log☇︎
600+ entries in 0.044s
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
asciilifeform: diana_coman: in re pythonisms -- for that matter i'm to ~this day~ ! using my orig v.py (w / phf's 'vdiff' , and your fixup patch.) also because never found in what to rewrite it.
asciilifeform: ( considering, it is in py , interpreted lang, and my pages are generated approx on par with speed of phf's , at the immediate shitting arse end (discounting pipe delay ) )
snsabot: Logged on 2019-08-27 11:35:18 asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-27#1931353 << phf yer missing the point , i need so that http://btcbase.org/log/2014-07-16#758070 AND http://btcbase.org/log/1945-07-16#758070 go to same thing !
asciilifeform: for thrd-completeness, also painfully missing are back/fwd linkages a la phf, but these need working date/era handling before i even dare to touch the q
snsabot: Logged on 2019-08-27 04:13:06 phf: fixing an old style link is then cheap e.g. for http://log.bitcoin-assets.com/?date=16-07-2014#758070 i parse out 758070, (entryref 758070), then render the url (which is standard url rendering function) which uses whatever date that's stored in the message record, http://btcbase.org/log/2014-07-16#758070 in this particular case they happen to be the same, but the point is that this is not a regex transform. essentially i
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-27#1931353 << phf yer missing the point , i need so that http://btcbase.org/log/2014-07-16#758070 AND http://btcbase.org/log/1945-07-16#758070 go to same thing ! ☝︎
mircea_popescu: phf, wouldja take it to #asciilifeform or whatever already.
a111: Logged on 2019-08-27 02:20 asciilifeform: atm i also suspect that a certain % of links imported from 'era 2' (phf's april 2016 -- aug. 2019) also do not go where expected, on acct of clock difference b/w the boxen.
asciilifeform would like to hear how phf dealt with this .
asciilifeform: ( the option 'hey just sync clock' dun do anyffin useful. for one thing, kako's clock was yet different. for another, both kako's and phf's drifted , just like mine... )
asciilifeform: atm i also suspect that a certain % of links imported from 'era 2' (phf's april 2016 -- aug. 2019) also do not go where expected, on acct of clock difference b/w the boxen. ☟︎
asciilifeform: meanwhile : phf's vintage log imports on testbed. however : certain uniturds decode strangely ( e.g. ru-style double-quote turns into accented A ) ; and link-dekakoizer will need db lookup to determine snsabot-correct date for given index # (else the links do not go where expected) .
asciilifeform: oh neato, ty phf.
asciilifeform: if asciilifeform ever figures out how to get a 6+digit usd there, will move in right next to phf .
asciilifeform: phf: where are you these days ? in ru still ?
asciilifeform: phf: i dun think you ever said, what was it that got you ? stray bullet ?
asciilifeform: phf: much appreciated. and happy to see that you've your strength back .
asciilifeform: phf: i'ma swallow yours and invite anyone who thinks he has gap fillers to compare , then.
asciilifeform: phf: oh hm loox like you actually have 2013 -- present in there ? oddly enuff when i thought you were dead and wget -m ... 'd the site, only saw 2016--present in /raw
asciilifeform: phf: where's 2014-15 ?
asciilifeform: ohai phf. ty.
asciilifeform: diana_coman: re older logs -- we'll sync these when mircea_popescu comes home to where he had the archive he had fed to phf in '16 . (or perhaps earlier, if phf comes back from grave..)
asciilifeform: lemme find in phf's , 1s
asciilifeform: iirc phf didn't bother with indexing, simply kept all of log in memory and did equiv. of grep. but this only scales to a point.
mircea_popescu: trinque, all of this is quite reasonable. what worries me, specifically, is the (by now sadly well documented) process whereby danielpbarron / mod6 / phf / etc 1. stop talking "because busy working", then 2. stop reading and then 3 "inexplicably" start "having problems", all very "varied" and "impredictable" subjectively, but from outside all very fucking transparently the same one thing.
mircea_popescu: the actual reality of the matter is this : that nicole could very well http://trilema.com/2018/meet-miss-piggy/ and yet her cunt didn't fall off ; eric could very well http://blog.lobbesblog.com/2019/08/on-relationships-or-how-pretense-kills/ and yet his cock didn't fall off ; but phf couldn't, and is butthole DID fall off.
mircea_popescu: one failure mode's not particularly better than the other ; nor is failure this magical and "nobody could have predicted"-y thing. your neglected blog's exactly like phf's neglected blog, and it drives the same sort of stupidity in both of you, for the same reasons, in the same ways.
trinque: phf: what the hell's the matter with you anyway, shake it off.
trinque: on the subj of phf, I suppose I'm just butthurt because warring over the null set is perhaps one of my fondest memories of this hall.
mircea_popescu: yes, but the idea is that "let whoever signed the genesis evaluate your patches, rather than do it for them through the venue of keeping them phf'd"
snsabot: Logged on 2019-08-22 21:51:18 trinque: bringing things in closer, apparently phf's in the hospital or something, and is thus persona non grata?
mircea_popescu: "Needless to say, I am unamused ; and, to answer the original inquiry in firmer terms containing no ifs or buts : no, I personally have no further interest in hearing what phf may have to say on any topic. The time for "ok then, I will get my logger to spec by X date and hope to have my blog up by Y date" came and went, sometime yesterday.
snsabot: Logged on 2019-08-22 21:51:18 trinque: bringing things in closer, apparently phf's in the hospital or something, and is thus persona non grata?
trinque: bringing things in closer, apparently phf's in the hospital or something, and is thus persona non grata?
snsabot: Logged on 2016-12-30 13:34:00 phf: true, but alf also "won't touch the web"
asciilifeform: phf's item ftr lives also in usa, somewhere in nyc region
asciilifeform: ( spoiler : rubbish-mirror actually loads slightly faster, from europistan, than phf's orig., for 'heavy page' test )
asciilifeform: ^ the log bot itself, for instance, thus far has only reconnected when asciilifeform manually cycled it, to update to new ver (last such time, to put phf's uniturd pill in production)
asciilifeform: lobbes: possibly it can even be used to bake a phf-style refill/sync system that eats znc.
lobbes: yes! phf's algo worx again
lobbes: In logotron testing news, (as probably expected) eat_dump.py still barfing on uniturd eating. Studying phf's algo to see if I can get it working
asciilifeform: lobbes: over time there was series of small changes to that awk, to work around various congenital defects of gnupatch; ended up with phf et al writing entirely new replacement proggy, which is the current vdiff.
lobbes: asciilifeform: with phf's uniturd fix in place, would you like me to repatch the following (with updated MANIFEST.TXT): http://blog.lobbesblog.com/2019/08/znc2tmsr-vpatch/ ?
asciilifeform has been digging through phf's output to see how he wrapped, initially i copied his stylesheet but had to break it to get 'checkerboard'
asciilifeform: trinque: 'as little as can' was my design principle, is how wrote ~400ln bot (and almost correctly -- did require phf's uniturd fix)
snsabot: Logged on 2019-08-09 08:33:52 phf: http://logs.nosuchlabs.com/log/trilema/2019-08-08#1926224 << perhaps quarter dead, but days go to going to hospital for daily wounds dressing, sleeping a lot and going for walks, i have another sugery on monday to close me up.
asciilifeform: ty phf , mp_en_viaje, lobbes .
asciilifeform: ty for reading the paste of bot, phf ; evidently not 1 other soul read!
asciilifeform: phf: i see blank line in both client and log
asciilifeform: phf-type (corrected) bot is live.
asciilifeform: phf: ty; appears to be the reason for the barf seen vs. lobbes's collection of barfolade. nao need to reproduce mp_en_viaje's
asciilifeform: the latin-1 fallbacks, per phf's algo, re-encode happily, python per se dun barf. postgres barfs when fed the result.
asciilifeform: phf: this -- worx . this is not the problem. problem happens when the re-encoded string is fed to postgres.
snsabot: Logged on 2019-08-15 20:52:35 asciilifeform: current variant (w/ phf's non-curative decode-fallback).
asciilifeform: the phf-style variant is currently sitting in #asciilifeform-test , btw, and can be diddled , put turd there and try to search for it, if it's in the test bot's log, search will return correct res
asciilifeform: phf: post plz how tested and which py.
snsabot: Logged on 2019-08-15 20:31:27 asciilifeform: meanwhile, update re unitardation -- phf's pill doesn't cure -- decoder successfully falls back to 'latin-1', if set up as described, and dun barf ; but ~encoder~ barfs in pg ( cur.execute(query, args) )
asciilifeform: phf: the issue seems to be postgres , i.e. mine is set to utf8 , so anything that is to sit down in it gotta validly re-encode to utf8
snsabot: Logged on 2019-08-16 09:52:44 asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-16#1929071 << the interesting bit is that i got 0 barf when importing phf's log (i.e. including all of the historic uniturds seen in log in 'era 2')
asciilifeform: ohai phf
a111: Logged on 2016-09-21 14:45 phf: mircea_popescu's xchat specifically still does that. so his latin-1 encodable text comes out as such, otherwise it gets promoted to utf-8
a111: Logged on 2016-09-21 14:44 phf: but not necessarily your special encoding. so client will guess the encoding of the line you're sending, and if it can be encoded as latin-1 it'll send it as such, otherwise utf-8
asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-16#1929071 << the interesting bit is that i got 0 barf when importing phf's log (i.e. including all of the historic uniturds seen in log in 'era 2')
asciilifeform: grrr apparently can't reproduce by pasting from phf's log, cuz -- surprise!111 -- it all turns into valid utf8 when displayed there
asciilifeform: however, the bot that is here , does ~not~ have the phf pill, that one is in #asciilifeform-test
a111: Logged on 2019-08-09 08:39 mircea_popescu: what can you do, alf just not as good at computing as phf :D
mp_en_viaje: now that we're talking about this, i recall a similar (if more restrained) experiment with phf, maybe cca late 2017, and he went away and came back with a properly wroking logger a week or so later -- though i dunno that he ever wrote a blog article detailing the problem and its solution
snsabot: Logged on 2019-08-15 20:31:27 asciilifeform: meanwhile, update re unitardation -- phf's pill doesn't cure -- decoder successfully falls back to 'latin-1', if set up as described, and dun barf ; but ~encoder~ barfs in pg ( cur.execute(query, args) )
asciilifeform: current variant (w/ phf's non-curative decode-fallback).
snsabot: Logged on 2019-08-14 07:42:01 phf: i even explain how to do it in python http://btcbase.org/log/2016-09-21#1546421
asciilifeform: meanwhile, update re unitardation -- phf's pill doesn't cure -- decoder successfully falls back to 'latin-1', if set up as described, and dun barf ; but ~encoder~ barfs in pg ( cur.execute(query, args) )
asciilifeform: i vaguely suspected that phf's proggy exists as currently-unreproducible (i.e. not fully represented by the source) 'save-lisp-and-die' ball
asciilifeform: phf: aha
asciilifeform: phf: my logbot genesis is some of the ugliest piece of shit yet put through 'v' . so dun be afraid to publish similar.
asciilifeform: phf: at the risk of 'broken record', i gotta say that all of this would be 9000x moar helpful if existed as genesis , rather than asciilifeform to sit and derive from old logs and phf's summary, how was done.
asciilifeform: right, assumed that phf meant 'znc' in abstract, i.e. any proggy that logs straight to txt , rather than concretely znc
asciilifeform: wb phf . just 1 znc, or multiples ?
asciilifeform: i'ma fix the uni-idjicy found by lobbes ( and apparently described in log by phf 2+ y ago ) , and add the 'did i join or am i pissing into wind' detector discussed 2d ago, to bot
snsabot: Logged on 2019-08-14 13:42:26 mp_en_viaje: asciilifeform, yes, but hard to like-phf when actually communicating things, on one hand. and on the other, hopefully multiple of these.
snsabot: Logged on 2019-08-14 13:41:36 asciilifeform: i'd like to avoid the situation where ' asciilifeform fixed on best-effort basis, and 1 time breaks leg' and then like phf.
mp_en_viaje: asciilifeform, yes, but hard to like-phf when actually communicating things, on one hand. and on the other, hopefully multiple of these.
asciilifeform: i'd like to avoid the situation where ' asciilifeform fixed on best-effort basis, and 1 time breaks leg' and then like phf.
asciilifeform: mp_en_viaje: i'm still waiting to hear from e.g. ben_vulpes re what he did when bot fell down, to sync the indices with phf's
asciilifeform: if anyone thought 'hey asciilifeform can do in 5 days what phf did in 1yr' -- this is naive.
asciilifeform: bot ate 100% of phf's log in testing phase w/out losing lines. but lobbes in fact discovered uniturds that break py's eater. i'ma implement the fix described by phf , but the q will remain re what, if anyffin, to do re the in-between.
asciilifeform: ( i.e. mine is behind by -- of last count -- 6 , vs phf's )
asciilifeform: ty phf , i'ma try said algo
a111: Logged on 2016-09-21 14:49 phf: but correct way of handling this in python, is to read() line by line, which returns whatever, ascii, then parse the [date] <nick> bits as ascii, then take the rest of line and try: msg.decode('utf-8') except: msg.decode('latin-1')
snsabot: Logged on 2019-08-14 01:00:50 asciilifeform: phf's bot, incidentally, ate. now would be a notbad time for the fella to come the hell out of his coma and say how.
a111: Logged on 2016-09-21 14:44 phf: http://btcbase.org/log/2016-09-21#1546322 << this behavior has nothing to do with znc (unless you explicitly configured znc to recode your messages), it has to do with the fact that irc is encoding agnostic. used to be you had koi8-r only channels, or latin-1 only channels etc. now utf-8 is the "standard" but from back in those days you have a peculiar hack surviging in a lot of clients. everyone's expected to be able to read latin-1,
asciilifeform: phf's bot, incidentally, ate. now would be a notbad time for the fella to come the hell out of his coma and say how. ☝︎
asciilifeform: ( from ~output~ of phf's , that is, to this day nobody knows how it birthed this output )
asciilifeform: to the extent that the renderer worx, is because lifted from phf's (and massaged empirically)
asciilifeform: iirc phf struggled with something not unlike this, when wrote his CL proggy
asciilifeform: and that phf did not use these, but hung up his logger straight on port80
asciilifeform: and now phf's log -- of 'heavy' pg -- via same. pretty interesting. total load time (to the tester author's box) -- 0.646s ; 1st byte -- 0.254s .
snsabot: Logged on 2019-08-12 16:44:50 bvt: asciilifeform: there is another thing that i just noticed that different between phf's logger and yours: single quote (') is encoded as &#039 in phf's logger, and as &apos in yours -- and &apos happens to be totally broken in links graphical mode (links -g). i can fix this on the links side, but i wonder if this is intended/acceptable behavior
bvt: well when i was re-signing another vtools vpatch at phf's request (for keccak regrind), i did the full re-testing as when releasing it the first time, to make sure that everything is ok.
bvt: asciilifeform: there is another thing that i just noticed that different between phf's logger and yours: single quote (') is encoded as &#039 in phf's logger, and as &apos in yours -- and &apos happens to be totally broken in links graphical mode (links -g). i can fix this on the links side, but i wonder if this is intended/acceptable behavior