#asciilifeform | 2020-08-04

newland0: dpb is here afaik because he subscribes to my isp .

11:30

newland0: what are you interested in learning ?

11:30

!w poll

11:30

Polling 12 nodes...

11:30

205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.082s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

108.31.170.3:8333 : (pool-108-31-170-3.washdc.fios.verizon.net) Alive: (0.107s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210 (Operator: asciilifeform)

11:30

205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.107s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

205.134.172.26:8333 : Alive: (0.142s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

205.134.172.27:8333 : Alive: (0.144s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210 (Operator: asciilifeform)

11:30

192.151.158.26:8333 : Alive: (0.152s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

208.94.240.42:8333 : Alive: (0.197s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

143.202.160.10:8333 : Alive: (0.236s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

213.109.238.156:8333 : Alive: (0.335s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

188.121.168.69:8333 : (rev-188-121-168-69.radiolan.sk) Alive: (0.348s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

103.36.92.112:8333 : (terebe.ns01.net) Alive: (0.585s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=642210

11:30

176.9.59.199:8333 : Busy? (No answer in 20 sec.) (Operator: jurov)

~ 6 hours 36 minutes ~

18:07

!q seen-anywhere whaack

18:07

whaack last seen in #ossasepia on 2020-07-30 15:52:31: erm...I ran `make ONLINE=1` and have a running bitcoind, I'm not quite sure that it is statically linked

18:07

!q seen-anywhere jurov

18:07

jurov last seen in #therealbitcoin on 2020-07-28 01:50:42: 59.199 restarted, this time with debug info for next time

~ 2 hours 2 minutes ~

20:09

asciilifeform: curious about cryptocurrency history and about the tech itself

20:10

like V which I asked a bit about here: http://logs.nosuchlabs.com/log/ossasepia/2020-08-01

~ 1 hours 2 minutes ~

21:13

newland0: i wrote the orig. 'v' in 2015. feel free to ask q's.

21:23

what is your take on why you wrote this? seems like it would lead to greatly increased fragmentation as compared to using something like git

21:27

newland0: what do you mean re 'fragmentation' ?

21:28

many different versions of everything, instead of one that is maintained by someone trusted and forks only happen when they really need to

21:30

newland0: are you seriously proposing that the 9000 forks of every insignificant piece of shit on shithub were made 'when really needed' /

21:30

?

21:31

lol not really

21:32

as for 'trusted', i wrote 'v' for use with trb ( see the initial patches here ) . where absolute attributability of all changes, followed by simplicity of the versioning system, were the primary design principles.

21:32

initially we simply posted signed unix .patch to the mailinglist.

21:32

following people instead of projects?

21:33

no concept of 'following' is built into v.

21:33

right, that's what i was trying to get my head around

21:33

newland0: see the orig. usage examples.

21:34

newland0: and, ideally, read ben's classic intro text re subj.

21:34

seen that, so you pick your trusted people and you have to know where to get new patches based on where those people usually publish them

↖

21:34

newland0: correct. how to get patches from their disk to yours is entirely outside the scope of vtronics.

21:35

can be by carrier pigeon, if you want.

21:35

there is no networking code in my orig. vtron.

21:35

this was a deliberate choice.

21:35

v can operate on a box w/out a nic installed. (and i frequently do.)

21:37

newland0: the simplest way to summarize 'v' is -- a vpatch is exactly the same as a standard unix patch (diff -uNr ... ) except 1) it is expected to carry one or more pgp signatures from author + those who fully understood it 2) contains 'before and after' hashes (in the present version, keccak) of EVERY delta. i.e. 'fuzz' is prohibited.

21:38

if publishing and fetching changes is harder then there will be less changes

21:38

all the liquishit traditionally found in versionatrons ('check out', 'check in', 'merge') are DELIBERATELY omitted.

21:39

newland0: didja catch the earlier para. re 'absolute attributability of all changes' ?

21:39

yep

21:39

what happens when you do want to merge two lines of development together?

21:39

so then will understand that i specifically did not give a damn re 'make easy for casual derps to contribute'. entirely the opposite.

21:39

merging is done 100% by hand.

21:41

trb (and all other projects asciilifeform involved with) ~entirely deliberately~ did not have a goal of 'mass participation'. any more than you want 'mass participation' in writing firmware for boeing 757.

21:42

participation was and is deliberately limited to competent folx ~willing to sign every line~ they contribute (and read every line they base contributions on, or deploy.)

21:43

sure

21:43

I am way too used to thinking "how will an idiot break this" in the stuff I work on

21:43

21:43

(trilema) 2014-11-14 asciilifeform: undata: do you know how soviet nukes and spacecraft were assembled?

21:45

^ ~that~ was the model on which i based 'v'.

21:45

imho -- to work w/ safety-critical software in any other way, is irresponsible .

21:47

returning briefly upstack -- 'how to pick trusted people' is a problem outside the scope of a versioning system. but the key concept is that all-deltas-signed ~makes this possible~ to begin with. (and nothing else can make it possible, unless, like sov. nuke base, you in fact lock all of the candidates in a jail , incommunicado, and authenticate their authorship

21:47

Logged on 2020-08-04 21:34:31 newland0: seen that, so you pick your trusted people and you have to know where to get new patches based on where those people usually publish them

21:47

physically.)

21:48

newland0: are you familiar with 'heartbleed' ?

21:49

yes

21:49

my opinion: something like that is only possible after years of fucking up software in the usual way, bloat and garbage making the whole thing uncomprehensible

21:50

newland0: then do you notice that no system other than 'v' can provide a button whereby 'i want NOTHING that nsa mole touched in my code EVER' ?

21:50

assuming that gpg and your hardware are trusted

21:51

the 'pain' of using 'v' is in fact a great incentive to keep code compact (the tradition is that patches must be short enuff for human-readability) and infrequently changes.

21:51

(trilema) 2016-01-18 ascii_butugychag: btw i hope everybody understands that life with 'v' is always going to resemble dark age blood sports like cvs, etc. far more than modern greased poles (e.g., 'git')

21:53

newland0: the gnarl of traditional gpg , and the garbage hardware it runs on, are problems that i've worked on for many yrs. the former, i am attacking by slowly baking a fits-in-head replacement ; against the latter, built a RNG , and continuing to work on yet-other irons even nao.

21:56

that would be some of the starting points

21:58

the non-negotiable starting points are 1) purge the fits-in-no-head liquishit 2) give attributability to all changes, and give a way for people to attest to having read & understood an artifact.

21:58

'v' is the solution (imho, all ~possible~ solutions are isomorphic to 'v') for (2) .

22:00

newland0: history of trb is a practical illustration of approaching 1+2 for a given artifact.

22:00

i.e. the transformation of a piece of shit, found at junkyard, into an item which can be understood and attributed.

22:01

otoh, e.g. ffa was built vtronically from birth.

22:04

http://btcbase.org/patches looks like good tooling for exploring one of those

22:04

but can't rely on it much or that becomes the weak point, right?

22:04

revisiting upstack -- indeed if you have a slave galley, with 500 indiancoaders, you will have a hell of a time using 'v'. but my position is that projects with >dozen or so participants, simply ~must die~.

22:05

if you have "outsourced" code then you are getting shit no matter what.

22:06

the problem isn't 'outsource', but the very idea that hundred+ people can 'write software together', is crock of shit

↖

22:06

it is fundamentally why e.g. linux (and all other current-day softs) are rubbish.

22:07

newland0: the btcbase tool was written by phf, a fella who afaik went off to live in the woods in ru and very rarely updates. he has so far taken care to keep the box running, however, you can use it to read historical vtrees .

22:07

unless what shows on that page is being intercepted and rewritten

22:08

is what the pgp sigs are for.

22:08

the intent was never 'cut&paste coad from phf's www viewer' at any point. it is a convenience tool for reading.

22:08

authoritative copies of $item are whatever you have not only downloaded to own box, but verified the sigs on.

22:09

( the q of 'how to get canonical public keys from the people you work with' is outside the scope of software, it is 'human' problem, generally resolved via combination of in-person meetings , WoT, and the simple act of following the work of a (pgp) identity over the years. )

22:09

right, but you have to make sure you read the same version that you verify the sigs on

22:10

naturally

22:10

this is what vtron is largely for to begin with.

22:10

so btcbase is for quick overviews, no more

22:10

aha

22:11

and only for when you're in no hurry (phf often takes coupla months to get around to adding a patch from someone's site to it)

22:11

another thing I noticed: vpatches don't have timestamps?

22:12

nope.

22:12

also deliberate.

22:12

timestamps are intrinsically bogus, in my lights.

↖

22:12

( gpg does bake timestamps into signatures, but i have no interest in them, they do not authenticate in any sense )

22:12

signer can make'em whatever he wants'em to be.

22:13

right, but trusted signers

22:13

and not for authentication but for a heuristic "is this thing still living"

22:13

newland0: trinque's 'deedbot' is the tool to use when you are interested in authenticable time.

22:13

http://wot.deedbot.org/312D6A6806B515864E3843C4FD6E2F17FF418772.asc

!!key newland0

22:13

deedbot

22:14

*

asciilifeform brb

~ 26 minutes ~

22:40

paraphrasing from some of the stuff in that web of links: "computers should be made of user-understandable, interchangeable parts, like ak's"

22:41

pretty much 'the' way to un-fuck computing

~ 17 minutes ~

22:58

thx for discussing/explaining. off for now

22:59

gregorynyssa

asciilifeform: I followed your link above and saw this. glad to finally encounter someone else familiar with New Lies for Old (1984). it was a revealing book.

22:59