<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to feat/tar-fmt: http://git.io/vZ0Vh
<ipfsbot>
go-ipfs/feat/tar-fmt 533a729 Jeromy: add sharness test for tar commands...
border has quit [Ping timeout: 255 seconds]
border has joined #ipfs
bedeho has quit [Ping timeout: 244 seconds]
hellertime has joined #ipfs
border has quit [Remote host closed the connection]
border has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
bedeho has joined #ipfs
hyperbot has quit [Remote host closed the connection]
hyperbot has joined #ipfs
border has quit [Remote host closed the connection]
border has joined #ipfs
<16WAASHPS>
[go-ipfs] jbenet closed pull request #1680: first pass at a tar importer (master...feat/tar-fmt) http://git.io/vZBwz
<7JTAANM7H>
[go-ipfs] jbenet deleted feat/tar-fmt at 533a729: http://git.io/vZ0oh
<arrdem>
what's the expected latency of the ipfs mount? I can't seem to cd to /ipfs/<me>
<whyrusleeping>
whats <me> ?
<arrdem>
my id
<whyrusleeping>
thats not a thing
inconshr_ has quit [Remote host closed the connection]
<whyrusleeping>
/ipfs/<hash> points to objects matching the hash given. your peerID is the hash of your public key
<whyrusleeping>
so if your public key were added to ipfs (which we currently dont do), you could cat /ipfs/<id> and see the raw binary blob of your public key
<whyrusleeping>
yeap, once ipns works, youll be able to cd to /ipns/<your id> and see things that you've explicitly published
voxelot has quit [Ping timeout: 250 seconds]
<jbenet>
note the N in ipns
<arrdem>
gotcha.
<arrdem>
sweet! that totally worked
<arrdem>
read > alpha quality software
<whyrusleeping>
yeap, lol
<whyrusleeping>
there are rough edges that we've yet to file down
NHAS has quit [Quit: Leaving]
<arrdem>
so ordinarily I'd whine about golang, but this is pretty awesome. howdoi help?
hyperbot has quit [Remote host closed the connection]
hyperbot has joined #ipfs
<jbenet>
arrdem what languages do you prefer?
<arrdem>
jbenet: I'm kidding, my background is C and Clojure. Been looking for an excuse to explore Rust/Go for a while now.
inconshreveable has joined #ipfs
hyperbot has quit [Remote host closed the connection]
hyperbot has joined #ipfs
<Leer10>
arrdem do you now a dn22?
<spikebike>
heh, I mentioned a ipfs name publish command I ran that was slow, it's still running 24 hours later
<arrdem>
Leer10: can't say I do
<Leer10>
okay then
<Leer10>
I knew a guy on another IRC channel that knew clojure and that channel had stuff like talking about dragons
<arrdem>
I also like dragons :P but I know not of such a place
<whyrusleeping>
spikebike: that means it hung, lol
jamescarlyle has joined #ipfs
<spikebike>
yeah, when you said slow I didn't think days ;-)
<ipfsbot>
[go-ipfs] whyrusleeping closed pull request #1644: import sub-lib of iptb so go doesnt whine about importing a binary (master...fix/godeps-1.5) http://git.io/vGQbt
<whyrusleeping>
yeeeaahh. slow like "it might take a minute tops"
<whyrusleeping>
(and a minute is overkill, it should just be a few seconds really)
jamescarlyle has quit [Ping timeout: 265 seconds]
<ipfsbot>
[go-ipfs] whyrusleeping deleted fix/godeps-1.5 at be14e5f: http://git.io/vZ0Mo
<ipfsbot>
[go-ipfs] whyrusleeping created fix-godeps-go1.5 (+1 new commit): http://git.io/vZ0MX
<ipfsbot>
go-ipfs/fix-godeps-go1.5 7496df7 Jeromy: vendor binary deps in test/dependencies...
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1686: vendor binary deps in test/dependencies (master...fix-godeps-go1.5) http://git.io/vZ0Mj
jhulten has quit [Remote host closed the connection]
<davidar>
whyrusleeping: pinbot still doesn't like me :(
<Quiark>
can you access QmZC2SeNHF4QxdL3jJdsC386ZMuxNj4KA5WWkdVCQ5yVY5 ?
rschulman_ has joined #ipfs
<whyrusleeping>
davidar, he needs to be redeployed
joshbuddy has quit [Quit: joshbuddy]
joshbuddy has joined #ipfs
jager has quit [Ping timeout: 250 seconds]
captain_morgan has joined #ipfs
jager has joined #ipfs
jager has quit [Read error: Connection reset by peer]
jager has joined #ipfs
kerozene has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman_ has quit [Quit: rschulman_]
<qgnox>
Quiark, no.
jhulten has joined #ipfs
keroberos has joined #ipfs
joshbuddy has quit [Quit: joshbuddy]
captain_morgan has quit [Read error: No route to host]
hellertime has quit [Quit: Leaving.]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
jhulten has quit [Ping timeout: 250 seconds]
orzo has quit [Ping timeout: 255 seconds]
orzo has joined #ipfs
atgnag has quit [Read error: Connection reset by peer]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
atgnag has joined #ipfs
atgnag has quit [Ping timeout: 246 seconds]
<Quiark>
so files from one of my machines are not accessible, how would I start debugging it? ipfs diag net prints a LOT of lines so it seems to be connected
<Quiark>
ipfs swarm peers also is non-empty
atgnag has joined #ipfs
<davidar>
Quiark: behind a nat?
<Quiark>
yes
rdbr has quit [Quit: WeeChat 1.3]
<davidar>
yeah, that's known to cause problems, ipfs doesn't have full nat-busting going yet
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<Quiark>
hm, last time I was told here that it does
<davidar>
whyrusleeping: nat status?
captain_morgan has joined #ipfs
border has quit [Remote host closed the connection]
Pharyngeal has quit [Ping timeout: 272 seconds]
<whyrusleeping>
davidar: nat works for the most part, just super restrictive nat will fail
<whyrusleeping>
Quiark: is it all files from your node?
<whyrusleeping>
try adding a new file and requesting it through the gateway
<whyrusleeping>
it could be network churn dropping records again
<Quiark>
whyrusleeping, yeah it seems like all files. Gateway didn't give me anything (context deadline exceeded)
jhulten has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
mquandalle has quit [Quit: Connection closed for inactivity]
bedeho has joined #ipfs
jhulten has quit [Ping timeout: 265 seconds]
<whyrusleeping>
Quiark: hrm, whats your peer id?
<Quiark>
PS C:\Devel\gopath\src\github.com\ipfs\go-ipfs\cmd\ipfs> ipfs id
<gendale_>
I have a custom list of hashes that I need to seed, but it's fairly extensive
<gendale_>
so I have my own refs-to-seed doc
<gendale_>
the issue is that some of the links have gone down, and I don't know which hashes they are, because the script is recursively dialing down
<gendale_>
I think we can repin the hashes so they run again, we just don't know which ones they are
ygrek_ has joined #ipfs
<gendale_>
but because the links have gone down, the pin script is hanging
<gendale_>
i'm going to file an issue for this, since i'm not sure where the best way to approach fixing it is
okket has quit [Remote host closed the connection]
okket has joined #ipfs
Encrypt has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
jamescarlyle has joined #ipfs
<atgnag>
gendale_: Would you like an anime?
Encrypt has quit [Client Quit]
<gendale_>
atgnag: pardon?
<_p4bl0>
hello :), so I'm playing with IPFS, but from what I understand, IPNS do not fully work yet as I can't directly make a request (via the http gateway for instance) to my identity name after having `ipfs name publish` something
<atgnag>
Or, does anyone in here like anime?
<_p4bl0>
however I already have a few question on how it will works when it fully does
<atgnag>
I'm adding Akame Ga Kill right now, so I was wondering if anyone would like to download, or pin it.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<_p4bl0>
first question: if I `ipfs name publish foo` where foo is the hash of a directory that I `ipfs add`ed recursively, will it work as expected? (i.e., will the entire website be available using IPNS at my identity name)
<_p4bl0>
I guess the answer is yes since if the files in the directory are recursively added <hash-for-directory>/real-file-name already works
<atgnag>
_p4bl0: Oh man, you know how shitty my internet is…
<atgnag>
Oh wait, nevermind.
<gendale_>
_p4bl0: I think as long as relative links all work correctly things should work fine
<_p4bl0>
okay, cool
<_p4bl0>
now what if I update a file in a subdirectory? I `ipfs add` it, and then how is that taken into account?
<gendale_>
well, when you change a file, there will be a new hash
<gendale_>
recursively as well
<gendale_>
so when you make updates, you would have to update the top level hash in ipns
<_p4bl0>
ah, it is a kind of merkle tree?
<_p4bl0>
ok, so when I update any file, I recursively `ipfs add` the root directory of the website and then `ipfs name publish` the new hash of the root directory
<_p4bl0>
that seems nice :)
<gendale_>
yeah it's kind of similar to git
<gendale_>
yeah. i think so. I haven't played around with ipns much
<atgnag>
If I torify the ipfs daemon, is it anonymous, or might it reveal some things?
<gendale_>
but that's my understanding up it
<gendale_>
atgnag: i don't know, but I would treat it like bittorrent or bitswap
<gendale_>
don't assume that ipfs is anonymous, or that it plays properly with tor
<atgnag>
Yeah, I figured.
<gendale_>
tor is not built for a protocol like IPFS anyways, and you would be harming the network
jamescarlyle has quit [Remote host closed the connection]
<atgnag>
Harming the network? As long as it isn't causing delays or outages, isn't it adding more hay to the haystack?
<_p4bl0>
thx for your answers :), so now here is my second question: is it (will it be) possible to have multiple identities with the same daemon?
<_p4bl0>
phrased differently: if I want to host multiple IPNS accessible website on the same server, how could I do that?
<gendale_>
I believe it will be
<_p4bl0>
gendale_: nice
<gendale_>
_p4bl0: ipns is really rudimentary right now but that's an obvious and desired outcome
<atgnag>
"The problem is that several popular Bittorrent clients (the authors call out uTorrent in particular, and I think Vuze does it too) just ignore their socks proxy setting in this case."
<cryptix>
Yup - you want to look for the keystore stuff
<gendale_>
atgnag: The first answer is "don't run Bittorrent over Tor". We've been saying for years not to run Bittorrent over Tor, because the Tor network can't handle the load; perhaps these attacks will convince more people to listen.
VegBerg_school has joined #ipfs
<gendale_>
tor doesn't necessarily need more hay (though that's not a bad thing) it needs more exit nodes
amstocker has joined #ipfs
<atgnag>
gendale_: Most (if not all, I haven't finished reading) of those attack scenarious don't apply to ipfs though.
<gendale_>
I would look into something like I2P or cjdns if you're really intent on this
<gendale_>
i2p probably best bet
bedeho has quit [Ping timeout: 255 seconds]
<atgnag>
gendale_: I2P can't connect to normal internet without a proxy though, and cjdns isn't anonymous.
<atgnag>
Although, I guess I'm ultimately trying to use IPFS for something it wasn't made for.
<_p4bl0>
:)
<gendale_>
i mean, it's not a bad idea. and I support anonymous transactions and i think tor integration would be great
<gendale_>
i would just caution you on whether it's prudent to burden two important, developing projects with... well, piracy, given the anime thing
<gendale_>
it places both a social cost and a bandwidth cost
notduncansmith has joined #ipfs
<atgnag>
Yeah, there's that.
notduncansmith has quit [Read error: Connection reset by peer]
<gendale_>
i'm not part of the project just a pro privacy anti censorship pro piracy pro creative commons guy
<atgnag>
I'd probably just use a VPN instead.
<gendale_>
realistically, your threat model is likely the ip scraping bots that hollywood tech startups use to track torrents
<atgnag>
gendale_: Yeah. And we probably won't have those for a while.
<gendale_>
right. and a good vpn will beat those pretty quickly
<atgnag>
Yeah.
<atgnag>
(muh money though)
<atgnag>
But I suppose outproxy owners also have to pay the bills.
ygrek_ has quit [Ping timeout: 268 seconds]
<gendale_>
you could always take a break from pirating and donate to tor. ;)
<atgnag>
True.
<gendale_>
or boycott copyrighted things and only consume creative commons. ;)
amstocker has quit [Ping timeout: 246 seconds]
<atgnag>
Well, i do that already, in addtion to piracy.
<gendale_>
<3
<_p4bl0>
:D
jamescarlyle has joined #ipfs
<atgnag>
:3
<_p4bl0>
I think IPFS should not focus on anonimity, otherwise people will react the same way they do to bittorrent "oh that's practical, i will use it to get content but I won't share anything because I don't want to get caught", or as with Tor "I use it but I'm not an exit node", and that not viable for what seems to be the goal of IPFS
<gendale_>
'
<gendale_>
_p4bl0: I don't know what the actual project stance is, if there is one
<atgnag>
_p4bl0: You really can't run an exit node on a residential connection though.
<atgnag>
In fact it's discouraged.
<atgnag>
_p4bl0: Also, why do you think anonymity would lead to more "hit and runs"? I'd expect the opposite to be true.
<_p4bl0>
true, but seeding on bittorrent is not, and even has to be forced by some tracker
<gendale_>
the bitswap protocol addresses some of the freeleech concerns
<atgnag>
Nevertheless, we must remember that enabling piracy is definitely not the goal of the project.
<_p4bl0>
atgnag: because people would fear it, because journalists will say it is the deepnet or whatshit
<atgnag>
_p4bl0: Maybe I'm cynical, but I kinda think that's an inevitably at this point.
<atgnag>
Anonymity network? Deepnet.
<_p4bl0>
gendale_: as for the stance of the project, I don't know either, but from what I've read, it is more the "permanent uncensored web" than "anonymity online"
<atgnag>
gendale_: Well… not really sure that affected its popularity at all.
<gendale_>
it literally is only scene movie releases
<gendale_>
no, but it has affected the development cycle
<atgnag>
Yeah.
<atgnag>
So, bad move, I guess.
<gendale_>
hah, it's a great move, but it is what it is
<_p4bl0>
gendale_: I'm not sure about that, it is technically possible to ask for every ipfs instance to "ipfs pin rm" and "ipfs rm" something, which would delete it from the network
<gendale_>
popcorn time exists, it probably will never unexist
<_p4bl0>
also, the right to be forgotten concerns for instance Google not displaying some results, not necessarily the removal of the incriminated page from the internet
<gendale_>
_p4bl0: yes, and that is largely a problem, but there is also the practical issue of trying to do that across jurisdiction
<gendale_>
_p4bl0: that's a good point
<gendale_>
_p4bl0: but, supposing hypothetically you built a decentralized google for IPFS websites, it would be hard to incorporate that into the results
<atgnag>
_p4bl0: Will I have a choice whether content can be removed or not?
<gendale_>
atgnag: yeah you can pin and unpin and garbage collect at will
<gendale_>
it's not like freent
<gendale_>
freenet
<atgnag>
gendale_: Oh, I see… nevermind.
<atgnag>
I thought you meant some could delete their stuff from the network.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<atgnag>
On that note, maybe we should remind people not to share /too/ much as adoption increases?
<_p4bl0>
gendale_: mh, I guess it would be possible for an ipfs search engine to censor its results yes
<atgnag>
Not people in general, but newcomers maybe?
<mafintosh>
jbenet: ipfs is getting a lot of attention recently - so cool!
<mafintosh>
jbenet: is there any way to know how many people are using it? (just curious)
<_p4bl0>
mafintosh: I guess the neocities announcement is what made people interrested
<gendale_>
_p4bl0: my point is that a ipfs search engine could be built in a way so that it might not be censored
<davidar>
mafintosh: how is hyperos/dat related to ipfs btw?
<gendale_>
*could not be
<_p4bl0>
gendale_: that would be interesting, I have no idea how that could be achieved
<mafintosh>
davidar: hyperos is gonna be able to use any replication engine using a very simple interface
<mafintosh>
davidar: (it just needs content-addressed files)
<_p4bl0>
mafintosh: i know that's why I decided to give it a try, if not for that announcement it would have stayed in the "cool project that I have to checkout every 6 months to know whether it is possible to play with" list in my head ^_^
<mafintosh>
davidar: if you haven't already join #dat as well
<atgnag>
multivac: I've already read it. It's a neat story. In fact, I'll pin that pdf.
<multivac>
Error at 0x08: Reference not found
<davidar>
mafintosh: yeah, I'm there, asked about dat+ipfs but nobody answered
<davidar>
atgnag: multivac is a bot btw :)
<atgnag>
multivac: Worked for me.
<multivac>
I don't know
<atgnag>
…
<atgnag>
I've never seen a markov chain so convincing.
<mafintosh>
davidar: ah see that now - i was sleeping
<atgnag>
Or maybe it's just my lack of sleep.
<atgnag>
Or maybe we're approaching a technological singularity. Who knows.
<davidar>
atgnag: feel free to teach him stuff to make him more interesting :)
<davidar>
atgnag: wiki.xkcd.com/irc/Bucket
<gendale_>
hahaha. i keep getting approached by bernie sanders campaign interns on twitter
<gendale_>
i'm pretty sure they're human, but they keep failing my turing tests. :P
<atgnag>
In case anyone wants it, my Akame Ga Kill folder is here. QmTR7uDE9mkU7jBScExzfQyo2dbbKK1i4mpPyebjx1meiu
<atgnag>
gendale_: lol
<davidar>
multivac: DAT
<multivac>
I'm sorry, there's currently nothing associated with that keyphrase.
<atgnag>
Right now ipfs is just used for sharing files and static webpages, but aren't there other things that IPFS can/will be extended to be capable of?
<gendale_>
atgnag: hahaha better question is "are there things that exist right now that can't be made in IPFS"?
<atgnag>
Okay.
<atgnag>
Can you elaborate?
<gendale_>
um
<gendale_>
well, i don't think there's things on the internet that frankly wouldn't be better suited to ipfs
ianopolous2 has quit [Ping timeout: 244 seconds]
kerozene has quit [Ping timeout: 265 seconds]
<atgnag>
Nice.
<gendale_>
it's working in concert with things like bitcoin, filecoin, ethereum, openbazaar
<gendale_>
bitmessage
akhavr has quit [Read error: Connection reset by peer]
keroberos has joined #ipfs
<atgnag>
Could we have a social network?
<gendale_>
sure
<gendale_>
openbazaar is kind of implementing that
<atgnag>
What about a video site? Like, with a comment system, rss feeds and all?
<gendale_>
alexandria (which is being built on IPFS) is currently working on a comments system that uses blockchain to store IPFS magnet links
akhavr has joined #ipfs
<gendale_>
video site
<davidar>
multivac: aggregation
<multivac>
Can't talk, zombies!
<ipfsbot>
[go-ipfs] jbenet pushed 2 new commits to master: http://git.io/vZu3h
<ipfsbot>
go-ipfs/master 65507cb Juan Benet: Merge pull request #1679 from pnelson/fuse-readonly-tests...
<ipfsbot>
go-ipfs/master 9ce3875 Philip Nelson: Fix test case for fuse/readonly....
voxelot has joined #ipfs
notduncansmith has joined #ipfs
ianopolous has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<atgnag>
Wow. Some people seem to be asking for the world. I hope the devs are ambitious.
<gendale_>
atgnag: juan wants to replace http. ;)
<atgnag>
Yeah. There's that.
<gendale_>
was zeronet not good?
<gendale_>
seems like a neat concept but i bet it'd be cumbersome
<atgnag>
gendale_: It was so crude.
<gendale_>
like, alpha software crude or...?
<atgnag>
It had websockets that couldn't connect to localhost somehow.
<atgnag>
gendale_: No, I mean "patchwork" crude.
<atgnag>
I mean, I never looked at the source, but that's what it felt like.
<atgnag>
It seems like the kind of project that doesn't have unit tests, etc.
<gendale_>
haha so, alpha software crude?
<atgnag>
I guess.
<compleatang>
anyone have a quick fix for `Error: merkledag: not found` on a fresh install (cleaned out ~/.ipfs prior to ipfs init) (on current master)?
<compleatang>
if not happy to drop an issue.
<compleatang>
nvmd
atgnag has quit [Ping timeout: 246 seconds]
atgnag has joined #ipfs
<davidar>
atgnag: aw, but the landing page was so pretty :)
<atgnag>
davidar: That's true.
<davidar>
if zeronet dies, can we steal it? :)
<atgnag>
Is there a way to view all the files I'm sharing?
<atgnag>
(not pinned files though)
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<davidar>
atgnag: looking through your local blockstore, maybe? :)
<drathir>
mornin...
<davidar>
morning
keroberos has quit [Quit: brb]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ei-slackbot-ipfs has quit [Remote host closed the connection]
notduncansmith has quit [Read error: Connection reset by peer]
Quiark_ has quit [Ping timeout: 255 seconds]
hellertime has joined #ipfs
rschulman_ has quit [Quit: rschulman_]
eater has joined #ipfs
rschulman_ has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
rschulman_ has quit [Quit: rschulman_]
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]
rschulman_ has joined #ipfs
atomotic has joined #ipfs
ianopolous has quit [Ping timeout: 240 seconds]
keroberos has quit [Max SendQ exceeded]
keroberos has joined #ipfs
jamescarlyle has joined #ipfs
Arpho has joined #ipfs
d6e has quit [Ping timeout: 240 seconds]
d6e has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
chiui has joined #ipfs
<xelra>
Sorry for asking again, but I'm still wondering. Is it possible to pin only parts of a file? Like let's say there are 50 chunks/blocks of data and I supply a list to pin only 3/50, 9/50, 16/50, 34/50 and 48/50?
rschulman_ has quit [Quit: rschulman_]
rschulman_ has joined #ipfs
inconshreveable has quit [Remote host closed the connection]
Arpho has quit [Ping timeout: 250 seconds]
keroberos has quit [Max SendQ exceeded]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
keroberos has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
<Bat`O>
xelra: you can pin block instead of the root file hash
<xelra>
Nice, that's great. So I can run a pinbot to distribute each file on 5 servers and then stripe those files. Nice.
rschulman_ has quit [Quit: rschulman_]
border has joined #ipfs
<_p4bl0>
hey, when I do `ipfs pin rm -r <hash>`, the recursivity applies to the DAG under <hash>, is that correct?
<_p4bl0>
i.e., if there is a new version of the file foo in <new_hash>/foo with hash <new_hash_foo> that differs from <hash>/foo with hash <old_hash_foo>, but there is also a file bar in <new_hash>/bar with hash <hash_bar> that is identical to <hash>/bar with <hash_bar> too, will my `ipfs pin rm -r <hash>` affect only <hash>/bar or will it also unpin <new_hash>/bar ?
<_p4bl0>
sorry my question is a bit complicated I guess :-S
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ianopolous has joined #ipfs
<Bat`O>
xelra: pinning the whole file on the five server would give you load balancing as well
<rendar>
there is a source where i can learn how the ipfs protocol works?
<xelra>
Bat`O: Yeah, it's just that then every server would need a complete copy. So when that's 50 TB, it will be a problem. But splitting 50 TB on 5 severs, each holding 10 TB and then striping from them for 5 times speed. That would be great.
jamescarlyle has joined #ipfs
dignifiedquire_ has joined #ipfs
dignifiedquire has quit [Ping timeout: 246 seconds]
dignifiedquire_ is now known as dignifiedquire
<daviddias>
jbenet: we fixed the Internetz!
<xelra>
Anyway, that's great. Could run a really sophisticated pinbot for load balancing and striping.
<Bat`O>
xelra: not that for this size of file, you will have intermediate node that aggregate block, so you can pin those instead of the final block
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<herch>
I am executing, go get -u github.com/ipfs/go-ipfs/cmd/ipfs
<drathir>
herch: version of go?
<herch>
go 1.5
Arpho has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
<drathir>
thats looks like compiler issue...
<drathir>
but im not any pro...
rschulman_ has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
mildred has quit [Ping timeout: 264 seconds]
Encrypt has joined #ipfs
ianopolous has quit [Ping timeout: 256 seconds]
jamescarlyle has joined #ipfs
mildred has joined #ipfs
<larsks>
What are the plans for being able to publish ipns names other than the node id?
<herch>
started the ipfs deamon with binary installation. Will figure out issues with compiler later.
mildred has quit [Ping timeout: 264 seconds]
<davidar>
larsks: I think that's coming, not sure on the specifics though
<jamescarlyle>
Bat`O: For a very large file: I see that objects/blocks can contain data and links - is there a limit to the number of data blocks listed as links in a parent object? i.e. under what circumstances (if any) do intermediate objects need to be created
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
mildred has joined #ipfs
<davidar>
jamescarlyle: blocks are limited to 256kb, so if the size of all the links is bigger than that
pfraze has joined #ipfs
Encrypt has quit [Quit: Apparently going to "clean" a Mac (<.<)]
jamescarlyle has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
voxelot has joined #ipfs
jamescarlyle has joined #ipfs
<lgierth>
wow the android modding community, and their file hosters
<lgierth>
i'm tryin to hunt down a RUU firmware for my htc desire z, and have found links to a dozen or so file hosters
<lgierth>
do you think any of those is still alive? of course not a single one
<davidar>
lgierth: yeah, android modding is really quite shady tbh
<davidar>
compared to desktop linux
<lgierth>
well at least i know the md5 hashes of the files i need
<svetter>
people still using md5 in 2015 :(
<lgierth>
so for the rooting tools, i was able to use google as a content addresses search engine :P
Quiark has joined #ipfs
<lgierth>
just haven't tracked down the RUU firmware yet...
<lgierth>
it's all quite frustrating
<herch>
how does ACL works on files shared on ipfs?
jamescarlyle has quit [Remote host closed the connection]
<herch>
if I dump some files on ipfs, does it mean it's available to anyone with the hash?
<Bat`O>
jamescarlyle: i'm not sure about what criterion is used to build these aggregate node
<Quiark>
herch, I think yes
<herch>
so how would one achieve something like dropbox on ipfs?
<herch>
I read the discussion on HN where people were talking about creating something like a dropbox using ipfs
hjoest has joined #ipfs
<herch>
but without ACL I don't know how will that work.
hjoest has quit [Remote host closed the connection]
<davidar>
herch: until ipfs gets permissions, i suspect syncthing would probably be a better bet
<Quiark>
herch, I think it's a feature for the next phase
<herch>
Quiak, ok
herch has quit [Quit: Page closed]
hjoest has joined #ipfs
voxelot has quit [Ping timeout: 250 seconds]
jamescarlyle has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman_ has quit [Quit: rschulman_]
null_radix has quit [Excess Flood]
null_radix has joined #ipfs
rschulman_ has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman_ has quit [Client Quit]
<cryptix>
hellow
cryptix has left #ipfs [#ipfs]
cryptix has joined #ipfs
VegBerg has quit [Read error: Connection reset by peer]
vijayee_ has joined #ipfs
iHedgehog has joined #ipfs
rschulman_ has joined #ipfs
Leer10 has quit [Quit: Leaving]
rschulman_ has quit [Client Quit]
jamescarlyle has joined #ipfs
jamescarlyle has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ianopolous has joined #ipfs
hjoest2 has joined #ipfs
manon has quit [Quit: leaving]
border has quit [Remote host closed the connection]
hjoest2 has quit [Read error: Connection reset by peer]
border has joined #ipfs
border has quit [Remote host closed the connection]
hjoest has quit [Ping timeout: 264 seconds]
<drathir>
hi...
rschulman_ has joined #ipfs
hjoest has joined #ipfs
border has joined #ipfs
border has quit [Remote host closed the connection]
<rossjones>
Does anyone know where I can find the hash (and instructions) for the video player? Vaguely recall seeing a demo of playing a video from ipfs using a player, on ipfs.
pfraze has quit [Remote host closed the connection]
<_p4bl0>
hello all, I have yet another question: don't you think the IPFS gateway should rewrite absolute links that starts with '/'?
<_p4bl0>
For example I think /styles.css on the page http://gateway:8080/ipfs/<hash>/foo/bar.html should automatically be rewritten to /ipfs/<hash>/styles.css
<_p4bl0>
ooooooooooh but hey, looking at the source code for the gateway I just saw that IPNS name can be resolved using /ipns/<hash>
<_p4bl0>
neat
Arpho has quit [Ping timeout: 264 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
_p4bl0: the gateway will not make any changes to the content it serves, if it did, it would have a different hash and you wouldnt be receiving the content you requested
hjoest has quit [Read error: Connection reset by peer]
<whyrusleeping>
cryptix: o/
<whyrusleeping>
long time no see!
<_p4bl0>
whyrusleeping: ah, and that is true even when using the IPNS indirection?
<whyrusleeping>
with ipns, it resolves the ipns record to an ipfs hash, and serves that
<_p4bl0>
whyrusleeping: okay, but now I'm trying to change what the IPNS record resolves to (and locally that works) but the gateway at ipfs.io still resolves to the old version, is that normal?
hjoest has joined #ipfs
<blame>
_p4bl0: I solve this problem by only ever using relative linking.
<_p4bl0>
Blame: yea I guess that's what I should do but it is very unpractical in my case
hjoest has quit [Read error: Connection reset by peer]
<blame>
_p4bl0: you could also pretty reliably write links as "/ipfs/hash/subdir"
captain_morgan has joined #ipfs
<_p4bl0>
Blame: in my case I would like the same files to be served via traditional HTTP as well, so that is not an option
<_p4bl0>
also, I would have to determine the hash before adding the file to IPFS, which would prove quite difficult :p
<blame>
You can run it the other way too. I serve ipfs via traditional http
rschulman_ has quit [Quit: rschulman_]
<blame>
that is how ipfs.io is hosted right now. it nginx proxies to ipfs.io/ipns/ipfs.io
<whyrusleeping>
_p4bl0: ipns resolving is verrry unpolished right now, I wouldnt rely on it until after we fix it
<_p4bl0>
Blame: hahahaha
simonv3 has joined #ipfs
<whyrusleeping>
its our main priority, but there is a lot of groundwork that needs to be laid before we can get to fixing it the right way
<blame>
_p4bl0: but yeah, circular links in ipfs just don't. You could solve it dynamically by having an origin argument in your query
<_p4bl0>
whyrusleeping: okay, but it actually works pretty well locally (i run version 0.3.8-dev)
<_p4bl0>
Blame: i think the cleanest solution is to use relative links
<whyrusleeping>
_p4bl0: yeah, its the networking side of it that needs some love.
Encrypt has joined #ipfs
<_p4bl0>
i wish i could duplicate myself, that way i would be able to have time work on that
Encrypt has quit [Client Quit]
<whyrusleeping>
_p4bl0: yeah... me too
<whyrusleeping>
mostly, we're waiting on certain specs to be fleshed out before i go and implement it all
<whyrusleeping>
its moderately 'easy' to fix the issue, but fixing it the *right* way is more complicated
<_p4bl0>
classic
<_p4bl0>
in any case, thanks a lot for developing IPNS :)
<_p4bl0>
i'm really enjoying playing with it
<whyrusleeping>
i'm glad you like it :)
<whyrusleeping>
i just realized i've been working on ipfs full time for about a year now
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<_p4bl0>
whyrusleeping: full time? how are you funded?
<whyrusleeping>
protocol labs is the company pushing forward the development of ipfs
<lgierth>
does have futuristic connectivity a ring to it?
<whyrusleeping>
it kinda does, yeah
<lgierth>
nearly a parse error
<lgierth>
cool
<lgierth>
i'll tell you about it later :)
rschulman__ has joined #ipfs
rschulman_ has quit [Ping timeout: 265 seconds]
rschulman__ is now known as rschulman_
<__uguu__>
another question but with a little bit of context first, i2p is packet oriented, the tcp stack used over i2p is homebrew and is rather suck due to latency and a poor choice of MTU, but... can send and receive messages up to 8KB before getting very droppy, now the question: does ipfs abstract the inter-node communication such that it doesn't care if it's stream based or not?
* __uguu__
is still rtfm-ing
<lgierth>
it needs to be stream based at the moment
hjoest has joined #ipfs
<whyrusleeping>
it does need to be stream based, but that is a limitation right now only due to our encryption scheme (XOR streams need to be ordered and contiguous)
<__uguu__>
i see
<whyrusleeping>
if we move towards a udp based protocol (which we *REALLY* want to) we will have to take another look at a better scheme for encryption that works for packets
<whyrusleeping>
we would love to avoid head of line blocking due to waiting for the next packet to decrypt
<__uguu__>
would something like nacl's crypto_box work?
<whyrusleeping>
crypto_box is just basic pki, isnt it?
<__uguu__>
does ipfs do something funky that isn't that?
<__uguu__>
crypto_box is just basic pki afaik
<whyrusleeping>
yeah, we use pki for the initial handshake
<whyrusleeping>
and use it to exchange a shared key for an xor stream
<__uguu__>
i see
<__uguu__>
what is the lifespan of that xor stream key?
<__uguu__>
lifetime*
devbug has joined #ipfs
<whyrusleeping>
the same as the lifetime of the connection between the two involved nodes
<Quiark>
crypto_box is just a simple easy to use correct package for encryption
<__uguu__>
ipfs reminds me a lot about how i2p is structured, i see you are using a bloom filter somewhere in there, what is the bloom filter trying to filter out in your case?
<__uguu__>
i like how there's a clear separation of components in ipfs
<whyrusleeping>
__uguu__: the bloom filter is currently not used. its been used in the past for caching sets and block tables
<whyrusleeping>
but its a decent impl
<__uguu__>
yes i can see, it's not using a crypto hash from what i can tell, which would slow it down a lot
<whyrusleeping>
__uguu__: on separation of components: we try really hard to keep that. We want people to use and give feedback on the components
<whyrusleeping>
we're working on extracting our peer to peer communication stuff as a library
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<__uguu__>
is there any roadmap/issue on the topic of utilizing a udp based inter node protocol?
Quiark has quit [Ping timeout: 252 seconds]
<__uguu__>
i see the libp2p repo, is that it?
<whyrusleeping>
__uguu__: there might be, cant remember off the top of my head
<drathir>
__uguu__: yea that correct security oriented, tor also in theory not so good in anonimity, bc of possibility to own and listen at exit nodes and what for sure nsa doin now too...
<pfraze>
we're thinking what we want/need is a way to distribute electron apps that can be relatively simple and secure
hjoest has quit [Ping timeout: 264 seconds]
hjoest has joined #ipfs
captain_morgan has quit [Ping timeout: 265 seconds]
devbug has quit [Ping timeout: 272 seconds]
jhulten has quit []
<FunkyELF>
can ipfs do live streams?... how does that work with p2p? each cunk points to the next like a linked list?
<pfraze>
Blame: what do you think IPFS would need for protocol plugins to work right?
hyperbot has quit [Remote host closed the connection]
hyperbot has joined #ipfs
<whyrusleeping>
FunkyELF: ipfs will be able to do live streams in the future, once we implement pubsub
<whyrusleeping>
pfraze: what do you mean by protocol plugins?
<pfraze>
whyrusleeping: if you put protocol-control in userland, your browser could adopt a new non-http protocol quickly
<blame>
Pfraze: ipfs IS a protocol. You would write a plugin to make a browser nativly handle is (fuse alone actually does a decent job of this)
<FunkyELF>
whyrusleeping: would those "live" streams then be persisted forever?
<blame>
*it
<pfraze>
last I checked, Chrome doesnt let you register new schemes. Fuse is also non-windows
<whyrusleeping>
FunkyELF: for as long as anything else in ipfs, yeah
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
hjoest2 has joined #ipfs
<pfraze>
you also have to get the browser respecting new security rules. What constitutes a domain, for instance, and what resources are shared between locations in a new protocol
<FunkyELF>
whyrusleeping: so like last night I was watching a stream of the thursday night football. Some worked for a while, some were even hosted on YouTube until they got taken down by copyright notices.
devbug has joined #ipfs
<FunkyELF>
whyrusleeping: ipfs, being a protocol, wouldn't really help content owners right... probably hurt them
simonv3 has joined #ipfs
devbug has quit [Read error: Connection reset by peer]
captain_morgan has joined #ipfs
hjoest has quit [Ping timeout: 264 seconds]
hyperbot has quit [Remote host closed the connection]
hyperbot has joined #ipfs
<__uguu__>
copyright hurts content owners
<whyrusleeping>
FunkyELF: yeah, limiting access to the live stream would be difficult
<whyrusleeping>
DRM and distributed systems dont like to cooperate as well as they could
<whyrusleeping>
you could share keys with your approved viewers, and hope they dont lose them
<FunkyELF>
whyrusleeping: yeah... ipfs (well probably just p2p in general) seems ideal for DirectTV's Sunday Ticket where you have massive amounts of people watching the same stream
<FunkyELF>
... but its a paid service
<whyrusleeping>
yeah... this is a problem we want to solve
<FunkyELF>
whyrusleeping: is it solvable in a decentralized way?
<FunkyELF>
I guess key distribution would have to be centralized or something right? but once a key has been leaked all security is lost
<whyrusleeping>
yeap
<whyrusleeping>
i dont know the answer to this problem yet
<FunkyELF>
dunno if its solveable, once the viewing keys are published and that data is persisted you're hosed
<FunkyELF>
you'd have to trust clients... which you cannot do
devbug has joined #ipfs
hjoest2 has quit [Ping timeout: 264 seconds]
<whyrusleeping>
you would basically have to ship your own 'viewer'
<whyrusleeping>
and rely on security through obscurity
marianoguerra has quit [Ping timeout: 252 seconds]
<FunkyELF>
cuz thats worked well in the past
<blame>
pfraze: What I am trying to say, while such a new browser could be rolled into IPFS, in generally it should be the other way around, we should make a new browser than can do MORE than just IPFS
<whyrusleeping>
yeah, lol
devbug has quit [Read error: Connection reset by peer]
<pfraze>
Blame: yeah I agree
<blame>
it could be as simple as forking an older version of firefox
<FunkyELF>
not trying to be debbie downer... I just don't see how it could be done. Can't even see how something like Facebook would be implemented where the application dictates who can see what
<FunkyELF>
and that can change over time... can't happen on "the permanent web"
<blame>
or building on webkit (which you are doing already really with the electron app)
devbug has joined #ipfs
<voxelot>
using a bitcoin protocol to check if funds are being delivered almost bit by it and sending a unique key with the output while funds come in
<pfraze>
yeah
<voxelot>
rediculous overhead though
<blame>
you could come up with a multi-key asymmetric key scheme.
<blame>
so the overhead is not as bad
<FunkyELF>
voxelot: is it a different key for each viewer?
<voxelot>
yeah would have to maybe check against IP or something
notduncansmith has joined #ipfs
hjoest has joined #ipfs
<blame>
ie: the stream is encrypted with a key that is the metric of a polynomial curve, given n any non-idential points on that curve you can solve for the curve
notduncansmith has quit [Read error: Connection reset by peer]
<voxelot>
or use the address of the cryptofunds in the keys going
<blame>
problem is then users can generate more keys once they have solved for the curve.
hyperbot has quit [Remote host closed the connection]
fc00bot has joined #ipfs
<blame>
but it is a lot less overhead to track which keys have been issues/used than it is to encrypt for 1000s of different keys
<voxelot>
interesting
<blame>
You can also do the same trick with the definition of a k-dimensional hyper plane which needs k+1 points to solve for it.
<blame>
generally it is used for k of n decryption such that k keys need to be combined to decrypt a message
<blame>
ie you define a 9^th degree polynomical, and then give 100 people different points on it. Then it takes any 10 people working together to decrypt the message.
<voxelot>
encrypting the stream going from node to node and attaching a decryption to the nodes with funds streaming in using your ussed key idea
<voxelot>
once used by a node the key is no good
<voxelot>
same encrytion, lots of ways to decrypt right?
hjoest has quit [Ping timeout: 264 seconds]
<blame>
yeah, but remember, essentially you are using your specific key to find the "master" key. Then you can just give away the master key.
<blame>
but also remember this is not about being perfect, only hard enough that nobody you cannot sue makes a tool to break it at scale
<voxelot>
ohh once solved its solved, got it. yeah that complicates things
<blame>
because it is content providers we are talking about here, honestly the 1 secret key per stream idea is fine
hjoest has joined #ipfs
<blame>
we just want to be able to DETECT when unauthorized people view the stream
<blame>
which is actually really easy so long as we control a meaningful fraction of the internal nodes involved in streaming
<blame>
IPFS is not a privacy tool. It can be shoehorned into one the same way you can run any protocol on TOR but it is not the explicit goal.
<voxelot>
right, coming up with privacy hacks for ipfs is fun
<blame>
That is why I think IPFS has a real chance and stuff like morph.is ect
<blame>
nothing we build that truely allows privacy will be supported by world governments and corporations.
qgnox has quit [Ping timeout: 244 seconds]
<voxelot>
they are going to have to outlaw encryption to stop us imho
<blame>
So we don't. We just build a useful tool that circumstantially slightly improves the privacy status quo that those governments and corporations value independently of the privacy problem.
<voxelot>
and companies have a vested interest to not ban encryption
<spikebike>
voxelot: heh, ya, it's touchy
<blame>
no they won't stop us, they just won't use it or support other people using it
<spikebike>
I was amused how usa wanted apple/microsoft to provide backdoors for content outside the USA
<blame>
and bluntly put, to solve the problem we want to solve we NEED them.
<spikebike>
but them apple/microsoft mentioned if they do that, then they have to allow russia and china to have backdoor access to files in the usa
qgnox has joined #ipfs
<blame>
So we walk the line, and do as much as we can to improve the status quo within the system.
<voxelot>
all we can do is present solutions, hope that users use them without even knowing it
<voxelot>
also the power of p2p and no central points of failure or places to put backdoors
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
fleeky has quit [Ping timeout: 240 seconds]
<voxelot>
but right now we need ethical companies, which for the most part i believe we have, just complying with what they have to
<voxelot>
nice little line in CALEA states "a company shall not be forced to provide decryption if they do not have the keys to decrypt the information"
fleeky has joined #ipfs
fc00bot has quit [Remote host closed the connection]
fc00bot has joined #ipfs
fc00bot has quit [Remote host closed the connection]
fc00bot has joined #ipfs
fc00bot has quit [Remote host closed the connection]
fc00bot has joined #ipfs
fc00bot has quit [Remote host closed the connection]
keroberos has quit [Max SendQ exceeded]
fc00bot has joined #ipfs
fc00bot has quit [Remote host closed the connection]
keroberos has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ianopolous has joined #ipfs
rendar has quit []
deltab has quit [Ping timeout: 246 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
domanic has joined #ipfs
bigbluehat has quit [Ping timeout: 252 seconds]
cleichner has quit [Ping timeout: 252 seconds]
ogd has quit [Ping timeout: 252 seconds]
jhiesey has quit [Ping timeout: 252 seconds]
ehd has quit [Ping timeout: 252 seconds]
yoshuawuyts has quit [Ping timeout: 252 seconds]
mappum has quit [Ping timeout: 252 seconds]
Sasha has quit [Ping timeout: 252 seconds]
Luzifer has quit [Ping timeout: 252 seconds]
henriquev has quit [Ping timeout: 252 seconds]
feross has quit [Ping timeout: 252 seconds]
risk has quit [Ping timeout: 252 seconds]
daviddias has quit [Ping timeout: 252 seconds]
jhiesey has joined #ipfs
mikolalysenko has quit [Ping timeout: 252 seconds]
kumavis has quit [Ping timeout: 252 seconds]
<whyrusleeping>
voxelot: lol, "turn this water into wine!"
<whyrusleeping>
the court demands it!
Luzifer has joined #ipfs
Sasha has joined #ipfs
Sasha is now known as Guest12310
risk has joined #ipfs
mappum has joined #ipfs
feross has joined #ipfs
ogd has joined #ipfs
yoshuawuyts has joined #ipfs
ehd has joined #ipfs
henriquev has joined #ipfs
daviddias has joined #ipfs
kumavis has joined #ipfs
dignifiedquire has quit [Quit: dignifiedquire]
bigbluehat has joined #ipfs
cleichner has joined #ipfs
mikolalysenko has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
devbug has quit [Ping timeout: 250 seconds]
therealplato has joined #ipfs
<qgnox>
Hi, one question how do I know/see what is being transfered to my node?, because with the "stats bw" command shows in one day I got 500MB bandwith in, but I didnt cat/get/download any hash since last restart.
<whyrusleeping>
qgnox: we have a tool called bsdash that you can run, it will show whats being transfered
pinbot has quit [Remote host closed the connection]
<whyrusleeping>
its on npm
pinbot has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]