Guest60594 has quit [Remote host closed the connection]
inconshreveable has joined #ipfs
O47m341 has joined #ipfs
jaboja64 has quit [Ping timeout: 248 seconds]
devbug_ has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
zootella has joined #ipfs
zootella has quit [Quit: zootella]
micxjo has joined #ipfs
_fil_ has joined #ipfs
<ipfsbot>
[js-ipfs] xicombd created refactor/http-api-routes (+2 new commits): https://git.io/vaczV
<ipfsbot>
js-ipfs/refactor/http-api-routes 419e916 Francisco Baio Dias: Register routes on the injected connection
<ipfsbot>
js-ipfs/refactor/http-api-routes 293765a Francisco Baio Dias: Add api running test for ipfs version
M-Cosmo has left #ipfs ["User left"]
zootella has joined #ipfs
<ipfsbot>
[js-ipfs] xicombd opened pull request #84: Refactor HTTP API routes (master...refactor/http-api-routes) https://git.io/vacz5
leer10 has joined #ipfs
zootella has quit [Quit: zootella]
ggoZ has quit [Ping timeout: 276 seconds]
reit has joined #ipfs
warner` is now known as warner
M-whyrusleeping has quit [Quit: node-irc says goodbye]
ipfsrocks has joined #ipfs
r04r is now known as zz_r04r
qbg has quit [Quit: Konversation terminated!]
qbg has joined #ipfs
computerfreak has quit [Quit: Leaving.]
mnp has joined #ipfs
inconshreveable has quit [Read error: Connection reset by peer]
tmg has quit [Ping timeout: 240 seconds]
<warner>
hey, is anyone here familiar with the way IPFS does TCP NAT-hole-punching?
<warner>
I'm trying to build something similar (but in python, with Twisted)
<warner>
is it basically: set SO_REUSEPORT, listen on 4001, make outbound connections with the local end bound to port=4001, ask everyone you connect to what address+port they see, then advertise that address+port to everyone?
<jbenet>
warner: we dont do an extensive amount yet. we've some expriments with natty and other things, but right now we do minimal stuff: we detect the outgoing ports and try to use them, advertise those addresses, and we try to use nat-pmp and upnp where available
* warner
nods
<jbenet>
and yes SO_REUSEPORT too.
rhalff has quit [Ping timeout: 244 seconds]
<warner>
SO_REUSEPORT was new to me, pretty crucial
<warner>
do you advertise port=4001, or port=(whatever your peers said they hear you on)?
<jbenet>
warner: this will need hooks with policy (like, dont want internal ip addrs getting out there necessarily). as well as known relays for trying to holepunch, etc.
<warner>
ok, so you gather externally-reported addresses and merge them into your advertisement
<jbenet>
warner: my takeaway from all NAT stuff is: **tableflip** -- we need a packetized transport at the application layer
<warner>
yeah
<jbenet>
warner: and we're building it-- more to come soon.
<warner>
I'm trying to figure out what's the simplest thing I can do that'll get me to 80%
<warner>
nice
dignifiedquire has quit [Quit: Connection closed for inactivity]
<jbenet>
simplest: support nat-pmp, nat upnp, advertise all known addrs, take observations.
<jbenet>
beware though, as cone nats may blow up your known addresses. make sure you see observations from multiple distinct peers
<jbenet>
errrr "fully symmetric nats"*
<warner>
I think it's gonna be SO_REUSEPORT, glean an external address from one connection, advertise localport, externalport, and maybe (localport + 2*(externalport-localport)) since some NATs allocate ports that way
<warner>
yeah
<jbenet>
yes SO_REUSEPORT is a must.
<warner>
so are you basically doing simultaneous-TCP?
<jbenet>
i'd advertise real observations. even then you'll get a bunch of useless addresses (symmetric nats)
* warner
nods
<jbenet>
we dont yet do simult open. we need a relay for that
<jbenet>
we've relaying in but we havent done it.
<warner>
according to the 2005 paper, most boxes allocate a stable port, but won't accept inbound connections unless they see an outbound SYN for the same 4-tuple
<jbenet>
(basically peer-mediated STUN)
<warner>
I've started to strace and pcap the IPFS process to figure out what's really going on
devbug_ has quit [Ping timeout: 248 seconds]
<jbenet>
haha, nice. im sure you'll find lots of wonkiness, please file issues! <3
<warner>
well, mostly I'm trying to figure out how it interacts with my (linux-based) NAT box
<warner>
so there's nothing that particularly coordinates my box making an outbound connection to peer X at the same time that X tries to connect to me, right?
<warner>
is there anything about the DHT "fingers" or "leaves" or whatnot that would cause connection attempts to line up?
<warner>
(e.g. if X wants to connect to Y, Y also wants to connect to X, so they're likely to be making attempts at the time time, which might cause their NATs to allocate compatible table entries)
<jbenet>
there's no simultaneous open no, and it's hurting us bad. we need it soon.
<jbenet>
oh nice observation, from dht there may be some accidental simultaneity. but not sure how likely it is
<warner>
ok, good to know
<warner>
so I wonder if the connections are only working on the NATs that allow inbound connections despite not seeing a SYN for the full 4-tuple
<warner>
but I've got a lot of port=4001 connections established, and I think my NAT doesn't allow those
<warner>
hm, well, the pcap will help figure that part out
<jbenet>
yeah, unclear. would love any intel you gather on this.
<jbenet>
thinks work ok, but not nearly as well as we want it to.
<rinpoo>
nice does conventional code work? Like if I wanted to use a phpBB could I use that with ipfs?
<kmq>
rinpoo: from what I understand, currently it's possible to have static(except for javascript) websites only
devbug has quit [Read error: Connection reset by peer]
devbug has joined #ipfs
<rinpoo>
ahh ok so it's like zeronet
<rinpoo>
would it be possible to use ipfs to host content on a normal website?
<Kubuxu>
Yup, ipfs.pics is PHP website and hosts pictures on IPFS.
<Kubuxu>
you have to just use nginx/apache rewrite rules to rewrite requests to /ipfs/... to local gateway on server site.
<rinpoo>
do I have to recode the php or can I just use something that has been done already
<rinpoo>
nice
<Kubuxu>
You just link to files under IPFS as /ipfs/Qmfdsdfsdfds/pic.png and then rewrite server site /ipfs/... to localhost:8080 or whatever port IPFS gateway on server runs.
kmq has quit [Read error: Connection reset by peer]
<yangwao>
but ipfs.pics is just writable gateway, literally :)
<yangwao>
but also, nice one!
<rinpoo>
thx I'll give ipfs a try then
<rinpoo>
I'm not too techy
<rinpoo>
more on the degin side
<rinpoo>
design
<yangwao>
:)
<Kubuxu>
When I make static sites I use IPFS to show them off to someone.
<deltab>
the current limits are 100 names per cert, 5 certs per domain per week; "There is no limit to the number of certificates that can be issued to different domains."
<deltab>
you were telling the shell to redirect from a file called 3
<rinpoo>
yes I realized
inconshreveable has joined #ipfs
<M-1>
hohai
pfraze has quit [Remote host closed the connection]
ugjka has quit [Quit: Leaving]
zz_r04r is now known as r04r
<rinpoo>
any reason why I need to read write my own blocks? I don't really get what the Dealing With Blocks is about
<deltab>
you don't need to; it's just to show you how it works at low level
kmq has quit [Read error: Connection reset by peer]
<rinpoo>
oh ok
<rinpoo>
I thought it was important
<rinpoo>
when I sue ipfs swarm connect /ip4/104.236.176.52/tcp/4001/ipfs/qmsolnsgccfuzqjzradhn95w2crsfmzutddwp8hxahca9z I get: Error: invalid peer address: failed to parse ipfs: qmsolnsgccfuzqjzradhn95w2crsfmzutddwp8hxahca9z failed to parse ipfs addr: qmsolnsgccfuzqjzradhn95w2crsfmzutddwp8hxahca9z multihash too short. must be > 3 bytes
<rinpoo>
use
mildred has quit [Ping timeout: 244 seconds]
<lgierth>
rinpoo: the part behind /ipfs/ is case-sensitive
kmq has joined #ipfs
<lgierth>
yours is completely lowercase while it should start with QmSoL, and there probably are a few more uppercase chars
<lgierth>
did you find it lowercase like that somewhere?
<rinpoo>
is it the peer ID I created upon generation?
<deltab>
no
<rinpoo>
ok them I'm completely lost
<rinpoo>
then
<deltab>
if you go to http://localhost:8080/ipfs/ plus the hash from the add command, you'll see the files you added
<deltab>
and if you go to the .md file, you'll see the raw markdown source
<deltab>
to have it rendered, you need to instead load a page with javascript in it; that script will read the markdown and convert it to html for display
<deltab>
that viewer is also distributed via ipfs, so it has its own hash
<rinpoo>
it's really confusing that the example viewer is in js and the example int go lang
<rinpoo>
-t
<voxelot>
haad: awesome list!
<voxelot>
i like seeing a title like why vector clocks are easy and then why they are hard haha
<rinpoo>
I'm kinda stuck I went trough all examples but the Simple JS Apps just confuse me. What's the next I should read trough?
<rinpoo>
+h
wiedi has quit [Ping timeout: 240 seconds]
<ipfsbot>
[js-ipfs] xicombd pushed 1 new commit to refactor/http-api-routes: https://git.io/vaC8x
<ipfsbot>
js-ipfs/refactor/http-api-routes 2f8b804 Francisco Baio Dias: Add api running test for ipfs version & fix #70
Guanin_ has joined #ipfs
wiedi has joined #ipfs
HostFat has quit [Ping timeout: 244 seconds]
pfraze has quit [Remote host closed the connection]
pfraze has joined #ipfs
ipfsrocks has joined #ipfs
<rinpoo>
https://ipfs.io is down for me does it work for you guys?
zorglub27 has quit [Ping timeout: 244 seconds]
<rinpoo>
just came back on
<voxelot>
Kubuxu, noffle, haad, : would you guys be interested in having a hangout just for pub/sub again, jbenet is doing a hangout for IPLD and maybe we could squeeze in after
<voxelot>
tomorrow, if dignifiedquire or someone could be kind enough to host
<Kubuxu>
I won't be available tomorrow at usual time as I have a inter-city bus, but you can do it without me.
<voxelot>
no worries, some other time is always an option too
hellertime has joined #ipfs
<noffle>
voxelot: sure. if you can org it I can be there :)
nonaTure has joined #ipfs
M-kode54 has quit [Quit: node-irc says goodbye]
rhalff has joined #ipfs
ipfsrocks has quit [Ping timeout: 248 seconds]
ygrek has joined #ipfs
ulrichard has joined #ipfs
inconshreveable has joined #ipfs
sikander has quit [Remote host closed the connection]
sikander has joined #ipfs
<ipfsbot>
[ipfs] RichardLitt pushed 2 new commits to master: https://git.io/vaCzr
<ipfsbot>
ipfs/master 83184fe Richard Littauer: Merge pull request #161 from slothbag/sloth_wip...
zorglub27 has joined #ipfs
leer10 has quit [Ping timeout: 252 seconds]
hellertime has quit [Quit: Leaving.]
mildred has joined #ipfs
orography has quit [Remote host closed the connection]
overdangle has joined #ipfs
M-cryptix has quit [Quit: node-irc says goodbye]
palkeo has joined #ipfs
micxjo has quit [Quit: Connection closed for inactivity]
Senj has quit [Ping timeout: 260 seconds]
jokoon has quit [Quit: Leaving]
patcon has joined #ipfs
Encrypt has quit [Quit: Quitte]
jaboja has quit [Ping timeout: 246 seconds]
hellertime has joined #ipfs
JasonWoof has quit [Quit: leaving]
_fe_ has joined #ipfs
compleatang has quit [Remote host closed the connection]
reit has joined #ipfs
reit has quit [Client Quit]
disgusting_wall has joined #ipfs
compleatang has joined #ipfs
inconshreveable has quit [Remote host closed the connection]
inconshreveable has joined #ipfs
Senji has joined #ipfs
ipfsrocks has joined #ipfs
Senji has quit [Ping timeout: 244 seconds]
Guanin_ has left #ipfs [#ipfs]
Guanin_ has joined #ipfs
zorglub27 has quit [Quit: zorglub27]
HostFat has joined #ipfs
_fe_ has quit []
Encrypt has joined #ipfs
tilgovi has joined #ipfs
Akaibu has joined #ipfs
HostFat has quit [Quit: Leaving]
mildred has quit [Ping timeout: 276 seconds]
hellertime has quit [Quit: Leaving.]
leer10 has joined #ipfs
M-jgrowl has quit [Quit: node-irc says goodbye]
hellertime has joined #ipfs
<mnp>
any go devs on?
silotis has quit [Remote host closed the connection]
silotis has joined #ipfs
nikor has quit [Quit: Lost terminal]
inconshreveable has quit [Remote host closed the connection]
<noffle>
mnp: what's your question?
<mnp>
hey. go noob here. i'd like to modify one of the godeps in go-ipfs. If I do that, is there a way to convince godeps or gx to propogate the changes into my workspace?
computerfreak has joined #ipfs
drwasho has joined #ipfs
<mnp>
i can do "go build" down in the submodule and it's happy with that. the changes are not visible up in root when i make build.
<drwasho>
Morning folks
<drwasho>
From Australia haha
compleatang has quit [Ping timeout: 252 seconds]
Tv` has quit [Ping timeout: 252 seconds]
palkeo_ has joined #ipfs
Ronsor` has quit [Ping timeout: 252 seconds]
spm_draget has quit [Ping timeout: 252 seconds]
ygrek has quit [Ping timeout: 244 seconds]
silotis has quit [Ping timeout: 252 seconds]
ipfsrocks has quit [Ping timeout: 252 seconds]
spm_draget has joined #ipfs
leer10 has quit [Ping timeout: 252 seconds]
palkeo has quit [Ping timeout: 252 seconds]
Ronsor` has joined #ipfs
ugjka has quit [Ping timeout: 268 seconds]
<drwasho>
Can someone point me to where I can find 0.4 build of IPFS?
<jbenet>
drwasho: hey-- i think you have to build it from src atm. cc am i right, whyrusleeping Kubuxu daviddias noffle ?
<jbenet>
drwasho: let's setup a time to talk next week, some news :)
<Kubuxu>
yup, that will be best way
<Kubuxu>
it was quite a time since last release
<daviddias>
drwasho: jbenet yeah, to get the latest and greatest, it has to be from src (master branch)
ipfsrocks has joined #ipfs
inconshreveable has joined #ipfs
leer10 has joined #ipfs
<drwasho>
Thanks guys, I'm keen to test out NAT traversal features
<drwasho>
Because in OpenBazaar we're talking more seriously about the possibility of using IPFS. Did you guys see Chris Pacia's Youtube video?
<drwasho>
jbenet: Lets definitely hook up a time to talk next week. Either mornings or late afternoons EST work well for timezone overlap
O47m341 has quit [Ping timeout: 260 seconds]
leer10 has quit [Read error: No route to host]
leer10 has joined #ipfs
zootella has joined #ipfs
pfraze has quit [Remote host closed the connection]
<mats_>
But I have problems with the 0.4.0-dev builds, they don't work with the Java and Scala APIs
<mats_>
Maybe they aren't OK?
<mats_>
I need to build from sources?
drwasho has quit [Ping timeout: 260 seconds]
drwasho1 has joined #ipfs
<dignifiedquire>
mats_: there were some changes in 0.4 vs 0.3.11 that might cause incompatability on the api layer. But I suggest try building master first and checking if that solves your issues
<dignifiedquire>
what exactly isn't working for you?
<mats_>
Unit tests are failing when I try to build the Java and Scala APIs
<mats_>
I have filed bug reports
<dignifiedquire>
can you give me a link?
<mats_>
The Java version seems to be maintained, but not sure about the Scala API
<mats_>
I have struggled with this for some time now, but I am not stuck, am doing other stuff and faking all IPFS calls in my code
<dignifiedquire>
mats_: the multipart errors are to be expected, as that part was rewritten entirely in 0.4 (it was actually broken before in some subtle ways)
<dignifiedquire>
mats_: but I see ianopolous has already commented and made somme commits a couple of hours ago
<dignifiedquire>
he is probably much more up to date on the state of the java api
<ianopolous>
dignifiedsquire: yep I'm looking into it, thanks.
<mats_>
Oh, I haven't checked that in a few hours, behind the curve, as always...
<mats_>
Thanks guys
<dignifiedquire>
for the scala api I'm not sure, I guess this hasn't been updated sind 0.3.9 which is quite old now
<mats_>
Is Go 1.5.2+ necessary? Ubuntu 15.10 has a standard package for 1.5.1
<ianopolous>
mats_: I will ping Chris to update the Scala API
<ipfsbot>
[webui] dignifiedquire closed pull request #297: Update eslint to version 2.4.0
Tv` has joined #ipfs
* noffle
is making good progress on his sunday hack for a distributed ipfs wiki
rendar has quit [Ping timeout: 264 seconds]
palkeo_ has quit [Quit: Konversation terminated!]
drwasho has joined #ipfs
rendar has joined #ipfs
silotis has joined #ipfs
drwasho1 has quit [Ping timeout: 276 seconds]
<mats_>
@ianopolous I don't think my program runs against the wrong IPFS, since it returns the correct version string, but it could be that I need to install IPFS from source
<mats_>
The build from sources descriptions is wrong:
<mats_>
OK, I see now. Oh well, too much for me now, past midnight over here
<ianopolous>
mats: I've built IPFS from source on 0.4.0-dev and all the tests pass except for "fileTest" which fails on the "ipfs ls" and I'm following that up with go-ipfs
inconshreveable has quit [Ping timeout: 264 seconds]
<mats_>
OK, I will do that tomorrow. I think the 0.4.0-dev binaries should be removed, I have wasted hours on testing them
computerfreak has quit [Quit: Leaving.]
r04r is now known as zz_r04r
tmg has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]