log☇︎
24500+ entries in 0.015s
snsabot: Logged on 2019-08-09 13:08:11 asciilifeform: btw re ddos -- mircea_popescu i think yer site also is being dos'd
mircea_popescu: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926886 << there was usage spike past few days, nfi. became popular somwhere
mircea_popescu: prolly take years for it to be surpassed, too. i guess there's that.
snsabot: Logged on 2019-08-09 12:54:00 asciilifeform: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926724 << all laffs aside, he had the seriously better proggy (and with 3y+ of massage, too)
mircea_popescu: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926867 << yeah, i agree, so far just about the record holder for "oh, but he showed so much promise" trope.
mircea_popescu: the extreme poverty of the contemporary ustard is only comparable to the extreme retardation of same.
mircea_popescu: "why buy oak chairs when you aren't going to do dishes ?"
mircea_popescu earlier stopped at gelateria. nicely furnished. had no table service and only paper cups. ONE SIZE. no plates, no nothing.
asciilifeform: not long ago asciilifeform went to bazar, and there were folx selling coffee. but instead of the usual signage saying what is name of the co, and from where coffee, there was... qr code. 'hey dontcha get it, aim pnoje camera here and will see little film clip of our spokesman..'
mircea_popescu: well, pretty much the only way they'll see any shops at all, soon enough
asciilifeform: i fully expect that reich will at some pt give up on maintaining street signage, and try to push ~this~ as 'replacement'
asciilifeform: mircea_popescu: funnily enuff, 1 of the more recent idjit fads in pnojeism etc is just this, 'augmented reality', where user is to try an' look at e.g. shops, roads, through peephole of pnoje, to see imaginary 'signs' imposed on the picture
mircea_popescu: sounds kinda fucking stupid and in any case more expensive than just painting some sheet metal and piking it in the ground by the roadside. but then again if mechanical engineering worked like software "engineering"...
mircea_popescu: and then what, have legions of gnomes running around painting/erasing that stop sign in correct perspectives at 20fps ?
snsabot: Logged on 2019-08-09 12:46:32 asciilifeform: the 'pc micro revolution' popularized an imho extremely braindamaged concept of what is 'comp lang'. specifically, where folx think that the front syntax is somehow logically glued to the internal mechanics ( whereas is merely happenstance, pc provides such broken bedrock abstractions that forces erry new lang author to write compiler 'from 0' as if year were 1959, but this is merely happenstance, rather than lo
mircea_popescu: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926862 << this is actually important point. the notion of "compiler", as ~different from os~ and somehow part of "lang" is not unlike the notion of road signage as differen tyfrom road and somehow part of all windshields and rearview mirrors.
mircea_popescu: im pretty sure every single item i nthat pic was there 25 years ago also.
bvt: asciilifeform: fished out some info re aws: http://p.bvulpes.com/pastes/koCFa/?raw=true (small patch for musl compat) and http://p.bvulpes.com/pastes/avKyY/?raw=true (build instructions) seem to be it
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.
BingoBoingo: mircea_popescu: !!key returns one for the fellow
snsabot: Logged on 2019-08-09 09:21:00 lobbesbot: BingoBoingo: The operation succeeded.
mircea_popescu: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926791 << for my cyruosity why did this go to trinque lol ? user dun have the key he reg'd ?
snsabot: Logged on 2019-08-08 22:24:24 asciilifeform: expects that before long, we'll find out that the pythonisms can't in fact be reliably reproduced, because 'ecosystem updated' or whateverthefuck rots
snsabot: Logged on 2019-08-09 09:14:38 PeterL: http://btcbase.org/log/2019-08-09#1926545 << python has the advantage that development happens on python 3, so python 2 has become "depcricated" (aka stable)
mircea_popescu: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926784 << yes, until you discover you can no longer get it to run.
asciilifeform: bvt: might be worth writing about how did this feat
bvt: iirc it is buildable without xmlada and all the crazy deps with some tweaks; at least i did build it once and run one of the examples
asciilifeform: bvt: it was 1 of the items i tried to build. no dice.
mircea_popescu: asciilifeform, nah, that was exact translation.
snsabot: Logged on 2019-08-09 04:45:56 spyked: also, I guess there's also the fact that despite all its problems, there's a pile of CL code (e.g. hunchentoot, cl-irc) that works... very similarly to how wordpress does the job, despite its size and other warts. maybe Ada could also find similar pieces of code for Ada, but tbh I haven't looked... yet?
bvt: http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926692 << there is ~20 years old 'AWS' https://github.com/adacore/aws
asciilifeform: it is possible that i dreamed it
mircea_popescu: bvt will have to believe this was also a ro piece years ago
mircea_popescu: asciilifeform, fuck, this is exact translation
mircea_popescu: i think it was more like 2013
asciilifeform: bvt: i think i read it in an earlier, ro piece. that grr cannot just nao find.
asciilifeform: ( 'magic' tends to end up in vehehery short supply, in wartime )
asciilifeform: ru army philosophy (inherited from sovok) holds that 'magic' is poison.
mircea_popescu: bvt, just about. btw, if you feel like reimplementing that use the hash-seed method
asciilifeform: iirc mircea_popescu had this vintage piece, re 'technology vs magic' . where the former is something you can meaningfully take apart and repair, making new piece with lathe if you must; whereas the latter... can only pray
asciilifeform: telex & friends have substantial wins over 'modern' -- broken telex can be repaired by sweaty d00d with screwdriver, and out of improvised parts. whereas what's it even mean to 'repair' e.g. pc, esp. in condition where the parts not even made outside of taiwan.
snsabot: Logged on 2019-08-08 18:58:36 mircea_popescu: asciilifeform, did the fg / kernel thing ever make sense to you ?
bvt: http://logs.nosuchlabs.com/log/trilema/2019-08-08#1926500 << i think get your point; though tbh, from my reading of linux it's not clear that urandom uses separate entropy pool, as i understood so far urandom uses the same pool as random, just ignores all 'entropy' measures (i still did not quite load that part in head, so this is not final info).
asciilifeform once saw a rare photoreportage of ru army hq and on what gear runs. 100% telex, made in ddr. )
asciilifeform: ( an' when output is to be made hardcopy anyway, to give to elderly general, why not telex, even ! )
mircea_popescu: early ro stock trading in the 90s used old commie era telexes. it was quite exciting
mircea_popescu: kinda reminded me of ye olde telex days
asciilifeform: mircea_popescu: recall the orig. pythonistic backend of phuctor, btw
asciilifeform: bvt: it's a headache that'll be around for good long time, hence worth to discuss
bvt: asciilifeform: until the day for tcp to die arrives, you'd still have to interact with tcp warts in that or other form
mircea_popescu: lol this logger's like ye olde teletype machines, feeds lines palpably
asciilifeform: problem isn't 'no way to expose', they're perfectly exposed, e.g. 'M' uses 100% raw syscalls for all i/o . problem is the fundamentally ill-conceived abstraction, where you don't ~only~ have the stream, but also a barely-known and large set of possible out-of-band eggog conditions
bvt: on posix, don't think there is way out of exposing fds and syscalls
asciilifeform: my 'udp' lib was orig. gonna be a 'udp and tcp' lib. but very quickly realized...
asciilifeform: could try to use ada's 'streams' model. but then must decide, how to represent ~all~ of the possible tcp hiccups.
asciilifeform: aand that's just pg. now consider how to deal with tcp.
bvt: re pg - yes, you'd have to implement the protocol; and serialization/deserialization is error prone and typically takes lots of code (i.e. too much), agreed
asciilifeform: sorta what i suggested to spyked at one pt
asciilifeform: i suspect the glue for pg alone (i.e. if pg disgorges eggog, to produce proper ada-compat. exception, rather than crash), not to mention utf8ism, would weigh like ten ffa's
bvt: true, but i don't find '80% of cl argument' too convincing; if want comfort, sure, use cl/python; want hard memory limits and gcc performance, can use ada, it won't be fundamentally dirtier (due to tcp and db stuff), just more boilerplate code
asciilifeform: ftr 95% of asciilifeform's sweat to make snsabot + logotron , went into debugging regexpolade.
asciilifeform: re general topic of 'string munging problems' -- they are plentiful and prolly inescapable, and i suspect there could be win from reviving an item like snobol (where ~sane~ abstractions specifically for stringism, rather than seas of regexp)
asciilifeform: ( tcp?! )
asciilifeform: and how else wouldja propose to talk to it.
asciilifeform: bvt: the c glue defo uses
bvt: i would actually expect that pg protocol does not use 0terminated strings. re 80% of CL -- inside of it's implementation you'd find same shit. dunno how it would be different from using heathen libs
asciilifeform: it is doable, and e.g. bolix's device driver coad was written that way. but is deep magic.
asciilifeform: this is the 'flip side of the medal' -- sometimes the given abstractions are not a good fit for $problem. picture e.g. attempt to ffa in commonlisp, where you'd have to bend over backwards to make sure you ~never~ cons
asciilifeform: nao if you throw out pg and sqlism, and make mmap-based db from 0... then possibly. but again goes back to '80% reimplementation of commonlisp' territory.
asciilifeform: but i strongly suspect that 'ada logtron' will consist mostly of non-humanreadable hex constants , and the pg glue will be eating null-termed c strings , and generally the result will not be a thing you 'would fly in it? i wouldn't walk under it!'(tm)(r)(h. hughes)
asciilifeform: damn near anyffin 'can be done'. full sized eiffel's tower from toothpicks, ~could~ be built.
bvt: i don't think there is a way out of treating utf8ism as raw bytes, other than finding a heathen library
snsabot: Logged on 2019-08-09 12:37:27 asciilifeform: i strongly suspect that 'logtron in ada' would inescapably contain the proverbial 'buggy 80% reimplementation of common lisp'(tm)
asciilifeform: my point , again, is not that 'cannot be done' , but that http://logs.nosuchlabs.com/log/trilema/2019-08-09#1926853 .
bvt: i did some minor nginx plugin development -- the linked list approach was not bad, the only op i had to do with actual buffer was splitting it across chains links to insert data between.
asciilifeform: and would not even be able to "<a>" + whatever + "</a>" etc, i.e. can't use the existing sugar for string constants at all
asciilifeform: ( afaik you cannot use ~any~ of the existing string ops for these... would have to handle as naked bytes )
asciilifeform: bvt: i do not presently know what would have to be done to use utf8istic strings in gnat, either
asciilifeform: there's an (ungenesis'd, cuz tripped on gnat bugginess) mmap glue also
asciilifeform: cuz the ancient ones in fact ~do not build~ anywhere that i could get hands on.
asciilifeform: i think i mentioned, considered writing one. but then realized that will have to write tcpism and postgresism glue from 0.
bvt: re logotron with arbitrary number of messages - can't you send data in a loop? otherwise the problem touches all levels of net stack -- can't have arbitrary sized packets either
asciilifeform: this is not to say that 'impossible to logotron in ada', but that result will be 1) gargantuan 2) ~unreadable 3) likely buggy despite herculean effort
asciilifeform: so imho lost cause to try and calculate max size of page , even in such simple item as logotron
bvt: well, for most of format strings you know the number of formatted elements, right. re memory allocator -- an arena allocator is typically used, and file data can be mmaped
asciilifeform: not to mention what is upper bound for # of msgs in day ?
asciilifeform: i suppose it is theoretically possible to calculate upper bound... irc msg is only N chars max; and if full of nuffin but <a></a><a>...... but who can do this and guarantee no mistake ?
asciilifeform: don't tell me, 'in heap'
bvt: you still need to format the numbers, etc, but for this can always know upper bound
asciilifeform: bvt: imho it's just barely usable, and would defo benefit from rewrite in compiled lang where can turn knobs in realtime (i.e. cl)
bvt: asciilifeform: you make a chain (linked list of buffers with content), pass it's head to kernel
snsabot: Logged on 2019-08-08 17:05:07 asciilifeform: snsabot and its www end are imho usably snappy, but you can tell that they run in interpreter, visibly slower than phf's
bvt: http://logs.nosuchlabs.com/log/trilema/2019-08-08#1926442 << the lag is noticable, but i'd say it is entirely usable.
asciilifeform: or do i misunderstand 'don't need to concatenate'
asciilifeform: bvt: how do i throw into a tcp socket a formatted fetch of log, consisting of unknown length of "<a ... " + blah + "</a>" etc, w/out string munging ?
bvt: asciilifeform: did not try myself
asciilifeform: ( cuz -- naturally -- wants dozen or so deps, each 1 weighing moar than sum total of what asciilifeform ever wrote, lol , and consisting of ??? )
snsabot: Logged on 2019-08-08 17:02:10 asciilifeform: diana_coman: principal weakness of ada in re this problem set (at least the ada 'specified' by asciilifeform) is stringism handling
bvt: http://logs.nosuchlabs.com/log/trilema/2019-08-08#1926433 << you don't actually need to concatenate anything as long as writev(3) is there. whether any gnat lib uses it -- dunno. at least nginx does use it (http://archive.is/QKjvD#selection-2735.36-2861.26), and imo this is a correct approach to the problem -- let kernel do the copying, if it needs to
asciilifeform: bvt: funnily enuff, i actually tried adacore's -- wouldn't build