciilifeform]: But that leaves me unsure as to what version of pest I should report in the packets
awt[asciilifeform|billymg]: and also how to handle version differences
billymg: awt: yeah, i use xbindkeys for mapping some of my function keys to things like screen brightness
billymg: the xmodmap thing works for remapping caps to control though
asciilifeform: oh hm, didn't realize it couldn't remap alt
asciilifeform: awt: btw if we go with this, defo will need to reject packets from >0xfa (breaks format)
asciilifeform still waiting for comments re proposed algo before writing it up
asciilifeform: awt: in general, the Right Thing re ver. to report is 'whichever ver. of spec nominally conforms to' i.e. w/out barfing
asciilifeform: version only becomes 'playing character' when actually breaking the format (as asciilifeform proposed today)
signpost[asciilifeform|billymg]: shinohai: grunting out this next one before long, just taking the time to do it right.
signpost[asciilifeform|billymg]: fuck was I infurated at the "hurrr hanbot can't run this, trash" after months of putting cuntoo together.
signpost[asciilifeform]: dunno what I expected.
signpost[asciilifeform|billymg] has vpatches for src and separate vpatches for musl compatibility for the items going into "pentacle"
signpost[asciilifeform]: incidentally still haven't named the thing to my satisfaction, ended up "pentacle" because asciilifeform noticed the ansi art in a readme of early cut.
asciilifeform thinks 'pentacle' is a mighty fine title
asciilifeform: 1337666
asciilifeform: http://logs.bitdash.io/pest/2022-01-26#1003094 << to asciilifeform loox like a 'smoking gun' gcc5ism (but unsurprisingly author dun fess up to it)
bitbot: Logged on 2022-01-26 22:50:27 signpost[asciilifeform]: farting around with a segfault in the flex build, details here https://developers.redhat.com/blog/2019/04/22/implicit-function-declarations-flexs-use-of-reallocarray
asciilifeform: ( see also similar !)
asciilifeform: http://logs.bitdash.io/pest/2022-01-26#1003090 << heh, but wasn't infuriated when asciilifeform was stumped in attempt to replicate..?
bitbot: Logged on 2022-01-26 22:48:32 signpost[asciilifeform|billymg]: fuck was I infurated at the "hurrr hanbot can't run this, trash" after months of putting cuntoo together.
asciilifeform iirc couldn't make it go on (what was to turn into) dulap-gentoo, ran into debianisms
asciilifeform didn't have a debian box handy, went to bake one but ran outta cycles
signpost[asciilifeform|billymg]: alrighty, will keep pentacle.
asciilifeform: this is a perennial plague btw that is quite difficult to escape -- even asciilifeform's 'rotor' iirc still bombs on some boxes, and not erryone got to the bottom of just why
signpost[asciilifeform]: asciilifeform: cuntoo script wasn't debian-based, but I did add a dulap bootstrap script for ya on pentacle.
signpost[asciilifeform|billymg]: environment variable leakage seems to have been one significant source of non-determinism in my latest item.
signpost[asciilifeform]: fixed all by reinvoking the scripts with env -i at top.
asciilifeform: ( the 'how the fuck do we get a repeatable environment on all boxen' thing was why asciilifeform ended up doing the 'M' experiment )
signpost[billymg]: indeed, M's this in the limit.
signpost[asciilifeform|billymg] atm drinking in the airport, otw back from new england excursion.
asciilifeform: even 'M' relies on sitting on x64 box w/ a linux from some time in last 15y or so
asciilifeform: signpost: oh yea iirc you were terraforming a pad in ne ?
signpost[asciilifeform|billymg]: yeah, haven't yet encountered an oddball problem with a statically compiled binary on any linuxen I use.
signpost[asciilifeform|billymg]: asciilifeform: still scouting a spot
signpost[asciilifeform]: likely to be new hampshire woods, quite enjoy them.
signpost[asciilifeform|billymg]: cheap enough acreage, and quick drive to higher elevation for proper snow.
asciilifeform: sounds a++
signpost[asciilifeform|billymg]: how are you finding your new place?
asciilifeform: signpost: still not moved in, slowly terraforming , almost '1 anchor bolt at a time'
asciilifeform: still gotta bolt ^ down, get the fiber in, various similar chores
signpost[asciilifeform|billymg] definitely emplacing a rack in the new place once acquired too.
asciilifeform had lulzy chat w/ local telco re 'biznis fiber'
asciilifeform: 'yes you can get static ip! it'll only cost 2.5x yer current rate!'
asciilifeform has this conv ~erry yr, lol
signpost[asciilifeform|billymg]: "yes but what about dhcp with infinite lease?" (crickets)
signpost[asciilifeform|billymg]: no idea why they're precious about this
asciilifeform: that's what asciilifeform has nao
asciilifeform: 'infinite', dies when they change battery in their rack by the train station
asciilifeform: ... or on no less than 2 occasions, when the psu in the fibertron they issued, gave up its bluesmoke
asciilifeform: monkey:'it must've been ac spike' asciilifeform:'mno, it's on this-here doubleconverting clean sine psu, and the other gear's happy'
asciilifeform: if the 'biznis' subscr. at least came w/ a spare fibertron, might take it. as it is, only offers some promisetronium re 'same-day repair'
asciilifeform: (that, and, promisetronically, +5 vorpal packet priority, and static)
asciilifeform: is actually 3x cheaper to subscribe to local cable shithole as fallback
signpost[asciilifeform]: can't bring own gear with these? (only ever used the ISPs fiber box myself, but often use own cable modem)
signpost[asciilifeform]: incidentally looks like there's a fiber ISP where I'm headed in south NH, lovely.
asciilifeform: signpost: own router, but the optic terminal is theirs (ethernet-to-glass thing)
signpost[asciilifeform]: yeah, I figure anything I'll be hosting at home, I'll just do the pirate-bay-style proxy from w/e boxen outside
asciilifeform: afaik this is typical
asciilifeform laying pipe to at least bring the fiber into machine room, tired of the nonsense where by default the fibertron mounted on exterior and gets rain, ice, wasp nests in it
asciilifeform opens, in spring, 'zzzz'
asciilifeform: last yr 1 burned out, tech brings new one, tears guts outta old, shoves in a box the size of 'pogo', duct tapes in. asciilifeform:'can get rack ears for this?' d00d:'what are those'
asciilifeform: at least they stopped asking 'where's yer tv coax'
asciilifeform: signpost: asciilifeform investigated, and lulzily seems that most 'biz' subscribers in 'human' streets are folx who were exiled into it from overeating mystery bw quota. but oddly enuff asciilifeform not stepped on this mine just yet
asciilifeform: (if yer in a 'biz' tower, they won't even give 'human' subsc.)
asciilifeform: nfi what you gotta do to overeat; e.g. asciilifeform regularly backs up multi-TB turds offsite & not eaten
signpost[asciilifeform|billymg]: yeah, I figure by now they're overrun with dorks torrenting from the obvious sources, anyone with even remotely unconventional bit-sources aren't worth chasing.
asciilifeform: may depend on the 'oversoldness' of yer particular sector (the upstream cabinet gets e.g. fixed 10g/s pipe, this in turn sat on by n streets)
signpost[asciilifeform|billymg]: "load" here is bitching from w/e copyright enforcement firms, not necessarily bw
asciilifeform: ^ oddly enuff not encountered to date
signpost[billymg|asciilifeform] gets about 980mbps consistently from coax in austin
asciilifeform: hypothetically similar here, tho the dusty old apu1/pfsense maxes out in 700s
signpost[asciilifeform]: (after new line ran to house; old was conducting partially through the bracket holding wire to the fence before)
signpost[asciilifeform]: 650mbps through nail!
asciilifeform: then again a ground is a ground
signpost[asciilifeform] running apu2 router on your recommendation
asciilifeform: great from asciilifeform's pov, but wouldn't hurt if could go over ~700, or at least not run at 65c
asciilifeform: by far warmest clothes iron in asciilifeform's rack atm
asciilifeform: (core atm at ~87c)
signpost[asciilifeform]: I might've gotten a later variant, pulling that 980 through him
signpost[billymg|asciilifeform]: not sure, could also be doing less packet fiddling
asciilifeform: pfsense?
signpost[billymg|asciilifeform]: pretty simple iptables setup
asciilifeform might replace w/ e.g. garden variety netbsd in new one
asciilifeform annoyed that to this day one can't buy an 'iron' router (i.e. w/ accel. nics) + clean linux
asciilifeform: or at any rate not in a 'fanless brick' shape
signpost[asciilifeform]: not terribly surprised that all of these are corrupted with cloudware, suspect chips, or etc., either by design or by way of fallen culture.
asciilifeform: ^ would be pretty handy basis for 'iron pestron', incidentally
asciilifeform: signpost: can buy 'iron router' so 'plentiful can eat with arse', but 0 afaik w/ clean support in (mainline or otherwise) kernel, hence shitware
asciilifeform: (concretely referring to 'not just nic, but can offload packet munging from cpu to this-here fpga' items)
signpost[asciilifeform|billymg] even bought a leet.company domain with the idea of doing sane router kits with cuntoo derivative.
asciilifeform: on heavy enuff (fan-blown) irons, entirely doable
asciilifeform: on apu1 will go like asciilifeform's pfsense goes
signpost[asciilifeform]: but didn't figure the market depth was there
signpost[billymg|asciilifeform]: neckbeards that might otherwise be customers seem to have huge boners for mikrotik and ubiquiti
asciilifeform could be persuaded to piss on that 1000volt rail again, but would take sumthing actually persuasive
signpost[asciilifeform] also open to future developments if worth it.
asciilifeform: atm asciilifeform can't picture what this'd look like, fwiw
asciilifeform: upfront cost of making even FG today would be prolly in 'mid mercedes' range.
asciilifeform: (i.e. for 1st board!)
signpost[asciilifeform]: when I'm done tilting at the "motherfuckers I will have an operating system where I can read all the parts in one lifetime that runs on obtainable iron", will grunt at w/e is the next most useful thing.
asciilifeform: aand that's with 'oh hey you want smt resistors? they'll get in q3 of 2023 maybe'
signpost[asciilifeform|billymg]: quite interested in the pestnet wareztron.
asciilifeform: luckily bakeable outta shelf irons
signpost[asciilifeform|billymg] bbl, hopping in the flugentube
signpost[asciilifeform]: http://logs.bitdash.io/pest/2022-01-26#1003094 << when in doubt, get older vintage. 2.6.2 worx fine. must've snagged 2.6.4 when carving off individual vpatches
bitbot: Logged on 2022-01-26 22:50:27 signpost[asciilifeform]: farting around with a segfault in the flex build, details here https://developers.redhat.com/blog/2019/04/22/implicit-function-declarations-flexs-use-of-reallocarray
signpost[asciilifeform|billymg] rebuilds all, sacrifices virgin chicken
signpost[asciilifeform|billymg]: asciilifeform: sometime, perhaps tomorrow, it'd be great to give the v "manifest" thing another look.
signpost[asciilifeform]: still feels like a kluge to me.
signpost[asciilifeform|billymg]: to restate the problem for the logs, extant v-trons walk backward from a given specified leaf node to genesis, then press forward along that path.
signpost[billymg|asciilifeform]: this is to avoid the fact that two given leaf nodes may both mutate the same "from" hash, and it's not decidable in current spec which takes precedence.
signpost[asciilifeform]: questions I think are worth revisiting: 1) do we in practice have any use for multiple press heads? 2) is the lack of v-tron enforcement of manifest contents matching the computed patch flow prone to unnecessary operator error? 3) do we care if we preserve vpatch compat with traditional posix patch?
signpost[asciilifeform]: (also ignore the GNS mention in the blog post. I was macroexpanding that to a distributed item, not a reboot of DNS with replaced crown)
signpost[asciilifeform|billymg] would like operators to have maximal control over which patches they choose to use, rather than an author-dictated ordering, where/if possible
signpost[asciilifeform|billymg]: asciilifeform originally envisioned v pressing all paths that can be pressed, iirc, or in other words, if the operator doesn't want a patch pressed why'd he place it in `patches` dir?
signpost[asciilifeform|billymg]: in that view, having same antecedent in two patches simply fouls the press, and operator is invited to regrind one of them.
signpost[asciilifeform] sees very little problem with this approach, but invites reminders if forgot other problems.
signpost[asciilifeform]: another approach would involve allowing patches to declare antecedents without requiring that those antecedents be modified.
signpost[asciilifeform]: perhaps this can even be done today with a no-op hunk. I'll have to try that.
signpost[asciilifeform]: hm, dropped a line fighting with mega-slow ping. will pick this up tomorrow. cya!
← 2022-01-25 | 2022-01-27 → ↑
Random(pest) | Download hourly DB snapshot | Get Source Code