notduncansmith has quit [Ping timeout: 256 seconds]
<jbenet>
kbala: does it pass for you locally?
<jbenet>
kbala: the tests sort-of passed on travis. notifs failed. so am running them again
<cryptix>
jbenet: o/ how is 0.4 coming along? :)
<kbala>
jbenet: oh sorry didn't see this, i'll pull and run them now
<jbenet>
cryptix: o/ good, coming!
<cryptix>
awesome :)
<cryptix>
trying to keep track of what you guys are up too but its hard currently
<kbala>
jbenet: they pass locally!
<jbenet>
kbala: ok let's see if they pass fine on travis this round
<jbenet>
cryptix: thanks! no worries :)
pfraze has joined #ipfs
<kbala>
jbenet i think notifications failed again
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to heems-master: http://git.io/vmRvw
<ipfsbot>
go-ipfs/heems-master 9f99dec Juan Batiz-Benet: better error for notifs...
<jbenet>
kbala: yeah that's annoying. not sure why this test keeps failing
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] jbenet closed pull request #1225: Rewrite pinning to store pins in IPFS objects (master...pin) http://git.io/vU00W
<cryptix>
jbenet: how confident+annoyed are you in the tests these days?
mdem has joined #ipfs
luca has quit [Changing host]
luca has joined #ipfs
<jbenet>
kbala: i think this works. squash my changes into yours? (i pushed another commit for debugging notifs)
<bret>
Great video https://www.youtube.com/watch?v=skMTdSEaCtA#t=331 lol at jbenet getting nerd sniped every time a technical inaccuracy is used as a demonstrative analogy/example :p
<bret>
don't worry about ;P ip people get on your back about it, its their issue not yours
<bret>
if*
<bret>
but yeah super good talk
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<Leer10>
is ipfs theoretically secure?
<bret>
thaaaats a tough question to answer
<jbenet>
Leer10 depends on the threat model.
<luca>
theoretically, everything's secure
<Leer10>
example: using IPFS on an unlocked wifi network
<Leer10>
so someone can use something like wireshark
<bret>
theoretically, linux is secure
<luca>
theoretically, TLS is secure
<jbenet>
Leer10 today our threat model does not include anonymity. the content is secure (authenticated by content addressing and/or cryptographic signatures, and the comm is encrypted.)
<Leer10>
anonymity isn't needed :)
<lgierth>
the comm is encrypted with 2048 bit rsa *cough* :)
<lgierth>
i need to keep nagging
<jbenet>
Leer10: they wont be able to see anything but the handshake. (with one caveat that our secio still needs to be audited professionally, or we need to switch to lib impls of TLS)
<bret>
if a copyright troll has a hash of a file they are concerned about, they could find peers with that hash... *right*?
* lgierth
off to bed
<Leer10>
bret: I'm pretty sure that is possible but then someone could flip a few bits
<bret>
Leer10: absolutely
<jbenet>
lgierth: the comm is setup with RSA (soon ECC) but encrypted with AES or Blowfish. (we should move to salsa/chacha)
<jbenet>
Leer10: please pre-encrypt any personal/private content you're concerned about though, for now. we'll have content encryption soon, but not there yet.
<Leer10>
okay
<Leer10>
FYI the most I've uploaded was 2 30 mb personal videos to share with a few friends
<bret>
how does hash obscurity compare to bitterness sync folder keys?
<kbala>
jbenet: make vendor gives me a bunch of package not found errors, how do i install all of these?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
kbala: go get all of them? maybe try `godep restore`
semidreamless has joined #ipfs
<cow_2001>
so what kind of irl stuff could i do with ipfs?
<Wallacoloo>
cow_2001: It's decent for some types of collaborative work. I write electronic music & use IPFS for maintaining a large sample library with my collaborators. Makes it way easier than manually tracking which samples I need to send to my team every time I make a change & it's more practical than putting the whole library on something like Dropbox due to the size and the fact that many samples don't need to ever actually be accessed.
<cow_2001>
how do you manage such a huge pile of files? is it added as a file hierarchy? was the whole directory tree addition memory hogging problem solved?
<Wallacoloo>
cow_2001: Sorry, I don't know much of anything about the internals of IPFS. The library isn't THAT huge - somewhere between 1000-3000 samples probably, many of which are never actually transferred to anyone. Yes, added as a hierarchy & never noticed any memory issues, but I didn't look out for that so potentially it is a problem. idk.
<whyrusleeping>
we've been really looking for someone who uses docker a lot to get some insight into how we can make their life easier
<Luzifer>
hmmm don't how ipfs could make docker easier…
<Luzifer>
imho docker is already quite easy…
<whyrusleeping>
okay
<whyrusleeping>
do you share docker containers with other people at work?
<whyrusleeping>
like, create one and have someone else pull and run it?
<whyrusleeping>
or better question, do you guys run your own registry?
<Luzifer>
just had the case today I needed an automated job in our main jenkins: I gave up after 4 hours of work, thats just impossible… 2min into docker: running;
<Luzifer>
yep we do
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<Luzifer>
currently the old and crappy v1 but there is already a PR for switching to v2
<Luzifer>
(crappy because I had to clean up messy images more than 3 times…)
<whyrusleeping>
lol
<Luzifer>
something went wrong while pushing and really _everything_ broke… no image based on ubuntu:14.04 was pushable anymore…
<whyrusleeping>
<.<
<whyrusleeping>
wat
<whyrusleeping>
thats impressive
<ipfsbot>
[go-ipfs] whyrusleeping created s3-0.4.0 from s3 (+0 new commits): http://git.io/vmgZu
therealplato has quit [Read error: Connection reset by peer]
<whyrusleeping>
sprintbot, worked on gx, now i'm making the datastore stuff work cleanly
Guest94496 is now known as shea256
www has quit [Ping timeout: 244 seconds]
<jbenet>
That's worse. Fragments the interfaces all over. No just ship it
<jbenet>
(Note though that it's annoying no matter how we dice it)
<shea256>
yo @jbenet
<jbenet>
shea256: hey! How's it going
<shea256>
jbenet: good good
<shea256>
first time in #IPFS
<shea256>
thought I'd see what's up
<jbenet>
welcome :)
<shea256>
thank you!
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<daviddias>
sprintbot: successfully move swarm to use stream-muxer abstraction and wrote a bunch of more tests for it. Switched Identify to use protobufs
tilgovi has quit [Ping timeout: 240 seconds]
<shea256>
jbenet: have you and muneeb discussed collaborating on the DHT layer since the last time we chatted?
<jbenet>
shea256: not more than that we should
<jbenet>
shea256: we're down to
<shea256>
ok let's do it
<shea256>
OpenBazaar and OB1 have also been making progress on their DHT
rht___ has quit [Quit: Connection closed for inactivity]
<shea256>
makes sense for us to all work on the same implementation
<jbenet>
Yep
<jbenet>
We can setup a time to discuss altogether. First thing to do is make sure it makes sense from everyone's requirements
<Tv`>
whyrusleeping: threadsafedatastore has plagued me too.. Though I'm not gonna blame Go on that one..
<shea256>
yes agreed
<whyrusleeping>
Tv`: also, with close in the interface, we can completely do away with the 'defaultDatastore' type
<Tv`>
whyrusleeping: well if you're doing something about it, let me know; I'm planning on redoing the Query method
<whyrusleeping>
the issue could be somewhat addressed in go if you could cleanly 'embed' multiple interfaces into a struct
<Tv`>
whyrusleeping: the method set of a type needs to be known at compile tone
<Tv`>
Time
<Tv`>
You can embed s combination of interfaces, but it needs to be a fixed set
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
whyrusleeping: before you make this change just send me an example of you're having a problem with? I suspect one interface type would fix the issue
<jbenet>
He io package cleanly separated all these interface pieces and with good reason
<jbenet>
Yes you do get some garbage interfaces as a result but the workload on implementing the core interface is very little. One needs only to implement Write or Read or Close as they want
<jbenet>
Changing the datastore interface and _all_ code on everyone every time we want a new kind of feature isn't?
<Luzifer>
Wow. I think I broke the universe. I'm calling a method expecting one argument with exactly one argument… Go compiler is telling me I'm not passing enough arguments… Wat?
<whyrusleeping>
Luzifer: have you tried turning it off and back on again?
<daviddias>
can connections be upgraded from ipv4 to ipv6 without us telling it to do so?
<Luzifer>
The universe? Let me press the reset button… Hope you all saved… *click*
<daviddias>
jbenet whyrusleeping ^^, is that a supposed behaviour?
<Luzifer>
someone might teach me something different by example but I'm quite certain this is not possible at all…
<daviddias>
Luzifer, I'm opening a tcp socket explicitly to a ipv4 address, but then when I console.log the remoteAddress I get:
<daviddias>
::ffff:127.0.0.1 IPv6 52864
<Luzifer>
O_o
<Luzifer>
okay the universe is broken. I'm going home.
notduncansmith has joined #ipfs
<Luzifer>
look like some ip handler is written to use ipv6 notation of ipv4 ips to connect
notduncansmith has quit [Read error: Connection reset by peer]
<daviddias>
ahah, let's fix it before going home :)
Leer10 has quit [Remote host closed the connection]
<Tv`>
jbenet, whyrusleeping: What has given me trouble is wrapping a Datastore that may or may not be ThreadsafeDatastore; you end up doing the wrapping on just Datastore, then in the code that does the wrapping make sure you have ThreadsafeStore, wrap it, then wrap it again to add the threadsafeness back
<Tv`>
Luzifer: you're calling a method on a type not on a value of the type
mdem has joined #ipfs
<Luzifer>
ahhhhh fuck
<Luzifer>
thanks
<whyrusleeping>
Tv`: yep! very much adreed
<whyrusleeping>
agreed*
<Luzifer>
Tv`: I'm searching that error for like 15mins… -_-
<lgierth>
sprintbot: finally wrapped up monitoring. i'm happy to help get you started whyrusleeping jbenet kbala krl cryptix. now looking into pinbot @ solarnet
<jbenet>
grawity: so the 157.60.0.1+ relevant udp port is embedded in the addr, correct?
<whyrusleeping>
the entire util/datastore2 package is terribad
<grawity>
yes
<daviddias>
grawity, so are the ::ffff:<IPv4> also teredo?
<grawity>
daviddias: no
<grawity>
daviddias: those are literally IPv4 addresses written in "v6 mapped" form
<grawity>
mostly so that you could use an AF_INET6 socket for both v4 and v6
<daviddias>
so teredo is just for tunnels, I see, thank you. So, does the IPv4 addrs mapped to IPv6 have a specially name?
<grawity>
which kind, the ffff ones?
<daviddias>
s/specially/special
<daviddias>
grawity: yes, the ffff ones
<grawity>
"IPv4-mapped"
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
patcon has quit [Ping timeout: 264 seconds]
<lgierth>
whyrusleeping: ok, looking at pinbot/main.go -- is the hosts file newline-delimited?
* lgierth
hopes ipfs-shell can deal with ipv6 addresses
<jbenet>
lgierth it should be able to-- multiaddr.
<lgierth>
ack
<jbenet>
if not it's a bug
Encrypt has quit [Quit: Sleeping time!]
<whyrusleeping>
lgierth: its just using a net.Dial("tcp", $HOST)
<whyrusleeping>
i beleive that should work just fine
<lgierth>
whyrusleeping: i need to s/localhost:6667/irc.freenode.net:6667/ eh?
<whyrusleeping>
ah, crap. i forgot to swap that out...
<lgierth>
wanna make it a flag or so?
<lgierth>
:)
<whyrusleeping>
yeap, one sec
<daviddias>
lgierth, IPFS should do IPv6 just fine, it is just the Node.js `net` module that translates automatically every IPv4 addr to IPV6, which is not great when I'm trying to tell the other peer which 'observed addrs' I have from it
notduncansmith has joined #ipfs
pinbot-test has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
pinbot-test has quit [Remote host closed the connection]
<lgierth>
daviddias: lol how helpful and considerate :P
<daviddias>
Ahaha :)
rschulman_ has joined #ipfs
<whyrusleeping>
lgierth: i pushed
<lgierth>
coolio
<alu>
alright lol
<alu>
which one of you is void4 ;)
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
therealplato has quit [Ping timeout: 240 seconds]
<alu>
Wow.
<jbenet>
alu: void's not here i think.
<jbenet>
alu: what's up?
<alu>
whoever void is, they just published a REALLY interesting project on GH
<alu>
out of the blue
therealplato has joined #ipfs
<alu>
which changes the workflow of janusVR
<alu>
now you can go directly from blender and publish a live VR room over IPFS
<alu>
I think its a good idea as I was just on the ipfs examples repo looking through
<alu>
i think they should be together
woahbot has joined #ipfs
<alu>
!mail yo void I saw your awesome project published on github! I have many questions but for now atleast ya deserve a cold drink, my treat :)
<woahbot>
alu: I'll give your message to yo the next time I see them.
<alu>
shit rofl
<alu>
!mail void yo I saw your awesome project published on github! I have many questions but for now atleast ya deserve a cold drink, my treat :)
<woahbot>
alu: I'll give your message to void the next time I see them.
<alu>
im gunna leave this bot in here is that okay?
<alu>
it wont bother anyone
<jbenet>
alu; that's fine
<whyrusleeping>
Tv`: i may be killing the ThreadSafe datastore thing
<alu>
!g test
<woahbot>
alu: http://www.speedtest.net/ - Test your Internet connection bandwidth to locations around the world with this interactive broadband speed test from Ookla.
<Tv`>
whyrusleeping: beware jbenet's wrath
<whyrusleeping>
Tv`: i have a good sword
<Tv`>
whyrusleeping: a "here, take this" association made me think that i should name a git server "finger"
<Tv`>
though i will never outdo "git push zen master"
<alu>
alright im headed to the coffee shop
<Tv`>
that was completely accidental
notduncansmith has joined #ipfs
<alu>
jbenet did you see the site i made earlier
<alu>
lol
notduncansmith has quit [Read error: Connection reset by peer]
<alu>
open the same link in janus for a multiuser private p2p virtual reality room
<jbenet>
Tv` i suggested that since the repo takes a threadsafe datastore, and then wraps it (knowing with that) it doesnt need to necessarily force every wrapper to be threadasfe (i.e. the repo is doing all the wrapping knowing that is or isnt ok)
<alu>
im hosting my neural network experiments on here tho so people can see the scale of things i do
<alu>
like upscaling from 256x256 to 4K
<alu>
i dont own a 4K screen but in VR you can understand the scale of things by not looking at a monitor but see it in 3D
<alu>
bbl, im craving a coffee or white russian now lol
<alu>
that void is having a drink on me
<jbenet>
alu: cool :)
qqqueue has joined #ipfs
patcon has joined #ipfs
qqueue has quit [Ping timeout: 264 seconds]
shea256 has quit [Remote host closed the connection]
rschulman_ has quit [Ping timeout: 256 seconds]
<lgierth>
whyrusleeping: still using the why01 and why2 droplets?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
hellertime has joined #ipfs
<whyrusleeping>
lgierth: yeap!'
<lgierth>
ack
<lgierth>
will leave them be then ;)
<whyrusleeping>
gratzi!
therealplato has quit [Ping timeout: 265 seconds]
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to s3-0.4.0: http://git.io/vmaVN