asciilifeform: ( for non-ru folx -- traditional grammar school example of why gotta have correct punctuation, 'execute cannot spare' , toggleable opposite meanings with comma . anybody know of an engl-compatible version ? )☟︎
asciilifeform: 'Discussion regarding article about apple and amazon denied that Chinese spies implemented backdoor chips into hardware' << uhm, 'казнить нельзя помиловать'☟︎
asciilifeform: aha. and it should be possible to operate a cuntoo box with sporadic, or no, net pipe.☟︎
asciilifeform: ( my primary objection to a hypothetical hybrid, is that i often do work on boxes with no nic, and imho builder must not rely on connectivity )
asciilifeform: even from existing gnu liquishit, it is fairly easy, turns out, to remove ( e.g. as i did with the mpi that diana_coman made into the 1st half of eucrypt )
asciilifeform: ( i have said before, would like to see the autoconf atrocity properly disappear into an unmarked grave )
asciilifeform: trinque: gprbuild is replacement for gnumake/autoconf liquishit, not so much for portagetron per se.
asciilifeform: trinque: i also have visions of integration with gossipd, where the thing, given a wot key, will know how to ask owner for $source item. but this is yet a bridge too far.
asciilifeform: trinque: trashed, i'd hope, eventually
asciilifeform: 1st thing is to freeze the old tars, then slowly can vtronicize.
asciilifeform: it was my understanding that it is eventually to happen. but i do recall that trinque's current item dun have it yet.
asciilifeform: trinque: was contemplating vtronic mechanism here, rather than 'cute' rename of the old emerge
asciilifeform: ( i do not know if trinque is going with the old gentoo-flavoured terminology, but expect that process will be roughly similar )
asciilifeform: way i picture it, is on cuntoo, equipped with wot keys, all you'd have to do is e.g. 'vmerge phf-v', 'vselect phf-v' and you get his latest.
asciilifeform: but even if not changes, but minor refinements, e.g. the graph plotter -- n00bs will end up with old vtron, and then gotta press a newer
asciilifeform: if it really is a certainty that the format will never change again -- then yes
asciilifeform: keeping 2 separate forms of a nominally-identical proggy in sync, is a pain
asciilifeform: diana_coman: not having to have a separate signed-tar distribution, really is all.
asciilifeform: ( and i suspect it'll be a while before e.g. cuntoo is distributed entirely without gnupatch )
asciilifeform: diana_coman: both old and new-form vpatch can be pressed manually with gnupatch, sorta like old cars could be started with crank if you really had to
asciilifeform: diana_coman: i suspect idea was 'make him manually gpg --verify ... and then press by hand-gnupatch a la pre-v trb, better than signed tar'. but i'ma let phf clarify.
asciilifeform: diana_coman: as i understand q was 'how to give n00b his 1st vtron'
asciilifeform: ( either its deletion, or its insertion -- therefore impossible to encode the contents of that tar, as a genesis )
asciilifeform: tangentially on-thread, it still bothers asciilifeform that he was unable to represent the diff b/w mircea_popescu's bitcoin-0.5.3.tar.gz and the genesis as a vpatch ( neither orig v nor current is able to represent the deletion of binariolade... )☟︎
asciilifeform: even setting aside 'if phf made mistake', it is not physically impossible for bit to flip on his hdd at the moment of signing regrind.
asciilifeform: cuz meat is a poor substitute for diff util. ( and diff util, in turn, is poor substitute for continuous v-sequence. )
asciilifeform: reread is great, but it is important for the work to actually be ~cumulative~
asciilifeform: ( phf's item, in particular, incorporates a heavily cut but still pretty crufty ball of gnu c. this is not a stab at phf , but imho is a pertinent fact re 'cost of reading' )
asciilifeform: this worx great for theorems, algos, but not so much for MB of coad soup☟︎
asciilifeform: recall the thread with the tabs & spaces.
asciilifeform: so from that pov , if it is possible to inexpensively avoid the situation, it is a win.
asciilifeform: asciilifeform does not have a stable of slaves to manually diff things. and so if at any point it appears to be necessary to hand-diff something, he is stuck doing it with own hands, and on that day, week, month, nuffing else of any use will come off conveyor.
asciilifeform: mircea_popescu: in re regrinds in particular, i am more thinking of mistakes, than lies
asciilifeform: but imho this loses something valuable.
asciilifeform: phf can, of course, release anyffing he likes, incl. to take an old proggy and make new genesis, and say 'any relation b/w this new and that there old, is happenstance'.
asciilifeform: maintaining hard-continuity of history is nearly whole point, as i see it , of vtronics.
asciilifeform: or does mircea_popescu like those german grenades without pins.
asciilifeform: what state. simple mechanism, like pin.
asciilifeform: phf can make mistake and sign a broken piece, just as asciilifeform on at least 1 occasion did
asciilifeform: mircea_popescu: a mechanically verifiable fact, when verifiable trivially, is superior to promise from anybody.
asciilifeform: if it is actually same, this is trivially verified. but gotta have the simple mechanism there, rather than 'go write bash script'
asciilifeform: and it will be. but good chunk of the win from v, is to avoid offloading the work of 'is this the same or not' to meat, when it can be done 100% reliably by the mechanism.
asciilifeform: mircea_popescu: 2 boot sequences -- one for machines with old v ( and for dedicated historians who want whole history ) and other for the rest.
asciilifeform: but also gotta have a tar, there is no reason to force newcomers to set up ancient v prior to getting hold of the current.
asciilifeform: mircea_popescu: iirc phf previously had a thing that pressed with old v, to create the new. imho this is proper.
asciilifeform: whereas it is the ~hindu~ who had the moar accurate picture -- that , sure, 'errybody soul', but some 70-90% have 'soul' from cockroach with 486dx upgrade pack
asciilifeform: mircea_popescu: imho simpler. the folx in question were labouring under ye olde christian delusion, where 'errybody got a soul'
asciilifeform: recall alan kay and his proto-ipad, 'if they get tabletron, children will learn to logic'
asciilifeform: the porcine idjicy of 'mass', was imho quite evident long ago, but 'tech' folx lived in denial.
asciilifeform: mircea_popescu: also entertaining , to read the thoughts of the folx baking early television, '20s, 'errybody will watch great theatre!'
asciilifeform: mircea_popescu: as i understand, it was actually ~roll piano~ that nuked 'entertain self' music, rather than the scratchy phonograph of the time☟︎
asciilifeform: http://btcbase.org/log/2018-10-06#1858813 << this, btw, is quite ordinary picture in recent ~decade of iron. given as even humble nic is nao a multicore arm thing, and has own kernel, and sometimes whole MB of internal state, fulla vendor shitware; so yes, can wedge .☝︎
asciilifeform: mircea_popescu: 'nuland help' was going pretty strong in 1980s.
asciilifeform: i.e. forgot commandment, 'first, do no bloat'
asciilifeform: imho can tie the two threads together -- when linux folx stopped giving a shit about kernel size, they went to the same hell as when the physics people ditto
asciilifeform: mircea_popescu: by my reckoning, some kinda pustule popped open at 2.7 and linus's merry men started taking liberties
asciilifeform: mircea_popescu: i dun know if i have the keys to where they keep the real semiology ( if mircea_popescu says that it exists , i'll believe him ) ; as it is, i hear the word and think 'lacan' and reach for luger
asciilifeform: http://btcbase.org/log/2018-10-06#1858800 << they switched the userland abi, but 2.4 continued to live ( i've found that i still have at least 1 device in service where not only 2.4, but 2.4 that post-dates the release of 2.6 ) for many years, as it was moar compact and suitable for small iron than 2.6 ( which had better support for e.g. smp , but overall hog vs 2.4 )☝︎
