a111: Logged on 2017-10-14 14:25 asciilifeform: apeloyee: iirc it was in middle of https://cr.yp.to/ecdh/curve41417-20140706.pdf ( which is largely about other, useless crud )
apeloyee: http://btcbase.org/log/2017-10-14#1725293 << lulzily, it mentions http://btcbase.org/log/2017-10-14#1725240 method on page 9: "we load the entire table into registers and perform the selection via arithmetic". wonder how you have read that paper if you didn't notice that.
shinohai: Don't open Slate this morning then, stupidity in form of "This Week, It Was Particularly Rough to Be a Woman" abounds.
mircea_popescu: lol pantsuit, they actually expect you can ban by reference.
mircea_popescu: moar "unity" and bullshit please, they totally work.
mircea_popescu: "Some of the emergency measures, notably beefed up police powers, are to be made permanent under a bill that has been approved by the lower house of parliament, the National Assembly."
asciilifeform: and the correctness of any variant (e.g. asmed) versions, will be judged against the canonical.
asciilifeform: it means the thing has to be gotten to acceptable performance sans asm.
asciilifeform: ( massive time is wasted deriving the carries )
a111: Logged on 2017-08-10 02:43 asciilifeform: for simplicity, tested the case that actually happens in practice: on a 64bit box, any ffa width over 512 bits gives a strictly 8-wide comba mult ocurrence
apeloyee: every other bolt << does this include an asm multiply?
asciilifeform: apeloyee: iirc it was in middle of https://cr.yp.to/ecdh/curve41417-20140706.pdf ( which is largely about other, useless crud )
apeloyee: so every time you calc an exp you calc 2 and 3 also ? << at the very start, once.
asciilifeform: i still think that it makes sense to do this only after every other bolt is as tight as physically possible -- bernsteinian karatsuba, unrolled comba, etc
apeloyee: no special cases. you precalc Base**0 = 1, Base**1 = Base, Base**2 and Base**3; assign 1 to result. then you take two next bits from exponent, starting with the most significant; take B to their power, as above, and multiply Result by B
asciilifeform: it still isn't clear to me how this algo can modexp to the power of 1.
apeloyee: asciilifeform: where do you see explicit indexing in the posted snippet?
mircea_popescu: apeloyee was your first sexual experience of the "just let me put the tip in" type ?
apeloyee: let's suppose B is a temporary, WindowLSB and WindowMSB are bits of 2-bit fixed window, One is, well, 1, BaseModSquare and BaseModCube are B**2 mod Modulus and B**3 mod Modulus. the posted snippet then assigns to B the power (2*WindowMSB+WindowLSB) of Base. See, we index the memory by a (small) secret!
mircea_popescu: not sure that buys you anything, now you got an FM carrier wave in the power supply instead of a simple levels thing.
a111: Logged on 2017-05-16 01:53 asciilifeform: now you store bit as ~phase~ of the square wave, rather than absolute steady state of the flipflop.
asciilifeform: prolly the only finalsolution to power variance is http://btcbase.org/log/2017-05-16#1656782 .
mircea_popescu: there is that.
apeloyee: you get the idea.
a111: Logged on 2017-10-14 12:06 apeloyee: it seems I see how to squeeze out one more FZ-sized temporary from FZ_Mod_Exp, besides the ^^ and using a proper barrett; it will exacerbate the above-mentioned physical leakage, though... http://p.bvulpes.com/pastes/XvDnd/?raw=true <- as usual, not tested.
a111: Logged on 2017-10-14 05:00 ben_vulpes: shinohai: do we have a from-zero "install gpg, ... now you're in the #trilema wot" ?
apeloyee: i'll have a square metre grid of whatever size i can buy, if i want. << not enough wires connecting them.
a111: Logged on 2017-10-08 00:20 asciilifeform: http://btcbase.org/log/2017-10-07#1722411 << 1 ) ffa is closed form. i.e. it CAN be written as a number of nand gates, with a 'funnel' at the top, to which you present a,b,c, e.g. 4096bit, numbers, and at the bottom in a little cup you get a^b mod c , and with NO UPWARDS FEEDBACK FLOW of information , i.e. answer comes after same interval of time always, and with strictly downwards signals.
apeloyee: lol even the exponentiation loop?
asciilifeform: fpga has a bit muxer and none of the loops will be loops, lol
a111: Logged on 2017-10-14 09:36 apeloyee: http://btcbase.org/log/2017-10-13#1724915 << http://www.ada-auth.org/standards/rm12_w_tc1/html/RM-3-5-4.html , paragraph 21: "the range of Integer shall include the range –2**15+1 .. +2**15–1"
a111: Logged on 2017-10-07 19:30 asciilifeform: the interesting imho discovery is that heathen bignumtrons don't win much (or even any!) speed by normalizing the ints being added/subtracted
apeloyee: it seems I see how to squeeze out one more FZ-sized temporary from FZ_Mod_Exp, besides the ^^ and using a proper barrett; it will exacerbate the above-mentioned physical leakage, though... http://p.bvulpes.com/pastes/XvDnd/?raw=true <- as usual, not tested.
apeloyee: http://btcbase.org/log/2017-10-13#1724915 << http://www.ada-auth.org/standards/rm12_w_tc1/html/RM-3-5-4.html , paragraph 21: "the range of Integer shall include the range –2**15+1 .. +2**15–1"
ben_vulpes: shinohai: do we have a from-zero "install gpg, ... now you're in the #trilema wot" ?
mircea_popescu: (in case anyone's curious, rassah is random faggot mildly famous for having stolen his bf's cash to buy right before a jump and then proudly selling it all and buying a car. 2013 iirc.)
mircea_popescu: somehow the stealing-by-name thing is really deeply ingrained in the muppet skull by now.
mircea_popescu: and in other hurr durrs, https://www.reddit.com/r/Bitcoin/comments/74fvvx/unpopular_solution_agree_to_2mb_hf_written_by/dnyfcig/
shinohai: Maybe Preet can get a job there lol
BingoBoingo: Which I offered in reply to his first mention of 10mbps hoping he got the letters confused
BingoBoingo: Looks like the Pinoys can't do it: "hello sir I'm sorry to inform you that it is not possible to met your requirement of 14Gigabits but our offer of 10mbps for 504 USD and we can upgrade that as your needs increase still stands." -jolo
mircea_popescu: apparently the black hat is black from pantsuit dust.
mircea_popescu: in other lulz - forum chick got warned for posting inappropriate content, in the shape of basically a rewrite of http://trilema.com/2017/remember-the-security-hole-automattic-refused-to-fix-back-in-2014/ (and a ton of bois pm-ing excitedly about "omg i saw that thread before it was deleted, r u real haxxor???").
asciilifeform: iirc there was a 'cpp ifaq'
asciilifeform: this is not bad time to ask the tuned-in folx, what parts of ffa they would like to see explained in moardetail.
mod6: perhaps -- i certainly think it's good to link l0gz in footnotes or whatnot. the what/why/how doc aught to summarize all of thinking/decision making that went into the code.
asciilifeform: ( otherwise same problem as writing 'faq' without anyone actually having asked any q )
mod6: i do also think there could be paragraphs even written around certain procedures in ffa. but not sure if that belongs in the code, or as a corresponding document outside of the code.
asciilifeform: mod6: i will confess, very much not satisfied with the comments, prolly comments should be written by someone else -- perhaps mod6
asciilifeform: hats off to apeloyee for the tip.
mircea_popescu: asciilifeform that last pic dun work for all the diff it makes
asciilifeform: prosecuted e.g. https://news.postimees.ee/3222207/the-assar-paulus-gang-finally-faces-court
asciilifeform: ^ head crown prosecutor of estonia ( at least at the time )
mircea_popescu: are the blades inside ?
asciilifeform was about to say 'no 1 spammer ever wrecked anything substantial' but then recalled usenet
mircea_popescu: an autistic child you didn't kill at the age of 7 will support a whole comicbook industry until age 70 ?
asciilifeform: exactly like the wp spammer discussed in http://trilema.com/2017/remember-the-security-hole-automattic-refused-to-fix-back-in-2014
asciilifeform: they're illiterates and chasing after imaginary moneys, while wrecking whatever they touch
mircea_popescu: "i tried stuffing all the stick and rock types i could find in all the holes that there were and even some that i made. why airplane not wash jocularstrap ???"
asciilifeform: this is what i think of sp4mz0rz of whatever type, incl. the ones with '1,000 servers' ( and previous years with 'million bots' ) etc
mircea_popescu: "he tried to trick them"
mircea_popescu: well, these fucktards are going around looking for chtulhu.
asciilifeform: there was a famous incident of a d00d who went around looking for bears
mircea_popescu: this has apparently become a major psychological liability for the crowd, they imagine that since everything they ever encountered was safely behind bulletproof glass, that's "just how life is".
mircea_popescu: there's a great gif with a stupid little girl confronting a lion from behind a glass pane.
asciilifeform: ( one difference, readily visible to naked eye, is that britney was made back when the sofas ~actually contained~ some pennies )
mircea_popescu: but anyway, apparently taking a few heads here and there creates exactly no conception of own vulnerability in the rest of the heads.
asciilifeform: gotta wonder how it was pictured to work -- 'if all the pennies in the sofas added up and made britney and coke, why not to make our shitcoin worth sumthing' ?
mircea_popescu: hey USG - take your useless sheep and drown them all for what anyone cares. seriously now.
mircea_popescu: basically the excel-powered secret agent strategy has become "try to disrupt bitcoin by creating fud on social media". which works to... disrupt usage by the usg's own muppets, the redditards, wikitards and assoerted other dust mongers ? which is somehow bad for bitcoin, or even noticeable or something ?
shinohai: This is the lulziest bit: " If you have transactions to make around late November, we suggest you buy some Ether in our wallet today. "
shinohai: http://archive.is/CbLsp <<< If adding ethereum, shapeshit, and BitcoinCrash wasnt enough ......
mircea_popescu: the problem is that they're not automatically diffedrent just because they could be.
a111: Logged on 2017-10-13 15:42 mircea_popescu: you'll end up with node lists, fingerprinting, mothership identification, the whole shebang.
asciilifeform: http://btcbase.org/log/2017-10-13#1725038 << every time i sit down to think about this megaproblem , inevitably come to the conclusion that the threeway cut was a mistake
asciilifeform: it is, to lift from another mircea_popescu article, 'a thrown brick, not airplane'
asciilifeform: the 'because you oughta!11' relaying can and does end up in same place as other 'because you oughtas' socialisms
a111: Logged on 2017-02-25 23:22 mircea_popescu: basically nodes are the digital equivalent of women : men fuck them so the state can have babies. hurr durr, pill plox.
a111: Logged on 2017-10-13 15:40 mircea_popescu: trinque one working model : all the nodes node 1 was connected to decided "it is spam", because they can do this now, so they didn't forward it but also didn't tell node 1. same did not hold for node 2.
asciilifeform: http://btcbase.org/log/2017-10-13#1725034 << the fundamental problem is the ~completely imaginary incentive structure for the expected node operations ( store, verify, relay)
shinohai: Me neither tbh
shinohai: I can't remember what I did with it lobbes .... but there was a plugin that let you announce a msg only to folks subscribed to it.
lobbes: That is a point. I can always tweak the announces to be less frequent as people see fit anyways
mircea_popescu: tbh, i don't think doing it in #trilema is either unfeasible or a bad idea.
mircea_popescu: lobbes yeah, we'll get to this in the coming weeks.
lobbes: Also gotta figure out where auctions would happen. I'd imagine #t wouldn't be keen on the auctions taking place in this channel. Would be a lot of added noise.
a111: Logged on 2017-10-13 01:55 mircea_popescu: trinque can directly push lobbes 's bot into it, what more is needed ? we have all the parts.
mircea_popescu: now THIS would be some interesting ecology, of the "keep the world safe for your children" kind.
mircea_popescu: human problem of meta is that humans can think outside the box. they end up choking each other with the unintentional by-products of that thinking.
mircea_popescu: (amusingly, this node-sybil-pie entirely exists BECAUSE IT CAN exist. think in GT terms : at t1 it doesn't exist. at t2, player 1 figures out it could exist, whether it can or not, and takes defensive meansures. at t3, some bug in player1's implementation is exploited (consciously or accidentally) and now... it DOES exist.)
mircea_popescu: you know, like the ahrefs guy.
mircea_popescu: you'll end up with node lists, fingerprinting, mothership identification, the whole shebang.
mircea_popescu: and EVEN IF maintaining a large (and, obviously, expensive) network of nodes etc, you can still run into surprises, as the bitbet issue very well illustrates.
mircea_popescu: trinque one working model : all the nodes node 1 was connected to decided "it is spam", because they can do this now, so they didn't forward it but also didn't tell node 1. same did not hold for node 2.
trinque: fed to another node, went fine
trinque: mod6: I did lose a transaction in a node's mempool the other day using it. was clearly there, sitting, going nowhere.
mircea_popescu: don't be doing that, not doing it is in the spec!
asciilifeform: ( based on the log link )
mircea_popescu: asciilifeform nevermind the actual implementation. the design/concept however you call it.
asciilifeform: sooner turn pig into a professor of mathematics, than cpp trb into sanity.
mircea_popescu: now, that they should be used sparsely is certainly true.
mircea_popescu: both theory and experience show this, sadly.
a111: Logged on 2017-09-14 05:30 trinque: in the process, realized I oughta publically denounce something in the funkenstein importprivkey/dumpprivkey patch
trinque: the introduction of new classes should be perma-banned
mod6: ah yeah, i do have a rawtx on sitting here -- just needs testing. also, still have some makefile changes that are needing some testing/review. and further, want to put in some requested changes to import/dump priv key.
mircea_popescu: in other tardstalk lulzgems, https://www.blackhatworld.com/attachments/fb_img_1507907065465-jpg.96283
mod6: iirc I had created a proper vpatch of the thing last year, but we never wrapped up all the testing. i'd like to revisit that.
asciilifeform: trinque, mircea_popescu : this is actually a variation on an ancient and imho interesting theme, the 'aeternity service' or what was it called, early attempt at 'bitcoinism' and p2p warez combo, 'proof of storage'
mircea_popescu: mod6 in any case don't completely abandon the trb-v castle, you're like our last guy there i fear
trinque: must also provide a web endpoint for the archive, etc
trinque: actually mircea_popescu's solution to forcing miners to hold all blocks could conceivably used to hire hosting for the archive from whoever has space. folks register with eg deedbot, must produce hash of all archived items at interval to receive a small bitcoin payout, flowing from subscribers to the archive service.
asciilifeform: or at any rate this is the condition for success
asciilifeform: in the end you will have 'written ffa' nearly same as asciilifeform had.
asciilifeform: and understand it the way a man packing a parachute understands.
asciilifeform: i intend for the whole lot of l1, to eventually read it
a111: Logged on 2017-10-13 15:14 asciilifeform: keep in mind that ffa ( esp. the more recent items ) will change.
mircea_popescu: http://btcbase.org/log/2017-10-13#1724903 << this is why i wouldn't recommend writing tests for it for any other reason than learning / checking. it's a prototype yet, only the growth spur should be checked as it were.
mircea_popescu: i am loath to put "and load this website" into the eg log bot spec.
mod6: it adds so much to the history though. being able to see wtf we were talking about.
mod6: <+asciilifeform> so there has to be a synthetic (i.e. Unsigned_32 ) type , that on archs that have narrower bitness than that, ends up a doubleword << ah, i see.
mod6: but for the ages.
a111: Logged on 2017-10-13 11:19 apeloyee: by the way, the bot which saved pastes on archive.is doesn't seem to do it anymore.
asciilifeform: note that this will result in exactly the same binary as builds now, on pc
asciilifeform: so there has to be a synthetic (i.e. Unsigned_32 ) type , that on archs that have narrower bitness than that, ends up a doubleword
mod6: The hope here is... when these changes are made, the given inputs still match the expected outputs.
asciilifeform: the current item is problematic on (hypothetical, i don't have one for which gnat builds) 8-bit archs
asciilifeform: mod6: there are serious ughs that need fixing, also. for instance to get rid of 'Natural' and 'Positive' and throw out the mess of conversions to and from Word
mod6: aha, this indeed may be the case - something found. but the main interest "does it do what it ~should~ do, given these inputs"
asciilifeform: and incidentally don't assume that the current forms, are the ideal and shortest/clearest possible
asciilifeform: keep in mind that ffa ( esp. the more recent items ) will change.
mod6: the fz_mux wasn't too bad, etc.
mod6: well, they're hopefully clear enough to be understood.
asciilifeform: 'now gotta write tests for the tests' etc
asciilifeform: the tricky bit is to avoid getting into 'quis custodiet ipsos custodes' trap.
mod6: i believe there to be a lot of merit to having unit tests around the specific procedures and functions in ffa.
asciilifeform: mod6: the key imho re a ffa tester, is that it gotta exist ~outside~ of ffa, and use a traditional, known-to-produce-correct-numberz ( and ideally, more than one ) arithm stack
mod6: again, these are unit tests, not functional, integration, or performance tests. meaning: i simply call a (so far public) procedure/function within ffa with specific parameters, and expect specific outputs.
mod6: im certain that even the tests that I do have are missing certain cases -- but its a starting point. and when I publish these, the hope is that people here can help fill in the gaps of the coverage so we can use this as a means to ensure that we do not regress as changes occur to the code.
mod6: im hoping to get some tests around the shifting procedures possibly today or this weekend. alf's latest changes from yesterday would fit into that as well.
mod6: i don't have a ton of coverage yet -- just mainly around the fz bitwise ops, fz_add, fz_mult, handfull of others.
shinohai: Awesum ... I'm going to take a look at the thing asciilifeform posted last night later, gotta get Excedrin for preemptive strike on headaches lol
a111: Logged on 2017-10-13 11:19 apeloyee: by the way, the bot which saved pastes on archive.is doesn't seem to do it anymore.
a111: Logged on 2017-10-13 11:18 apeloyee: e.g. the right shift main loop would look like http://p.bvulpes.com/pastes/XEvpD/?raw=true
apeloyee: e.g. the right shift main loop would look like http://p.bvulpes.com/pastes/XEvpD/?raw=true
apeloyee: http://btcbase.org/log/2017-10-12#1724608 << for some reason the subword shift is in-place, whereas the full-word shift isn't
deedbot: http://trilema.com/2017/remember-the-security-hole-automattic-refused-to-fix-back-in-2014/ << Trilema - Remember the security hole Automattic refused to fix back in 2014 ?
deedbot: http://trilema.com/2017/remember-the-security-hole-automattic-refused-to-fix-back-in-2014/ << Trilema - Remember the security hole Automattic refused to fix back in 2014 ?
mircea_popescu: aaand in the same female form, http://78.media.tumblr.com/70ef88d8decb2378fbd1a988b156ec79/tumblr_np0p0rapsi1shmurko6_1280.jpg
trinque: ^ lemme know if those look goofy. process that spat them out is brand new.
mircea_popescu: holy shit we're crashed all the way almost!
BingoBoingo: Ah not quite the same then, thing I am thinking of came in box same size and shape as "poptarts". Sold as "you know school sucks too much to teach this"
asciilifeform: here was explicit teaching of useful 13337ski111z for dyin'formotherland
BingoBoingo: Were sold at local bookstore. There was even electricity kit where lightbulb, battery, and wire was offered.
a111: Logged on 2017-10-13 03:11 BingoBoingo: So, anyone in the Repulic have any ideas on network switching equipment?
BingoBoingo: So, anyone in the Repulic have any ideas on network switching equipment?
saot: i think this is one of them
saot: some mental stimulation, cool corners of the internet are good to find unusual options
BingoBoingo: Maybe one of these days hitting a rock bottom will give him an opportunity for a spiritual experience? Anyways, what are you lurking for?
saot: Hi there
BingoBoingo: Still, the Asian cador continues to impress.
mircea_popescu: BingoBoingo eh i really don't need all of these.
mircea_popescu: trinque can directly push lobbes 's bot into it, what more is needed ? we have all the parts.
trinque: mircea_popescu: I'm gonna post a tutorial on deedbot.org tonight, and also hook up the newb nick+key conveyor
mircea_popescu: ccmtacks the important part is that unlike websites etc it's actually secure.
ccmtacks: Noted, I've read a lot of the help material, so I actually get it how it all works now, well a good part of it :)
mircea_popescu: trinque zinx ran me a grand or so so far, but i would way the fuck rather blow a grand on that than what the fuck is the alternative
trinque: sorry, jumped the gun.
mircea_popescu: make sure you get exact ammts though, otherwise there's trouble. if deedbot says 0.044236666 then make sure op sends exactly, all the way to the 6666.
ccmtacks: ok mate, no problem. I wanted to ask you regarding the fees, however, this time I've just been looking to do everything right
trinque: nbd. there needs to be a whole tx weighing mechanism in order to let people know what that amount will be in the response.
mircea_popescu: o darn i forgot the tx fee.
trinque: ccmtacks: actual amount will be less the tx fee. I'll fire that off in a bit.
mircea_popescu: say there you go.
mircea_popescu: ccmtacks say the !!v again, single space.
a111: Logged on 2017-10-12 20:52 mircea_popescu: trinque here's a thought : write me a branded single page "how to" somewhere, and ima modify the pay for your tits thing to only work with deedbot invoices. can then use it to push teh service, how's that sound ?
mircea_popescu: in other news, zinx sex got camhos lined up for http://btcbase.org/log/2017-10-12#1724491 already.
trinque: then from there, hot wallet, but one where refilling it is both human operated and quite easy
trinque: idea is to tighten the loop until it is a matter of <1min
mircea_popescu: ccmtacks sorry for all the hassle ; re-do once more, use a 1something address like god meant it.
trinque: asciilifeform: there's no hot wallet here; I'm marching my ass over to another box and making signed tx, then popping into a trb with polarbeard patch for sendrawtransaction
mircea_popescu: ccmtacks yeah but trinque runs it and if he says it won't then i tend to believe him.
trinque: lol, hi ccmtacks. I'm the operator of deedbot
ccmtacks: mircea_popescu, I've received a message from deedbot that the amount of coins will be transfered to desired btc address
mircea_popescu: you put an extra space after the v
ccmtacks: mircea_popescu, during the withdraw process, if I don't receive any response from deedbot, when I give him the encrypted OTP, does it mean that I have to do all the process again ?
mircea_popescu: or can keep it there, it's safe. DO NOT LOSE YOUR PGP KEY though.
mircea_popescu: this way you have a passive "your books and my books gotta match yo" extra accounbting check there.
trinque: yep, unique between you and other party
mircea_popescu: i daresay it's the right thing as it is.
mircea_popescu: trinque oh it counts PER NICK. this is very workable then.
trinque: don't want people sending 3 btc when they meant pay invoice 3
trinque: anyhow ccmtacks gotta pass the decrypted otp back to deedbot like with !!up
mircea_popescu: but still up in the air this, needs more human strokes.
mircea_popescu: the right thing would be for it to execute !!pay name if unambiguous and protest otherwise it needs an invoice number (or heck, just pay everything ?)
mircea_popescu: im sure they do.
trinque: I could have it burp the invoice number in logs if people don't care that it does
mircea_popescu: hm, there's no way to pay it without the number, huh. !!received-invoices step mandatory.
asciilifeform: whenever i think 'human mechanism' i think of the (terrible) film 'snowpiercer'
mircea_popescu: me too. it's the only right way to symbian.
trinque: I always start with human mechanisms and then speed 'em up
trinque: nah it's the good kind of pain
trinque: ok give it another crank
mircea_popescu: ccmtacks give it one moment, trinque has to fiddle the bot's guts.
trinque: there's an append only thing that doesn't exist yet
trinque: sec, lemme add his key to the right box
mircea_popescu: ben_vulpes i am entirely unsurprised. the "scale bitcoin" shitheads + the "exciting blockchain technologee" shitheads + the "we are marketing" dogeheads are pretty much === the cancer that has eaten foss.
ben_vulpes: asciilifeform: tiny set of scripts to get all of the source for a project from box a to box b
asciilifeform: mircea_popescu: anybody can make anything out of anything, esp. if the inputs are fictional.
ben_vulpes: believe it he proudly writes ethereum shits now
ben_vulpes: so i'm fiddling with some source delivery pipeline, and i discover that some asshole has put executable code into his cl .asd files; which i suppose is fine in some circumstances, but i'm doing Terrible Things over here and the asd's aren't getting concatenated with everything else and so just to add another entry to my list of folks whose decisions have made my life harder i look this dude up and can ya
asciilifeform: 'questions of a possible microwave or radio wave device .... If that were the case, the sounds wouldn’t show up on a recording' << and anyone should believe the clip is genuine article precisely why.
shinohai: I'll be back in a week when the vomiting, dizziness, and diarrhea subside.
mircea_popescu: ccmtacks hey there.
asciilifeform: and in the .ads, BitnessLog2 : constant Positive := W_Log2(Word(Bitness));
asciilifeform: http://wotpaste.cascadianhacker.com/pastes/s3As6/?raw=true << aaalmost forgot the log2 thing
asciilifeform: veteran ffaists will recognize the usual pattern -- do all of the possible ops that go into heaviest possible hamming weight of old-fashioned shiftness; but keep (via mux) only the required subshifts.
asciilifeform: they are almost certainly unnecessary on a pc, but can't put it past intel & friends to 'optimize' shl/shr, at the very least in the 'do nothing if operand is 0' sense.
asciilifeform: FZ_ShiftRight_Secret_SubW and FZ_ShiftLeft_Secret_SubW do exactly same thing as ye olde FZ_ShiftRight and FZ_ShiftLeft , but don't leak shift amt on machine w/out barrel shifter, at the cost of roughly 10x as much work
BingoBoingo: Maybe can count a bit on directionality of airflow provided by other fans, but... cable induced turbulence is a thing
asciilifeform: BingoBoingo: 'passive' boxes cool through the ears.
BingoBoingo: <asciilifeform> 'pcengines' racks a-ok << Gotta mind hot/cold aisle distinction for self defense from other's heat if otherwise passively coolable
a111: Logged on 2017-09-27 02:58 asciilifeform: http://btcbase.org/log/2017-09-27#1717726 << this reminded me of an ancient crackpotteriferous idea of mine: diskless, welded-shut box that boots into minimal sshiferous thing, privkey in boobytrapped sram etc. you fill it up remotely with os etc. and thereafter if power cut.