asciilifeform: BingoBoingo: i know for a fact that there exist pdp8 in actual cement, last i knew still processing inputs . so it aint entirely out of the question.
asciilifeform: ( ditto gnat. the troo adatron would be a ~50kloc bolix-style ada on sane iron support, rather than crusty old gnat )
asciilifeform: ( or for that matter any other unix, ditto )
asciilifeform: i cannot speak for others, but from my pov cuntoo is a wartime t34 , and defo not something i like to picture myself using 20y from nao
asciilifeform: characteristically of people who set out to solve the Wrong Problem , they will trip over own feet 9000 times and learn 0
asciilifeform: bvt: was this 'can' accompanied by a 'did' ?
asciilifeform: i.e. to go through the sweat of building an arch, only to then plant kernighan and poettering on top again.
asciilifeform: i suspect that the 'let's bootstrap c' thing comes from a narrow fixation on thompsonism to the exclusion of the other 9000 leprosies of the pdp age.
asciilifeform: ( on the other end of the barf -- picture, C is what they want to bootstrap. because somehow can have sanity while preserving GB of pointerfandangoandcrashlang in the stack, somehow . )
asciilifeform: the notion that 'all irons are the same' , while correct in a narrow mathematical sense , has done a good bit of damage . arguably turned entire field into garbage, at that.
asciilifeform: but author rarely cares to think about 'later'
asciilifeform: the result is ~always~ a massive ball o'shit ~later~
asciilifeform: all of these folx, afaik, do ~exact same thing, they take the school cpu from their kindergarten textbook and proclaim 'hey this fits in head, let's bootstrap on it'
asciilifeform: picture if i had made ffa deal in 32bit words, and then proclaimed 'implement bignum on that'. do you imagine the result would be in any sense 'fits in head' ?
asciilifeform: n iron -- you get 9000x the mass later.
asciilifeform: http://bvt-trace.net/2019/03/mes-part-1-stage0/#selection-29.94-29.340 << imho ~100% of the attempts on record , made exactly same mistake -- they assumed that 'architecture-specific aspects creep into the design of the boostrapping process' only concerns ~what is there~ in the arch, and not ~what is not there~ (e.g. sane memory management, type tags) . if you dun put the complexity of certain necessary sanities where it belongs -- i☟︎☟︎
asciilifeform: ( it's entirely ok as illustration of forthism, but absolutely terrible from 'bootstrap' pov , for the reason described in linked piece )
asciilifeform: mircea_popescu: entirely too soon. i'm not convinced that we want the flaglessness, for instance (mips buys with it , the pipeline, but do we even want pipeline)
asciilifeform: e.g. 'The first in San Francisco was co-hosted by CDC member Adam O’Donnell, an entrepreneur and a security engineer at Cisco Systems, and Alex Stamos, then the chief security officer at Facebook ... Chris Wysopal, a L0pht veteran who founded tech company Veracode ... Seattle fundraiser for the Senate race ... ' etc
asciilifeform: that one afaik 100% 'defeated self', ~100% of the 'graduates' turned into yarvins of 1 kind or another
asciilifeform: 'We weren’t deliberately looking for hacking chops. It was very much about personality and writing, really.' << 'rezistenta prin cultura'(tm)(r)!
asciilifeform: ( moar spoilers : the other added item, other than 'registers' (discussed previously) is 'verdict'. all tapes produce a verdict, it is returned to calling process under unixlikes )
asciilifeform: to use ,/; as 'goto' is stylistically prohibited, if tape leaves control stack nonempty at the end of the life, operator gets a warning.
asciilifeform: the applicable traditional mathematical model for peh is 'turing with finite tapes'(tm)
asciilifeform: this is 'forth-like' just like the rest of the mechanism.
asciilifeform: controlstack is over- and under-flow proof , similarly to the data stack from ch4-16.
asciilifeform: ( ';' is meant to be used with subroutines (ch18) but can also be used to make 'infinite' loop, as illustrated in http://btcbase.org/log/2019-03-14#1902388 . 'life' being typically finite, program still halts )☝︎
asciilifeform: btw , crib sheet for http://btcbase.org/log/2019-03-14#1902386 -- ':' puts current position on control stack; ',' pops a fz from datastack , a tape pos from controlstack, and if the fz was nonzero, control is transferred to the popped pos. ';' simply pops the next pos from controlstack.☝︎
asciilifeform: anyway if neither mircea_popescu nor anyone else files an objection today, it'll be 256 ( at least for the next few chapters ) .
asciilifeform: ( for n00bs/readers -- ffa has fully deterministic memory footprint, by design, and 100% of what it'll use, lives on the stack, e.g. under unixlikes, setbrk() is not invoked )
asciilifeform: at one time i was considering to make it adjustable on cmdline, but already have 4 params the user must give ( fz bitness, datastack height, max 'life' , and tapespace len ) and that's already much, calculating exact footprint for given invocation already requires thought
asciilifeform: orig. i had it set to 8, but decided that's unjustifiably spartan
asciilifeform: recursion is incidentally prohibited, so there aint much point imho in gigantic controlstack
asciilifeform: would like to discuss before i sign the ch
asciilifeform: i'd like to request 5min of mircea_popescu wearing papal regalia . ch17 ffa , as perhaps was clear from http://btcbase.org/log/2019-03-14#1902386 , has a control-transfer stack ( for loops & subroutines . ) it, like errything else, is nonheapistic, therefore has a fixed size. currently set to 256. this constant gotta be nailed down. is 256 enuff? too small? too big ?☝︎
asciilifeform: could try eating a bite of 'food' for instance
asciilifeform: BingoBoingo: he took iirc the 2 mo., as incidentally described by the 'ru argentina' d00d i linked to last mo. he wrote that ~100% of ru expats go from 'i want to be buried in recoleta' to 'burn it all' in ~2mo.
