Show Idle (>14 d.) Chans

← 2017-01-18 | 2017-01-20 →
mod6: << some very early, news to report: I've successfully implemented this fix into the forthcoming version of V (99994), and all original and new automated tests passed.
a111: Logged on 2017-01-14 01:07 mircea_popescu: a patch can only apply if ALL of its antecedents are present ; not if ANY of its antecedents are present
mod6: There's still a bunch more clean up and testing ahead, but a step in the right direction. o7
ben_vulpes: cool, mod6
mircea_popescu: is this good for rustcoin ?
deedbot: << Recent Phuctorings. - Phuctored: 48159739175968624148092767 divides RSA Moduli belonging to 'Christoph Giesel <>; '
mircea_popescu: mod6 nice.
mircea_popescu: << apropos of nothing at all, anyone else kill garond ?
asciilifeform: << so now we know which tumour has been chosen to kill emacs with.
asciilifeform: 'If you’d like to join us, there’s plenty to do. You could: Port a small C function in lisp.h to Port your favourite built-in elisp function to Rust.'
asciilifeform: and before long, it will be impossible to actually get normal-people emacs running on any typical os.
asciilifeform: 'We can leverage the rapidly-growing crate ecosystem. We can drop support legacy compilers and platforms (looking at you, MS-DOS).'
asciilifeform: turns out -- the fungus is in full bloom.
mircea_popescu: << mostly for lack of actual people.
a111: Logged on 2017-01-19 15:25 asciilifeform: and before long, it will be impossible to actually get normal-people emacs running on any typical os.
asciilifeform: noshit, fungus dun grow on healthy tissue
trinque: gotta just stop; I'm never installing another emacs version again
asciilifeform: trinque: recall, the drepperites are getting ready to break glibc so that no moar clasical emacs.
asciilifeform: which means living on specially-terraformed planet, if you want it to work.
trinque: I built on musl lots of times before
mircea_popescu: glibc is already frozen pre 5
mircea_popescu: well, gcc i mean
a111: Logged on 2016-12-28 19:23 mircea_popescu: which i suppose warrants a general warning : DO NOT UPGRADE YOUR GCC TO 5.0! SAVE YOUR COPIES OF 4.X AND PRIOR!
mircea_popescu: anyway, the "leverage ecosystem" ie, "someone else will do our work for us" is very much usg.dos style, but also very much hopeless. seriously, they'll leverage ?
mircea_popescu: turns out i can bit mit with one hand while watching porn. they'll leverage what, they got nothing.
mircea_popescu: der angriff steiners war ein befehl!!1
asciilifeform: mircea_popescu: the weevils do in fact leverage one another's weeviling.
phf: seems like freenode upgraded all their servers to letsencrypt, meaning that you can't just verify ssl's fingerprint once a year. instead each server has own ssl, updated once in 90 days.
mircea_popescu: you understand nothing they leveraged works ? they ain't got a browser. google is trying to rescue 20 years of weveling with dubious results. there's no python 3. there's no ipv6. there's no trb. there's no eth. there's nothing. NOTHING.
asciilifeform: mircea_popescu: 'works' ain't their objective. area denial, is.
mircea_popescu: problem being girls going "you can't go in here" only works on the other girls.
asciilifeform: phf: fleanode's 'security' was a joke afaik always.
mircea_popescu: phf it's entirely unclear what ssl is supposed to provide. it might have been of marginal utility prior to their nsa merger, but these days it's utter waste of time. treat all freenode connections as plaintext.
asciilifeform: 'According to the statement of facts filed with the plea agreement, Shames developed malicious software, known as a keylogger'
mircea_popescu: "sold to 3k users who used it on 16k victims" dude...
trinque: 3k sales and he couldn't pay for a semi-literate lawyer ?
asciilifeform: in related lulz, << krebs is back to doing what he was made for, elaborate witch-hunting for usg handlers
mircea_popescu: trinque "sales" here means usg sops style nothing, like obamacare isn't a tax. his thing was mentioned once on a forum with 3k "registered users".
trinque: these creatures must want to sign the papers.
mircea_popescu: hey, it's how you get the ssi rite ?
trinque: and you get to be an elite renegade hacker on the dole, no less
mircea_popescu: you know ?
mircea_popescu: anyway, so krebs has sob story about how republic successfuly denies empire comms ; and wants to add some sort of "but we pay to house some random bums for the rest of their lives so it's all ok"
mircea_popescu: if you believe this is how it works, all the better.
thestringpuller: trinque: good lawyer is hard to find too.
phf: in unrelated lulz a u.s. retail bank is unable to send an international wire transfer. i bring a piece of paper with all correct requisites. half the numbers "i don't know what this number is", so follow up with calls for numbers from the fucking paper that, who could've predicted, are required. write down the numbers on their end incorrectly. the last part of saga: they now claim one of the numbers is invalid, but if you simply google for it it comes up
phf: as a valid ifsc
mircea_popescu: yep. usg banks won't do intl wires, for a decade+ now
mircea_popescu: they very transparently pretend ineptitude, but it's pretty clearly fed instructions.
mircea_popescu: (and if you try to drive cash to mexico they'll try and steal it ; and if you eg open a foreign corp with a subsidiary in the us and send the money home they'll pretend you're engaging in money laundering, and in general expatriating your wealth is generally more difficult than it's worth, which is why sane people do not live in the us.)
mircea_popescu: also why eg apple doesn't wire to new york the proceeds of sales in paris.
mircea_popescu: conversely of course, every dollar you do take out of the us hurts the usg considerably ; i'd guess the going rate is 130k expatriated is about equivalent to shooting a nypd officer in the face.
mircea_popescu: about there.
phf: hmm
phf: might just have to fall back to cash, i was trying to be prim and proper, alas
asciilifeform: << worked great last i tried it ( 2013 )
a111: Logged on 2017-01-19 16:28 mircea_popescu: yep. usg banks won't do intl wires, for a decade+ now
asciilifeform: but i was sending small (few $k usd) amount.
mircea_popescu: well supposedly in moscow there also lived a fellow who always found fish at the meat shop and to this day has nfi what everyone was on about.
asciilifeform: note, i will not claim that there is necessarily still fish at that particular meat shop.
asciilifeform: or would have been if the planets had been aligned differently, etc.
mircea_popescu: incidentally this'd have been a great show, i imagine, guy going on street idly, happening on meat shop just as the truck unloads, and generally doing a grand blond a la chaussure noire but a la russe.
mircea_popescu: did this exist ?
asciilifeform: not afaik, but perhaps phf saw such a thing.
mircea_popescu: Mr. Serendipivse
mircea_popescu: almost sounds georgian neh ?
mircea_popescu: which yes, subtext :D how did you call these ? they were sopirle in romanian
mircea_popescu: ie lizards.
phf: so i called the new branch of the bank i'm transferring to, they confirmed the ifsc number. i'm not sure what i prefer, the fed theory or the incompetence theory. either...
phf: *new york branch
asciilifeform: << in other olds >> elaborate yarn re 'mysterious' habit of u.s. soldiers, of bashing in the heads of killed goatfuckers. presented as 'atrocious' and 'senseless' etc. and i read whole spittoon waiting for the obvious reason to be mentioned, and of course was not (ground beed the face, and you dun need to worry whether you executed the hit on exactly the right d00d..)
asciilifeform: lulzy, they even issued little axes for this work.
asciilifeform: (avoid wasting ammo)
phf: i thought little axes were standard issue item, sort of like a sapper spade
asciilifeform: claims -- not
asciilifeform: (certainly not in regular infantry)
asciilifeform: !~later tell trinque re: >> plz consider posting recipe for musltronic emacs build! i promise to test.
a111: Logged on 2017-01-19 15:55 trinque: I built on musl lots of times before
jhvh1: asciilifeform: The operation succeeded.
asciilifeform: (iirc all versions of emacs from past decade or so have some kind of perverse hardcoded reliance on glibc in particular)
phf: did anybody do a musl sbcl? i have a musl cmucl in the backlog, but way down the list (perhaps once i get this x60 going)
asciilifeform: phf: not afaik
asciilifeform: ... what does the thing even ~need~ libc for?!
asciilifeform: (i can clearly see that my copy is linked with it. but -- why?)
phf: sb-unix
asciilifeform: does it follow that libc can be omitted entirely if one nixes sb-unix ?
phf: but i think sb-unix exposes some kernel level shit, because need it for reading files and exec and such
asciilifeform: all of the crud that will have to be rewritten on (hypothetical) sbcl-on-iron, conveniently.
asciilifeform: phf: would you say that cmucl is more suited for 'iron' incarnation? (i.e., can it be build without gcc?)
phf: so i don't know how sbcl does it, but cmucl has unix.lisp, which kernel relies on early on in the operation, and yes, contains all the "talk to the world" crud. cmucl actually does some hack where only subset of unix is used for operations and then when you (require unix) it adds some of the userspace
trinque: phf: the gentoo musl overlay has sbcl patches iirc. gotta forgo threads
phf: there's was a wip fork of making linux-unix.lisp talk directly using syscalls without libc. that's probably lowest common denominator
asciilifeform: trinque: pretty strange, because threads work fine in musltronic trb
phf: who knows with threads, i wouldn't be surprised if sbcl touches them in very inappropriate, glibc specific ways
phf: one of the major reasons why porting cmucl might be easier, a lot more naive image construction (doesn't have that whole sb! sb- machinery), and lack of threads
phf: *porting cmucl to raw iron
asciilifeform can picture this.
asciilifeform: traditional headaches turn to wins, the fewer 'native xyz...' the better, for iron.
phf: cmucl is basically a very straightforward lisp machine port, so there's less unixisms built in. sbcl is modernized for unix, but with corresponding tradeoffs
asciilifeform: incidentally imho traditional unix-style scheduler is not appropriate for 'x86 iron lisps'
asciilifeform: instead ought to simply have 1 permanently-running instance of the runtime per cpu core
asciilifeform: with some means of communication between.
asciilifeform: phf: did you ever expand on why your gadget crashed daily when it sat on top of cmucl ?
asciilifeform: was it a unixism ?
phf: well, you were right, it's the lack of threads. you can't build for cmucl green threads like it's native.
asciilifeform: this i recall. but didja ever find out specifically why.
asciilifeform: as in, 'the locking system is defective in yet-undiscovered way, proggy fandangoes over own toes'
a111: Logged on 2017-01-19 17:16 asciilifeform: !~later tell trinque re: >> plz consider posting recipe for musltronic emacs build! i promise to test.
asciilifeform: ty trinque !
phf: no, i
asciilifeform: trinque: you tested, worx? also on x11 ?
asciilifeform: also it does seem to insist on dynamic linking.
phf: no, i'm not that far in the stack to tackle threading. because of goals, if i do, or once i do rather, i'll just work on making them better green threads, rather than native.
asciilifeform: phf: well yes, 'metal' cmucl would need own scheduler regardless.
trinque: asciilifeform: yeah works; I had it built against gtk2, -dbus, etc
asciilifeform: trinque: pretty neat.
asciilifeform: seems like a fully-musltronic, work-ready linux is possible, then.
asciilifeform: emacs, gcc4.9+gnat, build tools, trb, kernel.
phf: well, stali has a musltronic pdf reader and web (they have their own wrapper around webkit, called st i believe, which is literally just a frame with addressbar)
asciilifeform: (then perhaps some luxuries, e.g., 'midnight'. then x11 and ratpoison. is pretty much all.
asciilifeform: phf: webkit builds musltronically?!
phf: re cmucl threads i think that it doesn't always preempt correctly. like it has explicit yield, which you don't always have to call, but it being there implies. also hunchentoot wasn't working right without putting a yield somewhere in the scheduler. it's all very vague, because i've not spend any time looking at it
asciilifeform: sounds gnarly
phf: there's some strategy in building multithreading in a lisp that all the commercial lisps share with cmucl and that's different from sbcl, but i'm not quite sure what it is yet
asciilifeform: imho this is one of those things that will have to be demolished ~100% and rebuilt from scratch , for any reasonable 'metallization'
asciilifeform: also (and iirc i discussed this on my www at one point) the correct approach is to ditch the native compiler, in favour of the interpreter, hand-compiled to fit in L0 cache
asciilifeform: then thing has a chance of working correctly.
phf: so cmucl introduced this whole idea of vops (virtual operation)
asciilifeform: these are just asm macros neh ?
asciilifeform: or is it 'bytecode' a la tinyscheme?
phf: it's sort of "writing assembly with sexp", but really it's an abstraction layer for what used to be microcoding
phf: so cmucl's compiler interleaves vops, while cmucl's interpreter uses vops as bytecode instructions
asciilifeform: interleaves forth-style ?
phf: well, that's not the right term to use. depending on a vop it either jmps, calls or inlines
asciilifeform: so liquishit
asciilifeform: (this imho is quite sorry state, you oughta be able to go from the executable TO the lisp representation without ANY info loss)
asciilifeform: and -- necessarily -- vice-versa.
phf: so eval directly on source, without a vm even?
asciilifeform: not necessarily, though it is one way of achieving this.
phf: actually it's all a lot more clearer when you read the early cmucl papers. the mess grew in unix user space..
phf: you could maybe put cmucl's interpreter vm into l0, have vops as operations that you need but that you don't have in a vm (simd, floating point)
asciilifeform: the two major caltrops re 'iron lisps on x86' are 1) interrupts 2) dma
asciilifeform: they both violate whatever sane execution model you pick
phf: right
phf: fit cadr interpreter into l0 :P
asciilifeform: sorta what asciilifeform tried to do (then barfed.)
asciilifeform: thestringpuller: linked in the logs.
thestringpuller: btcbase search didn't show it from browser. weird.
phf: thestringpuller: which string did you use originally?
phf: hah
trinque quite interested to hear why asciilifeform barfed
asciilifeform: trinque: tldr: interrupts and dma.
trinque: ah ok, thoughts were connected
a111: Logged on 2017-01-19 17:31 asciilifeform: seems like a fully-musltronic, work-ready linux is possible, then.
asciilifeform: trinque: i must note: just because asciilifeform barfed, does not mean that a clean solution to the problem cannot exist. only that he did not find one.
mircea_popescu: << i always thought that's what yarvin was john smithing with his "jets"
a111: Logged on 2017-01-19 17:40 asciilifeform: or is it 'bytecode' a la tinyscheme?
asciilifeform: mircea_popescu: nah, 'jets' were a terrifying crackpottery where he promised to 'recognize algorithms' and 'transparently' substitute in optimized/c-tronic routines
mircea_popescu: yes but started (at least i think) as vops.
asciilifeform: well ~all interpreted programming languages use 'bytecode' or some variant (i.e. emulator for fictional cpu that fits the task)
asciilifeform: e.g., perl, python, tinyscheme, many forths
asciilifeform: name 'bytecode' simply came about because making the opcodes 1 octet long is convenient
asciilifeform: (yet unpublished) 'p' is a 'bytecode' system -- all ops are 1 byte long, and take no operands, and all bytes are valid ops.
mircea_popescu: << actually non-interrupt / non-dma (and possibly tagged ram) is about the reason to make own fpga cpu
a111: Logged on 2017-01-19 17:49 asciilifeform: the two major caltrops re 'iron lisps on x86' are 1) interrupts 2) dma
asciilifeform: well yes.
mircea_popescu: interrupts ffs. wtf.
asciilifeform: is how i got there.
mircea_popescu: asciilifeform i figured re p when i saw your padder attempt.
asciilifeform: interrupts, dma, are relics from dark age when there was 1 cpu and it cost most of what the machine cost.
asciilifeform: they make it possible to actually do any work on von neumann monstrosity.
asciilifeform: at the cost of destroying whatever conceptual integrity machine might otherwise have.
asciilifeform: (instead of 1 model of computation, you now have 3 -- that you know about. plus others, unwanted, resulting from interactions b/w the 3.)
mircea_popescu: oh and also - automated ring buffer allocation. if you declare 8 bytes and write 9, you overwrite your first. YOUR first. not someone else's.
mircea_popescu: which'd be interesting if all pointers live as mod math.
asciilifeform: mircea_popescu: ada's 'modular types' come to mind
mircea_popescu: yep. but cpu-enforced, all memory allocation is that.
mircea_popescu: your pointer is 0x88faf0:0a and god help you, if you wish to write 500 bytes in there all the better.
mircea_popescu: you'll be left with the last 10
asciilifeform: holy fuq why would you give the operator raw pointers.
asciilifeform: why not also raw voltages.
mircea_popescu: at some point raw pointers must be given to someone
mircea_popescu: not necessarily operator, but also none of the current straight on addressing bs.
mircea_popescu: cpu can not address memory in the sense of "from here to eternity".
mircea_popescu: cpu can only address memory if the correct size (0a) is provided for 0x88faf0.
asciilifeform: how cpu physically moves the bits, is quite separate question from what operator thinks about.
mircea_popescu: yah, but i want this quite exactly, for it to not be POSSIBLE for memory to be addressed as it currently is.
mircea_popescu: ie "i want to read from x onwards".
mircea_popescu: it must not be possible to read from x unless you correctly specify the ring size.
asciilifeform: the simplest way to do this using current off-the-shelf hardware is to not expose the pointers.
mircea_popescu: promise.
asciilifeform: aha, and know why ?
mircea_popescu: i want the cpu to not be physically capable of addressing it.
asciilifeform: because peripherals .
mircea_popescu: fuck peripherals.
asciilifeform: mircea_popescu: for so long as memory is sold as a dumb capacitor grid, cpu can necessarily address all of it (that is, all of the connected ones)
asciilifeform: but if you had something clever in mind, around this, i'm all ears
mircea_popescu: that doesn't mean the cpu die must be cast in such a way as to allow anyone else to do it.
asciilifeform: 'anyone else' being operator ?
mircea_popescu: if you fail to provide the buffer size, cpu reads 0 bits from address specified.
mircea_popescu: if you provide buffer size, and it is the correct buffer size for offset, you can read that many bits.
mircea_popescu: if it is incorrect, you read 0 bits.
mircea_popescu: same for writingt.
asciilifeform: mircea_popescu: in sane programming system (e.g., lisps) nobody gets to point to 'memory address'. they point to an item that necessarily and unseparably carries not only address (not visible to or alterable by the operator, normally) but also a type and -- when the thing is >1 machine word in size -- a size.
mircea_popescu: yes but that's an os.
asciilifeform: on a proper lispm (definitionally) it is done by the iron.
mircea_popescu: i don't want the functionality exposed.
asciilifeform: aha so mircea_popescu is simply asking for lispm, lol.
mircea_popescu: i guess.
asciilifeform: i don't disagree.
mircea_popescu: if it actually does that.
mircea_popescu: i suppose its tagged memory model does pretty much a superclass of the above
asciilifeform: it did, though not in the peculiar 'stick shift' way described by mircea_popescu earlier
mircea_popescu: i don't insist re implementation ; was just included as example
mircea_popescu: anyway. all pointer references to read must be offset:size or no dice ; all pointer references to write may be offset, and will write mod size ; all pointer allocation specifies size and receives offset.
mircea_popescu: and literally &p:sizeof(p*) calls throughout all c, none of this current bs.
asciilifeform: on a proper lispmtronic comp, there is no juggling of 'pointers' or any other naked, unannotated bitstrings.
mircea_popescu: i imagine it prolly also works. ianale
asciilifeform: if cpu wants to move, add, subtract, etc. a bitstring, it has a length glued to it, and a type (that indicates, among other things, what it means to 'add' it, say), also glued to it.
mircea_popescu: yeah and then interpreter bitches if wrong types, so it catches size issues by default.
mircea_popescu: my model is... well, aptly put, stick shift. ie, even if dudes wanna bitch about lisp, because "c is faster", the above STILL APPLIES
asciilifeform mutters something about naggum's bathtub catapult
mircea_popescu: i must say candi_lustt made it obvious lisp is workable development. way the fuck faster than figuring out which stdio to include and dumb shit like that
asciilifeform: mircea_popescu: and picture if it had been on your actual comp.
asciilifeform: (instead of uploading proggy repeatedly and waiting for output)
mircea_popescu: asciilifeform no i liked that part.
mircea_popescu: but this may just be me.
asciilifeform: mircea_popescu: it'd get quite painful for a large proggy.
mircea_popescu: large teams, small programs. not the other fucking way around.
mircea_popescu: good brothel = many whores, small building. not vice-versa. that's called "beachfront property".
mircea_popescu: incidentally, if the various dreppers etcs actually had any friends, the whole rot wouldn't exist in the first place.
mircea_popescu: it's quite evident from their gestalt they're coding with flashlight under the covers.
asciilifeform: maggots dun have plans, or objectives, they simply eat.
trinque: nah, the idea that these guys are using this as a substitute for healthy human interaction is spot on.
trinque: see the conferences, "social coding", hackathons, etc
mircea_popescu: "i'll get everyone to do X!"
mircea_popescu: dude... just get everyone to show up saturday night already, nevermind this dumb shit. buy some wine.
mircea_popescu: "i'm so terrifying of talking to that cute wallflower i'ma revolutionize the $name like $name for $name instead!11"
mircea_popescu: and in scandalous coincidence : me puts on pair of pants, at 3:30 pm. sticks hand in left pocket, out comes condom wrapper. sticks hand in right, out comes... different condom wrapper.
mircea_popescu: i guess i haven't worn these since i was last on roof.
asciilifeform: i thought mircea_popescu ends up with DShK shells in pockets when comes back from roof.
trinque: << yet shouldn't be overlooked that several folks came to play immediately when there was a shared repl available
a111: Logged on 2017-01-19 18:27 asciilifeform: mircea_popescu: it'd get quite painful for a large proggy.
trinque: it made an arena. good people like those.
trinque: and aside all other bullshit uses of "social coding", that ~was~
trinque: one can see what was useful in the experiment and not get fixated on the particular implementation
trinque has had plenty of good coding afternoons where one person drove keyboard and >=1 others hollered and pointed
deedbot: << Qntra - Tim Draper: Theranos Targeted By "Conspiracy"
mircea_popescu: nah, what'd i fire at
mircea_popescu: trinque yeah, like cs for thinking people.
ben_vulpes: for pete_dushenski on the topic of studded bike tires
phf: it uses the word "modern" a lot, but from cursory glance it doesn't look insane
deedbot: << Trilema - Cool it, Carol!
ben_vulpes: FAP FAP FAP
asciilifeform: phf et al: actually this is example of TERRIBLE 'retrocomp' design
asciilifeform: it offloads a megatonne of work onto a modern micro
asciilifeform: is in no real 1980s sense 'fits in head'
phf: well, it's a lispm kind of architecture, with a single central controller, and extension boards of various degrees of dirty. though i guess the real "core" here is the opaque STM32
asciilifeform: phf: aha. the z80 is ~decorative on such a box.
asciilifeform: folx keep makin' these nonsense abortions.
phf: i didn't know what stm32 was, i thought it was some PIC variation
asciilifeform: and i can see why they do it, even buying ~one~, e.g., UART, today, is not easy
asciilifeform: phf: arm
thestringpuller: asciilifeform: is NSA-TRB down? started up an old node to sync... and get:
asciilifeform: thestringpuller: it's up but blackholed
asciilifeform: (my log tail looks quite like thestringpuller's paste)
asciilifeform: megatonne of 'socket no message in first 60 seconds, 1 0' etc.
asciilifeform: phf, ben_vulpes : << much better z80ism
asciilifeform: (0 anachronisms)
asciilifeform: and no, cf card is not anachronism, it runs in ata-compatibility mode
shinohai: They were SO CLOSE
asciilifeform: (d00d ~does~ offer a schematic for 'keyboard and vga' using modern micro, at the end of the page, though)
asciilifeform: shinohai: jet ?!
asciilifeform: reminds me of idiot joke from 1990s
asciilifeform: ukr ministry of culture says: 'ancient ukrs had radio! we have proof'
asciilifeform: interviewer: 'what's the proof'
asciilifeform: ukr: 'our archaeologists have turned up no wires...'
trinque: lol
shinohai: To me appears most of hull is cloth stretched over a frame, or are those wrinkles?
asciilifeform: looks like honest attempt at a glider, possibly
asciilifeform: 'we found no propeller! MUST BE JET'
asciilifeform: why not ufo.
asciilifeform: with antigrav motor.
asciilifeform: phf: << lulzy, ALL of the building instructions for cmucl -- 404 !!
asciilifeform: this thing makes ~my~ horrors look well-documented.
asciilifeform: srsly, all 4 links under the 'Developer info' category, are duds
asciilifeform: why the fuck even bother with having the other docs
asciilifeform: if the motherfucker cannot be BUILT without whole day of reverse-engineering
asciilifeform: 'To be written.'
asciilifeform: y'know, this is how folx end up on a scaffold with mircea_popescu pissing down their neckstump
asciilifeform: and i have 0 sympathy.
phf: what "folks"?
asciilifeform: 'developers'
asciilifeform: 'open sores' folx.
phf: again what "folks"?
asciilifeform: #ifdef DEBUG_BAD_HEAP
asciilifeform: * For some reason x86 has a heap corruption problem. I (rtoy)
asciilifeform: * have not been able to figure out how that occurs, but what is
asciilifeform: * happening is that when a core is loaded, there is some static
asciilifeform: * object pointing to an object that is on a free page. In normal
asciilifeform: * usage, at startup there should be 4 objects in static space
asciilifeform: * pointing to a free page, because these are newly allocated
asciilifeform: * objects created by the C runtime. However, there is an
asciilifeform: * additional object.
asciilifeform: * I do not know what this object should be or how it got there,
asciilifeform: * but it will often cause CMUCL to fail to save a new core file.
asciilifeform: ^ save.c. this looks terrifying.
asciilifeform: phf (or anybody else?) -- dare i ask, how the fuck does one BUILD IT
asciilifeform: is the idea 'read the code for a month and then MAYBE you will realize in what order to build'
ben_vulpes: dog if that's a jet fighter i'm the ceo of apple
asciilifeform: fart-propelled
asciilifeform: << phf if this is current, then i am officially cured of all desire to have anything to do with the thing
asciilifeform: travail arabe.
ben_vulpes: i was certain that we'd done the 'omg cmucl compilation' lolz thread before; is this a time loop? where are picard and riker
asciilifeform: !#s cmucl build
ben_vulpes: ah yes, 'twas adlai who brought do
a111: Logged on 2016-09-17 02:43 adlai: this is funny: but this is funnier:
a111: Logged on 2016-09-17 02:43 asciilifeform: ## Recompiling CMUCL is not an exact science... ##
asciilifeform: looks like same result.
a111: Logged on 2016-09-17 02:57 mircea_popescu: how can you lot put up with these idiots ?
asciilifeform: in fact we had the thread!!
mircea_popescu: honeslty, should probably go with whatever either phf used to run btcbase or ben_vulpes used to run candi.
asciilifeform: iirc both are on sbcl
asciilifeform: which is ~unsuitable for 'metallization', see today's earlier thread.
asciilifeform: (sbcl runs well today precisely because it is 'sharpened' for linux)
mircea_popescu: "metallization" is way premature in any case.
asciilifeform: aaaactually if sbcl, cmucl, or whichever lisptron, could be whittled down to 'get char', 'put char', 'malloc' -- could be on iron tonight.
phf: nyef actually spent a lot of effort trying to put sbcl on raw hardware, and he's a competent hacker (he also did sbcl port to arm for example)
mircea_popescu: but they can't.
asciilifeform: phf: is this experiment documented somewhere?
asciilifeform: mircea_popescu: what i wanted to establish was the ~whynot~
mircea_popescu: because dma and interrupts ; and bad memory addressing.
asciilifeform: you can lose dma if you're willing to lose nic and video.
mircea_popescu: the problem with "string" isn't there just to fuck with mp's ai bot.
mircea_popescu: what "char" ?
asciilifeform: octet from/to uart.
mircea_popescu: if only.
asciilifeform: it's enough to operate a very basic comp.
asciilifeform: (quite enough for a forth, say.)
mircea_popescu: this octet from/to uart doth not exist as such on extant hardware.
asciilifeform: theoretically you can, e.g., host www on just an uart.
phf: asciilifeform: there are text files (he makes adequate technical logs of most of his efforts. i'm using his notes to recreate arm port for cmucl), but i won't be able to find them at the moment (i'm not on my main machine)
mircea_popescu: theroetically you can lick your own anus.
asciilifeform: mircea_popescu: it exists, i have run number crunching routines this way.
phf: i'm pretty sure he was doing around the same time as you were doing loper, so i'm surprised you haven't seen it. i think he might've been doing it under TUNES umbrella
asciilifeform: boot sector loads a few 100 kB of hand-rolled code, and talks solely via uart.
asciilifeform: you can do actual work this way.
mircea_popescu: you will not "whittled down to 'get char', 'put char', 'malloc'"
asciilifeform: mircea_popescu: in point of fact, tinyscheme does.
asciilifeform: but is unsuitable for different reason (it has no compiler, is a c proggy, cannot compile self)
mircea_popescu: i don't recall, what was the objection to tinyscheme as the root ?
asciilifeform: see above. it needs gcc.
mircea_popescu: ahahah. so "in point of fact tinyscheme doesn't"
phf: so far only person who wrote anything for tinyscheme is myself, and tinyscheme is more of a PoC than anything else. bignums for example were ~unreasonably~ expensive
asciilifeform: phf: this is theoretically solvable. but it's a c proggy.
mircea_popescu: phf was abortive experiment in eulora too.
phf: "<nyef> (Speaking as someone who wrote his own standalone Forth and then used it to load a stripped-down hacked SBCL core on bare metal x86.)"
mircea_popescu: was this in #lisp ?
phf: actually i should probably track his notes down, but it might be easier to just ask him
phf: yeah, 2010
asciilifeform: ~somebody~ has to have 1) attempted this 2) not burned all of his work in fireplace in disillusioned despair
mircea_popescu: #lisp is pretty fucking dead.
mircea_popescu: for one thing who are the people there even ?
ben_vulpes: occasionally i ask a question and someone answers
ben_vulpes: "bike" is useful, "beach" is responsible for some amount of CLIM...
jurov: asciilifeform: i think these somebodies went on to make their own lisp dialect, using python build system or jvm or llvm or somesuch
asciilifeform: jurov: they may also have gone to the bottle, for all i care
asciilifeform: jurov: i am interested strictly in the record re the relevant bits.
mircea_popescu: i have nfi why it has to come in scattered 2 page text files
mircea_popescu: they should hang kids in jr high by their ballsac until they learn to blog every day.
phf: he used to have a top level domain with artifacts that he was producing. txt files of notes, diffs for sbcl etc
mircea_popescu: yes but it's gone.
mircea_popescu: << there's a log, insanely shitty with too much js to run, but i think it may go back usefully.
asciilifeform: poor sad nonvtronic patch. go and try to find WHAT he patched.
mircea_popescu: phf that shitty "" thing that spits the most obnoxious errors, really bitch, "no such key" ?
mircea_popescu: !#s haverbeke
mircea_popescu: !#s bridgewater
mircea_popescu: (nyef = alistair bridgewater, from the pw:rn subdirectorate of intel)
ben_vulpes: pw:rn?
deedbot: << Daniel P. Barron - FUCKGOATS For Sale
mircea_popescu: !~google pw:rn
jhvh1: mircea_popescu: No matches found.
asciilifeform bbl -- meat
mircea_popescu: you knwo you should eat some bread and vegetables on occasion
phf: all that meat's what making you so angry
mircea_popescu: he's angry ? seems more like incurably disdainful
phf: hmm, i thought russian disdain is usually a lot colder, refined. sort of like nabokov talking about what make lolita appealing to the american public.
mircea_popescu: anger's from the blood. what he's got, it's definitely bile.
mircea_popescu: lol check that out alf, FUCKGOATS reselling as collector itamz lel
phf: if you visit mit's open courseware, first link is on "Gender & Media: Collaborations in Feminism and Technology"
phf: which brings me to my question, what's a good book on inorganic chemistry (or general chemistry)
mircea_popescu: !~google costin d nenitescu, chimie anorganica
jhvh1: mircea_popescu: Chimie generala nenitescu - Cumpara cu incredere de pe <>; Costin D . Nenițescu - Wikipedia: <>; Costin D . Nenitescu - Chimie anorganica - manual pentru licee de ...: <
mircea_popescu: ^ best ever.
mircea_popescu: prolly saw 20 editions
mod6: evenin'
mod6: how goes?
mircea_popescu: not bad! hutin' elusive snails.
mod6: aha nice!
mod6: they're out there.
mircea_popescu: #lisp is more fucken dead than #bitcoin-otc ffs.
mod6: lol
mod6: we're creapin up on 450`000 blocks.
mircea_popescu: "Mircea has got 8.000m ECu worth of remarkable eps" << go me.
mircea_popescu: you're like a good luck charm mod6
mod6: haha, nice!
mod6: that is a LOT of eps, Sir.
mod6: btw, i wanted to note that hanbot v. random .ar derp would have been an awesome battle to watch.
mod6: why would any one care who feeds what to a street-cat?
shinohai: /me waves to mod6 o/
mircea_popescu: it's a woman thing.
mod6: hai shinohai
asciilifeform: mircea_popescu: lol, i ~did~ wonder what he wanted with that crate.
asciilifeform: neato danielpbarron .
asciilifeform: << as (perhaps folx suspected) -- 'meat' is not necessarily to eat.
a111: Logged on 2017-01-20 00:04 mircea_popescu: you knwo you should eat some bread and vegetables on occasion
mircea_popescu: even so. some fruit, something.
mod6: i just had this impression that alf was on atkins or something
mircea_popescu: which one is that one ? no pasta ? all pasta ?
mod6: all meat, no carbs type thing.
mircea_popescu: wait you mean to tell me the keto dorks "reinvented" a fad diet from the 60s ?
asciilifeform: 'meat' in asciilifeform's shorthand is simply any nonmaskable interrupt from meatspace.
mircea_popescu: tsk tsk, and i thought it was so original
mircea_popescu: asciilifeform no dude, it's literally roast beast, can't fool us.
asciilifeform: can be!
mod6: i was at the deli the other day, buying up all the meats... and just remembering, and getting even pissed about the thought of some asswads substituting soy in for fat.
mircea_popescu: this is a thing ?!
asciilifeform: what next? solidol ?
mod6: we talked about how some this happened once upon a time in a socialist hell hole.
mod6: i'd get medieval on someone if they did that to my salami
mod6: don't fuck with a guys meat.
phf: sometime in the early 2000s my family basically lived on what we called wolf diet. small bits and pieces, like nuts and fruit for two three days, but then once in three or four days we'd cook a giant steak or roastbeef or such. there was no reason for it, just sort of gravitated towards..
phf: also the sbcl os patch, applies cleanly against sbcl_0_9_14, but is missing the forth bootstrap that nyef's talking about
asciilifeform: that'd be problem.
phf: indeed
asciilifeform: (why did he need a whole forth?! why not just a few 100 ln of asm)
mircea_popescu: yesterday was a 2 gallon pot of bolognese with smoked mozarella. today was 7 sushi rolls. this isn't counting the pastries, cookies, cocktails etc.
asciilifeform implodes.
mircea_popescu household goes through >100k calories / week and all the girls are dieting.
mod6: omg. i bet the bolognese is awesome.
phf: girl couldn't figure out how i can eat so much (now) and not gain any weight was a mystery. few weeks ago a lightbulb "ooooh, it all goes to thinking!"
asciilifeform: phf: i've spent last hr or so reading cmucl code, more or less at random, and am struck by the sheer proportion of scar tissue
asciilifeform: (things that are there to work around $bug, which in turn stems from workaround of $bug2, etc)
asciilifeform: makes trb look good.
mircea_popescu: phf enjoy it while it lasts bub, i had the same thing in my 20s/early 30s
asciilifeform: now, it isn't as if i had a better cltron to offer.
mod6: oh, yeah. almost forgot... having the feather stems in the pillows is the worst. i refuse to use them for that reason. nothing worse than those things poking you in the face.
asciilifeform: also the thing is chock-full of os-specific conditionals
phf: asciilifeform: i wonder how much of scar was removed in sbcl. i know while newman was still hacking, he removed a lot of gunk. i vaguely suspect that cmucl can't be improved without tackling the bootstrapping problem, which is probably going to look very similar to sbcl's solution
asciilifeform: lol holyshit sbcl is still on soresforge?!!!
mod6: is there any big hang-ups about chicken scheme?
asciilifeform: gotta ditch gcc.
asciilifeform: given as it only runs on unixlikes. (and is a massive pig)
mod6: well, yeah, but arn't all scheme interpreters built with C?
asciilifeform: not, e.g., 'dream'
phf: no in scheme land you have mit scheme and scheme48
asciilifeform: and conceivably one could hand-compile tinyscheme.
asciilifeform: (ideally scheme-on-iron would be a very few kb of heavily-massaged asm)
mod6: that'd be neat
asciilifeform: mod6: likewise, there are cltrons (e.g., sbcl) that -- at least in theory -- can compile selves to x86
mod6: 'cl' is like 'lisp' formal right?
asciilifeform: common lisp specifically
mod6: ah
asciilifeform: (there are many different lisps, starting from the 1958 one)
mod6: <+phf> no in scheme land you have mit scheme and scheme48 << with these, which one is more preferred? or why choose one over the other?
phf: there's a lineage of lisp, which is significantly mutually compatible. LISP, maclisp, interlisp, zetalisp. common lisp is a standardization attempt on top of those. for example if you take eliza code (written for LISP 1.5) you can make it run on common lisp without any transformations (need to provide some missing forms though)
phf: this is not really the case with scheme, T, clojure, etc.
asciilifeform: ftr phf is the one and only serious lisp d00d i ever ran across who capitalized.
mod6: capitalized in the sense that most lispers live in the park in san fransisco?
phf: kek
asciilifeform: capitalized LISP
mod6: ah
asciilifeform: phf: to be fair, the main (other than lisp1-ness) semantic diff b/w scheme and lisp1.5 was static scope. which made its way into cl.
asciilifeform: (which, to nitpick, has selectable scope. but defaults to static)
mod6: what's the meaning there, is LISP the "proper" lisp for an actual lisp machine?
asciilifeform: mod6: nah
phf: well, i think, if you're trying to communicate clearly, LISP means thing that mccarthy wrote, and then spinoffs of that, LISP 1.5, 1.6
asciilifeform: mod6: it was from early literature, and from dark ages machines with no shift key. and so folx who learn of the existence of lisp from encyclopaedia in school, tend to LISP
phf: then you have a version of LISP for project mac called MACLISP, also capitalized. but it shares so much with LISP that you can call it LISP too
mod6: ahh
mod6: what is the general feeling of 'common lisp' within the republic?
phf: these days when you say "lisp" you usually either mean an equivalent of "algol" or "common lisp" specifically
mod6: (i ask all of the questions because im simply curious, and it seems to be such a broad range of different dialects, if that's even a way to put it)
asciilifeform: mod6: i have nfi what 'general feeling' would mean, there is no homogeneous writhing mass here
asciilifeform: but you can ask specific people.
asciilifeform: to asciilifeform commonlisp is rather like i imagine an old, nagging wife is for old man
asciilifeform: defective in all kinds of ways but it isn't as if another is to be had.
mod6: hmm. ok.
asciilifeform: (nor does the old far necessarily deserve anything else)
asciilifeform: mod6: scheme is neat on paper (~50 pg. exhaustive spec! at least before the poetterings got to it in rev. 6) but you will find that it is ~impossible to program in the language that appears in that document
asciilifeform: without megatonne of 'libraries' that make the overall mass of the system quickly approach and exceed that of cl
phf: scheme48 comes closets to sane lisp imo, it was written on a lisp machine, and comes with a lot of very sane batteries (the library is clearly inspired by cl)
mod6: would it be a worthy project for someone to start to write some essential libs for tinyscheme?
asciilifeform: mod6: theoretically. but, as phf observed earlier in this thread, certain entirely necessary things are impractical to retrofit to it (e.g., bignum)
mod6: i did see that. and yah, that sucks. for lack of a better way to put it.
phf: if you take any of the canonical bignum-in-lisp implementations and port them to tinyscheme, multiplication of one 5 digit number by another 5 digit number takes ~seconds~. optimization is left as an exercise. i suspect some of the routines can be implemented as opcodes.. you could also take that numeric thing asciilifeform extracted out of gnupg and bolt it to tinyscheme as a PoC
mod6: would that alone be worth experimenting with, just to see how expensive it really would be?
asciilifeform: phf: you could, but it would be terrible
asciilifeform: megaturd of c crapolade.
asciilifeform: mod6: i can't speak for others, but i specifically am interested in a ~0-c~ box.
phf: mod6: i did a poc while experimenting with, but i only gotten as far as addition and multiplication. i think taking a stab at writing a canonical bignum-in-scheme is a good idea
asciilifeform: so, if it's an x86 box, this means that no opcode executes on it, from cradle to grave, that didn't come out of (hypothetical) compiler with mandatory bounds/type checks.
mircea_popescu: mod6 there';s a bot you can fuck around with if you wanna try it out.
mircea_popescu: very far from any sort of anything.
mod6: yeah, i saw some of that. neat stuff.
asciilifeform: phf: scheme48 1.9.2 is >300,000 line of c!
mod6: phf: yeah, might be a worthwhile project.
asciilifeform: 100k ln.
mircea_popescu: phf the problem with the bolt-ons being that why are you bolting on, use the language of the boltons instead.
mod6: these automated tests could use some serious cleanup and refactoring -- they work fine, just not very pretty, etc.
mod6: but I'm so glad I wrote these.
deedbot: << Recent Phuctorings. - Phuctored: 1415...7823 divides RSA Moduli belonging to ' (ssh-rsa key from (13-14 June 2016 extraction) for Phuctor import. Ask asciilifeform or framedragger on Freenode, or email fd at mkj dot lt) <>; ' (Unknown DE)
deedbot: << Recent Phuctorings. - Phuctored: 1530...0617 divides RSA Moduli belonging to ' (ssh-rsa key from (13-14 June 2016 extraction) for Phuctor import. Ask asciilifeform or framedragger on Freenode, or email fd at mkj dot lt) <>; ' (Unknown DE)
BingoBoingo: !~bcstats
jhvh1: BingoBoingo: Current Blocks: {"blockcount":449062} | Current Difficulty: 3.36899932795E11 | Next Difficulty At Block: 449567 | Next Difficulty In: None blocks | Next Difficulty In About: None | Next Difficulty Estimate: None | Estimated Percent Change: None
BingoBoingo: !~ticker --market all
jhvh1: BingoBoingo: Bitstamp BTCUSD last: 893.3, vol: 10688.44298373 | BTC-E BTCUSD last: 875.0, vol: 5522.72558 | Bitfinex BTCUSD last: 891.0, vol: 13106.35779268 | BTCChina BTCUSD last: 885.889845, vol: 330382.74660000 | Kraken BTCUSD last: 889.969, vol: 2631.19989002 | Volume-weighted last average: 886.156920849
asciilifeform: re earlier thread, << somewhat interesting
← 2017-01-18 | 2017-01-20 →