<alu>
im at a hackerspace scanning the hackerspace and making a vr site for it
MDude has quit [Ping timeout: 265 seconds]
infinity0 has quit [Remote host closed the connection]
seafood has joined #ipfs
infinity0 has joined #ipfs
tilgovi has quit [Remote host closed the connection]
muvlon has quit [Ping timeout: 240 seconds]
Oatmeal has joined #ipfs
tilgovi has joined #ipfs
seafood has quit [Ping timeout: 250 seconds]
brianhoffman has quit [Ping timeout: 264 seconds]
brianhoffman_ has joined #ipfs
kevina has quit [Ping timeout: 260 seconds]
kevina has joined #ipfs
muvlon has joined #ipfs
<whyrusleeping>
nice!
<whyrusleeping>
one in LA?
dignifiedquire has joined #ipfs
Winter_Fox has joined #ipfs
rendar has joined #ipfs
tilgovi has quit [Quit: No Ping reply in 180 seconds.]
tilgovi has joined #ipfs
Oatmeal has quit [Ping timeout: 265 seconds]
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
<interfect[m]>
See, won;t people then have to go to the hackerspace and check out the VR headset, to experience the virtual hackerspace?
lacour has quit [Quit: Leaving]
<dignifiedquire>
good morning interplenetary people
<dignifiedquire>
soo many dependencies in that bitswap pr :O
<alu>
I wanted to show others how fast it is to build and collaborate in the metaverse with IPFS
<dignifiedquire>
looks like we have our work cut out for us today daviddias
<interfect[m]>
Is bitswap finally coming? Will my node go out and trade for the hottest bits?
<dignifiedquire>
interfect[m]: bitswap is already there, the pr is about improving it and addig ipld features to the js implementation
<dignifiedquire>
but bitswap is still using pretty simple strategies for block distribution
<interfect[m]>
Isn't IPLD a layer on top? What does it do for Bitswap?
fjl_ has joined #ipfs
<dignifiedquire>
daviddias: do you want to start setting up those interop tests using the new ipfsd-ctl? then I can take care of all your cr comments in the meantime
ulrichard has joined #ipfs
<dignifiedquire>
it is not full ipld integration, more bitswap needs to understand cids to make it possible to make it easy to transfer all the new blocks that are coming with ipld
steefmin_ has joined #ipfs
<daviddias>
dignifiedquire: sounds good :)
Zaibon_ has quit [*.net *.split]
fjl has quit [*.net *.split]
steefmin has quit [*.net *.split]
steefmin_ is now known as steefmin
<dignifiedquire>
but do you know what I do before that? ;)
<dignifiedquire>
you have one guess
<dignifiedquire>
but do you know what I do before that? ;)
<dignifiedquire>
you have one guess
Zaibon_ has joined #ipfs
<daviddias>
dignifiedquire: gym?
<dignifiedquire>
💯
s_kunk has quit [Ping timeout: 250 seconds]
<daviddias>
eheh :D
Zaibon_ has quit [*.net *.split]
ylp has joined #ipfs
wkennington has quit [Read error: Connection reset by peer]
palkeo has joined #ipfs
tilgovi has quit [Ping timeout: 258 seconds]
kenshyx has joined #ipfs
Guest104947[m] has joined #ipfs
tilgovi has joined #ipfs
Zaibon_ has joined #ipfs
Akaibu has quit [Quit: Connection closed for inactivity]
palkeo has quit [Quit: Konversation terminated!]
s_kunk has joined #ipfs
tilgovi has quit [Remote host closed the connection]
schmidta has joined #ipfs
schmidta has left #ipfs [#ipfs]
Oatmeal has joined #ipfs
lkcl has quit [Ping timeout: 265 seconds]
infinity0 has quit [Ping timeout: 260 seconds]
hpk has joined #ipfs
infinity0 has joined #ipfs
lkcl has joined #ipfs
jerryisme[m] has joined #ipfs
<jerryisme[m]>
hi
PseudoNoob has joined #ipfs
lkcl has quit [Ping timeout: 268 seconds]
espadrine has joined #ipfs
Encrypt has joined #ipfs
seafood has joined #ipfs
gmoro has quit [Remote host closed the connection]
Winter_Fox has quit [Ping timeout: 268 seconds]
slothbag has left #ipfs [#ipfs]
Winter_Fox has joined #ipfs
seafood has quit [Ping timeout: 248 seconds]
seafood has joined #ipfs
faenil has joined #ipfs
seafood has quit [Client Quit]
john1 has joined #ipfs
<alterego>
Has anyone got an example of getting a file from ipfs using js-ipfs-api from browser?
<alterego>
I can only see node example using process.stdout.
<alterego>
Nvm, figured it out
apiarian has joined #ipfs
bastianilso_____ has joined #ipfs
elico has quit [Quit: Leaving.]
bastianilso_____ has quit [Client Quit]
Kingsquee has quit [Quit: Konversation terminated!]
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
Encrypt has quit [Quit: Quit]
ulrichard has quit [Remote host closed the connection]
infinity0_ has joined #ipfs
infinity0_ has quit [Changing host]
infinity0_ has joined #ipfs
infinity0_ is now known as infinity0
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
hpk has left #ipfs ["WeeChat 1.5"]
ivo_ has quit [Remote host closed the connection]
gmoro has joined #ipfs
PseudoNoob has quit [Remote host closed the connection]
<Mateon1>
I'm wondering... How well does the DHT work in network separation scenarios (i.e. LAN/WAN but no inet)? It seems that some content will 'disappear' from (the network's view of) the DHT (even if the content exists in the network). Is that the case?
<Kubuxu>
in LAN/WAN with little peers you will be connected with all/almost all of them
<Kubuxu>
so DHT isn't really used then for the blocks
Encrypt has joined #ipfs
ppham has joined #ipfs
<dignifiedquire>
Kubuxu: where are the error codes that the json response in the http-api sends defined? I can't seem to find them, do they just match the cli exit codes, i.e. 1 = error?
<Kubuxu>
dignifiedquire: commands/response.go:18, yes they map to CLI exit codes
<dignifiedquire>
thanks
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
anonymuse has joined #ipfs
elico has joined #ipfs
<Kubuxu>
dignifiedquire: they don't map to exit codes, sorry
<Kubuxu>
exit codes: no error - 0, some error - 1
ppham has quit [Remote host closed the connection]
PseudoNoob has joined #ipfs
<dignifiedquire>
okay, but still error = 1 ?
<Kubuxu>
on which layer?
Winter_Fox has quit [Ping timeout: 258 seconds]
cemerick has joined #ipfs
<dignifiedquire>
if the http-api returns a json formated error, what should the "Code" be?
pfrazee has joined #ipfs
MrManor[m] has joined #ipfs
ashark has joined #ipfs
<Kubuxu>
I think commands/response.go:18 so 0 to 2
<Kubuxu>
I think, not sure
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
ppham has joined #ipfs
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
<richardlitt>
ipfs.pics is down :(
ppham has quit [Remote host closed the connection]
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
flyingzumwalt_ has joined #ipfs
ppham has joined #ipfs
Akaibu has joined #ipfs
ppham has quit [Ping timeout: 258 seconds]
anonymuse has quit []
anonymuse has joined #ipfs
JustinDrake has joined #ipfs
<flyingzumwalt_>
davidar are you the main one working on the code that indexes files in-place instead of copying their content into the .ipfs repository?
<JustinDrake>
Hi all. What is the status of CIDs? Are we definitely getting CIDv1 for IPFS 0.5?
<Kubuxu>
JustinDrake: CIDs are supported in to be released 0.4.5
<lgierth>
flyingzumwalt_: that'd be kevina
<lgierth>
flyingzumwalt_: codename filestore
<flyingzumwalt_>
lgierth++
<flyingzumwalt_>
kevina are you around?
<JustinDrake>
Kubuxu: Would you recommend OpenBazaar (to be released in April 2017) use CIDv1 instead of CIDv0?
<JustinDrake>
By "use" I mean "recommend"/"use as default"
<Kubuxu>
I think yes, we plan to transition to default CIDv1 in 0.5, with backward compat with CIDv0
<Kubuxu>
IDK if there is a way to have `ipfs add` operation to return CIDv1 instead of v0, this is something that needs a bit more thought.
ylp has quit [Quit: Leaving.]
pfrazee has quit [Read error: Connection reset by peer]
pfrazee has joined #ipfs
shizy has joined #ipfs
MDude has joined #ipfs
<dignifiedquire>
daviddias: I don't get it :(
<dignifiedquire>
before you started merging things the tests in js-ipfs were passing
<dignifiedquire>
and now they don't anymore
<dignifiedquire>
ohhhh
<dignifiedquire>
daviddias: I know the issue in js-ipfs
kenshyx has quit [Remote host closed the connection]
espadrine has quit [Quit: Konversation terminated!]
espadrine has joined #ipfs
<dignifiedquire>
daviddias: we have a problem :(
<dignifiedquire>
I'm afraid we have to change the pubsub http-api interface
<dignifiedquire>
because we can't cancel fetch requests in the browser, resulting in subscriptions not being canceled, which is why the browser tests are failing
<dignifiedquire>
I warned about this multiple times when you mentioned this api for the first time :(
<daviddias>
well, looks like we are going to do what I was proposing in the beginning, have a /feed?topic=<topic-name
<daviddias>
and a /unsubscribe
<daviddias>
hmm sorry if I didn't understand what you were saying, but I believe I was on your side
Mizzu has joined #ipfs
<dignifiedquire>
this is going to be tricky
<dignifiedquire>
because a standing connection is pretty ideal, but ideally you would have sth like a websocket with heart beats
<dignifiedquire>
which the standing http-request comes pretty close to
<daviddias>
a quick patch would be to send a 'FIN
<daviddias>
packet on that open connection
<dignifiedquire>
there is one trick I can still try on js-ipfs-api
<daviddias>
the connection would stay open, but it would tell go-ipfs that we are no longer interested
ulrichard has joined #ipfs
f[x] has joined #ipfs
<dignifiedquire>
daviddias: how are the interop tests coming along? I think we can already work with the current release of ipfsd-ctl for them, as the new features are just improvements but I don't think anything is blocking in terms of the interop tests
lacour has joined #ipfs
JustinDrake has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<flyingzumwalt_>
anyone want to assemble a list of reference materials or tutorials on "How to Sign a Hash and Publish It"?
infinity0 has quit [Ping timeout: 256 seconds]
cubemonkey has joined #ipfs
infinity0 has joined #ipfs
<whyrusleeping>
gmornin
infinity0 has quit [Remote host closed the connection]
s_kunk has quit [Ping timeout: 252 seconds]
<whyrusleeping>
Kubuxu: theres no current way to make ipfs add return cidV1 without recompiling
<whyrusleeping>
but the really easy code change you can make is to change the default prefix in the merkledag package to be version 1
<lgierth>
path is there because path are the primary interface to everything, and cid is there for efficiency (to avoid additional conversions/resolutions/lookups)
<whyrusleeping>
lgierth: because the path shouldnt require us to resolve the whole thing before i can have a path object
<whyrusleeping>
turning /ipfs/QmFooBar/a/bc/d into a Path object shouldnt require me to fetch a bunch of objects
<whyrusleeping>
otherwise i might as well just get the node we're referring to and move on
<lgierth>
it doesnt -- once you need it resolved, you're gonna do CoreAPI.ResolvePath(path)
<lgierth>
and the return is the same path, but with a cid
<whyrusleeping>
is the cid QmFooBar?
<whyrusleeping>
or is it whatever 'd' refers to?
<Kubuxu>
whyrusleeping: responded
<lgierth>
yeah what 'd' refers to -- completely resolved down to what it refers to
<lgierth>
mh there's probably something about IPLD to consider here
<lgierth>
i still don't understand how you can come up with conflicts in what a path refers to (you mentioned that last week or so but didn't say how)
<whyrusleeping>
Kubuxu: its not actually a breaking change
<Kubuxu>
yeah, I've forgot that protobuf uses IDs internally
<whyrusleeping>
lgierth: so that means that i have to fetch a bunch of objects from disk before i can get a path object?
<whyrusleeping>
thats not what that thing is for
<whyrusleeping>
also, the ambiguity i'm talking about revolves around *single step* resolutions
<whyrusleeping>
if i say "resolve one step of QmFooBar/a/b/c/d"
<whyrusleeping>
its unclear if i mean resolve "a" under QmFooBar
<whyrusleeping>
or resolve "b" under "QmFooBar/a"
<whyrusleeping>
we can define this strictly to mean "attempt to resolve a/b/c/d under QmFooBar"
<whyrusleeping>
in all cases
<whyrusleeping>
but the reason i'm making this distinction, is because you might want to try and resolve "b/c/d" under "QmFooBar/a" at some point, and it becomes difficult to specify thats the action you want to perform
anonymuse has quit [Remote host closed the connection]
<lgierth>
the only way of coming up with that conflict is allowing forward-slashes in keys within ipld objects, from what i understand. we don't have the links/data distinction that merkledag has anymore, right?
infinity0 has quit [Remote host closed the connection]
infinity0 has joined #ipfs
<lgierth>
Qmfoobar/a can be a link to something, or it can be more data right there, but not both at the same time
anonymuse has joined #ipfs
<lgierth>
i can see though that "d" isn't neccessarily an object of its own (i.e. with its own cid)
<whyrusleeping>
larssssssas
<whyrusleeping>
the key here is *single step* resolution
<whyrusleeping>
one step
<whyrusleeping>
an ipld object can have longer paths within it, so b/c might be resolvable within the object, but "d" is part of the next object
<lgierth>
i feel like i'm missing something big
<whyrusleeping>
but the user doesnt really know that
<lgierth>
yeah that's what i meant in my last line "i can see though that [...]"
<whyrusleeping>
so the user says "try to resolve "b/c/d" in this object"
* ansuz
tries turning it off and on again
<ansuz>
hi ladies
<whyrusleeping>
and we can only actually resolve b/c in this particular object
<whyrusleeping>
ohai ansuz
<ansuz>
ohai whyrusleeping
<whyrusleeping>
lgierth: so the key here i think is that one path element does not necessarily correlate with one object
<whyrusleeping>
multiple path elements may be resolveable within a single object
<lgierth>
yeeah got that part. so that's the conflict you meant
<lgierth>
and it hints that we either need multiple cids in that path object, or come up with something new
<lgierth>
passing around strings sucks because it makes it harder to use the api for anything more than one-off things
<pjz>
victorbjelkholm pinbits.io is back up and working. I ran into an issue with my deployment automation :(
<lgierth>
or have a path->[cid...] cache within :)
<whyrusleeping>
the conflict i'm worried about is really a separate issue from the paths thing though
<pjz>
victorbjelkholm: but it's fixed now, including the bug you found
<whyrusleeping>
but i think that paths should be {cid, /path/under/that/cid}
<whyrusleeping>
so that turning a string into a path object doesnt require any lookups
<lgierth>
doesn't require any lookups right now either. the resolution down to the cid is on-demand
<whyrusleeping>
but you were saying that you wanted the cid in the path object to refer to the final object in the path
<whyrusleeping>
which requires lookups
<lgierth>
at the point where you actually need to pass it into e.g. n.DAG.Get() -- but yes i haven"t really taken ipld into account
<whyrusleeping>
its not an ipld thing
<Kubuxu>
whyrusleeping: updated the multibase PRs
<lgierth>
well yes eventually things require a lookup so you can grab the stuff from the dagstore
<lgierth>
or whatever store is appropriate :)
<whyrusleeping>
okay, here, what will path.ParsePath(/ipfs/QmFooBar/a/b/c/d/e) do?
<whyrusleeping>
this is what i'm concerned about
FrankPetrilli has joined #ipfs
<lgierth>
won't do a lookup -- lookups/resolutions happen whereever a coreapi func needs a cid, but only gets a path without a cid
<whyrusleeping>
okay
<lgierth>
this extended path object is really just kind of a cache for resolutions
<lgierth>
and right now i'm thinking, let's just have the coreapi impls silently have a cache there or something like that
<lgierth>
yeah except right now it doesn't have a the root cid in there
<whyrusleeping>
hrm... because it could be ipns
<whyrusleeping>
well, ipns paths could be a different object implementing the same interface, but that doesnt need to happen now
<lgierth>
and it kinda assumes that the cid refers to the last path segment, which it doesn't neccesarily do with ipld
<whyrusleeping>
right
<whyrusleeping>
it could be referring to the subobject
<lgierth>
yep
<whyrusleeping>
hrmmm
<whyrusleeping>
whats the cache for exactly?
<lgierth>
i'm thinking: 1) burry this caching within the coreapi itself, hidden from the interface, 2) cache all CIDs that occur within a path (step by step)
<lgierth>
just to avoid extra lookups
<whyrusleeping>
ah
<whyrusleeping>
i see
<whyrusleeping>
how many times will a path be used more than once?
<lgierth>
coreapi func returns "/ipfs/Qmfoobar/a/b/c", and you pass that right into the next coreapi func
<whyrusleeping>
ah
<whyrusleeping>
okay
<lgierth>
and without a kind of cache, it'd do the same shit all over again :)
<whyrusleeping>
riiiiiiight
<whyrusleeping>
okay
<whyrusleeping>
so lets go ahead with your path interface
<whyrusleeping>
the caching will be hidden from the user, i think
<whyrusleeping>
either within an implementation of the path interface, or within the coreapi itself somehow
<whyrusleeping>
(the resolver could cache things like this perhaps?)
ulrichard has joined #ipfs
<lgierth>
yeah that might be a better place
<lgierth>
in addition to the lookup, i was thinking that the pathstring<->cid conversions are also wasted cpu time, but that's less severe than the lookups
rendar has joined #ipfs
<lgierth>
i have to run so they'll still let me swim a bit, but i'll update the proposal with the above
<lgierth>
when i'm back
<lgierth>
thanks for bearing with me
<lgierth>
:]
<whyrusleeping>
haha, of course <3
<whyrusleeping>
communication over the internet is hard
<whyrusleeping>
heading out for a bit, i'll be sporatically available on irc for the next few hours
JustinDrake has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
JustinDrake has joined #ipfs
JustinDrake has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
JustinDrake has joined #ipfs
rc1 has joined #ipfs
ylp has joined #ipfs
<Yatekii>
hey guys, if i'd like to use ipfs to sync my music and dotfiles, would you advise to not do that or if you think this is okay, can I somehow force my server to pull files so my laptop can fetch them even if my pc is not down at home?
<Yatekii>
*if my pc is down
chris613 has joined #ipfs
MDude has joined #ipfs
pcre has joined #ipfs
<deltab>
Yatekii: it's designed for public data, and asks around for other sources, so it's not good for private data
<jbenet>
i need a name for the ipld / ipfs importers project
fzxi has quit [Ping timeout: 258 seconds]
<deltab>
Yatekii: I'd suggest rsync, bup, or bittorrent sync
<deltab>
Yatekii: maybe sparkleshare
<jbenet>
deltab: that's not quite true. Yatekii can use ipfs for private content by (a) making a private network, (b) pre-encrypting the content. so requires some work right now, but all of ipfs _is_ designed for private content too. a lag in implementation does not mean it's not designed for it.
espadrine has quit [Ping timeout: 248 seconds]
<deltab>
okay
<deltab>
at present then, it's not well-suited for it
<jbenet>
its important to have the aims clear, so people do not misunderstand goals.
<jbenet>
deltab: i use personally use ipfs for all sorts of personal content. i pre encrypt it.
<alterego>
I actually encrypt a lot of stuff on ipfs.
<deltab>
ah, okay
<jbenet>
its definitely cumbersome to not have this in the main tooling yet, but it's a hard problem that requires really good UX.
<jbenet>
(i.e. time consuming and we have important things to nail first)
<deltab>
Yatekii: how do you feel about encrypting data, or adding it when updated?
<Yatekii>
uhm depends :S I don't need to encrypt music nor dotfiles
<Yatekii>
I don't mind it being public
<deltab>
copyright might be an issue?
<Yatekii>
uhm well maybe if I seed ... not the downloadpart
<Yatekii>
hmm
Aranjedeath has quit [Quit: Three sheets to the wind]
ppham has joined #ipfs
<Yatekii>
jbenet: how do you encrypt?
ppham has quit [Remote host closed the connection]
<Yatekii>
question is if it adds a lot of overhead
ppham has joined #ipfs
chris613 has quit [Quit: Leaving.]
<deltab>
external tools such as gpg? an encrypted filesystem?
Mizzu has quit [Quit: WeeChat 1.6]
galois_d_ has joined #ipfs
galois___ has joined #ipfs
galois___ has quit [Remote host closed the connection]
galois_dmz has quit [Ping timeout: 245 seconds]
galois___ has joined #ipfs
galois_d_ has quit [Ping timeout: 240 seconds]
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
galois_dmz has joined #ipfs
ppham has quit [Remote host closed the connection]
galois_d_ has joined #ipfs
galois___ has quit [Ping timeout: 258 seconds]
espadrine has joined #ipfs
<jbenet>
i personally use a makefile and `senc` -- a simple encryption thing I wrote). it's not a particularly good setup.
<deltab>
Yatekii: also, yes, you can tell a node to pin a directory (or other content) to fetch and keep it available
ppham has joined #ipfs
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
arpu has joined #ipfs
ppham has quit [Remote host closed the connection]
arpu has quit [Max SendQ exceeded]
ppham has joined #ipfs
anonymuse has quit [Read error: Connection reset by peer]
anonymuse has joined #ipfs
ylp has quit [Ping timeout: 260 seconds]
ylp has joined #ipfs
ZaZ has joined #ipfs
ulrichard has quit [Ping timeout: 248 seconds]
chris613 has joined #ipfs
ylp has quit [Ping timeout: 260 seconds]
Oatmeal has joined #ipfs
ppham has quit [Remote host closed the connection]
anonymuse has quit [Remote host closed the connection]
anonymuse has joined #ipfs
<interfect[m]>
Can git annex encrypt stuff before ipfs-ing it?
<interfect[m]>
Or does git annex only pull from ipfs?
ppham has joined #ipfs
wak-work has quit [Ping timeout: 258 seconds]
anonymuse has quit [Ping timeout: 268 seconds]
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
seafood has joined #ipfs
ppham has quit [Remote host closed the connection]
ylp has joined #ipfs
ulrichard has joined #ipfs
bastianilso_____ has joined #ipfs
cemerick has quit [Ping timeout: 260 seconds]
tilgovi has joined #ipfs
ppham has joined #ipfs
anonymuse has joined #ipfs
Encrypt has joined #ipfs
ppham has quit [Ping timeout: 252 seconds]
eater has quit [Ping timeout: 250 seconds]
eater has joined #ipfs
Akaibu has joined #ipfs
<Yatekii>
deltab: ok, taht's really cool :) will try that ty!
<Yatekii>
first I need to find a tool for organizing music tho :P
ylp has quit [Ping timeout: 260 seconds]
<alterego>
Hrm, anyone know why I keep getting these errors: dagNode.data.copy is not a function
<alterego>
Using js-ipfs-api and ipld-dag-pb
<alterego>
Trying to add a link to an object.
maxlath has joined #ipfs
infinity0 has quit [Remote host closed the connection]
JustinDrake has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
infinity0 has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
Encrypt has quit [Quit: Quit]
<victorbjelkholm>
daviddias, man, webrtc is kind of stupid, trying to connect to localhost and if there is browser reports it's offline, I get "InvalidStateError: Can't create RTCPeerConnections when the network is down"
<victorbjelkholm>
even though everything happens from/to localhost...
ylp has joined #ipfs
infinity0 has quit [Remote host closed the connection]
<interfect[m]>
victorbjelkholm: It's the Internet Supremacists, who believe that persistent connection to the Internet through a single default gateway is the only valid network connectivity state.
seafood has quit [Quit: seafood]
ulrichard_ has quit [Client Quit]
Oatmeal has quit [Ping timeout: 260 seconds]
<lgierth>
victorbjelkholm: that condition sounds similarly naive as chrome's test for whether to use ipv6
ylp has quit [Ping timeout: 260 seconds]
<lgierth>
victorbjelkholm: would be great to figure that out and write down
<victorbjelkholm>
lgierth, what they do?
<victorbjelkholm>
not sure there is much to figure out, I mean, even if offline, just try to connect anyways? :p
<victorbjelkholm>
ooh, just ~15 seconds? That's good, considering I'm on a bus in the middle of sweden currently
chungy has quit [Ping timeout: 258 seconds]
<lgierth>
hahaha that's nice
<lgierth>
!botsnack
<sprint-helper>
om nom nom
<pinbot>
om nom nom
<lgierth>
victorbjelkholm: yeah but what i mean is, do other browsers do the same, is the behaviour part of the specs, etc.
<lgierth>
it's worth checking out so we can push vendors and w3c in the right direction
<lgierth>
same for the really-nice-to-have-in-the-future WebUDP api
<victorbjelkholm>
lgierth, ah, I see what you mean... Yeah, just quickly tested firefox and chrome, they acted the same, didn't even try to connect
chungy has joined #ipfs
<lgierth>
the chrome+ipv6 issue is just plain annoying :) if it can't ping6 some google server, it'll just refuse to do any AAAA dns lookups
<victorbjelkholm>
wow, sounds like a very good way of making sure your software won't survive if your company don't survive
<lgierth>
so you basically can't point a domain name at a cjdns fc00::/8 address, unless you have ipv6 internet too
<lgierth>
that too hehe
ZaZ has quit [Quit: Leaving]
anonymuse has joined #ipfs
bsm117532 has quit [Ping timeout: 268 seconds]
<victorbjelkholm>
daviddias, once you get online, put bunch of new previews of libp2p website in the issues, please take a look when you can
<lgierth>
just noticed a request to ipinfo.io on that libp2p.io preview ^
<lgierth>
is that part of external address discovery?
<lgierth>
(just saying it's a commercial service)
<victorbjelkholm>
lgierth, yeah, something I'm trying out to be able to get the country of the visitor
<victorbjelkholm>
idea is that the nodes would bear the flag of the current country you're in
<victorbjelkholm>
if you know a better way? :)
pcre has quit [Remote host closed the connection]
<whyrusleeping>
when will I be able to not load unsafe scripts for js-ipfs?
seafood has joined #ipfs
<interfect[m]>
Not load unsafe scripts?
<lgierth>
victorbjelkholm: eeh i'm not a big fan of nation state flags -- but anyhow you can use the geoip data which the webui uses too
<lgierth>
if that makes i simpler
<victorbjelkholm>
whyrusleeping, when we have dns in multiaddr or daviddias I think had a alternative solution. We need secure webrtc when pages uses https, and I think secure webrtc needs a domain
se3000 has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
<lgierth>
also i figure some of your people in barcelona might not be too happy about being represented by the spanish king's flag :P
<victorbjelkholm>
lgierth, well, it's a cool effect, even though I might not agree with all nations :) And true, catalans will strike!
<lgierth>
whyrusleeping victorbjelkholm: yes it's about both /dns and /tls
<victorbjelkholm>
I think what the webui uses is based that the multiaddr includes the actual ip of the node, in browser land (afaik, right now), we don't get that
arpu has joined #ipfs
<victorbjelkholm>
with webrtc-star, it seems the address is the webrtc-star address + node ID
<victorbjelkholm>
otherwise I would have used that
<lgierth>
ah yeah for the geoip data you need to find out your external address first
<victorbjelkholm>
yeah, hence the call to ipinfo.io ... Would love to avoid it though, but not sure how