341500+ entries in 0.212s
mircea_popescu: juist like numerically
there is no way
to distinguish between a
transaction
that was signed low and one
that was signed high and malleated
to low.
mircea_popescu: numerically,
there is no way
to distinguish between a
transaction
that was signed high and a
transaction
that was signed low and malleated
to high
ben_vulpes: i create a
transaction with a low s value. someone listening
to broadcast
transactions can rewrite my
tx
to have a high s value.
mircea_popescu: if you mine yourself, on your own coin, it is possible
to do anything.
ben_vulpes: if you have access
to
the gradebook...
mircea_popescu: in what sense ? paper will carry anything.
the social practice of
throwing curves won't. etc.
mircea_popescu: "is it as possible
to
turn a grade from 50%
to 75% as it is
to
turn it from 75%
to 50%" ?
ben_vulpes: in
that it's possible
to malleate in either direction. ease i suppose doesn't figure into
the question.
ben_vulpes: mircea_popescu: meaning
that
transactions can only be malleated from high->low order?
mircea_popescu: phf does
the archive.is implementation work for yo u ?
ben_vulpes: can someone confirm
that it's just as possible
to malleate a
transaction low->high s as it is
to malleate it high->low ?
phf: fwiw stock selection magic doesn't work out of
the box on safari
phf: put everything in
tar containers?
ben_vulpes: anyways, mod6 you gotta upgrade v.pl
to press *all* reachable leaves.
that's
the problem i'm running into with asciilifeform's recent shiva rebake
☟︎ mircea_popescu: we're going
the other way :
taking gpg out, not putting more of it in.
jurov: but
they are completely unreadable
assbot: Logged on 31-01-2016 21:27:52; jurov:
thay would be much cleaner solution
that
trying
to invent some "container" for patches
to prevent whitespace/newline munging
jurov: like he does with
the email client
jurov: i'm willing
to bet
than when
there will be
turdatron www upload form, asciilifeform will use
the most broken browser in existence
to
try
to upload
the files
mircea_popescu: nevertheless, a large chunk of what mod6 is doing is running
through
the ever-expanding ball of patches and making a curated set.
assbot: Logged on 31-01-2016 20:28:00; phf: huh so with original v design, mod6 could sign all
the patches
that are deemed releasable, and
then all you need is mod6.asc in your wot and
trb will press
ben_vulpes: aha i begin
to see
the wisdom of pressing all reachable leaves now
trinque: felipelalli: I
thought you were
talking about assbot
kakobrekla: felipelalli ask
trinque, i dont run deedbot-
felipelalli: kakobrekla, do you have an idea why deedbot- didn't accept my renewed key?
thank you.
trinque: jurov │
trinque: does deedbot allow submissions with detached signatures? like "deedbot- <url1.sig> <url2>" << it does not, but is not rocket surgery
to add. seems useful for posting scripts
phf: ascii_rear: you
tried fastmail?
ascii_rear: but also don't have 2hrs/day
to hand-sift
ascii_rear: gotta be able
to receive from not-seen-before addrs
assbot: Logged on 31-01-2016 20:48:02; jurov: asciilifeform: can't you just get rid of gmail? or it's
the same situation as you and usg?
jurov: thay would be much cleaner solution
that
trying
to invent some "container" for patches
to prevent whitespace/newline munging
☟︎ ben_vulpes: asciilifeform: if i understand
the shiva rebake correctly, applying
tinyscheme_genesis_vpatch,
then
tinyscheme_crud_snip,
then
tinyscheme_repl_fix_fixed bangs
the
tree into
the correct shape?
assbot: Logged on 31-01-2016 14:42:22; mircea_popescu: just, if you want
to add a patch, should be able
to dump it as dpaste also.
adlai: ben_vulpes: archive.is
the best -ception
felipelalli: deedbot- gives some feedback after
the add-key?
jurov: oh yes!
try resubmitting your pubkey
to deedbot.
adlai: assuming or suggesting
that people use automated key upgrade mechanisms is... somewhere between negligent and criminal. at least, criminally negligent.
jurov: yes you uploaded,
that part is fine, but bot hasn't downloaded it
jurov: you must wait for
the botmaster (trinque afaik)
to fix it
jurov: well,
then both mine and deedbot-'s keyring wasn't updated with your update
adlai: what is expired? i was able
to load it without any error about
this
jurov: if you want you can send it
to btc-dev, it ignores key expirations
felipelalli: jurov, and how can I workaround
this? Or
this is out of my control? :)
adlai: felipelalli: nope, but
the symptom is reproducible!
jurov: to mirror
turdatron, only clearsigned
text is checked if it's someone known
to gpg, anything other is passed unchanged
jurov: punkman you're welcome
to contribute a script
that checks web submission (upload form for clearsigned
text + any number of "attachments") and
turns it into email
jurov: asciilifeform: can't you just get rid of gmail? or it's
the same situation as you and usg?
☟︎ mod6: meanwhile, im gonna work on getting
this SoBA completed. :]
punkman: web-based
turdatron woulda saved you a lot of debug
time :/
mod6: asciilifeform: <+mod6> so... i've integrated asciilifeform's fix into a new patch
that will replace
the old one:
http://dpaste.com/2KCXN8A.txt << <+mod6> ahh, and
then
there is
this now -- when compiling with alf's fix incorporated: <+mod6>
http://dpaste.com/1AVK5ZP.txt << any
thoughts on how
to fix0r
this?
jurov: (example: [BTC-dev] Some
test patches and my V implementation)
jurov: ah i see. but you send
the clearsigned part as Content-Type:
text/plain
punkman: pub@extemporized.com , I manually set octet-stream for all my attachments, but it's possible
the crappy email-sender I use overrode it for
the email body
jurov: did you ever sent anything
to ml?
punkman: everything I've sent has been octet-stream, I
think
jurov: which caused
turdatron
to skip it while searching for clearsigned part
mod6: well, i missed a few, need
to clean
those up :P
mod6: some of you out
there who are following along will be like, "yeah, what about line 426, or 484
then!! eh!?"
mod6: yup yup, i put a 0x0a in
there just
to preseve 80 cols. but perl deoesn't care nor does it add a '\n'
to
the command string
mod6: notice how
there is a space after '=long '
ben_vulpes: well
there is a line break between
those
two strings
mod6: <+ben_vulpes> so
the '.' is both a line-break escape and string concatenation? << its not a line break. its just a string concatenation.
mod6: and execute $import and set
the result (output) of
the command
to $res
mod6: then in
the next line i do
this:
ben_vulpes: so
the '.' is both a line-break escape and string concatenation?
mod6: "command part A" . "command part B" and set
that equal
to $import
mod6: ok so whats going on
there is
that
to preserve 80 cols, i create a command spanning
two lines
assbot: Logged on 31-01-2016 17:53:34; phf: mircea_popescu: hey
trinque jurov phf : would it be possible
to
take
that very nice graph phf showed,
turn it into a proper svg and have
the nodes clickable ? << like
http://104.131.72.249/patches/? (is in no way bulletproof, so i expect it
to be down by
the
time i come back from breakfast)
assbot: Logged on 31-01-2016 17:53:21; mod6: <+ben_vulpes> yo mod6 what does it mean in perl
to do my $var = "string" . \n "otherstring" is
the period a linebreak escape in perl, and
the variable implicitly a concatenation of
the
two strings << where do you see
this?
this is not a
thing.
that would fail
to compile/pass-interpreter-smell-test.
phf: perhaps
then
trb instructions should be "download mod6.asc into ~/.wot" which is a
trust starting point
phf: huh so with original v design, mod6 could sign all
the patches
that are deemed releasable, and
then all you need is mod6.asc in your wot and
trb will press
☟︎ mod6: so
that's not
the only
thing
that needed
to change.
mod6: ahh, and
then
there is
this now -- when compiling with alf's fix incorporated:
mod6: i can't even
take
the
time
to jump down
that rabbit hole right now.
mod6: it's not hurting anyone, and its not going in
the release.
mod6: like i said, lets just hold off on
that.