dansmith_btc: you mean HTTP response headers? What are you lookign at?
belcher has quit
MrMoneyBags has quit
MrMoneyBags joined the channel
MrMoneyBags has quit
MrMoneyBags joined the channel
proslogion joined the channel
MrMoneyBags has quit
MrMoneyBags joined the channel
dansmith_btc
moo-_-, hi, yes http headers if SO proxied by CF show no indication of SO.
waxwing
dansmith_btc: maybe you were getting one of those cloudflare captcha street sign things? :)
(see earlier "raging argument" :) )
proslogion
which makes it convenient to add the more compact Schnorr multisig types, which allow further scale by making smaller multisig and (singleSig) multiple input transactions which improve also fungibility and are coinJoin compatible.
proslogion: what specific type is interesting? in general, they're no different from non-CJ transactions.
proslogion
so you put all the signatures in the scriptsig field and that's it? how are the signatures ordered then, or no order is required?
waxwing
each input has its own scriptSig
proslogion
okay then i get it
waxwing
you just fill in the signature in the redeemscript that satisfies the requirement of your input
proslogion
it's unfortunate
waxwing
i PMed you on just for the hell of it :) but only testnet
proslogion
hmmm...then how is Segwit going to work for a CJ or a geneal multi-input tx?
waxwing
i haven't looked at it yet. let me know if you find out something interesting. my first guess is it shouldn't really be different in function.
proslogion
k so compressed keys start with 03
proslogion has quit
proslogion joined the channel
waxwing
proslogion: or 02
the 02/03 flag tells you which of the 2 possible y coordinates for this x coordinate is the correct one
proslogion
dansmith_btc: waxwing sent
MrMoneyBags has quit
MrMoneyBags joined the channel
mkarrer_ joined the channel
lol waxwing there is a decentralized communication network junseth just reminded me, the bitcoin testnet
waxwing
proslogion: true, but unauthenticated right
proslogion
you could figure it out like ThomasV
waxwing
proslogion: wrt your key, looks fine, but ofc we can't tell much from looking at it; how did you generate it? otoh we can do tests better than before :)
yeah bitcoin's p2p layer can be very light because you don't need auth
ofc still sybil/dos problems, but seems that's ok
proslogion
it's the one we used the last time, basically this time i go for the second half
of the 65 bytes
waxwing
so is the second half the x or the y? and were you able to check the parity?
proslogion
if you don't think it's good i will generate a new one anyway
waxwing
no that's fine, i used the same as last time also; either let's make sure the method you're using is sound, or else we can do a test spend anyway.
proslogion
what wallet do you usually used for generating a pubkey?
nah i just want to do it properly
waxwing
i just use pybitcointools because it's what i've used a lot before
you could use any wallet i guess? or any similar tool
e.g. run electrum and make a brand new wallet, and just take one pubkey. should work i think.
come to think of it i'd be surprised if bitcoin's own rpc doesn't have everything you need, although i haven't checked
proslogion
just trying to figure how much work could it take for something as simple as this
waxwing
it's no work at all to do it; it's some work to be sure you're doing it right of course (at least, the first time).
you could play on testnet or regtest of course.
the latter is the best, but a pain to set up first time.
if i didn't want to deal with it, i would make a fresh electrum wallet and store the private and public key somewhere. or, i'd bet you can do it with the rpc api with bitcoin-cli if that's easier.
proslogion
no i don't mean it's difficult for me :)
just how much work it needs to go from 0 to get it done
MrMoneyBags has quit
MrMoneyBags joined the channel
waxwing
dansmith_btc: the 'way too fast' error can happen if you're not on the right server, correct?
dansmith_btc
waxwing, yes whenever theres no response
waxwing
someone tried on twitter and got too fast, i told him that could mean wrong (old) server. seems to be fine from here.
i mean, someone on twitter tried.
dansmith_btc: i just tried a stackoverflow page and it worked for me.
dansmith_btc
worked as in got correct headers? cause n10n worked for me too
waxwing
hang on, will come back to that, got interesting question:
"would corporate firewalls make a difference?"
i guess the obvious point is they would if they block you connecting
possibly they're filtering on ports?
and it's not an http connection right
hmm, well, http responses but .. seems quite plausible
i guess given your firewall work dansmith_btc this would be a good Q for you :)
dansmith_btc
well the easiest one can do is wget http://52.91.68.11:10011 , if it says connection established than no firewall is blocking you
waxwing
right, good idea
dansmith_btc
it is probable that exotic ports like 10011 get blocked
waxwing
ah, let me go back to that headers question
yeah i agree
yeah, i see no reference to stackoverflow in the headers
dansmith_btc
also if u enter http://52.91.68.11:10011/ in the browser, u ll immediately get "no data received" error
waxwing
well, it's in the title of course :) but cloudflare is evil, i think we've established :)
dansmith_btc
title can be faked by any other site hosted on CF :)
waxwing
sure, "irrefutable as long as you trust the public key" etc etc
bitfinex = incapsula is same deal except set-cookie: _bfx_session
but i must have missed a point; are the headers really special? couldn't a site spoof that too? i mean, we're ultimately trusting cloudflare somehow in that example, aren't we?
i should go back and reread that keyless ssl idea
wow bitfinex.com serves an incredible amount of text on its homepage
dansmith_btc
the headers are sent by CF,so they can be trusted
waxwing
yeah, i get the distinction, and it still isn't great; but even so, are we sure CF doesn't allow clients to add to headers themselves?
if they always promised a standardised thing like 'one particular header always contains the domain name' then at least that would be something.
belcher joined the channel
so here's the thing, the certificate for the bitfinex.com page is for CN *.bitfinex.com
who owns the private key? gotta go back and read keyless ssl