Framedragger: hmm, certainly not. do you see the ip?
asciilifeform: Framedragger: unrelatedly is it you who has the 1/minute wget on phuctor ?
asciilifeform: registered by a certain van der Wansem of 'btcsoccer' ( https://bitcointalk.org/index.php?topic=544115.0 ? https://github.com/tomasvdw/btcsoccer ? or some other, unrelated crackpottery ? or plain disinfo..? )
asciilifeform: the glossy www, for instance, is new -- domain bought on apr. 1 of '17
asciilifeform: Framedragger: tbh i don't actually give enough of a shit to try to cut apart the ball of yarn in shitlang -- what part stolen, when, from which logs ( our discussion of block/tx storage has been running for ages )
Framedragger: (also, my mp-emulator is saying "why do you spend energy challenging these and not challenging the empire", and my response is that i don't give the privilege to the empire of even *challenging* their shit; they're not worth it imho. this all just to say that take it easy)
Framedragger: you could of course say that 'git doesnt prove anything', and i'll have to agree. but if you allow that, then it's reasonable to say that 'unsigned irc logs dont prove anything, either', from the point of view of the counterparty. i won't say more, only that this 'thief!!' is very unconvincing
Framedragger: from what i take it, 'spent tree' is the relevant notion here. tell me if i'm wrong tho (very much could be). but if i'm not, then - 22 dec 2016: https://github.com/tomasvdw/bitcrust/commit/ac4f4eb12c7a202a3971f782f92b741431ff75ed (and commits thereafter).
Framedragger: i see that part of the related turd is here (5 march): https://github.com/tomasvdw/bitcrust/commit/01001d9f6033426624fba570da437a0a316993c0
Framedragger: do you recall when you described the storage of nqb?
asciilifeform: what part of the linked item verifiably existed in aug '16 ?
Framedragger: did you describe the idea before, say, august 2016?
a111: Logged on 2014-08-13 14:40 asciilifeform: 'They tend to show up late at parties because they figure they can always steal the cake anyway, so why bother go early ? Let the suckers figure out first - on their own dime - where the good cakes are, then just swoop in and collect... This means the only way they can get in is if you let them get in. Dont let them get in for cheap - they have no business here.' (mp's http://trilema.com/2014/people-us-dollar
Framedragger: trinque: does `!!deed` work when issued over privmsg? i just did that and then realised that may it's not supposed to work that way?
mircea_popescu: first it was "get a college degree doing no work -- just on the basis of your current experience!11" now it's "draw a salary doing nothing other than recounting the psychogenic hum in your otherwise empty brainbox".
BingoBoingo: And don't even get started on the peacucks
a111: Logged on 2017-04-07 22:15 trinque: by the nato precedent set by turkey, I'd expect russia can down an american plane or two without risking a nuclear exchange, eh?
mircea_popescu: http://btcbase.org/log/2017-04-07#1640387 << ~everyone can do just about anything without risking such a wonder seeing how the shit dun werk.
asciilifeform: https://www.orau.org/ptp/collection/consumer%20products/magazines.htm << in other lulz
pete_dushenski: that's exactly my scheme to bid up the prices of g5s. bwahaha!
shinohai: Now you make me want to buy one of these ppc's pete_dushenski
pete_dushenski: "Well I was a major miner for over two years and I lost out because other people had a lower CoC. I had to say goodbye to a business that I built from the ground up. I’m not calling fowl play and throwing my toys out of the pram. Its called business and its what enables people to invest in bitcoin." ~sam cole (tm). so fowl. so chicken. cluck cluck!
pete_dushenski: https://medium.com/@samcole_74219/asicboost-655a73d48ae4 << more of the same sadness as above. strap double gas mask on for this one. 'sam cole founder of kncminer' is functionally illiterate.
pete_dushenski: in other boosts, asicboost is back as the segwit saviour : http://archive.is/1BYAX (gmaxwell)
trinque: by the nato precedent set by turkey, I'd expect russia can down an american plane or two without risking a nuclear exchange, eh?
trinque: http://www.reuters.com/article/us-mideast-crisis-syria-russia-medvedev-idUSKBN17928I << in other concurrency issues
trinque: mircea_popescu: reads block nobody, never block. I understand the where clause of an insert statement to be a condition on a read query that gets fed into a write, just as my "common table expressions" ('with' keyword) feed into each other.
pete_dushenski: but the replacement is still hush-hush. stay tuned.
shinohai: And selling the mercedes! O.o
mircea_popescu: ie, s5 is NOT waiting to do EITHER its insert or its where check because S3 is doing EITHER its insert or its where check ?
mircea_popescu: suppose S = "insert if not exists where a = b" prototype. suppose S takes time T composed of ti for the insert and tw for the where clause. Suppose thjere's 5 concurrent S presented to db, S1..S5. are you saying that IN NO CASE will Sn wait for a Twi where i != n ?
trinque: if you and I are trying to update the same table your update may block until after mine
mircea_popescu: trinque so an insert if not exists where a = b does NOT block a different select while it checks the where clause ?
trinque: the thing forks the database upon each write and reads happen in the snapshot of state they began
trinque: data produced by concurrent transactions performing updates on the same data rows, providing transaction isolation for each database session. MVCC, by eschewing the locking methodologies of traditional database systems, minimizes lock contention in order to allow for reasonable performance in multiuser environments."
trinque: mircea_popescu: https://www.postgresql.org/docs/current/static/mvcc-intro.html << "Internally, data consistency is maintained by using a multiversion model (Multiversion Concurrency Control, MVCC). This means that each SQL statement sees a snapshot of data (a database version) as it was some time ago, regardless of the current state of the underlying data. This prevents statements from viewing inconsistent
BingoBoingo: <asciilifeform> mircea_popescu: on contemplation, cars may not be a very good hiding place, they ( esp in usa ) routinely move through chokepoints - toll archs, and these are already equipped with camera and radio receivers << Depends on region and car
mircea_popescu: http://btcbase.org/log/2017-04-07#1640078 << are you sure this is equiv to select * where something=something and if nothing is selected insert, ie, does the insert take read locks while it looks for the where ?
asciilifeform: in other noose, swedish lulz.
trinque: reminds me I gotta buy the HD screws for g5 cheese-grater
asciilifeform: well of course they will
ben_vulpes: it's not strictly speaking hung afaict; cursor blinks, newlines and other chars write to term
asciilifeform: i've had it hang on other occasions, never found out why.
asciilifeform: thank the nice folx who created autoconf.
mircea_popescu: the immediate (not even spoken, because they don't want to say it because they don't want it discussed because they want it to be forever true) retort from the empire'd be that "well, this doesn't work". except... it does work.
a111: Logged on 2017-04-07 16:11 Framedragger: alright. i will grant you that the end system wouldn't be "oh so elegantly simple", because you would have to have a submission queue (maybe something that trinque had in mind). user submits key; gets permalink (immediately); meanwhile key gets sent to master (immediately), and master puts it into "to be inserted" queue. under normal loads, the insertion happens ~immediately,
mircea_popescu: http://btcbase.org/log/2017-04-07#1639964 << one of the (~the~ ?) advantage of tmsr work for intelligent folk is that there isn't a braindamaged/choiceless manager in the loop going "i know it's stupid but it must be done".
ben_vulpes: customers complain about random outages, att/verizon/$provider wastes a pile of money doing infrastructure testing and then calls in fcc vans for triangulation rolls
ben_vulpes: bill /for the state/
mircea_popescu: one could also rack up quite a bill paying the wife for blowjobs.
asciilifeform: mircea_popescu: on contemplation, cars may not be a very good hiding place, they ( esp in usa ) routinely move through chokepoints - toll archs, and these are already equipped with camera and radio receivers
asciilifeform: i.e. one where 1) they cost less than the retail model 2) linearly less with quantity
Framedragger: ah the english word is 'successor', forgot
asciilifeform: Framedragger: i never found a bulk souce of the ~old~ one
Framedragger: btw supposedly there's a new iteration/model of esp8266, anyone has an opinion?
mircea_popescu: what, there's a rule pigs don't bleed ?
mircea_popescu: and get shot in the face.
phf: you get sloppy with opsec and that 3 people team that's been drawing paycheck on your assignment for the past 4 years is going to close in
asciilifeform: but they can be deniably planted elsewhere (e.g. roofs)
asciilifeform: phf: aha, if civilian traffic stops -- goodbye to the car transmitters
BingoBoingo: <jurov> coinbr public announcement: there are some stuck orders, and it came at worst possible time. During the weekend I should be able to fix. << Have some thank you cake https://s-media-cache-ak0.pinimg.com/564x/bd/cd/db/bdcddbf71f1966610be20506b082eb61.jpg
mircea_popescu: eminently the sort of thing resisting civillian population can do with little risk to disrupt obama-gestapo.
mircea_popescu: get a network of these thick enough in urban setting, an' farewell to arms.
mircea_popescu: the idea of getting signal picked up/repeated by the old "item attached to someone else's car" is the golden ticket.
asciilifeform: ideally ( and this was in the gps jammer thread ) your transmitter won't be a point source, either -- but a chorus of ~synchronized~ transmitters far apart, in polygonal pattern
phf: well, they didn't print this simple fact in "young electronic" so i wouldn't know, i like the idea of msec bursts though
mircea_popescu: if it "obeys the law" i pointedly don't want it, and FOR THAT REASON.
ben_vulpes: wake me up when they work over shortwave
asciilifeform: phf: all of the linked examples thus far, fit this pattern
phf: i was excited about scuttlebutt for couple of minutes. usually hacks that come out of "digital nomad" circles are at least worth looking at since they have a lot of uncommon requirements built in (must work over 14.4 dialup, etc.). until i realized it's written in javascript by a hacker with "200 packages in npm"
asciilifeform: 'Matrix provides state of the art end-to-end encryption in beta using the Olm and Megolm cryptographic ratchets, and ensuring that only explicitly authorized devices can participate in a conversation. Based on the Double Ratchet Algorithm popularised by Signal...'
ben_vulpes: while we're surveying and lest anyone call us princeton, there is also http://matrix.org/
asciilifeform: tardanoization, ever the favourite idiot gambit.
mircea_popescu: i find wordplay a fine substitute for competent engineering. it betrays such a wordliness in the cunning linguist responsible.
ben_vulpes: there is also 'discord'
mircea_popescu: ben_vulpes i find it's very important to establish aforehand what each barbie doll will wear, do or say at the dollhouse teaparty.
a111: Logged on 2017-04-07 16:05 asciilifeform: which, note, replaces, e.g., sks and other heathen idiocies.
asciilifeform: ( in the minds of the perpetrators, probably the operative historical example is hurd )
asciilifeform: i asked the fella who sent me the link, 'how does this beat irc?' and answered, 'but it has LOGS!'
a111: Logged on 2017-04-07 16:04 asciilifeform: EVERYBODY thinks they're soooo much cleverer than dumb old asciilifeform
mircea_popescu: i would characterise it as "they got confused on their own terminology and backported orphan chains into the story"
Framedragger: i wonder how you'd categorise the 'affect heuristic' as kahneman et al. call it.; i.e., when you make decision with at least partial influence of current emotion. it's very *fast* (thereby falls under kahneman's type 1). ridiculous scenario where it's useful: i'm chased by tiger, see two paths in woods, one has sign 'DANGER' in red, i don't have time to even parse word without making decision,
mircea_popescu: it's iffy, human body comes out of the hackpile, inc. labs
mircea_popescu: trinque yeah, a proper reflex closes in the spine. there's opf course conditioned reflexes which close in the "brain", but not a very conscious part thereof afawk
mircea_popescu: all those shots of people both facing the camera grimacing ? 18month old baby fare.
trinque: comes from the spine doesn't it, or even more locally
mircea_popescu: anyway. heuristics are not involved in watching daytime drama, which is what the type 2 described above is.
Framedragger: kk. yeah i won't argue further without doing some decent mp1 thought here :)
Framedragger: they are mp2, the one with lower ordinal, and my point was that it's still hella useful. can you employ the methodical and analytical mp1 when you touch hand on hot stove? no time to route through brain.
mircea_popescu: Framedragger neither of those fall under mp1.
a111: Logged on 2015-08-24 18:01 mircea_popescu: previous flight lost an engine foil, concorde ran it over, it cut a wheel, wheel exploded, large fragment of tyre hit underwing, shockwave broke a fuel tanker, fuel gushed out and caught, engineer cut the engine next, at which point they had 2km worth of runway left and needed 3 to abort.
mircea_popescu: asciilifeform often happens, which is why people insist they follow manual and procedure, which is what opens the whole thing up to mp going http://btcbase.org/log/2015-08-24#1250375
mircea_popescu: expand on these merits when you will.
mircea_popescu: you propose we also sort planets with earth in center and main sequence stars from the sun onwards ?
mircea_popescu: SORTING is the most important activity of the human brain. and you should see how much trouble well experienced slavegirls encounter when their words are interpreted on a first-priority basis.
mircea_popescu: moreover wtf was he doing he sorted them backwards!
mircea_popescu: if you're wondering, mp doesn't use the reference because the introduction de novo is a paragraph, the discussion of differentiation is a book.
asciilifeform: ( according to the b00k, most folx die in airplane because of ~thinking~ )
mircea_popescu: which is why prussians came up with it. fought a helluva lot of recruit idiocy to arrive to the concept.
mircea_popescu: and to link it to blogging : the point of blogging (daily!) is that it forces an easy and painless transition from type 2 to type 1 in the heads of the practitioners. there's no promise as to the rate of conversion, but then again that's the best you can get for painless.
mircea_popescu: to go back to the machine model : until you're running symbolix, don't pretend like you're computing.
mircea_popescu: as a general rule, people who can't produce genus/difference definitions for all the symbols in their own output are better off not thinking yet, as what they think thinking is isn't.
mircea_popescu: nate the latter ; under the (broadly correct) assumption that a quiet head works and a loud one doesn't. this isn't so different from, eg, buddhism.
mircea_popescu: no, let's also de-equivocate think. there's two kinds of think, one's a forge/reflow/examination of trees resulting in analytical consumption of inputs with actionable outputs guaranteed ; the other is a neurotic behaviour perhaps best described as spinning, whereby specific emotional triggers / detriggers are visited in succession. the prussian model was never concerned with the former in any sense, but merely aimed to elimi
asciilifeform: but imho the logical end of it is proper mechanization.
mircea_popescu: it's important to de-equivocate dumb ; there's two kinds of machine that may be casually referred to as dumb : one that is very slow (eg, z80) / one that is very demanding on the programmer (eg, windows). good programmer wouldn't really call z80 dumb, bad programmer wouldn't really call windows dumb. this directly translates to humans, there's a kind of dumb the incompetent identify mostly disjunct from the kind of dumb the c
asciilifeform: ( wouldn't be the first time )
mircea_popescu: unrelatedly, http://trilema.com/2017/the-lordship-list-fourth-year/ << last call for this, prolly getting done weekend/early next
a111: Logged on 2017-04-07 18:08 Framedragger: mircea_popescu: i thought one of the parts was a c proggy. quite certain it is, asciilifeform mentioned it multiple times. python is the web backend afaik
asciilifeform: http://btcbase.org/log/2017-04-07#1640141 << Framedragger is correct, there is a wwwtronic (and rfc4880-parsing) front end, that is in python; and a wholly separate werker that actually does the number crunching, in c.
mircea_popescu: you may snag some decidability issues (depending exactly how unique id is generated -- remember keys are not unique in this sense, kinda the point of phuctor)\
Framedragger: mircea_popescu: re "nothing here" the idea would be that upon user submitting gpg key, page reloads *instantly* (vs. near-instant page load on current phuctor - reloads only after key inserted into db), shows permalink and "waiting for insertion" msg; html could have property to auto refresh every second until master responds with any links to other keys. but i understand asciilifeform's reasons against this
a111: Logged on 2017-04-07 16:34 asciilifeform: imho to even suggest such a thing, betrays a very serious misunderstanding of the concept. dumb humans are in every respect an inferior version of the machine. the only thing more agonizing than programming comp, is programming dumb humans.
a111: Logged on 2017-04-07 16:33 Framedragger: (i'll just note that the *permalink* can be derived on the slave box, and nowhere did i say something contradicting this)
mircea_popescu: ah yeah there is a part
Framedragger: mircea_popescu: i thought one of the parts was a c proggy. quite certain it is, asciilifeform mentioned it multiple times. python is the web backend afaik
a111: Logged on 2017-04-07 16:18 asciilifeform: understand, if i paste a gpg key into phuctor, and cannot then ~immediately~ link to it in-chan -- phuctor is broken!!
trinque: danielpbarron: seems ok, maybe my tar or gzip has the HIV
trinque: then also danielpbarron this project with ISOs might take rewriting the downloader, will have to look at it
trinque: I'm getting on the previous batch of deeds that I have insufficient funds, which is bullshit.
trinque: got 10s of gigs over there free, nah
asciilifeform: ty for taking the sweat to puzzle over this, trinque .
trinque: does this query look up the mod IDs to append their factor arrays, or does your worker already know those
trinque: the insert statement can return its id to a calling update statement
asciilifeform: for some of whom -- we may already know it, and have it in the indices
asciilifeform: thing is, it also has to update the respective mod's factors indices array
trinque: or thereabouts
trinque: I'm of course shooting in the dark as to what happens when in what you have.
jurov: coinbr public announcement: there are some stuck orders, and it came at worst possible time. During the weekend I should be able to fix.
trinque: the "unique" is just an index which yes eggogs if insert happens and is already present. the eggog is just a matter of how do you want to find out about this
trinque: well one sec, so you're trying to establish relationship between keys and factors in the db, yes?
asciilifeform: is this in point of fact any faster than running a search and ensuring that there is no eggog query submitted ?
asciilifeform: do to the db, i mean.
asciilifeform: factors get created by the werker process (in c, and unable to do anything whatsoever other than to search for whether a particular factor already exists)
trinque: you have a keys table, factors table, factors has a unique constraint on value column of factors, there is a join table called key_factor which joins key and factor
asciilifeform: Framedragger: it means that the user doesn't immediately learn the 'do we know this key?' answer.
Framedragger: last note before i fuck off: the "user submits key; gets permalink (immediately); meanwhile key gets sent to master (immediately), and master puts it into "to be inserted" queue" + "all inserted + updated rows get sent back to slave via streaming replication and pg trigger" doesn't look like petrocheese to me. is all.
asciilifeform: and yes, if you were to throw out the realtime query ability, you could run phuctor as a 'newspaper'.
asciilifeform: this presupposes that the task can be cut.
trinque: you conflate the paper delivery kid with the printing works and the writers and, and, at your own design
asciilifeform: maybe i misunderstood the idea ? what was meant ?
asciilifeform: i do not need 'stable boy' for anything. there is no conceivable room, imho, for any such thing in any of my systems.
asciilifeform: imho to even suggest such a thing, betrays a very serious misunderstanding of the concept. dumb humans are in every respect an inferior version of the machine. the only thing more agonizing than programming comp, is programming dumb humans.
asciilifeform: but meanwhile i also must address the 'stable boy' thing
asciilifeform: trinque: i'ma reread the trinque lines from log.
Framedragger: (i'll just note that the *permalink* can be derived on the slave box, and nowhere did i say something contradicting this)
asciilifeform: this in re the dual-db optimization, yes
asciilifeform: the dual db thing
asciilifeform: i must point out, that i am unwilling to solve it by redefining the problem.
trinque: I find the notion that asciilifeform here is the only guy capable of adhering to own standards lulzy
asciilifeform: the performance wins described in this thread by Framedragger and trinque can only come at the cost of 1) massive increase in complexity, and in particularly state 2) loss of immediate linkability of freshly pasted keys.
Framedragger: asciilifeform: yeah i said some stupid things, apologies for depressing you there
asciilifeform: the apparent willingness of, e.g., Framedragger , to introduce petro-cheese is quite depressing
asciilifeform: given that i found that various otherwise reasonable people are willing to make all kinds of compromises
asciilifeform: at one point i thought that giving it to other folx could be a reasonable thing. now i ~definitely~ do not
asciilifeform: or the cost of getting a 2nd box with five-nines uptime.
asciilifeform: this is not even to mention the complexity cost.
asciilifeform: while retaining what i see as the non-negotiable features.
asciilifeform: i don't presently see them as cleanly cuttable apart
trinque: not let them seep all through
trinque: http://btcbase.org/log/2017-04-07#1639899 << imma seriously start calling this russing over the issue if we get one more of y'all in here doing that.
asciilifeform: and if ANY of these exist previously, incl. from a submission 5 seconds ago -- they gotta be correctly linked
Framedragger: it is updated in ~realtime from the master via streaming replication
asciilifeform: and the cached copy is updated when ?
Framedragger: it reads from the cached copy, damnit.
Framedragger: but you no longer reading from the front (in the sense of sql queries), which in db-land is good for performance.
asciilifeform: understand, if i paste a gpg key into phuctor, and cannot then ~immediately~ link to it in-chan -- phuctor is broken!!
asciilifeform: you are STILL stuck writing new submissions to the front.
asciilifeform: having two databases sounds superficially like a great thing, until you realize that the read-write thing applies just the same to the 'front' one as it does to the existing single.
Framedragger: people are telling you that there is a setup which is better but not multi-mil. it does require planning, etc.
Framedragger: you are presenting a false dichotomy consisting of {current phuctor; $multi-mil db setup}. you propose to escape it by writing new db as the only way to escape it.
asciilifeform: it is -- quite obviously -- doing work that dun need doing. behind the curtain. 'for your own good.'
asciilifeform: the disk can push 150 MB/sec. if the db cannot do 150 MB/sec, it is retarded and must burn.
Framedragger: (again, i did realise that the "oh very simple" angle isn't strong here)
Framedragger: can you not picture this setup with a read-only slave, with a *separate* "give new key to master" interface (separate so that the replication doesn't become write-write but stays write-read)
Framedragger: and you get your 5 seconds. if the db is loaded, it gets inserted "fast", and user is able to see results upon refresh (because when it gets inserted, update gets immediately sent to slave via trigger rule.)
Framedragger: alright. i will grant you that the end system wouldn't be "oh so elegantly simple", because you would have to have a submission queue (maybe something that trinque had in mind). user submits key; gets permalink (immediately); meanwhile key gets sent to master (immediately), and master puts it into "to be inserted" queue. under normal loads, the insertion happens ~immediately,
asciilifeform: it's the only serious long-term pill.
asciilifeform: what i'ma have to do, eventually, is to replace all of the idiocy. ditch the python, ditch the sql.
Framedragger: yes that's the tricky part. my natural answer to that would be to "drop it, have user be able to come back right away - to a permalink - but results only displayed when the upstream db actually processes and inserts the new entry. (this entry would then get fed into the slave via streaming replication)"
asciilifeform: including the guy who submitted 5 seconds ago.
asciilifeform: it isn't merely 'hash is computed', it shows if anyone else at any point in history ever had the mods.
Framedragger: you can give permalink to user on the www read-only box.
Framedragger: hold on, hold on. first of all, regarding the permalink: you can have this anyway, because the hash is computed in python anyway (i ~recall the procedure that you once gave me). so,
asciilifeform: the job where it takes you from a base64 gpg key you found on some godforsaken usenet post, to a permanent link in phuctor
asciilifeform: Framedragger: it ain't about 'demoocracy', it is about actually doing the job
asciilifeform: 'but does it do the same job ?' 'ummmmm didn't think of it'
Framedragger: because gotta serve all the our democracy right here and now?
asciilifeform: EVERYBODY thinks they're soooo much cleverer than dumb old asciilifeform
Framedragger: are you absolutely unwilling to have a delay there?
Framedragger: asciilifeform: aha, right, that. i literally forgot about that lol. but wait, first we need to clear up the question about whether you want an immediate result to be shown to user upon gpg key submission.
asciilifeform: if the 'www db' isn't writable
asciilifeform: does he IMMEDIATELY get a bookmarkable link based on the key's hash ? that he can come back to next hour, next day, next decade ? if so, how ??
asciilifeform: under your hypothetical system
asciilifeform: theoretically this can be done on one box. but -- for my enlightenment, Framedragger , describe to me :
Framedragger: well, i'm not that certain, but i am assuming you have more experience there with me. i will only remark that you merely need a *read slave*, not an actual mirror db which can handle writes and sync state. the syncing would go one way only. (hence the multiple references to pg streaming replication.)
Framedragger: now, complexity management-wise.. maybe; having same person manage both boxes may not be best idea (and the alternative has its own advantages). but the current "phuctor is down, i don't know why, it's a black box" isn't the greatest example of current setup, either. (this may be a red herring, i'm not sure)
asciilifeform: Framedragger: you're talking about replication. replication, generally speaking, doesn't work. certainly not on posgres or any other free db, and certainly not without titanic sweat and inevitable increases in attack surface.
Framedragger: bandwidth required for the www box would be ~same as what's currently needed from phuctor.
Framedragger: bandwidth? bandwidth required for phuctor box would be "what's needed to send new rows to this other box".
asciilifeform: not to mention their bandwidth.
asciilifeform: and the work of shepherding 2+ machines.
Framedragger: of course being able to point to a working prototype would do so much more than arguments. unfortunately that would most likely require the needed modifications on the phuctor box, so a bit of chicken-and-egg.
asciilifeform: Framedragger: why are you regarding the added complexity (100x what i have now!) as 'free' ?
Framedragger: asciilifeform: hold on. the idea was to separate reads from writes. having a separate box for www which gets updates from phuctor box, and having pg indices on it for quick search is *not* resource-intensive. i can cite examples but basically i'm quite certain that a <= 16gb memory box would suffice. phuctor box is 256 gigs yes, but it does *so much more*.
asciilifeform: Framedragger: think about how much complexity in the liquishit we call 'modern comp' comes from these attempted modularizations.
asciilifeform: ( ever read the s.nsa broadcast ? )
asciilifeform: Framedragger: now you're talking about having 2 boxes ? do you have an idea what the ~one~ costs ?
shinohai: The conspiracy theories begin, Trump leaves the runways intact: https://pbs.twimg.com/media/C8z7QViXYAIrhtc.jpg
Framedragger: (granted, these are just nice abstract words.)
Framedragger: asciilifeform: the parallel to "this is how winblowz blew up" breaks, imho, if you consider the splitting-off of www not as an addition, but as actual splitting-off, i.e., the box with phuctor on it may no longer have a www interface (just an option, i know you may be against it). if you picture it that way, it's more about modularisation vs. fixing and inflating a single monolithic thing.
a111: Logged on 2017-04-07 14:48 Framedragger: the 'www side can live on another server, even' degree of separation may be most easily achievable (given lack of resources to rewrite everything as of now) via pg notify / streaming replication. but maybe asciilifeform would insist that this is 'marrying the db'
asciilifeform: it is one of the main reasons for phuctor at all.
a111: Logged on 2017-04-07 14:45 trinque: without this separation the idiocies of www will creep in, such as "must respond to user accurately and *now*"
asciilifeform: and this requires guaranteed consistency. right now i have this guarantee because there is one db, and not a multilevel morass.
a111: Logged on 2017-04-07 14:44 trinque: mircea_popescu: more clearly stated, I do not see a www as part of the algorithm of phuctor. it is one source of input where there could be many, and one output idem. with a clearly defined line between www and phuctor (even allowing for that www may require cached copies of phuctor data to operate properly), this gives you something you can nuke later and replace.
asciilifeform: http://btcbase.org/log/2017-04-07#1639880 << imho this is a mistake. the presentation of results is not ~cleanly~ separable from the generation, because certain situations (dupes, or the omission of a concordance factor/modulus/key linkage) must be HARD-guaranteed not to happen
a111: Logged on 2017-04-07 14:02 asciilifeform: so why then is this not standard? why locks exist at all ?!
mircea_popescu: no i don't mean that. i just mean, yes in principl;e doing something with the plebs is not a bad idea ; but the exact what and wherefore coulo use more conteplation.
Framedragger: the 'www side can live on another server, even' degree of separation may be most easily achievable (given lack of resources to rewrite everything as of now) via pg notify / streaming replication. but maybe asciilifeform would insist that this is 'marrying the db'
trinque: thus the notion of sticking someone in the www toilet, who manages commands of specific outputs and inputs, and above all keeps the shit in the shithouse
trinque: without this separation the idiocies of www will creep in, such as "must respond to user accurately and *now*"
trinque: mircea_popescu: more clearly stated, I do not see a www as part of the algorithm of phuctor. it is one source of input where there could be many, and one output idem. with a clearly defined line between www and phuctor (even allowing for that www may require cached copies of phuctor data to operate properly), this gives you something you can nuke later and replace.
trinque: I'll close with the observation that asciilifeform's "fits in head" serves "build to iterate and throw away" very well.
trinque: nor the www part of your thing
trinque: asciilifeform: I threw away first bot, and second bot, and we sit here with the 3rd