log☇︎
208800+ entries in 0.154s
mircea_popescu: screen -x results in a hanged terminal, ctrl-c does nothing, supposedlt active terminals won'\t spawn windows etc. box is otherwise fine though. fucking linux,.
mircea_popescu: nah, it's the box that produced that debug log last week
asciilifeform: ^ over wires, that is
asciilifeform: (afaik nobody tested this yet)
asciilifeform: so this'll test the sync-from-0
asciilifeform: btw i assume this is a brand-new box, there is no blockchain on it
mircea_popescu: hm, i dun think this workt.
asciilifeform: and woah, record-breaking liquishit in mempool today
asciilifeform: (everybody gets the same 1)
asciilifeform: and the user is 'tunnel' just like in the recipe.
asciilifeform: and the ip is dulap's
mircea_popescu: also, architecture-wise : i plan to get my boxes in a ring, and link 1 of you fellows to each. unless objections to the arrangement.
mircea_popescu: slave supposedly ready to go, what ip do i tell addwire ?
trinque: all servers are toilets; ass needs to be able to lay turds elsewhere at a moment's notice
asciilifeform: that being said, you can use 'wire' with anything that can maintain a ciphered tcp pipe between two boxes. dun have to be ssh.
asciilifeform: if i had a ticket to the mythical planet where programmers aren't retarded and 'the shame of poverty' is not smelled, we would not even be having the trb conversation.
mircea_popescu: what can i say. trinque has it.
asciilifeform: trinque: don't put wire-trb on your icbm-controller box. the skull is there for a reason.
trinque: your whole planet being poor does not erase the shame of poverty
Framedragger: yupyup. easy to shame but nothing to replace it with
asciilifeform: mircea_popescu: with what to generate ~ssh~ key ?
mircea_popescu: me too.
Framedragger: i expect most of those broken openssh keys were generated by ssh-keygen
mircea_popescu: "hey, we didn't say it is EXPERIMENTALLY secure now did we! door's that way and don't forget your social security contributions."
Framedragger: mircea_popescu: heh feel free to convert it to openpgp (http://siphnos.mkj.lt/datadrop/crap-from-scans-to-be-sorted/ssh-to-pgp.py) and submit to phuctor.
ben_vulpes: but the pseudo-useful key generator is /cryptographically secure/ mircea_popescu, i don't know what you're complaining about
mircea_popescu reviews teh log.
asciilifeform: it is not married to ssh
mircea_popescu: ok, so now with the built trb with ssh patch...
asciilifeform: because ssh is a product of 'satoshis', people who were dropped as children, and cannot be made to self-reignite.
asciilifeform: you gotta have something that will re-ignite ssh if it drops.
asciilifeform: ( re the recipe : 'autossh' is a c proggy, from heathendom, i have not audited it. this is one of the reasons for the skull'n'crossbones. there is a similar utility called 'RSTunnel' that consists solely of bash script. but i have not tested it . )
asciilifeform: true, it opens you up to 'hm which one do i have on THIS box' etc. which is why i like to keep the buildtron on each.
asciilifeform: there's nothing promisetronic re copying a static bin that you built ~yourself~ to >1 box. (for one thing, it demonstrates that it is genuinely static and will run on heterogeneous boxen.) but you can do it either way.
mircea_popescu: half the point of the original effort that started this whole thing was to avoid just that bad habit
asciilifeform also likes to keep the whole orchestra on his nodez
asciilifeform: then you can. (and for bonus point, compare the binaries, they oughta be exactly same)
mircea_popescu: i guess this'd be it. i intend to rebuild on all of them.
asciilifeform: (and on any other box that will be a wire slave)
asciilifeform: (at the risk of pedantry) first put the newly-build 'bitcoind' binary, on the box that will become the first slave.
asciilifeform: now can proceed to the recipe itself
jhvh1: mircea_popescu: The operation succeeded.
mircea_popescu: !~later tell mod6 list of items for mod6's convenience : 1) nice job! 2) successive presses overfill .gitignore and possibly other files 3) deps directory should prolly be same level as patches and .seals
jhvh1: asciilifeform: The operation succeeded.
asciilifeform: !~later tell phf i was thinking' of buying that 'ivory' and making a signal interposer . any idea whether the magic cpu was soldered down, or LIF socketed..? ☟︎
deedbot: http://phuctor.nosuchlabs.com/gpgkey/B92BDB8947B78A414F8FE580BCFBCE7E1171228C1B5DD26AA5E5BD716796AF08 << Recent Phuctorings. - Phuctored: 1618...4153 divides RSA Moduli belonging to '210.185.87.179 (ssh-rsa key from 210.185.87.179 (13-14 June 2016 extraction) for Phuctor import. Ask asciilifeform or framedragger on Freenode, or email fd at mkj dot lt) <ssh...lt>; ' (210-185-87-179.intrapower.net.au. AU)
deedbot: http://phuctor.nosuchlabs.com/gpgkey/B92BDB8947B78A414F8FE580BCFBCE7E1171228C1B5DD26AA5E5BD716796AF08 << Recent Phuctorings. - Phuctored: 1431...4833 divides RSA Moduli belonging to '210.185.87.179 (ssh-rsa key from 210.185.87.179 (13-14 June 2016 extraction) for Phuctor import. Ask asciilifeform or framedragger on Freenode, or email fd at mkj dot lt) <ssh...lt>; ' (210-185-87-179.intrapower.net.au. AU)
mircea_popescu: stop squirming and don't make trusting noobs rm-rf-nuke the wrong things!
asciilifeform: anyway it's entirely a timesaving measure, you entirely can load them from somebody else each time, the thing won't proceed unless the hashes match
mircea_popescu: what needs fixing is a) for srs! and probably also b) /deps/ dun belong that low.l
mircea_popescu: yes, and they are frozen forever in the place online=1 sucks them from. so this part is correct, and needs no fucking fixing.
mircea_popescu: this is not lazy, this is sane. not every slant of roof gets a tower, because things aren't all equal, and something must be the bottom.
asciilifeform: and you aint backing up 'this box' but the trb deps. that got frozen 4evah.
asciilifeform: mircea_popescu: 'lazy man does the work twice' or how did it go
mircea_popescu: asciilifeform understant : not everyone treats the boxes like his mother. i am of the people who treat them like his socks. no fucking way am i going to backup this. let the box suck it.
mircea_popescu: mod6 that said, /deps/ should prolly be level with seals and patches, not part of /bitcoin/
asciilifeform: mircea_popescu: once you have the contents of 'deps' again, i recommend to back it up to another dir, and to burn a cd, those turds ain't ever changing -- there will only get ~fewer~ (when we ditch, e.g., bdb)
mircea_popescu: ben_vulpes yeah, i'm just giving him a hard time. because for srs!
ben_vulpes: if the makefile is intact, make ONLINE=1 i believe is the flag for autoprocure
mircea_popescu: well, ima start over now. isn't that fucking wunderbar.
asciilifeform: (for some reason mod6 put them inside the dir that gets pressed)
asciilifeform: if you still have the deps somewhere, copy'em over, otherwise gotta download'em
asciilifeform: so cd to it and then 'make'.
asciilifeform: and then run make
asciilifeform: mircea_popescu: now all you gotta do is to move the subdir 'bitcoin' inside this 'trbfoo' to replace the 'bitcoin' dir in trb54
mircea_popescu: so that out of the way,
asciilifeform: mircea_popescu: looks like this worked. confirmed that mod6's vtron doesn't check for file-exists, appends. (i won't even blame him, as such, this is the default idiot behaviour of gnudiff !)
asciilifeform: mircea_popescu: you pasted the .gitignore from inside the resulting dir 'trbfoo' ? make sure this was so ?
ben_vulpes: mircea_popescu: if you press to the makefiles head do you get a concatenation as well?
mircea_popescu: as best i can tell.
asciilifeform: is that what came out of the new press also !
mircea_popescu: well it's what's in there.
asciilifeform: mircea_popescu that looks like the doubled-up one
asciilifeform: ( if you examine bitcoinrpc.cpp, you will see the wire option , for instance )
asciilifeform: should looks like the expected one.
mircea_popescu: let's now examine the .gitinfo
asciilifeform: it should make the dir and press to it
mircea_popescu: you want me to run that ?
asciilifeform: ./v.pl p v trbfoo asciilifeform_wires_rev1.vpatch
asciilifeform: mircea_popescu, let's backtrack, and tellme what happens when
mircea_popescu: hm are we talking at cross purposes ?
asciilifeform: ( there'dve been nothing to build, if it had not pressed )
asciilifeform: it had to , for the build to proceed
asciilifeform: yeah but his vtron pressed to that dir once.
mircea_popescu: i made a special directory for it, that was empty at start.
asciilifeform: nope, remember, you built the binary
mircea_popescu: asciilifeform this was a virgin space
asciilifeform: and then press again.
asciilifeform: rm -rf trb54
asciilifeform: looks like this is what mod6's vtron does instead of overwriting files
asciilifeform: it happened on the 2nd press attempt
mircea_popescu: i have no idea how the binary would have come out then ?!
mircea_popescu: yeah but it has to be teh correct file then :D
asciilifeform: ( and where it won't erase unless the hash of the victim matches! )
mircea_popescu: asciilifeform in fact i can't actually patch it out as it won't match the correct file huh, it does line by line removal
asciilifeform: we have the improper treatment, where the thing gets -------'d
mircea_popescu: which is of course bound to hit the situation where we have no proper treatment of file removal in vtron because old horrors.
mircea_popescu: meanwhile /me will go research how to write a patch to take out .gitignore altogether.
mircea_popescu: yeah, but no, no gains in forking the process. let the foundation item be apical
asciilifeform: ( which iirc mircea_popescu and also hanbot successfully used to build trb )