notduncansmith has quit [Read error: Connection reset by peer]
inconshreveable has quit [Ping timeout: 248 seconds]
domanic has quit [Ping timeout: 264 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
patcon has quit [Ping timeout: 250 seconds]
patcon has joined #ipfs
dread-alexandria has joined #ipfs
dread-alexandria has quit [Client Quit]
nanda has joined #ipfs
Leer10 has joined #ipfs
<Leer10>
I wonder if someone could make a reddit-like platform run on ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
anshukla has joined #ipfs
<prosodyContext>
Leer10: Sync www.getaether.net to IPFS? Their community has been exploring how to extend the desktop client to web browsers, while maintaining security.
<Leer10>
Hmm that'd be cool
<Leer10>
and then maybe somehow by that time IPFS could be housed within an extension
<Leer10>
I heard that people were thinking of making IFPS run in js so an extension wouldn't be too far off
dread-alexandria has joined #ipfs
<Leer10>
I'm gonna try out aether
<prosodyContext>
Leer10: it would be fantastic is someone developed a way to store node lists on IPFS perhaps. Peer auto-discovery is still very manual/unfinished.
<Leer10>
grrr 32 bit problems
<Leer10>
sucks having only older hardware
<Leer10>
looks like I can compile it though
bren2010 has quit []
<Leer10>
well once they find that out then I guess one should spread a variety of extensions each with a different list of seeds
<spikebike>
auto-discovery is still very manual/unfinished?
<spikebike>
ah, yeah, I've got issues with twisted+python
<prosodyContext>
Apparently Matrix.org uses a similar stack fwiw spikebike
<spikebike>
and GiL... and pycrypto
<spikebike>
thus switching to go... but diffent strokes for different folks
<prosodyContext>
Leer10: Yes, but I am still learning how I want to use PGP/Bitcoin Signed Messages and maybe run a full linode server. Small communities...
dread-alexandria has quit [Quit: dread-alexandria]
anshukla has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<Leer10>
it'd be cool if there was a website that didn't technically exist anywhere
<Leer10>
I guess one could do it with IPFS if the website asked its members to "pin" the parts of the website that makes it run
<spikebike>
Leer10: ya, seems like a couple projects are heading in that direction
<spikebike>
for now the closest is a tor secret service or whatever it's called
bren2010 has joined #ipfs
<Leer10>
okay now I hope that it doesn't download 20 GB of stuff I don't have the space for
<Leer10>
ooh pretty intro
<Leer10>
hmm the license agreement is forcing me to agree to make all of my content CC BY-SA and say it's by aether
<prosodyContext>
Leer10: Their CC licence is pretty well received, given the nature of their distribution and publishing system it already is very creative commons by default.
<Leer10>
*agree*
<Leer10>
can I use you as a bootstrapper prosodyContext?
<prosodyContext>
208.54.37.159:51978. (Sadly I may have firewall issues?) I need to setup Dyndns like the old Hotline Communications days. O:)
timgws has joined #ipfs
<Leer10>
okay it's syncing
<Leer10>
this actually looks pretty cool
<Leer10>
if only I could have gotten this as a binary
<Leer10>
It's very pretty
<prosodyContext>
Nehbit/aether is a designer as much is not more than developer. That's why benet's suggestion of an 'Cross-Org Open Working Group' is so desperately needed, I've dropped their twitter link to that request a couple times but nobody has committed to creating a shared safe space, I feel matrix.org would be perfect for a live chat between leaders that all can
<prosodyContext>
read free.
<Leer10>
hmm maybe I should have just used default bootstrap
knivsta has joined #ipfs
<knivsta>
What is this?
bren2010 has quit []
<Leer10>
IPFS is a type of network where content is addressable not by IP address but rather by a hash of it
www has joined #ipfs
<prosodyContext>
knivsta from eff, static.benet.ai/t/ipfs.pdf may be a good start, and to learn about how it addresses your privacy concerns you stated in #eff
<knivsta>
Leer10, omg Really? that you really make the Internet a nice cool and potential anonymous place to interact?
<Leer10>
I'm not the best person to answer questions but I'm sure someone like whyrusleeping (pingpong) would be capable
<knivsta>
how do we access it?
<Leer10>
There's this application that you can install from http://ipfs.io
<Leer10>
the name of the network that is formed by ipfs nodes is called the merkledag
<Leer10>
there's also an Internet->IPFS gateway that lets you see what's on there but there isn't that much content yet
nanda has quit [Remote host closed the connection]
<knivsta>
Ok
<knivsta>
:)
<Leer10>
the most interesting concept is that content no longer has to travel all the way across the world if someone else nearby already accessed it
<Leer10>
if someone from NY watches a video from IPFS that was originally located in London, a next door neighbor's IPFS node will instead stream it from the neighbor instead as it's already cached there
<Leer10>
that means the slashdot effect is effectively nil
can0e has joined #ipfs
<prosodyContext>
Someone needs to submit IPFS urls to Slashdot, that might get decent recog.
<knivsta>
prosodyContext, or HN?
<knivsta>
or Torrentfreak?
<knivsta>
Some popular site where people actually look for news and new content
<knivsta>
Torrentfreak.com is the perfect place for this network's publicity
<prosodyContext>
I read it as deliberate how ipfs and matrix and aether are reserved x-posting to keep a developer focus. Sad for end-user networkers for me who want to see their friends feel free to beta test, but probably a signal ve. noise issue.
* prosodyContext
thnks how to write about social media issues as an end-to-power-user.
<prosodyContext>
thlinks*
<Leer10>
wait what do you mean by that first sentence prosodyContext?
<Leer10>
knivsta: there isn't much on IPFS yet. It's a lot more similar to Freenet in terms of how things are served.
<Leer10>
I suppose one could set up a messenging system, but nobody has done so yet
<knivsta>
Leer10, ok :)
<knivsta>
I see
<knivsta>
Thanks for information guys
<knivsta>
:)
knivsta has left #ipfs ["Leaving"]
<Leer10>
hmm it's still syncing with default bootstrap
dread-alexandria has joined #ipfs
dread-alexandria has quit [Client Quit]
manu has quit [Ping timeout: 256 seconds]
manu has joined #ipfs
e-lima has joined #ipfs
<prosodyContext>
Leer10: that's why the fragmentalary nature of web r&d needs to change, so many disjointed tool race conditions. :
<cryptix>
gmornin
<Leer10>
yeah it's a shame that many of these programs and apps slowly stop development cause of a lack of momentum
<Leer10>
that's why it's a nice change of pace to see this channel be fairly active a lot of the time
<Leer10>
morning cryptix
<Leer10>
in fact tox is still in development even though drama surges every once in a while
mildred has joined #ipfs
reit has quit [Ping timeout: 276 seconds]
faleidel has joined #ipfs
<Leer10>
still syncing...
<prosodyContext>
Leer10: I am pretty sure Nehbit blocks all of Baltimore because they do not read me weller doing @cryptoempathy bitwism related to aether. So I do not get more than a few nodes a lot of the time. These tools need a better voice for the commmunity.. .
<prosodyContext>
Their blocking my twitter/cryptoepathy/ accounts coincided with me going from 5-8 nodes to 1 maybe 2 nodes fwiw
<Leer10>
*shrugs*
faleidel has quit [Ping timeout: 246 seconds]
kbala has quit [Quit: Connection closed for inactivity]
<prosodyContext>
EFF faces a lot of shrugs too for the shame ;) Literacy, biteracy, I get _why... They're ready to be a Open Working Group if the war cries died down.
Tv` has quit [Quit: Connection closed for inactivity]
flugsio has joined #ipfs
www has joined #ipfs
<fleeky>
is there a way to 'auto-pin' when someone is downloadng a file with ipfs , in such a way that anyone else downloading that hash would automatically connect to the swarm downloading it ?
<lgierth>
jbenet: we're still having cron restart ipfs on the gateways -- can you think of a metric that'd show a daemon needs a restart?
<cryptix>
hey lgierth :)
<cryptix>
iirc they froze at some point. ie not respond to commands but i'm not sure
domanic has joined #ipfs
<lgierth>
cryptix: spikes in gateway responses times, i think
<lgierth>
yeah
<lgierth>
we can graph that
atomotic has joined #ipfs
<lgierth>
or rather, spikes in ipfs-add -> gateway-get roundtrip times
<cryptix>
jbenet: can i continue nicerIndex under the assumption that you will merge 1419? :))
<jbenet>
fleeky: what do you mean by this "in such a way that anyone else downloading that hash would automatically connect to the swarm downloading it ?"
<jbenet>
fleeky: nodes already connect to each other
chrisr_ has joined #ipfs
chrisr_ has quit [Changing host]
chrisr_ has joined #ipfs
carver has quit [Ping timeout: 252 seconds]
<fleeky>
jbenet: so a group of nodes all downloading the same hash will create a torrent-like swarm ?
ed_ has quit [Ping timeout: 276 seconds]
<jbenet>
fleeky: You can think of it that way. but the "swarms" overlap on one.
<jbenet>
whryusleeping: we should still be able to _list_ indirect pins.
<jbenet>
whyrusleeping the "indirect pin" conecept is not gone, it's just implemented differently.
<whyrusleeping>
uhmmmm
<whyrusleeping>
hrm, okay
carver has quit [Ping timeout: 246 seconds]
<jbenet>
whyrusleeping: i think this would be easier to understand + CR as 2 PRs/changesets. if that saves you time, cool. if not dont worry about it.
<whyrusleeping>
youre saying reintroduce the indirect pins on the command in a followup PR?
<jbenet>
whyrusleeping: no, i'm saying this PR does many things: {changes the blockstore interface to add the RWLock, changes pins, adds a mark and sweep GC}
<jbenet>
whyrusleeping: i think the blockstore interface changes can be separated out and done in isolation first, to make sure the RLocks happen everywhere they need to.
anshukla has quit [Remote host closed the connection]
<jbenet>
whyrusleeping: i think it's not clear who between you and lgierth would be taking care of it, or how. maybe you two can discuss what needs to be done there?
<whyrusleeping>
alright, infra meeting on monday?
www has quit [Ping timeout: 246 seconds]
patcon has quit [Ping timeout: 276 seconds]
patcon has joined #ipfs
<jbenet>
i just tagged you in an issue, just write out there all the necessary considerations. wanted to do this 2 weeks ago, so would be nice if it landed before next sprint, or at least have the necessary changes specced out.
anshukla has joined #ipfs
pfraze has quit [Remote host closed the connection]
<lgierth>
grml now it's mercury acting up? or pluto itself which runs prometheus?
<vandemar>
does ipfs add -n <blah> do anything other than show the hashes? as in, it won't make anything accessible by somehow remembering the locations of files instead of copying the blocks to its own storage, right?
<cryptix>
alu: did you make that video?
<alu>
nope but I find it useful to explain whats going on when i talk about neural networks
<cryptix>
vandemar: yup :) we recently added -n to have something like 'md5sum/shasum that just produces the hash
<cryptix>
alu: ah yea :)
<cryptix>
i think they recently released the deepdream sotware
<alu>
yeah i been playing with it soon as it came out! :D
<alu>
learning a lot in the process
<cryptix>
im trying to find ppl that are into this kind of visuals or something like http://sub.blue/fractal-lab but its a rare trait :)
<cryptix>
jbenet: re ne index: hrm.. all of the sharness tests crashed in t0080-repo
<cryptix>
is there some kind of assumption of whats inited?
<cryptix>
jbenet: yup.. there are more pins in the actual output - should i just add those?