<cryptix>
btw could it be that sometimes wantlist entries arnt removed? if i stop a node during refs -r and bring it up again, i get spammed with blocks until i manually unwant them
<whyrusleeping>
cryptix: thats strange
chriscool has quit [Ping timeout: 268 seconds]
<cryptix>
i think i will try to add a bucket counter into the dup stat thing and print a warning when a block was received un-asked-for 10 times - its quite the nethog
<cryptix>
but sleep now - l8ters
<achin>
today, in achin's adventures in ipfs+rust, we've learned how to generate an IPFS private key, and export it to an ipfs config file \o/
ianopolous2 is now known as ianopolous
<ion>
gj
jamie_k_ has quit [Quit: jamie_k_]
<whyrusleeping>
achin: nice!
<whyrusleeping>
not that i advocate doing so, but it should be possible to use your ipfs private key for ssh
<whyrusleeping>
just have to export it
<achin>
could use it for all sorts of things! could also public your public key, and have people send you encrypted cat pictures
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/log-hang from 4cb826f to 823df45: http://git.io/vlJ3U
<ipfsbot>
go-ipfs/fix/log-hang 823df45 Jeromy: fix log hanging issue, and implement close-notify for commands...
<achin>
whyrusleeping: how long did it take you guys to find the keys for your sol nodes?
<whyrusleeping>
achin: a few days mining
<achin>
each? or for all?
<whyrusleeping>
hardest part was getting enough entropy
<achin>
in either case, i guess that doesn't seem too bad
tilgovi has quit [Ping timeout: 268 seconds]
<whyrusleeping>
we just mined a bunch of keys, and picked ones that looked good
<ion>
whyrusleeping: Unfortunately publishing in the test suite is still broken (https://github.com/ipfs/go-ipfs/issues/1941), but I verified the tests to pass in a modified environment where the daemon has access to other peers.
<whyrusleeping>
hrm..............
<whyrusleeping>
ion: you could add some resolve tests to t0130
<whyrusleeping>
it has multiple ipfs nodes
<achin>
M-davidar: brute force and fast machines!
amstocker has joined #ipfs
simonv3 has quit [Quit: Connection closed for inactivity]
<M-davidar>
achin (IRC): I mean, is it the same as the bitcoin address mining tools, it simmering special?
<M-davidar>
Er, something*
<achin>
i generate an RSA key, see if its ipfs peerID looks interesting. if not, i toss it out, and generate a new one
<achin>
here, "looks interesting" means "contains some text i want"
<M-davidar>
cool, anything interesting yet?
<achin>
depends on your definition, i guess :) since you asked for archive-stuff, i plugged in a few letters and found the key QmQKeGJYAhaArchJgNfL1jBBMGRTLdnt9aXhS7GBLRRt75 (has "Arch" in the middle)
<achin>
i'm too impatient for prefix matching at the moment, but i'll let this run overnight
wHATEver_ has joined #ipfs
wHATEver_ is now known as myp
<ion>
whyrusleeping: Added some tests to t0130.
<ion>
achin: There is a fireplace downstairs (with other tenants) but it’s mostly electric.
<achin>
for someone living in the northeast (USA), it's hard to imagine a form of heating that doesn't involve a flame
<achin>
ion: want a new ipfs key? :D
<ion>
achin: Accepting a private key generated by someone else would be bad practice even though I trust you. :-P
<achin>
well, let me tell you what you're missing!
<achin>
turns out "ion" is a fairly common occurence in keys (ignoring case). the best one i found was QmamdUyC9nZKSs3QgwDgwrwY465wvtGm8xUmHWentD2ion
<ion>
nice
<ion>
>amd
<achin>
i also found QmeYb6oXqfrbFKUDpVFvqehEJZVjp3CRiN4SKjK4aJiPfs
<ion>
heh
semiconfident has quit [Killed (Sigyn (Spam is off topic on freenode.))]
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/log-hang from 823df45 to c5b133a: http://git.io/vlJ3U
<whyrusleeping>
(i'm not actually sure if that maps over neatly to the libp2p spec or not)
<whyrusleeping>
but the host is what we use to actually open new streams, and set stream handlers and stuff
<whyrusleeping>
go-ipfs has two implementations of that, the first implementation relies on the addresses you feed into it to be able to find and dial peers
<whyrusleeping>
you can connect to someone if you dont know their ip and port
<whyrusleeping>
we give that host to the dht, and then use the dht to create an enhanced 'host'
<whyrusleeping>
when you call 'connect' or whatever on that one, and you dont have addresses for the peer, it will use the dht to search for addresses
<amstocker>
hmm okay
<ion>
Does go-ipfs store a list of previously connected peers on disk to be able to bootstrap in case there is a problem with all the bootstrap nodes?
<amstocker>
so the first implementation uses kademlia to do the routing
Oatmeal has quit [Ping timeout: 260 seconds]
<amstocker>
*some variant of kademlia
<amstocker>
but then whats the point of having the dht protocol on top of libp2p?
telharmonic has joined #ipfs
<amstocker>
I think I get the gist of what you said tho, thanks
dd0 has quit [Read error: Connection reset by peer]
<achin>
ion: seems like you could make a backup list yourself by saving the output of "ipfs swarm peers"
<achin>
(i think that's already in the right format for feeding into "ipfs connect")
Oatmeal has joined #ipfs
tilgovi has joined #ipfs
dd0 has joined #ipfs
tilgovi has quit [Ping timeout: 268 seconds]
<atgnag>
I hope ipfs can replace the annoying file-sharing sites…
sseagull has quit [Quit: Lost terminal]
<whyrusleeping>
atgnag: i already use it to replace them
<whyrusleeping>
its sooooo much easier for me to just add a file to ipfs and give a link to someone on ipfs.io
border0464 has quit [Ping timeout: 246 seconds]
border0464 has joined #ipfs
<atgnag>
whyrusleeping: Right. Unfortunately, there's still so much content that's uploaded to megaupload clones, and the like.
<whyrusleeping>
yeah...
<whyrusleeping>
the times they are a changin
<amstocker>
can't wait for the multiverse
<whyrusleeping>
cryptix: Blame question
<whyrusleeping>
if we fail to dial a peer, should we remove their addresses from our local addressbook?
<Blame>
There are differing opinions
<Blame>
In what i've done i do that and rely on maintenance to re-add the peer if it was temporarily unavailable
<Blame>
There it also an approach of 'three strikes' or similar
<whyrusleeping>
hrm... okay
<Blame>
I liked the first method because it minimized the state i had to keep
<whyrusleeping>
ipns (and other dht ops) are slow right now because we wait on dials to dead peers to timeout before finishing the query
<whyrusleeping>
agreed
<whyrusleeping>
i think tossing dead peers is totally fine
<Blame>
Just make sure you don't get into a situation where you keep dropping and re-adding dead peers
<ion>
How about an exponential backoff?
<whyrusleeping>
hrm, we do exponential backoff for something
<whyrusleeping>
now to figure out what that actually does
<whyrusleeping>
looks like we do have backoff for dead peers
<whyrusleeping>
but its obviously not working...
telharmonic has quit [Killed (Sigyn (Spam is off topic on freenode.))]
Matoro has quit [Ping timeout: 255 seconds]
Cole has joined #ipfs
ygrek_ has quit [Ping timeout: 252 seconds]
Matoro has joined #ipfs
Cole has quit [Remote host closed the connection]
Oatmeal has quit [Ping timeout: 272 seconds]
<whyrusleeping>
you guys are going to love this
<ipfsbot>
[go-ipfs] whyrusleeping created dht/speed (+2 new commits): http://git.io/vljMN
<ipfsbot>
go-ipfs/dht/speed ec06ecf Jeromy: Add in some more notifications to help profile queries...
<ion>
whyrusleeping: It might be nice for some applications if you could ask for the first valid record the daemon can find and then get notified if the daemon finds something better.
stoopkid has joined #ipfs
<ion>
whyrusleeping: Perhaps even have the daemon poll the IPNS name periodically as long as the API request is active.
<ion>
whyrusleeping: But yeah, I realize pubsub will cover most of this at some point.
reit has quit [Quit: Bye]
<zignig>
ion: looking forward to pubsub, opens quite a few nifty applications.
reit has joined #ipfs
* zignig
melts his brain reading CRDT wikipedia and papers.
<ion>
whyrusleeping: Looking at swarm_dial.go, there are many Unlock invocations not in defer statements. I don’t doubt the author checked that whatever was executed before the Unlock should be unable to panic, but verifying that when reading the code requires work whereas you can just glance at “defer …Unlock()” and instantly know it’s alright.
<Functional>
Any fun one liner examples using cURL I can run straight from the command line (Linux)?
mildred has joined #ipfs
amstocker has quit [Ping timeout: 255 seconds]
anshukla has quit [Remote host closed the connection]
rendar has joined #ipfs
<haadcode>
daviddias: thanks for merging, and also updating the npm package! I can finally get rid of my custom ipfsd-ctl and ipfs-api node_modules in my repo \o/
Functional has quit [Ping timeout: 246 seconds]
<daviddias>
thank you for that PR :)
mikee has quit [Ping timeout: 268 seconds]
<daviddias>
always feel welcome to submit your changes to the repos
border0464 has quit [Ping timeout: 264 seconds]
border0464 has joined #ipfs
domanic has quit [Ping timeout: 246 seconds]
niekie has quit [Ping timeout: 260 seconds]
<myp>
if i add directory to the first node, check the time, run on second node `ipfs ls /ipfs/hash_of_directory`, check the time again, may difference between that times be the time of directory files transferred to second node?
<davidar>
myp: i'm not sure what you're asking?
<myp>
davidar: i want to measure speed between two peers
<davidar>
myp: ls won't necessarily transfer the whole directory, if that's what you're asking
<davidar>
use `ipfs refs -r` instead
besenwesen has quit [Quit: ☠]
davidar_ has quit [Quit: Connection closed for inactivity]
<cryptix>
gdmrnng
besenwesen has joined #ipfs
besenwesen has quit [Changing host]
besenwesen has joined #ipfs
<myp>
davidar: thanks
<whyrusleeping>
ion: yeah, thats kinda what i'm trying out
<whyrusleeping>
i'm not sure what the goal of the original implementation was
s_kunk has quit [Ping timeout: 240 seconds]
NightRa has quit [Quit: Connection closed for inactivity]
bedeho has quit [Ping timeout: 260 seconds]
bedeho has joined #ipfs
niekie has joined #ipfs
bedeho has quit [Ping timeout: 268 seconds]
mildred has quit [Ping timeout: 244 seconds]
s_kunk has joined #ipfs
Tv` has quit [Quit: Connection closed for inactivity]
<ion>
whyrusleeping: Re: GitHub, making a public constant mutable just for random code to be able to use it in a table with an optional field seems to me like the worst workaround. :-\
Dorsobranchiata has quit [Ping timeout: 240 seconds]
<cryptix>
we currently have no CI system for our tests on windows so _here be dragons_ :)
<glumpy`>
And can people host regular site with ipfs? with html5, durpal, wordpress, wikimedia, etc?
<glumpy`>
forums etc?
<cryptix>
you have to rethink the application model of dynamic sites (like wikimedia, forums, etc) to apply them properly to ipfs
<glumpy`>
You mean the registration process of the users?
<ion>
Specifically, any centralized design will gain little from IPFS.
<ion>
(But distributed designs can gain a lot.)
<glumpy`>
I'm wondering why no one has created a torrent search engine with reviews and all (like kickass, btjunkie, etc) on a decentralized system. Like what ipfs is trying to offer or maelstrom (if it really exists) etc.
<glumpy`>
Hmmm I was under the impression that in ipfs everyone holds a copy of the site. From what I read on issue 8, that's not the case. How then does this protect the site from losing data if one of the nodes disappears?
<cryptix>
ipfs gives you the tools to create the content in a way that it is very easy to verify and distribute because it is immutable
<cryptix>
to make it permanent we need service like ipfs-cluster or later filecoin
<cryptix>
(ipfs-cluster and 'openipfs' are working titles of ipfs nodes cooperating to keep content availalbe)
<cryptix>
but every user also requests and hosts content if they run ipfs locally (or soon in the browser even)
<cryptix>
it's a bit of a rabbit hole at first... :)
<cryptix>
glumpy`: tl;dr - everyone _who requested it_, holds a copy of the content
<M-davidar>
cryptix (IRC): the cothority thing looks interesting
<M-davidar>
You could try bringing it up in the aggregation thread?
<cryptix>
M-davidar: :) nice - i so its not just my monkey brain
<cryptix>
M-davidar: the 300 page CRDT issue? i can try to make a stand :)
* cryptix
needs to try matrix.org again
<M-davidar>
cryptix (IRC): yeah... that was my concern too
<glumpy`>
cryptix, Hold a copy of the content he requested or of the whole site? I guest the first option. Because you would like to d/l a full site just to read an article
<M-davidar>
Yeah
<M-davidar>
cryptix (IRC): use vector.im/beta
<cryptix>
M-davidar: where do you link to the latest vertex viewer?
<cryptix>
vecotr*
<cryptix>
heh guess you answered that :)
<glumpy`>
:)
<M-davidar>
cryptix (IRC): I haven't updated the ipfs hosted copy in a while, waiting on some upstream changes
<glumpy`>
So then only the altruistic people or the ones who get paid will hold a full copy, right?
<M-davidar>
cryptix (IRC): it seems kind of related to the CRDT discussion, so maybe those people will have something to say about it? :/
<M-davidar>
glumpy` (IRC): basically, yeah
NeoTeo has joined #ipfs
<M-davidar>
Economically the situation isn't that different to the current system
<cryptix>
M-davidar: re crdt: i think so, too - they might have other ideas how to apply those principles
<M-davidar>
Although ipfs does allow a more competitive market
<cryptix>
M-davidar: i'll try to make a summary that fitts into the discussion
<M-davidar>
cryptix (IRC): cool, that would be great
<glumpy`>
I read a few months ago about an ex reddit employee who's trying to create a decentralized reddit with tips for moderators and holders of content. Have you heard about it?
<cryptix>
nothing concrete.. the SNR around meta-reddit went way down for me
<glumpy`>
SNR?
<cryptix>
signal to noise ratio
<cryptix>
to much bad stuff, to much news.. have lost track of where it is heading
<cryptix>
but it's definetly an interesting proposition to build a reddit ontop of ipfs
<glumpy`>
When do you think ipfs will be ready for the public?
<davidar>
glumpy` (IRC): depends on your definition of public
<cryptix>
+1 - i think even for novice web-devs the time is right now, if you are unafraid of new concepts :)
<cryptix>
(and bugs ^^)
<glumpy`>
people that are able to install it as a plugin on firefox/chrome, if it's ever possible to do it like that.
<davidar>
glumpy` (IRC): you won't even need a plugin with ipfs.js
<davidar>
daviddias: ^ timeline?
<daviddias>
for full IPFS node in JS?
<glumpy`>
Since it's partially based on bittorrent, I think the public is people that are able to use bittorent. It's not 80% of the people, but it's probably 10-20%
<davidar>
daviddias (IRC): yeah
<daviddias>
depends a bit on the go-libp2p and its refactoring, if we don't stumble into any hard constraint, I hope to be around December or early January
<cryptix>
rough beta for end of december would be epic
<cryptix>
32c3
<glumpy`>
And why did you say "you won't even need a plugin"? How will people install it then?
<daviddias>
glumpy`: because it will be 100% JS
<davidar>
glumpy` (IRC): they won't, it will just be loaded like any other js
<glumpy`>
hmmm
<glumpy`>
So whenever I go to an ipfs site I wouldn't know the difference?
M-cryptix has joined #ipfs
<cryptix>
glumpy`: right
<cryptix>
there are plugins for firefox and chrome right now, that can use the local go-ipfs instance already
<cryptix>
so when you visit ipfs.io/ipfs/$contentHash it redirect through localhost:8080 to pull the content
<cryptix>
but ipfs.js will be much less friction
<davidar>
cryptix (IRC): y u irc? :(
<glumpy`>
And how do you solve the problem of site take downs?
<davidar>
glumpy` (IRC): as in DMCA?
<glumpy`>
yeah
<davidar>
kyledrake and lgierth are working on it
<davidar>
There's going to be a blacklist
<glumpy`>
of what?
<davidar>
Of taken down content
<glumpy`>
But how will a simple user be able to get it? if for example a site like secrets.org was taken down, you try to enter it and you get a message that no site exists
anticore has quit [Ping timeout: 265 seconds]
<ion>
The people who run the public HTTP gateway will inevitably receive DMCA demands and they will prevent the gateway from serving that content to comply with the law. You get to make the decision about what you download and upload using your personal node.
<M-cryptix>
davidar: habbits... :)
<davidar>
@cryptix:matrix.org: :)
locusf has quit [Ping timeout: 255 seconds]
<glumpy`>
But couldn't there be an internal registrar like .onion, opendns or other systems?
<davidar>
glumpy` (IRC): yeah, ipfs will be compatible with onion addresses, etc, soon enough
<davidar>
glumpy` (IRC): ipfs isn't going to officially sanction illegal content, but of course you're free to do whatever you want :)
<glumpy`>
onion addresses without the need to use tor? I mean without tor's slow speeds?
<davidar>
glumpy` (IRC): like namecoin?
<glumpy`>
namecoin is one option
<davidar>
At the end of the day, you can't take down a hash without taking down all the nodes serving it
<davidar>
So, take from that what you will
<davidar>
And you can connect to ipfs peers over whatever transport you want
<glumpy`>
illegal is sometimes based on culture and regime.. in some places telling what the government has done is illegal, in others sharing copyright material is illegal, and in some reading news and information from other parts of the world is illegal
<davidar>
I'm not sure what we're arguing about
<davidar>
If you want to do that stuff, nobody's stopping you
<davidar>
But the official nodes run by protocol labs are going to comply with the appropriate laws
Deadly has quit [Remote host closed the connection]
trisyllabical has joined #ipfs
TheWhispery has joined #ipfs
TheWhisper has quit [Ping timeout: 246 seconds]
<glumpy`>
I'm not arguing, just pointing out that the classification of something as illegal is not global.
<davidar>
I know, it's wherever that particular node is located
<davidar>
glumpy` (IRC): for example, a nice based in the US might not want to distribute 1984 because it's still under copyright, but a node in Australia would be perfectly OK with it since it's public domain here
<davidar>
*node
<glumpy`>
Really? How come it's a public domain in Australia?
<davidar>
We have shorter copyright periods
<ion>
Mickey Mouse hasn’t managed to bully or bribe your politicians?
<davidar>
Not yet, but I suspect it won't be too long :(
<davidar>
Especially with the tpp :/
<glumpy`>
So one is able to download/read 1984 from an Australian site without this content being flagged as pirated?
<glumpy`>
I thought the copyright craziness is an anglo-saxon related issue. It's surprising that Australia has lax views about it
OutBackDingo has quit [Quit: Leaving]
OutBackDingo has joined #ipfs
Zuardi has quit [Remote host closed the connection]
Zuardi has joined #ipfs
<M-davidar>
We also have public health and (to some extent) tertiary education ;)
<glumpy`>
hehe
mildred has quit [Ping timeout: 240 seconds]
cemerick has quit [Ping timeout: 250 seconds]
hellertime has joined #ipfs
anticore has joined #ipfs
OutBackDingo has quit [Remote host closed the connection]
OutBackDingo has joined #ipfs
anticore has quit [Remote host closed the connection]
<achin>
overnight, i generated about 7 million keys. one of the strings i was looking for was "ipfs", and i found about 200 of those
mungojelly has quit [Ping timeout: 246 seconds]
<cryptix>
achin: mining key prefixes? :))
dansup has quit [Ping timeout: 240 seconds]
<achin>
yep. also found this key: QmQVKcRyptix76pvw6Ugn5UeuiTGU2fR3cMDVSKaH4vUSS
<cryptix>
hahaha <3
<cryptix>
can i haz?
<achin>
yep :) let me know how you want it delivered (maybe a encrypted GPG message?)
<cryptix>
keybase.io/cryptix
<achin>
ok! i'm heading into work now, but i'll send you something soon.
<cryptix>
also: can you link me again to your irclogs? botbot cant fnd it somehow..
<achin>
and then once you have it, i will permanently delete the key from my system
<cryptix>
(how ironic, right?)
<cryptix>
perfect
<cryptix>
no rush
<achin>
let me find the irc thingus
<cryptix>
quick q if you have the time: how confident do you feel about the format of the logs (not the viewer)
<achin>
not sure what you mean
<nicolagreco>
achin do you randomly generated keys or proceeded in a sequence
<M-cryptix>
if i would start to traverse and index it, would the traversal change much? havent looked into it alot tbh thinking out loud
<achin>
the format of the logs is just however weechat dumps them to disk. my tool is hard-coded with this format in mind, but can easily be adapted to any file that includes a timestamp (or other sequence number) on each line
jamescarlyle has quit [Remote host closed the connection]
jamescarlyle has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
jamescarlyle has joined #ipfs
hellertime has quit [Read error: Connection reset by peer]
hellertime has joined #ipfs
NightRa has joined #ipfs
anticore has joined #ipfs
rombou has joined #ipfs
<cryptix>
doodle.com 503.... siiigh
<cryptix>
one more week of flu and ill make that myself
<kpcyrd>
on ipfs!
<cryptix>
of course :)
diffalot has quit [Ping timeout: 250 seconds]
drathir has quit [Ping timeout: 244 seconds]
koo7 has quit [Ping timeout: 250 seconds]
<cryptix>
also: the sad moment when you realize the unix tool you wanted to wrap into a api service doesnt support stdio because it uses file extensions from cli args...
glumpy` has quit [Read error: Connection reset by peer]
diffalot has joined #ipfs
diffalot has joined #ipfs
glumpy` has joined #ipfs
mue has joined #ipfs
substack has joined #ipfs
__konrad__ has quit [Ping timeout: 250 seconds]
mue_ has quit [Ping timeout: 255 seconds]
ThomasWaldmann has quit [Ping timeout: 255 seconds]
M-nated has quit [Ping timeout: 250 seconds]
elasticdog has quit [Ping timeout: 250 seconds]
cblgh has quit [Ping timeout: 250 seconds]
substack has quit [Ping timeout: 250 seconds]
M-cryptix has quit [Ping timeout: 250 seconds]
M-whyrusleeping has quit [Ping timeout: 250 seconds]
nicolagreco has quit [Ping timeout: 250 seconds]
rollick has quit [Ping timeout: 250 seconds]
JasonWoof has quit [Ping timeout: 250 seconds]
cblgh has joined #ipfs
vakla has quit [Ping timeout: 250 seconds]
JasonWoof has joined #ipfs
__konrad__ has joined #ipfs
vakla has joined #ipfs
elasticdog has joined #ipfs
koo7 has joined #ipfs
rollick has joined #ipfs
drathir has joined #ipfs
ThomasWaldmann has joined #ipfs
ThomasWaldmann has quit [Changing host]
ThomasWaldmann has joined #ipfs
M-nated has joined #ipfs
M-cryptix has joined #ipfs
M-whyrusleeping has joined #ipfs
JohnClare has joined #ipfs
JohnClare has quit [Remote host closed the connection]
JohnClare has joined #ipfs
JohnClare has quit [Remote host closed the connection]
rombou has quit [Ping timeout: 240 seconds]
JasonWoof has quit [Quit: why isn't this in screen?]
JasonWoof has joined #ipfs
JasonWoof has joined #ipfs
jamescarlyle has quit []
sonatagreen has joined #ipfs
rombou has joined #ipfs
sseagull has joined #ipfs
pfraze has joined #ipfs
cemerick has joined #ipfs
ashark has joined #ipfs
<ion>
Depending on file extensions to determine the format is evil.
rombou has quit [Ping timeout: 264 seconds]
<ansuz>
integers with an even number of bits are evil
dansup has joined #ipfs
mildred has joined #ipfs
guybuddy has joined #ipfs
<guybuddy>
is it possible to upload content to ipfs swarm that is password/identity protected??
<guybuddy>
so that nodes cant look at it or someone browsing the hashes on the chain cannot see it?
<ion>
guybuddy: For now, encrypt it with PGP first. IPFS implementations are going to have built-in support for key management and on-disk encryption in the future.
<guybuddy>
okay so eventually it will be built in, thats cool... whats the policy and theory behind hosting illicit content then?
<guybuddy>
is it just content blind or are there going to be choices?
<ion>
The policy: your law enforcement won’t like it if you do that. The theory: uh, I have no idea.
<ion>
IPFS is just a protocol.
anticore has quit [Ping timeout: 240 seconds]
<achin>
recall that your ipfs node won't serve/contain data unless you downloaded it, or you added it
<achin>
or in the case you run a public gateway, if someone requests hash via your gateway
<ion>
But in that case you are protected in many legislations (automatically or by implementing a procedure like the DMCA in USA).
anticore has joined #ipfs
<achin>
you *might* be protected in many legislations
voxelot has joined #ipfs
voxelot has joined #ipfs
<ion>
Google successfully serves cached copies of web pages and images in most countries.
<achin>
yes, and google likely has a team of lawyers to make sure they are legally protected
anticore has quit [Ping timeout: 240 seconds]
<achin>
(you would probably be OK, but i don't think anyone in here is a lawyer)
bedeho has joined #ipfs
<kpcyrd>
achin: I'd post the link again but jbenet told me not to
<ipfsbot>
[js-ipfs] diasdavid created jsipfs (+2 new commits): http://git.io/v8UZF
<ipfsbot>
js-ipfs/jsipfs bbe6d1c David Dias: jsipfs roadmap
<ipfsbot>
js-ipfs/jsipfs ae86582 David Dias: add commands for ronin
water_resistant has joined #ipfs
<ion>
“the link”?
<kpcyrd>
accidentially distributed drm keys from other peoples nodes
<achin>
whos keys?
mildred has quit [Ping timeout: 250 seconds]
<kpcyrd>
achin: google "AACS controversy"
<alterego>
jbenet, why not use runtime.GOMAXPROC(runtime.NumCPU()) ?
Encrypt has joined #ipfs
rombou has joined #ipfs
simonv3 has joined #ipfs
glumpy` has quit [Quit: Leaving]
<achin>
davidar: daviddias: either of you interested in the novely IPFS key QmfMPH7ZJYNwWThaaVjhaevEoqNjfpN95DD5j64HrDAViD
<demize>
alterego: Isn't that the default with Go 1.5?
<alterego>
demize: I would have thought so, wondering why runtime.GOMAXPROCS is set to 3 then in ipfs
<demize>
Good question.
<demize>
Probably just not been changed since it defaulted to 1
<kpcyrd>
achin: basically, ipfs is built around the assumption that nobody can trigger a GET to the gateway, while most applications are built around the assumption that a GET must not do anything so it's all ok if there's a GET here and there
anticore has joined #ipfs
<kpcyrd>
there are way too many ways to cause a GET to somewhere, the browser being the most notorious one
<achin>
im' not sure what you're saying here
munkey has joined #ipfs
<ipfsbot>
[js-ipfs] diasdavid opened pull request #35: Getting jsipfs out the door (master...jsipfs) http://git.io/v8Ua7
domanic has joined #ipfs
<kpcyrd>
I totally understand why it's built the way it is, but you need to be careful with "you don't seed unless you chose to". Especially Germany is fun with copyright laws.
Tv` has joined #ipfs
<achin>
in the current version of ipfs, there are only a very limited set of commands that will cause your ipfs node to serve data
bedeho has quit [Ping timeout: 260 seconds]
besenwesen_ has joined #ipfs
M-amstocker1 has joined #ipfs
bret has quit [Ping timeout: 264 seconds]
<kpcyrd>
achin: how good are you with http?
M-oddvar1 has joined #ipfs
ike___ has joined #ipfs
<achin>
pretty good?
haadcode_ has joined #ipfs
dvn- has joined #ipfs
sbruce has joined #ipfs
kumavis_ has joined #ipfs
gendale__ has joined #ipfs
eaterof has joined #ipfs
Stskeepz has joined #ipfs
besenwesen_ has quit [Changing host]
besenwesen_ has joined #ipfs
besenwesen has quit [Disconnected by services]
domanic has quit [Ping timeout: 246 seconds]
<kpcyrd>
achin: if you request a file from 127.0.0.1:8080/, you send a GET, ipfs gets the file, caches it and redistributes it
<achin>
right
jfred__ has joined #ipfs
<achin>
now your node will contain to make that file avaialble over http (via the gateway), or via ipfs, to other ipfs nodes
<M-cryptix>
altergo re gomaxprocs: that's from pre 1.5 times and the idea behund not using ncpus() is, to not kill/overload the system
besenwesen_ is now known as besenwesen
<M-cryptix>
should prob. set to 3*ncpus()/4 or something like that
lohkey_ has joined #ipfs
rombou has quit [Ping timeout: 250 seconds]
tiago_ has joined #ipfs
yosafbridge` has joined #ipfs
water_resistant1 has joined #ipfs
<kpcyrd>
achin: now imagine you run a browser on the same box and click random links on irc
<kpcyrd>
achin: nope, because GET is considered harmless
rendar has joined #ipfs
<kpcyrd>
you won't get the response back, but it still makes the request to check if there's CORS enabled
<kpcyrd>
but basically that, yeah
wowaname has joined #ipfs
domanic has joined #ipfs
bedeho has joined #ipfs
lohkey_ is now known as lohkey
ogd has joined #ipfs
donpdonp has left #ipfs ["WeeChat 0.3.8"]
yoshuawuyts has joined #ipfs
jhiesey has joined #ipfs
<achin>
i just tried something similar (using jquery via the javascript console), and i got net::ERR_BLOCKED_BY_CLIENT
kumavis_ is now known as kumavis
Oatmeal has joined #ipfs
NightRa has joined #ipfs
oleavr has joined #ipfs
eaterof is now known as eater
cemerick has quit [Ping timeout: 244 seconds]
* achin
-> afk; lunch
anticore has quit [Remote host closed the connection]
munkey has quit [Ping timeout: 240 seconds]
<kpcyrd>
achin: only tested this in firefox
<kpcyrd>
achin: but there are /some/ other implementations of http which don't check this
<noffle>
morning
bret has joined #ipfs
<kpcyrd>
achin: maybe xhr isn't the best example for this, try <img> instead
Stskeepz is now known as Stskeeps
Stskeeps has quit [Quit: Reconnecting]
Stskeeps has joined #ipfs
Stskeeps has joined #ipfs
chriscool has quit [Ping timeout: 272 seconds]
zeroish has joined #ipfs
<clever>
achin: yeah, the browser will block any cross-origin where the JS can access the reply, but for img/css/script, the JS cant so it just allows it
<clever>
a common bypass is to <script src="example.com/page.php?data=body&cb=cb123"> which returns cb123(reply);
anshukla has joined #ipfs
domanic has quit [Ping timeout: 268 seconds]
guybuddy has quit [Ping timeout: 246 seconds]
s_kunk has quit [Ping timeout: 240 seconds]
<achin>
good point about images
pfraze has quit [Ping timeout: 255 seconds]
amade has joined #ipfs
pfraze has joined #ipfs
domanic has joined #ipfs
sonatagreen has left #ipfs ["Leaving"]
HoboPrimate has joined #ipfs
cow_2001 has quit [Quit: ASCII Muhammad - @o<-<]
HoboPrimate has quit [Remote host closed the connection]