log☇︎
17000+ entries in 0.006s
asciilifeform: it's a sisyphian proposition.
asciilifeform: ( or, alternatively, marrying intel 4evah )
asciilifeform: bellard himself abandoned it, not the least reason , near as i could tell, was that he did not relish idea of maintaining 9000 backends
asciilifeform: nao it resembles minix, 100x the mass of the original, and ~less~ function
asciilifeform: seems to have disappeared. and tcc was 'picked up' by fungi.
asciilifeform: but worth a shot ( tho asciilifeform's track record for 'write 'im a mail' is currently abysmal )
asciilifeform: iirc old
asciilifeform: afaik then he sank into the sea, never continued.
asciilifeform: he had a kernel-building compiler in coupla 1000 loc .
asciilifeform: http://btcbase.org/log/2017-03-23#1631830 ☝︎
asciilifeform: !#s bellard
asciilifeform: btw do you recall fabrice bellard's compiler ?
asciilifeform: a la mpi etc
asciilifeform: aand , i suspect, by cutting coupla MB of entirely dead mass
asciilifeform: nursing old gnat aint optional, currently, just like nursing trb aint optional
asciilifeform: granted. sorta why i said 'it's a trb'
asciilifeform: q is, do you actually ~want~ a thing that shits out 200kB of peepholed x86 ??? for 'hello world' .
asciilifeform: it's at the mpi confiscation stage atm.
asciilifeform: ( btw if anyone else sees a dead link, plox to write in , i do fix )
asciilifeform: will surely update.
asciilifeform: luckily asciilifeform has last-known copy. will mirror if anyone wants.
asciilifeform: loox like moved to shithub
asciilifeform: RIP
asciilifeform: mircea_popescu: seaforth dead ??
asciilifeform: or the 'bus width is handful of bytes' nonsense
asciilifeform: for instance -- the expectation that ram contents vanish when plug is pulled ( entirely avoidable since, what, late '90s or so ) ; or that machine is permitted to crash ( subj discussed in detail in log, crash oughta be seen as same calamity as catching fire , and this pov is entirely practical )
asciilifeform: mircea_popescu: asciilifeform's observation is that all historic attempts at 'pure (b)' were gimped by low expectations from iron, on acct of author's lifelong crippling experience with shit irons
asciilifeform: and fails to provide locking in iron primitive, which gives the fungal bloom of software threading deadlock etc
asciilifeform: in that e.g. it imposes ridiculously small bus widths, gimping rsa;
asciilifeform: currently could almost say 'iron fails math'
asciilifeform: 'proper primitives' is not independent of a) what backing iron b) what operator expects to ride on top (langs, memory management style) tho.
asciilifeform has entire site about this..
asciilifeform: aha, exactly so
asciilifeform: so happens that i have a page re subj : http://www.loper-os.org/?p=256 .
asciilifeform: aha!
asciilifeform: the unix 'tower of layers', nominally separable, is spurious ( but mircea_popescu knew this ).
asciilifeform: ( can has boat hull that dun leak ? )
asciilifeform: the central imho open q is, how much of the iron's retardation can be kept out of such a system.
asciilifeform: mircea_popescu: in pctardation land, it's a consequence of the (massively failed, historically) attempt to hack around the fact that erry peripheral has free run of entire bus
asciilifeform: ( has : translator, from various preferred lang syntaxes to the actual machine coad, but without massive expansion of the ast )
asciilifeform: in asciilifefor's cosmography, sane machine has neither 'linker' nor 'compiler' in the customary sense
asciilifeform: right
asciilifeform: rather than '1 moar layer'
asciilifeform: asciilifeform's hypothesis, is that it is only worth to add a layer if you can remove dozen by doing so
asciilifeform: moar concretely, e.g. memory allocator -- attempts to put sane one on top of linux's, all caught fire in very similar way, i.e. impedance mismatch ( whether libc's or java's or commonlisp's )
asciilifeform: it's a forced failure, in the sense that it is imposed by trying to be a layer on top of os, rather than os per se
asciilifeform: i'm not currently certain that there is any meaningful overlap.
asciilifeform: ( glibc & co. do not handle devices, impose memory organization scheme , etc )
asciilifeform: mircea_popescu: it isn't entirely same problem
asciilifeform: the appeal of 'emulate sane iron' is partly 'now can has readable disasms for errything', and partly 'can then bake actual iron around same'
asciilifeform: mircea_popescu: in practice, given static link, it's duplicated 9000 times, rather than single.
asciilifeform: i suspect not only cheaper in manpower, but faster in wall clock. cuz can hand-massage emulator, whereas compiler barf , even with peepholing etc, is invariably bloated.
asciilifeform: consider from pov of 'what is total mass of binary that has to be audited'. in which case is the mass smaller : where there are 9000 bins on the machine, each consisting 90% of a coupla MB of 'bounds check this array' and 'propagate this exception' ? or if the 1 runtime does these, and the bins -- invoke.
asciilifeform: rright but you want to put the sweat into item that gets leveraged, rather than duplicated 9000 times, is the notion.
asciilifeform: i suspect cheaper than 1 sane gcc likewise.
asciilifeform: ( see e.g. http://btcbase.org/log/2019-02-17#1897833 ) ☝︎
asciilifeform: i suspect that it's a ~smaller~ problem than 'bake 17 sane gcc-like back-ends'
asciilifeform: it's a quite serious set of open problems, many documented by asciilifeform in the log (e.g. how to drive nic?)
asciilifeform: there is no magical stone.
asciilifeform put ~decade into various attempts at $subj, but ~0 forward motion until made the sage breakthrough and gained ability to debug bios. but has been on back burner in preference to moar urgent tmsrisms ever since
asciilifeform: ( specifically -- is Right Thing ? and if yes, then how much sweat is worth to put into keeping ye olde gcc alive, and how much sweat instead for same ? and what oughta be in it ? )
asciilifeform: at some pt before upstack thrd gets lost in the sands of time -- would like to see mircea_popescu's ex cathedra pov re http://btcbase.org/log/2019-02-20#1898145 , when he has the time ☝︎
asciilifeform: i suppose if suit already brought, likely there's already a thiel.
asciilifeform: aa hah
asciilifeform: mircea_popescu: wouldn't this require a thiel to show up to bankroll $kid's suit ?
asciilifeform: mircea_popescu: thiel killing gawker, sure ; rando kid vs volkischer beobachter ? will be odd if it dun get 'appealed' for 25 yrs, until $kid dies of hunger
asciilifeform: i suspect also that erry sargeant in erry unit on planet, also could summarize compactly.
asciilifeform: re the penicillin -- also compactly summarized in http://trilema.com/2014/the-all-american-asshole-in-his-own-words-with-my-own-notes/#selection-3501.0-3501.226
asciilifeform: ( cabbage -- much moar toxin by mass, but we have the enzymes )
asciilifeform: fulla poison. animal kingdom dun sit still tho, which is why we think of e.g. belladonna as poisonous but not cabbage
asciilifeform: BingoBoingo: virtually errything that grows on satan's green earth
asciilifeform maintains a kind of purgatory for these , in #a, and it is quite cheap on acct of the fact that usually only takes'em a coupla hrs to realize that no, this aint the fungal farm, fungal farm is three doors down, and then go to where their place is.
asciilifeform: the benighted folx who 'wai is mircea_popescu so eager to hit us with spiked club', i can tell they never cell cultured.
asciilifeform: funnily enuff asciilifeform was just recently describing to pet , how mircea_popescu's pesticide worx, and finally clicked when i 'recall cell culture ? when yer growing eukaryotes, you add antibiotic, or guess what dish will be full of next day' ☟︎
asciilifeform: http://btcbase.org/log/2019-02-20#1898163 << could be quite interesting read. didjaknow, for instance, that the javatards got showered with enuff printolade that they baked chips ? and that none of these were in any sense commercial success ? ☝︎
asciilifeform: also matters to actually plow the field, and even matters with what (iron beats wooden stick) and so forth.
asciilifeform: http://btcbase.org/log/2019-02-20#1898157 << i dun even disagree that nuffin's worth doing if you don't kill the locusts first. but also imho gotta other things , the result of 'pesticide uber alles' agriculture is monsantostan. ☝︎
asciilifeform: mircea_popescu: i also ignored. but imho it's an interesting entomological case.
asciilifeform brb, teatime
asciilifeform: or the deathbed bolix people, who did ~99% of the job but sank on the titanic of the dec alpha, the only arch they managed to massage into compliance before ran out of dough
asciilifeform: or gabriel_laddel, who thought that linux kernel is an acceptable item to build the '256kB' out of, through power of magical thinking
asciilifeform: various folx attempted something roughly like this, but succumbed to own idiocy -- e.g. the java people, who 1) wrapped their 'vm' around the nonsensical shape of their tardlang 2) stopped short of 'and no this doesn't run on your tard os, is IS the os, fuckyou' ; ☟︎☟︎
asciilifeform: nao tbf, this approach was not available to the gnu people, on acct of sheer poverty of 1980s, when erry byte & cycle counted, and 'must not waste' , yer entire memory space in 1985 might be 256kB. but today plenty of 'what to waste', why not 'waste' for actual gain ( like bitcoin 'wasted' mains current ) , rather than microshit/gnu-style.
asciilifeform: i strongly suspect that any other approach to the problem (i.e. 'we'll do exactly what the gnutards did, but they were mice and we are men' ) is certain to degenerate into ~same rubbish as what the gnutards ended up with.
asciilifeform: yer still stuck with certain idjicies of iron (e.g. dma) but then when finally bake new iron, you have a ready compiler & userland to plant on it, that actually behaves.
asciilifeform: the Right Thing will look roughly like a 256kB chunk of asm that the box boots straight into, and afterwards forgets that it's an x86, arm, etc. ☟︎
asciilifeform: as in http://www.loper-os.org/?p=165#selection-79.0-85.510 .
asciilifeform: it was this very front that asciilifeform published the sage thing for.
asciilifeform: aaha
asciilifeform: absolutely that.
asciilifeform: i.e. http://btcbase.org/log/2018-10-29#1867124 . ☝︎
asciilifeform: i hold that the answr to the http://btcbase.org/log/2019-02-20#1898121 quandary depends on whether the right cut of that knot is made. and imho ^ is that cut. ☝︎☟︎
asciilifeform: rather than bloating erry single binary on the box with identical chunks of sanitycheck
asciilifeform: all the bounds checks, exception handling liquishit, etc. that gnat shits out, oughta be happening in either silicon ( which we aint baking yet ) or failing that, in the 'sane arch emulator'
asciilifeform: this'd cost, yes, some % of horsepower, but in exchange gets a 100klok toolchain instead of millions of ??? loc
asciilifeform: really oughta have ~1~ compiler, and for a sane arch; and if yet cannot afford to actually siliconize the sane arch, then the sad iron oughta be booting straight in bios to a compact (asmistic) emulator of same ☟︎
asciilifeform: i also suspect that on modern irons, 'retargetable compiler' is plugging the wrong end of the funnel.
asciilifeform: it's a standard 'spittoon in one strand' : for so long as you retain ~any of the c liquishit, yer stuck with gcc ( which is the only thing that builds it , the 'portability' of the lang is a fiction from day 1 )
asciilifeform: i fully expect ^ adds up to ~100% of the c opensores mass, yes
asciilifeform sees linux kernel, gcc, ftr, as 'life support' item, rather like trb -- worth freezing and maintaining until proper replacement, but not really items with a serious future