BingoBoingo: Maybe, but there just wasn't more than 5 to 10-ish square mm of metal surface exposed on them for air to do much heat exchanging
asciilifeform: even these, i suspect, would've lived 4x longer w/ air.
asciilifeform: presently i blame 100% of the headaches in pilot plant, on lack of proper air
asciilifeform: the primary win of closed 1u chassis for rk is proper, civilized forced air
asciilifeform: BingoBoingo: item will actually slide, given as will have only 2 wires going to it (mains & 1 ethernet) and not 15
asciilifeform: so i'ma omit the relays ( granted the 1u flat setup will be considerably easier to swap pieces in, than in old plant )
asciilifeform: rright, and the thing will still have to be slid out an' opened to swap
BingoBoingo: asciilifeform: I can't remember reboot requests for the rockchips that didn't involve disk swaps
diana_coman: asciilifeform: thank you; I haven't yet set the cron jobs, yes; will do it later today.
asciilifeform: diana_coman: i noticed that your db snapshot link dun go yet. here is example, how mine worx. and here is the crontab. can adjust paths for yours.
mircea_popescu: keks, no moar "parliament". only took 1k or so years, the albiontards are slow.
diana_coman: http://logs.ossasepia.com/log/trilema/2019-08-31#1932624 -> thanks for posting the versions that worked on your testbed, they saved a lot of futzing around!
lobbes: annd I'm off to bed. Got LAMP stack up on heathen server yesterday; tomorrow gonna try logotron proper
lobbes: in other words, make a post my friend. If anything I'll read it and comment
lobbes: mod6: take it from me, please, the best thing I ever did was start blogging/writing things down/*expressing* myself. Only through this kind of reflection (and by this I mean analyzing the reasoning you used in the past for action and analyzing your reasoning for future action), and through the critical eye of others can that fog even start to clear
lobbes: (while this lasts, of course. The 'sword of damocle' hangs over ALL of us still sucking blood from the reich, "9-5" or no.)
lobbes: Note, too, that my saltmine right now IS a "9-5", only I lucked out finding one that a) "pays well" and b) I don't, actually, work a full "9-5". I'm bezzling the bezzle, in a sense
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932520 << ftr, THE thing that helped me the most in the last month was removing a broken homefront partnership from my life. My saltmine job (as an "excel jockey") has remained the same for the last 4 years
snsabot: Logged on 2019-08-30 12:36:27 mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932106 << what, the later tell delivery ?
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932283 << I do understand the 'spammy-ness' of it; perhaps could have two commands: a !Qlater chan and !Qlater pm
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932253 << congrats diana_coman! I tip my hat to your beating me to the finish line
asciilifeform: will be lulzy if we even end up with bigger net than fleanode, simply from the junk boxen people ~already~ have.
asciilifeform: ( and when we teach'em to talk to each other -- then can kiss goodbye to fleanode... )
asciilifeform: i expect the load times will tell us something interesting, just as before.
asciilifeform: trinque: iirc you had some box in asia -- consider to put logger there. the moar, the merrier.
asciilifeform: then looks entirely like i expect.
diana_coman: asciilifeform: ah, you mean that curl doesn't follow the redirect, right.
asciilifeform: diana_coman: look in reader.py to see why, i made it default to selected chan ( for if folx want to substitute it for phf's on their blogs )
asciilifeform: gotta use full url w/ chan, to get the page. curl normally doesn't walk 301 ( there's a toggle but i fughot )
diana_coman: asciilifeform: you are right; let's redo then
diana_coman: ftr the isp guys were very prompt and willing to do as asked; although one has to ask for the most basic of things, they at least do it
diana_coman: for completeness, here's the timing from UK, apparently Moldova wins by far: http://p.bvulpes.com/pastes/U5Fnx/?raw=true
asciilifeform: which i guess was at least part of the objective of thrusting it upon the illiterates.
asciilifeform: segshitness seems to make these easier to produce -- nonexistent sigs make room for moar past tx references.
asciilifeform: re blox -- loox like ..456 is another 'heaviweight'.
mod6: im up to 490 on the foundation's node @ piz.
asciilifeform: the ominous bit is that miners themselves are apparently letting these go w/out actually verifying (or they'd sit as long as zoolag is sitting, per)
asciilifeform: i expect that the o(1) tx db will become needed at some pt for realtime operation, rather than a 'wouldn't it be nice one day' item.
asciilifeform: in re trb -- zoolag ~still~ chewing on a block (perhaps the genuine ..450, perhaps not) even nao.
asciilifeform: ( for which will need boxes just the same )
asciilifeform: imho an orchestra of log mirrors / bot boxen is a notbad prelude to the eventual gossipdification .
asciilifeform: mod6: it's a nearly starvation-bare logtron atm, missing all kindsa luxuries (e.g. phf's backlinks system) . but atm it's the only 100% working ~published~ one
mod6: i use the page daily, works well.
asciilifeform: mod6: depending on whether you have a box where can run python & postgres, it is maybe 1-2hrs of sweat.
mod6: i saw in the logs that others were standing one up - immediately thought "ok, yup do this too." but before I embark on yet another thing, need to ensure that i have proper time. but i'll add it to list for sure.
asciilifeform: press asciilifeform's vtree and follow the readme ( with diana_coman's nitpick from 12h ago , re the createdb step )
asciilifeform: re further trb lulz : 24.93.104.14 spams liquishit blox at ~gb/s
mod6: the important thing for me is that i'm resolved to not just sit around and wait for things to happen to me, re: 9-5 tarpit. im starting down a path of emacipation from that. might take some time, but not too much time.
mod6: but now, need to re-invent myself a bit. need to move on from 9-5ism, and be my own man. will be a "project" for me, but working on the deets/plans atm.
mod6: *nod* in the early days, pre-tmsr, I had a similar thing, was much easier to stay focused/up-to-date.
mod6: but yeah, 9th circle. but now, I'm committing myself to getting rid of the salt-mine, somehow. I've got to do some thinking, and talking to peoples, etc. but I want out of that life.
mod6: BingoBoingo: ok, banners should be there. want to try again? (otherwise, can move to blog post)
BingoBoingo: mod6: I suddenly hit up to 453 all within the last minute
lobbesbot: mod6: Sent 3 hours and 6 minutes ago: <BingoBoingo> the links to your pizarro banners appear to all be 404 on mod6.net, is there any way you could assemble all the banners together in a blog post or something?
asciilifeform: BingoBoingo: last time this q was raised, asciilifeform sat and wrote in fact 80% of a 'tmsr node explorer' proggy, to try an' find 'where are the humans' instead of continuing to rely on chance. but never got chance to finish.
BingoBoingo: asciilifeform: Apparently now blocks have a size and a stripped size. Looking for a raw stripped 450 in the wild
asciilifeform: ( if yer lucky, it'll reorg when the actual 450 shows up )
BingoBoingo: Hence the poking
BingoBoingo: Or there's some segshit attached which should be shaved off
mp_en_viaje: anyways, i'll be off for the night. cya all laterz
asciilifeform: the network is in prolly saddest shape it's ever been in
asciilifeform looks at 'zoolag' log, it quite depressing picture, 900000 'heathen command, banned' erry hr
asciilifeform: BingoBoingo: btw for that heathen dump, the conversion is xxd -r -p liquishit.txt > liquishit.bin .
asciilifeform: e.g. there.
snsabot: Logged on 2019-08-01 18:31:03 asciilifeform: at this pt, seems quite evident that someone is throwing around crafted wedge chains (i.e. mined after-the-fact , with backdated timestamp, going from older block) specifically to wedge syncing folx.
BingoBoingo: After a long stretch of not needing to give a day or two and seeing the whole trb-iverse visible from my chair stuck on the same block, I figure why not try to forcefeed the problem block
asciilifeform: always did the 'let's give it a day or 3' thing
BingoBoingo: Pretty sure my copy-paste buffer munged the hex: Block decode failed (code -22)
asciilifeform: ( it's either that or there's a 3000km spool randomly sitting in brazil , that it goes through !! )
girlattorney: there are two cables: one from 2000 that afaik carries only phone calls and govt activities (atlantis 2) and the 2nd that is still not operational (ella link)
asciilifeform: girlattorney: i'm quite certain presently that you're right, and the packets from south a. go through north a. always. simply based on delay measurements to date.
girlattorney: BingoBoingo i think you aren't aware that currently there isn't an IP route from fortaleza to eu
girlattorney: the game changer should arrive with "ella link"
asciilifeform: the (consistent) delays you can measure with own hands, are quite revealing of the actual physical topology.
asciilifeform: this may be so . and given the delay b/w e.g. piz & europistan, i'm quite certain the packets go a few undocumented 'extra' 1000 km .
girlattorney: won't change much politically, but this cable has the potential to lower the overseas link prices
asciilifeform: girlattorney: there's supposedly a fiber from brazil to portugal, iirc, but still owned by reich-controlled conglomerate afaik.
asciilifeform: girlattorney: correct. in usa there are a great number of 'televison'-flavoured nominally ip-based services, and they mostly do not use the actual net at all, but instead connect to a cabinet in the isp's local house.
girlattorney: but then when you go outside, you instantly go from a 10 lanes road to a single lane road
asciilifeform: when you get out of the reich space, it begins to very much matter where you are vis-a-vis your peer.
girlattorney: read the bingo bongo post about cdn and local intranets with load of bandwith
asciilifeform: girlattorney: isps in the nato reich seem 'fast' , but only for so long as the connecting peer is also in the reich. connectivity to the free world is rather more limited.
asciilifeform: but sometimes this falls to 50, or worse, when 1 or more of the piz inhabitants is at high load.
asciilifeform: currently we do not have precise measurement in realtime. but on most days i can get 100-500kB/s to my box there (from usa)
girlattorney: if i can ask, what is the oversell rate of a pizarro grade isp?
asciilifeform: real pipe is quite expensive. at e.g. piz it is substantial % of the operating cost.
asciilifeform: net connectivity only seems cheap because residential isp 'oversell' their capacity.
asciilifeform: girlattorney: trying to get trb box for 20 is rather like trying to get luxury automobile for 1000
girlattorney: 100 a month is too much, im in the 20 range atm
girlattorney: i must disclose that at the current value of BTC i'm haven't got so much to justify multiple nodes in expensive colos
asciilifeform: girlattorney: re colo -- talk with diana_coman , who has been surveying various colos. ( piz is pretty packed in re trb, has no fewer than 3 operating nodes . and really these benefit from being dispersed geographically, rather than concentrated in 1-2 cages ! )
asciilifeform: girlattorney: usg takes the 'rats in a barrel' approach to ipv4, they hog 80+% of the ip space and wait for the remaining people to choke from crowding and eventually submit to centralized control
girlattorney: i've read that with ripe ncc you have your loa with your own ip space, if you "fucks up" 3 times announcing blocks, the peers connected to you have to disconnect
BingoBoingo: Anyone else waiting for one of their nodes to eat and spread block 592450?
asciilifeform: slowly the 'soft' is going out, and instead 'hard'
asciilifeform: rather than to compete with arpa's
asciilifeform: girlattorney: the washingtonization of the net was 'soft-pedalled' so that the few actually independent kingdoms , e.g. ru, would allow their indigenous nets (e.g. 'fido') to fully decay and fall apart
girlattorney: from what i've understood iana is the big boy that decided everything in the beginning, then created some other orgs to give the idea that every continent got it's own independent association
girlattorney: thanks for the insights
BingoBoingo: girlattorney: BGP is held together by routers on the edges of individual networks holding the entire route table in memory. In part there's folks allocating IP address blocks. In the other part there's the routers accepting and/or rejecting updates to the routes they already know.
snsabot: Logged on 2019-08-21 00:00:24 asciilifeform: ( often enuff there's a concealed -- if not always well-concealed -- roundtrip to london / washington, sitting in 'between austria and switzerland' etc )
girlattorney: i'm in europe, so in my case the middleman should be ripe ncc
asciilifeform: they ~would~ like to turn it into something resembling dns.
snsabot: Logged on 2019-06-25 16:35:23 asciilifeform: BingoBoingo: usg ministry of lulz is trying to push usgtronic 'pki' bgp. so, 1st gotta 'make weather', vandalize a bit, drum up interest.
girlattorney: is not like DNS where there is a precise hierarchy, correct?
girlattorney: and if i have understood correctly in BGP you do advertise your own routes and eventually someone go to your home and say "you cannot advertise these routes"
asciilifeform: girlattorney: the major weak point is propagation from miners, who are living in own parallel chinese universe and between whom and trb there is a thick layer of garbage.
girlattorney: so as long as trb doesn't care about the peers, and there isn't a single point of failure (aka DNS and root servers) we are sound, correct?
asciilifeform: and if running with default config they are invisible to the heathen 'node list' www's.
asciilifeform: the more properly working trb nodes there are -- the less payoff to anyone for monkeying with their connectivity. at present in fact i do not know how many there are. not erryone bothers to advertise .
asciilifeform: girlattorney: bitcoin net presently ~worx simply because 'impossible to fool ~all~ of the people ~all~ of the time'(tm)(r) via ip hijinks
asciilifeform: girlattorney: observe that there is no attempt at authenticating peers in the existing trb. you could easily be connecting to washington's node when thinking yer connecting to e.g. mp_en_viaje's. at one time i published an experimental patch where can route to people you personally know via ssh pipes, but currently not in use anywhere afaik.
girlattorney: in particular i'm referring about BGP hijacking and the general fact that IANA is a third party of a government, that a day could decide to limit the user freedom
girlattorney: btw, after stripping out DNS on my important applications (such as TRB) my question now was about IP space assigned from IANA, could this be in the future an attack vector?
asciilifeform: girlattorney: as you have found now to own satisfaction, w/out having to take asciilifeform's word for it -- the actual # of ~working~ (i.e. stores and serves the entire chain) public nodes on the net, is very very small.
girlattorney: basically most of the juice (chunks of 50-60gb) camed from your node asciilifeform and bingo boingo
asciilifeform: girlattorney: aa, so you had the 'who-gave' patch, neat. unfortunate that lost. would be very interesting to read.
asciilifeform: girlattorney: i have 1 going on a residential pipe since '15, but it is a rather costly pipe
girlattorney: i also had a nice graph made with darkstat about the peers that helped most on syncing from 0
asciilifeform: girlattorney: indeed these are very sensitive to quality of net pipe.
asciilifeform: ^ i in fact have not had to spoon-feed mine yet. but nao there is from where ! if needs.
asciilifeform: at some pt we will have to think of a way to mechanize these.
diana_coman: asciilifeform: the gap, yes
asciilifeform: diana_coman: how didja fill in the lines ? /raw/ and by hand ?
diana_coman: not sure if there's some algo to it; my choice was so it's easy to remember/match
asciilifeform: so thereby could actually make it so that the diff bots live at any given time on diff ones
diana_coman: I checked at http://lobbesblog.com/static/tmsr_bot_directory.html and !o didn't seem taken but it's not among the "next in line" either
asciilifeform: diana_coman: observe btw that the bot config allows to put ~list~ of fleanode endpoints. so it is not necessary to use the default rotation dnsism, can in fact do like mp_en_viaje does and put list of favourites
snsabot: Logged on 2019-08-30 08:31:43 mp_en_viaje: ~most commodity classes are more monetized than new cars, which is universally what the argentines do : they also make heaters that are worse heaters than scrap wire. and worse coputer cases than tree trunks and so on.
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932215 << that comp case, btw, may have been an inept clone of 1 of my current ones, where ps goes in bottom and ~intake~ fan sucks cold air through filter in the bottom panel. but the orcs did not bother evidently to think about why, and... decided to put ~hot~ end there !
asciilifeform distinctly recalls a thread where asciilifeform was unable to try some heathen proggy because 'the correct' tcl would not build in his gentoo. but sadly cannot find in o(1)..
snsabot: Logged on 2019-08-30 07:00:57 mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931978 << pretty much thast was the idea yeah. iirc we even discussed tcl in ancient thread re this same matter
snsabot: Logged on 2019-08-30 07:21:35 mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932043 << as far as we know this is probably needed in the general.
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932202 << ada imho is uniquely unsuited to 'munge 9000 strings' type of proggy. so not sure whether would ever want pg glue in it. any attempt to use, will be tremendously ugly
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932193 << it's exactly like the infamous 'ppsh' submachine gun factory, where in early models parts in fact were not interchangeable, cuz pre-selected from piles to work 'in set'
snsabot: Logged on 2019-08-30 06:59:30 mp_en_viaje: trying to maybe gwt out of it, hence the work to spec "blog" in the abstract ; but mp-wp stack works for trilema TODAY ; i ain't taking a break from publishing untul "better world". god knows the week long #trilema outage was long enough.
BingoBoingo: mp_en_viaje: Well, its the Argentine Navy firing in the general direction of Chinese boats. And they did sink one Chinese boat in 2016.
snsabot: Logged on 2019-08-30 12:03:30 BingoBoingo: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932212 << No, the Argentines are dumb enough to have their navy fire upon the Chinese boats fairly routinely to the point China's papers report on it http://archive.is/WzLiZ
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932251 << yes ; but a) that's not chinese paper and b) i dunno what naval warfare this is from, but the "fired and nothing happened" seems to confirm my comment.
mp_en_viaje: it;s not "just the tunnel", bout 30-40km each way, radio dead zone.
mp_en_viaje: i honestly dun see the prioblem with in-chan delivery.
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932106 << what, the later tell delivery ?
diana_coman: aha, the db; kk
diana_coman: asciilifeform: which raw dump link? the db dump you mean?
asciilifeform: diana_coman: i'ma cycle the dumper manually . go and load the 'raw dump link' in 30s or so. (and if anyone speaks between now and when you import it, can fill the hole with the raw export knob and eater.py )
diana_coman: asciilifeform: re vtree I considered it but atm the source is your blog so there it is; when I get around to wrap it up I'll sign + publish etc
asciilifeform: btw i have found that the banners spill the screen on some exotic browsers ( crapple's, in particular ) but did not bother yet to try and fix somehow
diana_coman: asciilifeform: indeed it does! and yes, I'll have to look into something for younghands probably but pizarro will stay there too
asciilifeform: re banners , i expected that folx will want to put own , rather than use mine. ( plus there's no practical way to put any such thing in a v genesis ) so i omitted'em
asciilifeform: ( in the form orig. sent in by mod6 iirc )
asciilifeform: the larger banner is on phuctor www
lobbesbot: BingoBoingo: The operation succeeded.
BingoBoingo: !Qlater tell mod6 the links to your pizarro banners appear to all be 404 on mod6.net, is there any way you could assemble all the banners together in a blog post or something?
BingoBoingo: diana_coman: http://logs.nosuchlabs.com/static/piz.jpg Here's one. I'll try to dig for where the rest of the mod6 banners are. Searching the #pizarro logs returned 404'd links
feedbot: http://thetarpit.org/posts/y06/09d-hunchentoot-vib.html << The Tar Pit -- Hunchentoot: requests and replies [b]
diana_coman: asciilifeform, BingoBoingo can I get that pizarro banner to hang in there?
BingoBoingo: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932212 << No, the Argentines are dumb enough to have their navy fire upon the Chinese boats fairly routinely to the point China's papers report on it http://archive.is/WzLiZ
BingoBoingo: "Creo que va a venir el préstamo del FMI, pero estarán esperando el respaldo por parte de la oposición" << Argentard of apparent import in the Argentardocracy discussing pending, yet to be revealed currency exchange controls. Dude just assumes more dole fodder is on the way so long as Argentines all agree to the right ritual handwashing.
asciilifeform will eat the remaining log after tea; brb
asciilifeform: init the db (via the supplied script) and let reader & bot run in background . then see how behaves.
asciilifeform: diana_coman: i tested the bot by setting chan to e.g. #asciilifeform-test . can do similar, then you can use whatever unregged nick etc (dun fughet to set pw to empty then)
diana_coman: asciilifeform: the config knobs were set (easy part!!)
asciilifeform: diana_coman: dun fughet to set up the config knobs (per readme)
diana_coman: and ofc I need to see if the *bot* also works, lolz
diana_coman: asciilifeform: what's the incantation to do with apache to check now?
diana_coman: i.e. so far I could at least actually run the reader and it's waiting; ran the eater on my own ossasepia deeded logs and it ate
snsabot: Logged on 2019-08-30 04:19:52 spyked: re. http://logs.nosuchlabs.com/log/trilema/2019-08-25#1931011 <-- /me has been thinking a bit about this after reading the discussion, and it occurs to me that one of the reasons that might keep people from signing items is that there's no way to "unsign" them. nor that is desirable imho, but "what if" some sign piece of coad turns out to burn your house to the ground. in any case, this reveals to me the (
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932162 << there of course is no such thing as 'unsign'. next best thing is 'antipatch', i.e. where you fix the errata. and imho all signatures oughta include human-readable annotation
snsabot: Logged on 2019-08-29 22:51:04 lobbes: For clarity, on testbed I got the eater.py and the reader.py working fine. Was able to view loglines and search with no issue (on localhost only, didn't test port forwarding). Did *not* test the bot.py, however. For reference, my version info: Flask 0.12.2; psycopg2 2.7.4; Python 2.7.15
snsabot: Logged on 2019-08-30 04:12:47 spyked: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931955 <-- recall, /me already wrote a toy prototype as his introduction to ada. other than the fact that it looks noobish and requires work to integrate with the c-isms, writing logger on top of it would also require something along the lines of a. mod_adalisp and b. handling other wwwistic s
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932159 << mod_xyz is only really needed with ancient horrors like php, which otherwise sink 1e9 cpu cycles on revving up the interpreter with ~each~ www req. for items that serve themselves up on a port, dun need mod_xyz, can simply fwd port
snsabot: Logged on 2019-08-30 04:09:03 spyked: and the only *quick* solution that I can think of here is signing the binturd that I've been using for the last few years; if anyone has a better idea, pl0x to chime in
BingoBoingo: I'm thinking something like http://trilema.com/2016/the-tatu-party-bus-and-other-wasps/ but clean
snsabot: Logged on 2019-08-30 06:34:02 mp_en_viaje: anyone wanna guess the make ?
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-30#1932170 << I mean cuntoo exactly. testbed is the exact iron in this post
mp_en_viaje: anyway, point being argentine is ~equiv innocent black youth : pure entropic process, exactly like yeast fermentation. "industry" pointed misnomer for the activity.
mp_en_viaje: ~most commodity classes are more monetized than new cars, which is universally what the argentines do : they also make heaters that are worse heaters than scrap wire. and worse coputer cases than tree trunks and so on.
snsabot: Logged on 2019-08-29 21:36:05 BingoBoingo: asciilifeform: China doesn't even particularly like Argentina. The Argentine Navy shoots at their fishing boats.
snsabot: Logged on 2019-08-29 21:29:43 BingoBoingo: Well, probably waiting for local labor prices to drop enough to outsource the spear work to the Paraguayo immigrant population.
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932060 << certainly how the tutsi massacre thing worked irl.
snsabot: Logged on 2019-08-29 21:28:21 asciilifeform: BingoBoingo: the chinese can show up with the slave galleons and start loadin' whenever they feel like. ( and it is a deep enigma, far above asciilifeform's paygrade, to explain wtf they're waiting for )
snsabot: Logged on 2019-08-29 21:22:37 asciilifeform: atm i have a distinct unpleasant feeling that the ratchet presently does not ratchet very well.
mp_en_viaje: i was here many times before (though not in the past... well, 20 or so yeaes) but i never knew this is what i wanted to make it complete!
snsabot: Logged on 2019-08-29 21:09:24 asciilifeform: diana_coman: you'd need to bake pg glue for ada, there aint any
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932043 << as far as we know this is probably needed in the general.
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932041 << i'm sure there's a library/tool for this/that.
snsabot: Logged on 2019-08-29 21:02:32 diana_coman: asciilifeform: so then what, I'm still better off writing the bot in C or what?
snsabot: Logged on 2019-08-29 20:59:22 asciilifeform: diana_coman: i vaguely suspect that this is the thing that discouraged past folx who made loggers, from genesising. 'dafuq is the point, likely no one can even replicate this'
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932035 << possibly the case indeed ; but we gotta start actually documenting it with a view to ever fixing it, there's no way the fuck out alresfy.
mp_en_viaje: you can build ok from gentoo work if in there, but it won't work in centos world and vice0versa, "version number" irrespective.
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932021 << she has a point i nthat "version numbers" are utterly meaningless (which is why one's 2 something, the other 0 sometiing, yet used in same pile). all that matters is vintage, like in wines, DOC for code.
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931978 << pretty much thast was the idea yeah. iirc we even discussed tcl in ancient thread re this same matter
mp_en_viaje: trying to maybe gwt out of it, hence the work to spec "blog" in the abstract ; but mp-wp stack works for trilema TODAY ; i ain't taking a break from publishing untul "better world". god knows the week long #trilema outage was long enough.
snsabot: Logged on 2019-08-29 18:38:11 asciilifeform: ( mp is fond of old php , but it suffers from 100% of same headaches , and in fact even slower , esp. if the job is even slightly moar complicated than wp ; he arrived at it same way i did at 'flask' -- picked up decade ago and 'it worx, i'ma pour cement on this' )
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931963 << pretty sure this was in the og alread, lolz. poor guy, wth is h to do. gotta use somethign
snsabot: Logged on 2019-08-29 18:09:05 diana_coman: centos 6; because gentoo problem as you just described + the poor moldavian guys anyway pretty much @kukuruz
mp_en_viaje: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931937 << keks. you should see the poor romanian guys of sibiu. pretense to cntrary notwithstanding, @sameplace.
snsabot: Logged on 2019-08-29 13:10:29 BingoBoingo: asciilifeform: I'm working on organizing a break down of the situation. It's a very map-heavy exercise
mp_en_viaje: anyone wanna guess the make ?
mp_en_viaje: hi all from the comfortable insides of florimund, my european mobile mping station.
spyked: diana_coman, yeah, well I suppose that's another reason why the manifest mechanism is useful: if I signed a patch P at t1, and later I found out that it does something stupid or subtly malicious or whatever; then at t2 I can sign the "inverse" to P and give context to that signature in the manifest (whoever tries to press that has to know why particularly I made some change)
snsabot: Logged on 2019-08-29 22:50:59 lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932021 << indeed, testbed on trinquean cuntoo. Yeah, did not encounter same horrors (besides the need for commenting out the cache thing)
diana_coman: ofc there's alway 4. give up on it at this time
diana_coman: re logger atm I am undecided as my options so far seem to be: 1. do another round of madness with flask until it works on this old (but stable at least) centos 2. replicate environment aka burn down centos and have fun installing remotely on the machine cuntoo 3. simply run irssi (as I'm otherwise running this code anyway as my client) with logging to db into an mp-wp database and be done with it (possibly each line a one comment - will end
diana_coman: spyked: there's no way to undo what one did, yes; make amends maybe, undo no.
spyked: perhaps obvious) fact that the signature says something about the signatory *at the time* when he/she has signed, not necessarily at the time that I'm using the object in question.
spyked: re. http://logs.nosuchlabs.com/log/trilema/2019-08-25#1931011 <-- /me has been thinking a bit about this after reading the discussion, and it occurs to me that one of the reasons that might keep people from signing items is that there's no way to "unsign" them. nor that is desirable imho, but "what if" some sign piece of coad turns out to burn your house to the ground. in any case, this reveals to me the (
spyked: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931955 <-- recall, /me already wrote a toy prototype as his introduction to ada. other than the fact that it looks noobish and requires work to integrate with the c-isms, writing logger on top of it would also require something along the lines of a. mod_adalisp and b. handling other wwwistic s
spyked: and the only *quick* solution that I can think of here is signing the binturd that I've been using for the last few years; if anyone has a better idea, pl0x to chime in
snsabot: Logged on 2019-08-23 05:47:14 mircea_popescu: this understanding is current as of cca 2016. meanwhile we agreed that because a) it is preferrable to work with republican rather than imperial items and to prevent more imperial seepage than needed ; and because b) there's no limit to signature count as per long standing observations and discussions (with a very early asciilifeform cca 2013 maybe) then therefore the correct approach is to sign things early, to get them i
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932106 << I agree this is probably the proper behavior. But it ain't changing anytime soon. The heathen coad lobbesbot runs on [i.e. it doesn't sit on the logbot tree like auctionbot] is slated to be discontinued just as soon as I get some more important things complete. Not sinking any more time trying to staple that dead horse
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1931971 << not yet; I have a vintage php search for the classic #e logs, but that is all atm
snsabot: Logged on 2019-08-29 20:54:34 asciilifeform: i recall lobbes was using some kinda system for sucking down py libs that worked through the deps chain, but can't recall what was
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932030 << on the testbed I got both flask and psycopg2 via portage
lobbes: For clarity, on testbed I got the eater.py and the reader.py working fine. Was able to view loglines and search with no issue (on localhost only, didn't test port forwarding). Did *not* test the bot.py, however. For reference, my version info: Flask 0.12.2; psycopg2 2.7.4; Python 2.7.15
lobbes: http://logs.nosuchlabs.com/log/trilema/2019-08-29#1932021 << indeed, testbed on trinquean cuntoo. Yeah, did not encounter same horrors (besides the need for commenting out the cache thing)
asciilifeform fwiw suspects that most ada proggies dun even use preemptible threads -- they spawn N threads and they run until the warhead detonates..
asciilifeform: the vendor's heat sink is pretty spartan, but worx, and snaps in with bare hands, convenient.
asciilifeform: it runs rather hot when 'naked'
asciilifeform: dun fughet the heat sink
asciilifeform: no storages on pcb (is major reason why i picked it) , needs sd to boot from and then usb3 stick to mount root
trinque: link to the exact model? and I'll stand up my crossdev stack for it