asciilifeform: BingoBoingo: will come in handy when they switch from yakyak to gladiatorial ring, i suppose
asciilifeform: BingoBoingo: lol! d00d loox just like bezos
asciilifeform: BingoBoingo: recall the infamous wedding cake thing
asciilifeform: ( and if it requires baking a board, process takes coupla months anyway, so why not get the ball rolling )
asciilifeform: phf: i prolly won't get a chance to actually put it to use in near term, hands pretty full with ffa; but would like to have it. if it's there to be had.
asciilifeform: iirc it's been a year+ since phf 1st mentioned it
asciilifeform: ( i half-expected that corlan in '90s would have discovered the non-pointeristic/non-heapistic style asciilifeform works in today -- but apparently nodice )
asciilifeform: yea the astonishing part wasn't the months, but that it finally went
asciilifeform: wonder if something like this, was the culprit
asciilifeform: BingoBoingo: in usa, erry so often one hears of a defrocking of a postman who stashed letters in his cellar instead of delivering, 'will do'em later!', sometimes years' worth
asciilifeform: BingoBoingo: for comparison, round-trip of misaddressed FG to argentina, from asciilifeformistan and back, took slightly under 3 months
asciilifeform: BingoBoingo: it went as message-in-a-bottle and oceanic currents finally changed, or wat..
asciilifeform: ( rsa on z80, 6502, etc. is ~nonstarter, even if one 'banks' the address lines to give enuff mem, the 8bit buggers lack a multiplier, so you get 'egyptian' speed )
asciilifeform: it's prolly the oldest chip, not counting 'heavies' like vax, that can rsa on sumthing like human timescale.
asciilifeform tried to source 68k back in the original draft cardanoism days. found that the only source, aside from vintage collectors, is some goldentoilet co that makes spare parts for usg rockets
asciilifeform: ( iirc BingoBoingo is even today aficionado of 68k )
asciilifeform: 68k was imho a pretty great all-around allpurpose cpu. and -- mega-unsurprise -- ~unobtainable today...
asciilifeform: 'i know how to build chopper, but no steel yet, grr'
asciilifeform: they remind me of old man leonardo and his chopper
asciilifeform: 'A long time ago, in the late eighties, I myself tried to develop such a system, on a home-made 68K with a cp/m executive (the file system was on a second, z80 processor, that had a floppy drive). It's name was 'li' meaning a half lisp (but it had a simple, non-optimizing JIT compiler that directly generated 68K code)...' << this is at least the 3rd fella i've corresponded with who 'i tried to build lispm in '80s out of 68k'
asciilifeform: living off da fat O'da << d00d oughta consider new career, of beggar an' petty thief in e.g. calcutta. at least there could steal actual food, and not 'protein bar'
asciilifeform: i dun recall talmud prohibiting having 'normal sized yard' lol
asciilifeform: (i.e. recycling dead man's name is a-ok)
asciilifeform: then again iirc only if the elder is alive
asciilifeform: BingoBoingo: interesting; it was my understanding that the j00z frown on I,II,IIIism
asciilifeform: diana_coman: it is, and still appears in modern ru, e.g. for pope Иоанн Павел II etc
asciilifeform: 'this letter is for john' 'which' 'VIII'
asciilifeform: also lulzily, some of the groups in usa have this custom, but they also affix numbers, e.g. i knew a few folx who 'III', and at least 1 fella who 'IV'
asciilifeform: ( possibly result of 'acute orthodoxy of the central nervous system'(tm) ? )
asciilifeform: mircea_popescu: 'ioan' sounds like an attempt to oldslavonic
asciilifeform: i was almost at the point where 'it's 5 people and mircea_popescu knows'em all' lol
asciilifeform: mircea_popescu do you know this fella ? http://dan.corlan.net 'Alexandru-Dan Corlan, MD, PhD' 'stochastic models of patient evolution and pharmacologic response' 'high performance computing and programming in R, Ada and CommonLisp'☟︎
asciilifeform: btw, diana_coman were you able in the end to make sense of the http://www.loper-os.org/?p=2675 proof ? ( mircea_popescu's s.mg broadcast implies that yes -- but thought i oughta ask conceretely )☟︎
asciilifeform: type cast is the single most dangerous operation there is ( second only to pointerism, and notice how ada makes the latter extremely painful ) hence the thing forces you to 'measure 7 times, cut 1ce'
asciilifeform: when you do this kind of thing, you gotta make sure that the proggy behaves correctly regardless of what rubbish may have been put into the octet form that gets transformed to the record. ( possibly this is obvious, but imho bears restating )
asciilifeform: diana_coman: it's safe (i.e. endianism-proof) if you actually nail down the layout of the record (see barnes re how)
asciilifeform: there's exactly 2 ways to do it -- the above conversion mechanisms, or the stream (read nqb to see how) . cuz transmutation of data types is a specifically dangerous act, that is deliberately not made easy in ada
asciilifeform: hence in 'nqb' , i used a (custom-sewn) stream
asciilifeform: according to the 'rationale' document, serialization is ~the~ reason why streams are even a thing
asciilifeform: but as for the q of 'can haz this without any smell of c sulfur' , the answr is no, because entire concept is slightly antithetical to adaism, and so you gotta 'cheat' at least a little in order to permit it.
asciilifeform: i suppose the 3rd way would be to write simple serializer ( that uses System.Address and yourthing'Size ) to turn your record type into (and out of) the requisite # of octets, and compartmentalize all of the System.Address'ism in 1 module. i have something quite similar to this in my mmap lib.
asciilifeform: diana_coman: either this, or the 'unchecked conversion' you had earlier (given that you specify a fixed size for the record, you will always end up with the given number of octets, it has no failure mode)