notduncansmith has quit [Read error: Connection reset by peer]
Eudaimonstro has joined #ipfs
voxelot has quit [Ping timeout: 260 seconds]
amstocker has joined #ipfs
<whyrusleeping>
cryptix: <3
otherchas__ is now known as cleichner
amstocker has quit [Ping timeout: 246 seconds]
ianopolous has quit [Ping timeout: 246 seconds]
<jbenet>
Tv`: yeah we need a kv abstraction. Could be something like the pinset, or something like a compressed trie (patricia tree), or a b-tree etc
* davidar
wonders if redis-on-ipfs would work
<Tv`>
jbenet: any hope of using sha512 for ipfs object identity, on request?
<Tv`>
that would make the keys align
<Tv`>
and it sounds like the multihash thing might agree with that
<Tv`>
sorry gotta go, will catch up later
gordonb has joined #ipfs
<jbenet>
Tv`: yep sure can. Right now have to craft your own objects or recompile.
gordonb has quit [Client Quit]
<Tv`>
crafting my own is fine
<Tv`>
i think ;)
<Tv`>
will explore later
<Tv`>
thanks!
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
infinity0 has quit [Ping timeout: 272 seconds]
voxelot has joined #ipfs
tjgillies__ has joined #ipfs
infinity0 has joined #ipfs
<jbenet>
whyrusleeping did we fix the bug with gobuilder imports?
<whyrusleeping>
jbenet: i was hoping that adding // +build ignore would work
<whyrusleeping>
but it didnt
<whyrusleeping>
do you think it would be okay if we put our binaries in /test/bin ?
<whyrusleeping>
or something like that
<jbenet>
binaries?
<jbenet>
is that go-ipfs/test/bin ? or really /test/bin
<whyrusleeping>
like iptb and the other tools we use in test
<jbenet>
and they still need to be built-- because we dont have builders of every arch for them.
<whyrusleeping>
oh yeah, we will still build them
<whyrusleeping>
i'm just saying to move the source out of godeps
<whyrusleeping>
so we can get around this issue
<jbenet>
(the https://github.com/ipfs/distributions stuff reminded me we can have a very simple format of binaries (that works beyond go) where we could ship an installer that downloads the right bin for your arch. (you know. without encroaching too much on pkg mgrs)
<jbenet>
whyrusleeping: why dont we just use gx?
<jbenet>
gx doesnt work well yet-- but it would def improve it.
<whyrusleeping>
i dont want to do some things in gx and some things in godeps
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
whyrusleeping: we cant switch completely until gx "just works" and thats not reality.
<jbenet>
whyrusleeping: i tried to use it to vendor the stuff in go-multicodec and failed.
<voxelot>
think we are still working on documentation tho
<Tv`>
sorry for being lazy
<Tv`>
huh it's hanging
<voxelot>
no it's a god question
Eudaimonstro has quit [Remote host closed the connection]
<voxelot>
right now it just says to use the cli commands as endpoints i believe
<Tv`>
oh right, i should point at the gateway not at the api service
<Tv`>
okay then
<jbenet>
Tv` what functionality do you want?
<whyrusleeping>
Tv`: whatcha workin on?
<Tv`>
i'll tell you if it works ;)
<voxelot>
:)
<davidar>
lgierth: thanks, fixed
<jbenet>
whyrusleeping: we need to fix gobuilder. can we either fix it with gx, or fix it with godeps? (we can vendor it in another dir if we need to, like go-ipfs/vendor/ or something
<whyrusleeping>
or we can move the source for the binaries we are using into test/tools or something
<jbenet>
whyrusleeping: that's what i mean, vendor those there
<jbenet>
whyrusleeping: i dont think we should just put the whole source in a "not-clearly-a-vendor-dir" and will need to track the hashes.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
jbenet: what if you had a single tree shared amongst a bunch of nodes?
chriscool has joined #ipfs
<jbenet>
davidar: do you mean, "multiple writer problem" ?
<davidar>
jbenet: I guess so? (sorry, distributed computing isn't exactly my area :)
fleeky has quit [Remote host closed the connection]
fleeky has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ygrek has joined #ipfs
amstocker_ has joined #ipfs
ygrek has quit [Ping timeout: 255 seconds]
notduncansmith has joined #ipfs
<davidar>
jbenet: I think what I'm trying to ask is, suppose you had some global tree for the whole network (so you don't know all of the nodes in advance), and every node publishes their own subset of the tree, and you need to merge them all together to perform lookups on the global tree. How would you coordinate that?
notduncansmith has quit [Read error: Connection reset by peer]
amstocker has quit [Ping timeout: 246 seconds]
<davidar>
(The application I'm thinking of here is where every node publishes a reverse index for their local blocks, and you want to perform lookups across all of them)
<davidar>
Presumably you want to merge them recursively, but without having to impose a hierarchy on the network
notduncansmith has quit [Read error: Connection reset by peer]
sharky has quit [Ping timeout: 272 seconds]
domanic has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
sharky has joined #ipfs
voxelot has quit [Ping timeout: 272 seconds]
<davidar>
Eh, service discovery + lazy merging would probably be the easiest
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
what's the current status of the dmca blocklist?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ygrek has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
reit has quit [Ping timeout: 244 seconds]
reit has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Tv` has quit [Quit: Connection closed for inactivity]
reit has quit [Ping timeout: 245 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Leer10 has joined #ipfs
chriscool has quit [Ping timeout: 260 seconds]
chriscool has joined #ipfs
<cryptix>
gmorning
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
morning cryptix
<cryptix>
hey davidar, how r u?
<davidar>
cryptix: not too bad, you?
<cryptix>
ntb either, waiting for coffee to kick in :))
reit has joined #ipfs
<davidar>
can't say I'm much of a coffee person, but perhaps that's what my problem is :)
chriscool has quit [Ping timeout: 260 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
* davidar
is looking through old bookmarks, rate of bitrot is depressing
<zignig>
/me hands davidar a shiney new IPFS as anti-bitrot solution.
<davidar>
zignig: that's why I'm involved now :)
notduncansmith has joined #ipfs
Encrypt has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<zignig>
davidar: what wre your thoughts on that dht indexing paper.
<zignig>
it melted my brain for a while , re-read it a few months later and it made more sense.
<cryptix>
ohi zignig
<zignig>
cryptix: ohhai 2u2u
<davidar>
zignig: yeah, it looks pretty cool, but can't work out the best way to implement it on ipfs, especially if you assume nodes are untrusted but the tree is too big for any single node to store
<davidar>
zignig: but I'd really love ipfs to support a distributed store that you can throw random key-value pairs into
<davidar>
(from what i can tell, the technique in that paper is resistant to peers going offline, but not malicious nodes)
<davidar>
it also doesn't say too much about how to handle updates, other than pushing them to the relevant nodes and hoping they accept them
<davidar>
zignig: what were your thoughts?
Encrypt has quit [Quit: Quitte]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
was also thinking about how recursive merging could work, without priviliging specific nodes
<davidar>
perhaps you could do the recursive partitioning multiple times, so that malicious nodes can't subvert all of them
ygrek has quit [Ping timeout: 246 seconds]
<davidar>
then since merging radix trees is commutative, you can just check for equality of the root hash at the end
<davidar>
(i.e. kind of like what i think crdt does, but without each node having to replicate the data fully)
domanic has quit [Ping timeout: 246 seconds]
<davidar>
jbenet: any of that make sense? :/
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Encrypt has joined #ipfs
<davidar>
or maybe crdt already does all that and I'm just trying to reinvent the wheel :)
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<daviddias>
whyrusleeping: could you remind me where is that PR you did to make your 'drag-drop' browser app work, regarding Browser Buffer type? (the Buffer.isBuffer) thing?
<daviddias>
whyrusleeping: got it :) thanks anyway
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
Encrypt has quit [Quit: Quitte]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<cryptix>
whyrusleeping: hrm go shell, it looks like FileList and List each are having a bug (empty Names in FileList and List cant unmarshal into LsLink because type is an int in the json, not a string)
<cryptix>
we should rig up testing for that thing - with the new docker services we could run it like i run git-remote-ipfs tests
<cryptix>
i.e against a known ipfs image, nodejs later too i guess
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
dignifiedquire has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
therealplato has quit [Ping timeout: 240 seconds]
dignifiedquire has quit [Ping timeout: 245 seconds]
<cryptix>
whyrusleeping: if we resolve the shell issue i can push a working version
<cryptix>
(or i could vendor...)
therealplato has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
reit has quit [Ping timeout: 246 seconds]
Encrypt has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rawtaz has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
zignig: six weeks until mandatory metadata retention...
<cryptix>
travis.. Y U NO detect godeps
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<cryptix>
Luzifer: uhm... why do you 'godep restore' if _workspace/src is present?
datagrok has quit [*.net *.split]
celehner has quit [*.net *.split]
cryptix has quit [*.net *.split]
zml has quit [*.net *.split]
cryptix has joined #ipfs
datagrok has joined #ipfs
celehner has joined #ipfs
zml has joined #ipfs
<Luzifer>
cryptix: why not? `go build` will not automatically detect gopaths…
<cryptix>
Luzifer: GOPATH=$whereEver:$(godep path) is another option if you don't want to 'godep go build'
<cryptix>
my porblem with godep restore is, that you cant add commits that aren't published at the repo root
<cryptix>
like hot fixes, id need to fork and rewrite import paths
<cryptix>
also, godep restore clones && checkouts iirc - wasteful if _workspace/src is already included in the build repo
<cryptix>
whyrusleeping: whyrustillsleeping
<cryptix>
</rant> :)
<Luzifer>
yeah… and more different handlings of paths, godeps, vendoring… slowly stops making fun
<cryptix>
hehe how about gb? ;)
<Luzifer>
yeah. and go1.5 experimental vendoring and someone surely finds a way to break everything…
<cryptix>
the 'go list' thing was quite careless... :/
<Luzifer>
I think I will rebuild the build-script in Go and then implement feature-flips… disable-godeps, use-vendor-path and $stuff
<Luzifer>
but not this weekend.
<cryptix>
<3
<cryptix>
jbenet: how do i get ipfsbot to rt me? :)
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
tjgillies__ has quit [Quit: Connection closed for inactivity]
theguy159 has left #ipfs [#ipfs]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
reit has joined #ipfs
ebarch has quit [Quit: Gone]
ebarch has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] rht opened pull request #1623: Fix and refactor daemon api test (master...fix-check-api-test) http://git.io/vGZ3T
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Tv` has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
voxelot has joined #ipfs
Encrypt has quit [Quit: Going outside!]
<rschulman>
cryptix: git-remote-ipfs is cool!
<rschulman>
Is there any way to get it to use ipns? For a stable upstream url?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
gozala has quit [Ping timeout: 256 seconds]
gozala has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
kord has joined #ipfs
<cryptix>
rschulman: thx :) supporting ipns to republish is the idea
<cryptix>
rschulman: now that i have the mechanics of push done, just publishing the the new hash is easy
<cryptix>
but there are corner cases, too.. where do you push if you cloned from an ipns entry that you dont own?
<cryptix>
and i'm sure there are bugs/missing features in the push code
<cryptix>
and i need to figure out how to to the initial publish, i'll most likly just port whyrusleeping's git-ipfs-rehost and hook it to something like ipfs://new
bedeho has joined #ipfs
<whyrusleeping>
internet is out...
<whyrusleeping>
I hate my new place
<cryptix>
whyrusleeping: thats bad
<whyrusleeping>
yeah, it is :(
<cryptix>
neighbours wifi?
<whyrusleeping>
anything I can do for you from my phone?
<whyrusleeping>
neighbors WiFi is down too
<cryptix>
yea.. why is go-ipfs-shell such an undocumented untested %*!@?
<whyrusleeping>
all three of them
<whyrusleeping>
Lol
<cryptix>
FileList() and List() both have bugs
<whyrusleeping>
hmmm, did you file them?
<cryptix>
FileList() doesnt poplulate names and List() errors out because on of the fields is returned in int, not string by the server
<whyrusleeping>
or PR thm?
<cryptix>
not yet
<cryptix>
i finished git-remote-ipfs push :]
<whyrusleeping>
ooooh, that is something I've never tester
<whyrusleeping>
awesome!!!!!!!!!!!
<whyrusleeping>
:D
<cryptix>
yup. sure there are corner cases but 'add > commit > push' works and is fetch/cloneable :D
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<cryptix>
regarding testing: we should rig up testing for that thing - with the new docker services we could run it like i run git-remote-ipfs tests
<cryptix>
i.e against a known ipfs docker image, nodejs later too i guess
<rschulman>
cryptix: Awesome, looking forward to using git + ipfs in the future. :)
<cryptix>
rschulman: yup :) all those ddos attacks on github really pushed me to it
<cryptix>
we could setup a webhook on gh that pushes to ipfs when master is updated or something
pfraze has joined #ipfs
<cryptix>
i wonder what the bast way to pre-fill an ipfs daemon is.. i guess i just prepare an .ipfs dir and unpack that
<cryptix>
relying on travis networking is really meh
<rschulman>
cryptix: Why not just have travis call ipfs init first thing?
<cryptix>
rschulman: that's what i do now
<cryptix>
but when the 'clone ipfs://$hash' test runs, it actually fetches that from another node across the network
<cryptix>
timeout galore
<rschulman>
oh, interesting.
<cryptix>
i want to stuff in the needed data from the repo so that it doesnt have to hit the network.. travis is slow enough already :x
pfraze has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
eternaleye has quit [Ping timeout: 265 seconds]
notduncansmith has quit [Read error: Connection reset by peer]
<zml>
cryptix: the travis cache might be able to be rigged up to store the data you need -- you can make it cache an arbitrary directory
<cryptix>
zml: travis cache? got a docu link maybe?
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
finally, internets back
<whyrusleeping>
and then i get a notice that the air outside is hazardous
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
cryptix: could you provide a repro or something for the ipfs-shell issues?
<cryptix>
whyrusleeping: 20mins
<whyrusleeping>
cryptix: mmkay
<Tv`>
whyrusleeping: "air is hazardous" - are you in China or something?
<whyrusleeping>
nope, washington state.
<whyrusleeping>
its apparently all on fire
<cryptix>
friend tricked me into playing path of exile, now im doomed
<whyrusleeping>
cryptix: lol, my roommate is trying to suck me into playing borderlands
<whyrusleeping>
"dude, its saturday morning, get some coffee and play borderlands with me" <- quite the convincing argument
<Tv`>
i find my decision to only game on a projector+sofa setup is very good.. it's only dark enough after sunset, so i don't waste too much time there ;)
bedeho has quit [Ping timeout: 255 seconds]
uhhyeahbret has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
therealplato1 has joined #ipfs
therealplato has quit [Ping timeout: 250 seconds]
notduncansmith has joined #ipfs
eternaleye has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Eudaimonstro has joined #ipfs
<Tv`>
there's something rotten in ifps daemon
<Tv`>
yesterday, i tried to curl a block from the api service, and just got block not found
<Tv`>
today, i start the daemon, and re-run the curl, and it's 200 ok
<Tv`>
i did not touch go build etc
<whyrusleeping>
did you get the block?
<Tv`>
it seems to work now
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Encrypt has joined #ipfs
amstocker has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<cryptix>
oh wow.. that was not 20 minutes
t3sserakt has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
bedeho has joined #ipfs
Eudaimonstro has quit [Ping timeout: 265 seconds]
pfraze has joined #ipfs
Eudaimonstro has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<cryptix>
simplest hack would be just to add a 2nd type for List()
<cryptix>
not sure where it goes wrong for FileList()
<cryptix>
i'll open issues too
Eudaimonstro has quit [Ping timeout: 252 seconds]
bedeho has quit [Ping timeout: 244 seconds]
<Tv`>
whyrusleeping: btw just to add some fuel to the fire: because ipfs daemon hijacks connections (to do its own chunking), it closes the connections, and that means its not in general safe to just do http against the api, one needs a retry loop. and go doesn't have one, yet: https://github.com/golang/go/issues/4677
<cryptix>
(opend issue #8 and #9 on gh)
therealplato1 has quit [Quit: Leaving.]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Eudaimonstro has joined #ipfs
therealplato has joined #ipfs
bedeho has joined #ipfs
mildred has quit [Ping timeout: 264 seconds]
mildred has joined #ipfs
Eudaimonstro has quit [Ping timeout: 256 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
t3sserakt has quit [Quit: Leaving.]
t3sserakt has joined #ipfs
t3sserakt has quit [Client Quit]
montecarlo has joined #ipfs
Eudaimonstro has joined #ipfs
mildred has quit [Ping timeout: 265 seconds]
Eudaimonstro has quit [Ping timeout: 255 seconds]
pfraze has quit [Remote host closed the connection]
jatb has quit [Ping timeout: 244 seconds]
jatb has joined #ipfs
Eudaimonstro has joined #ipfs
ygrek has quit [Ping timeout: 250 seconds]
<whyrusleeping>
Tv`: yeah, i would really like to have the changes from go1.5...
<whyrusleeping>
there are so many problems caused by us not using the stdlib stuff
Eudaimonstro has quit [Ping timeout: 246 seconds]
ygrek has joined #ipfs
bedeho has quit [Ping timeout: 246 seconds]
Eudaimonstro has joined #ipfs
therealplato1 has joined #ipfs
therealplato has quit [Ping timeout: 246 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
therealplato has joined #ipfs
tjgillies__ has joined #ipfs
therealplato1 has quit [Ping timeout: 260 seconds]
Eudaimonstro has quit [Ping timeout: 260 seconds]
<lgierth>
whyrusleeping: me looking at hellabot now
Encrypt has quit [Quit: Sleeping time!]
kord has quit [Quit: Leaving...]
Eudaimonstro has joined #ipfs
<amstocker>
is there a way to test if an object added to the local repo has been shared to other nodes?
<amstocker>
without combing the logs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
bedeho has joined #ipfs
bedeho has quit [Remote host closed the connection]
bedeho has joined #ipfs
bedeho has quit [Remote host closed the connection]
<whyrusleeping>
amstocker: not really an easy way right now
<amstocker>
ok i was just curious
<amstocker>
when I run repo gc does it clear object that i've added locally or just objects that I've downloaded off of the network?
<zml>
clears anything that's unpinned
<jbenet>
amstocker: clears things not on your pinset
<jbenet>
(ipfs add automatically pins)
<jbenet>
(but can add more with: ipfs pin add -r )
<amstocker>
oh ok
<amstocker>
good to know
montecarlo has quit [Quit: leaving]
<zignig>
jbenet: how is it all going ?
<zignig>
what are your thoughts on private swarms ?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
pfraze has joined #ipfs
<whyrusleeping>
zignig: private swarms work pretty well as far as i've heard
<zignig>
how do you keep them private if one one connect to a public node will it join up.
<zignig>
I was thinking a shared key on the secure connection.
<zignig>
with a default of IPFS_IS_AWESOME.
<zignig>
that way the the swarm you not et out and rnadom nodes could not get in ...
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
zignig: yeah, right now to run a private swarm, you would have to firewall your nodes off
<whyrusleeping>
the join key is an interesting idea
<whyrusleeping>
could be fairly easily implemented in the handshake
<zignig>
using a hashed transaction does the hash of shared_key + nonce concur.
<whyrusleeping>
well, how would you avoid sending your shared key to unwanted nodes?
<whyrusleeping>
would only the dialer send the key?
<zignig>
don't send the key... ;P
<zignig>
both nodes would have to have the key before hand and you would compare hases of the key.
<whyrusleeping>
so we connect, you hash you key and send it
<zignig>
if all the nodes have a key of zignigs_swarm they could connect.
<whyrusleeping>
now i have the hash, and i can send it to you or other nodes in your net
<zignig>
so the server reciever would deny any connection that did not have the correct hash.
amstocker has quit [Ping timeout: 244 seconds]
<whyrusleeping>
but i have your hash, you sent it to me
<whyrusleeping>
and your receiver would accept me
<zignig>
that's only becuase I put it up on irc...
<whyrusleeping>
nonono
<whyrusleeping>
my node connects to your node
<whyrusleeping>
we exchange h(shared_key)
<whyrusleeping>
now my node knows h(shared_key), and can send it to your node on another connection
<zignig>
nope, compare nonce + hash of shared key.