<brendyn>
Ah thanks. There was no man page installed on my computer, so I didn't realise I could do that :)
<cow_2001>
it isn't readily available here as well
<cow_2001>
i had to go around searching for "multihash" and stuff
<cow_2001>
the word was actually just "hash"
<brendyn>
Yeah, it's different to the output of sha256sum too, since its in a different base. I tried looking for commands to convert arbitrary bases but that was too bothersome
<brendyn>
I'm looking for ways to have files in IPFS and in my normal file system without duplicating anything. It seems I have to use ipns and then create symlinks between /ipns/Qm..... and the original filename.
<cow_2001>
ipfs is still kinda slow
<cow_2001>
wait, no
<cow_2001>
my computer is very slow
<brendyn>
I notice it uses quite a bit of RAM
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
stoogenmeyer has joined #ipfs
tg has joined #ipfs
zorglub27 has joined #ipfs
stoogenmeyer has quit [Ping timeout: 264 seconds]
Foxcool has quit [Read error: No route to host]
Foxcool has joined #ipfs
<dawuud>
lgierth: thanks!
<dawuud>
lgierth: ah ok so only hardcoded ipfs node addrs then?
<dawuud>
ok i obviously need to read the spec
<dawuud>
i am currently reading the libp2p spec
j12t has joined #ipfs
<dawuud>
after a year of deliberation i think we can finally write the tor/onion integration
egareva has joined #ipfs
<dawuud>
i'd like to release the tahoe-lafs tor integration at around the same time and then perhaps mirror a tahoe-lafs onion grid to ipfs... although that just makes me want to write a cryptographic capability layer on top of ipfs...
<dawuud>
yeah... otherwise mirroring them would break the cryptographi security model
j12t has quit [Ping timeout: 260 seconds]
Encrypt has joined #ipfs
xenic has joined #ipfs
egareva has quit [Ping timeout: 240 seconds]
somebody1234 has joined #ipfs
slothbag has joined #ipfs
cketti has joined #ipfs
somebody1234 has quit [Ping timeout: 250 seconds]
uvgroovy has joined #ipfs
somebody1234 has joined #ipfs
<somebody1234>
is anyone here active?
palkeo has joined #ipfs
<lgierth>
dawuud: i don't know what you mean with hardcoded addresses -- there's default bootstrap nodes configured, but they are just config defaults ;)
<lgierth>
dawuud: i agree we're at a good point for onion integration
<lgierth>
btw when you say "tor/onion" are these the same, or subtly different?
<somebody1234>
i have a few questions
<somebody1234>
like is the entire ipfs network one big blockchain with git
<somebody1234>
meaning that once you put data there, it stays forever?
<somebody1234>
stays forever meaning undeletable
<brendyn>
No, I think that is inaccurate
<lgierth>
somebody1234: there are no blockchains involved -- it's a store for data, with links between objects. files and directories are a particular type of data on ipfs.
<lgierth>
you can easily store blockchain-like data in ipfs though
<brendyn>
If you want to make comparisons, it's more like bittorrent, except you torrent every single file individually, that way if two different "torrents"/directories contain the same file, they will seed among each other
<lgierth>
dawuud: so is onion the protocol, and tor a specific impl?
<dawuud>
lgierth: jbenet: hey i noticed that go-multiaddr-net Dial and Listen require thin waist protocols and tor/onion isn't thinwaist so what to do?
<lgierth>
ah mh i think we had that discussion somewhere, but not sure we had a good result
<dawuud>
lgierth: tor clients can connect to non-onions such as ip addrs via exit nodes
<lgierth>
aah got it
<jbenet>
hey dawuud! o/ -- need different dial/listen methods, which is fine
<dawuud>
jbenet: well we need wrappers
<dawuud>
those methods are there because they are written correctly to use the standard golang interfaces
egareva has joined #ipfs
<jbenet>
Yep, we need similar stuff for onion dial/listen
<jbenet>
And/or improve that lib too to take onion into account
<dawuud>
hmm ok wait. you ware saying that a tor/onion plugin for ipfs would need it's own dial/listen methods, yeah that sounds correct. and we wouldn't use the ma-net Dial and Listen? correct?
<jbenet>
dawuud: we have a workshop in Lisbon this week. We could do a session/discussion on this to figure out next steps / work on it some
<jbenet>
dawuud: we could do it either way, I'd want a separate onion.Dial/Listen, and maybe also plug it into the lib, but I'd like to be able to JUST use onion and not the other stuff
somebody1234 has quit [Ping timeout: 250 seconds]
Foxcool has quit [Read error: Connection reset by peer]
<dawuud>
jbenet: yeah ipfs needs to be able to enforce a connection policy such that non-tor connections are disallowed
<lgierth>
also need onion-aware content routing
<dawuud>
if this is not yet implemented then maybe your workshop could discuss putting that on the roadmap in addition to some other higher level concerns that come up when discussing tor/onion integration for ipfs
<dawuud>
it seems the next small step is to write the plugin for tor... but then to use it properly for ipfs i would need guidance
<lgierth>
i can be your liaison for implementing it once we have an agreed approach
<lgierth>
the others are pretty busy in Q3 but i managed to evade all the "MUST" milestones haha
<dawuud>
ok. i suspect the plugin will be far easier to write than actually using it with an enforced policy etc
<dawuud>
i'll try to push some code soon that somewhat mimics the websocket plugin using the bulb dial and listen
<dawuud>
and then we can enter a code review feedback loop until it's done
reit has joined #ipfs
<dawuud>
does that sound OK for now? given that there are higher level design decisions that need to be made i think we may work on those in parallel
HastaJun_ has quit [Ping timeout: 250 seconds]
blaze_pascal has quit [Ping timeout: 240 seconds]
blaze_pascal has joined #ipfs
<dawuud>
previously we discussed how the information in a multiaddr would not be sufficient to start an onion service
<lgierth>
sounds good to me!
<lgierth>
happy that you have time for it again
<dawuud>
the onion listener not only needs an network address to contact the tor control port but also the onion service key material
j12t has joined #ipfs
somebody1234 has joined #ipfs
<dawuud>
therefore a higher level wrapping api would then abstract these details... and provide the multiaddr dial listen interface implementations
<lgierth>
dawuud: i think it'd be okay to treat the control-port info out-of-band, similar to how the kernel figures out which network interface to hand packets to
Combined2857 has joined #ipfs
xenic has quit [Ping timeout: 264 seconds]
<somebody1234>
sorry, i had problems with freenode, so 1. you can access ipfs via gateways (like proxies), right? and 2. can ipfs be used for dynamic content
<lgierth>
somebody1234: yes (go-ipfs includes the http-to-ipfs gateway), and yes (ipns is a mutability layer on top the immutable datastore)
j12t has quit [Ping timeout: 250 seconds]
<dawuud>
lgierth: yes of course the multiaddr should not contain control port or onion key material if that's what you mean by out of band
<rendar>
lgierth: is this about the use of ipfs in UDP?
<dawuud>
i meant the multiaddr string
<lgierth>
rendar: tor/onion
<rendar>
i see
<lgierth>
dawuud: i'll have to look up what we discussed last year -- i'm lacking context haha
<lgierth>
dawuud: i'm all into packet switching in libp2p at the moment
xenic has joined #ipfs
<somebody1234>
help, is there any documentation on how ipns works? the ipfs whitepaper doesn't seem to mention anything about how it works/what it does
apiarian has left #ipfs ["zoom"]
apiarian has joined #ipfs
<somebody1234>
also, is there a standalone gateway somewhere? would requiring clients to use their own gateways when possible be more performant?
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
<whyrusleeping>
somebody1234: every ipfs daemon runs a daemon on localhost:8080
<somebody1234>
whyrusleeping: ?
<Kubuxu>
runs a gateway
<Kubuxu>
whyrusleeping, lgierth: my morning was a bit sad.
<Kubuxu>
browsers only support HTTP/2 over TLS
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
<somebody1234>
Kubuxu: yeah ik, but i have no idea what whyrusleeping is replying to/how it relates to my question
<Kubuxu>
we plan extracting gateway from go-ipfs itself
<Kubuxu>
lgierth is working on it
<somebody1234>
oh ok
<whyrusleeping>
somebody1234: sorry, you were wanting a standalone gateway. The daemon provides a standalone gateway
egareva has joined #ipfs
<whyrusleeping>
i might be not thinking clearly right now though, coffee hasnt yet kicked in
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
* somebody1234
is confused (i meant standalone as in a distribution that includes just the gateway executable)
Foxcool has joined #ipfs
pguth has quit [Remote host closed the connection]
<somebody1234>
also, i still don't know what ipns does/how it works
pguth has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
<lgierth>
the gateway binary can't do much without an actual ipfs node that it can utilize to fetch objects from the network and the datastore
egareva has quit [Ping timeout: 244 seconds]
dmr has quit [Ping timeout: 250 seconds]
<somebody1234>
right
<somebody1234>
is ipfs being versioned a good thing?
<whyrusleeping>
depends on what you mean by versioned?
<somebody1234>
meaning you can access the state as it was from any point in time (git-like)
Foxcool has quit [Remote host closed the connection]
dmr has joined #ipfs
<whyrusleeping>
ah, immutable
<somebody1234>
shouldn't it be an option?
<somebody1234>
*instead
<somebody1234>
whyrusleeping: lgierth: is it possible to run a server with dynamic content e.g. a search engine? does IPNS enable that?
<brendyn>
I see a couple other people have mentioned libgen here. I think libgen couple benefit from using IPFS greatly since they use ftp/http/torrents together inelegantly. Has anyone done any envangelising toward them yet? :p
<somebody1234>
hmm
<somebody1234>
so ipns basically points to files?
<somebody1234>
basically generate a file -> share -> reroute ipns request to file link?
<apiarian>
yup
<somebody1234>
is that a good idea?
<apiarian>
why not?
Encrypt has quit [Quit: Quitte]
<apiarian>
and what might be some alternatives?
Encrypt has joined #ipfs
<somebody1234>
idk yet, but: can you have two servers serving the same client for ipns?
<apiarian>
well the entry isn't just hosted on your single node
<somebody1234>
but if it's dynamically generated, it's unlikely new servers will have any other nodes with the same entries, right?
<apiarian>
i think there's some confusion here about "servers" and "nodes"
<apiarian>
we're all nodes
<somebody1234>
yeah, i know that
<somebody1234>
but i think i may still be confused about how ipns works
<ipfsbot>
[go-ipfs] lgierth force-pushed feat/v0.4.3-changelog from cd7ddb6 to ff70266: https://git.io/vKM73
<ipfsbot>
go-ipfs/feat/v0.4.3-changelog ff70266 Lars Gierth: changelog: add v0.4.3...
<somebody1234>
if a server generates files dynamically, there's a pretty high chance the file will be completely new, right?
<apiarian>
sure
<somebody1234>
wait a sec, who/what recieves/what happens to an ipns request?
<ipfsbot>
[go-ipfs] whyrusleeping pushed 2 new commits to master: https://git.io/vKyE7
<ipfsbot>
go-ipfs/master ff70266 Lars Gierth: changelog: add v0.4.3...
<somebody1234>
not an ipfs search engine, i mean something more like google (say, it might search over a specific section of files/files in a list of its own)
j12t has quit [Ping timeout: 276 seconds]
stoogenmeyer has joined #ipfs
<davidar>
that's what that issue is about
Combined2857 has quit [Quit: Leaving]
<davidar>
you can't literally run google on ipfs, you need a distributed search engine
stoopkid has quit [Ping timeout: 250 seconds]
<somebody1234>
is there any preferred method yet (out of the three)?
<somebody1234>
i'm not entirely sure, but the magic way looks like it's not possible yet, and the static way may be slow on searches with few matches
Akaibu has quit [Quit: Connection closed for inactivity]
M-mubot was kicked from #ipfs by whyrusleeping [M-mubot]
M-mubot has joined #ipfs
M-mubot was banned on #ipfs by whyrusleeping [*!mubotmatri@gateway/shell/matrix.org/x-jsdhloowdlchdkmm]
<ipfsbot>
[go-ipfs] lgierth created feat/remove-gobuilder (+2 new commits): https://git.io/vKyuX
<ipfsbot>
go-ipfs/feat/remove-gobuilder ec68ef0 Lars Gierth: cleanup: remove mentions of gobuilder...
<ipfsbot>
go-ipfs/feat/remove-gobuilder 3975213 Lars Gierth: docs: update release procedure...
Foxcool has joined #ipfs
<zero-ghost>
does ipfs run inside i2p yet?
<somebody1234>
would dynamically generated (one time/with user credentials) files clutter the global index?
<lgierth>
zero-ghost: no -- there's been a few notes in the ipfs/notes repo
<zero-ghost>
rough guesstimate?
<lgierth>
no
<lgierth>
:)
<lgierth>
check it out and see if there's anything you can do to help -- for me it's not a priority, i've never used i2p and have no particular interest in it, but it'd be great to have
<whyrusleeping>
Hey everyone! We just put out the first release candidate for ipfs v0.4.3
<whyrusleeping>
Please give it a try and test it out with everything youre using ipfs for
<whyrusleeping>
it should be better in pretty much every way, the changelog is huuuuuuge and we have put a ton of effort into making this the best release yet
<lgierth>
in addition to what whyrusleeping said, if you are running go-ipfs built from master between *after* v0.4.2 please update to v0.4.3-rc1 immediately, as there was a security vulnerability on master for about 3 weeks
egareva has quit [Ping timeout: 258 seconds]
xenic has quit [Ping timeout: 276 seconds]
vavovi has joined #ipfs
xenic has joined #ipfs
espadrine_ has quit [Ping timeout: 260 seconds]
apiarian has quit [Ping timeout: 252 seconds]
Oatmeal has joined #ipfs
apiarian has joined #ipfs
j12t has joined #ipfs
jaboja has joined #ipfs
j12t has quit [Ping timeout: 250 seconds]
bearbin has quit [Ping timeout: 244 seconds]
neurrowcat has joined #ipfs
egareva has joined #ipfs
egareva has quit [Ping timeout: 258 seconds]
zz_r04r is now known as r04r
uvgroovy has quit [Quit: This computer has gone to sleep]
ianopolous has joined #ipfs
<brendyn>
I will upgrade and see if it works :)
jaboja has quit [Remote host closed the connection]
palkeo has joined #ipfs
zorglub27 has quit [Ping timeout: 260 seconds]
vavovi has quit [Ping timeout: 264 seconds]
egareva has joined #ipfs
ruby32 has quit [Remote host closed the connection]
egareva has quit [Ping timeout: 244 seconds]
<brendyn>
I'm wondering, would it be possible to write some system sorta like BOINC in that it would allow people to work together to seed a huge database. For example if there was 100TiB of files and 1000 people, the system could dynamically allocate each person 100GiB worth of files to pin in IPFS, being updated as new files were added?
<apiarian>
where is this data coming from?
<brendyn>
Like Library Genesis for example.
<brendyn>
Any arbitrary file sharing project.
<Kubuxu>
brendyn: it would be essentially ipfs-cluster (something that is being worked on) but with handling of a hostile environment
<brendyn>
There could be a central application that allowed the user to select how much hard drive space and bandwidth to donate to a project
<brendyn>
Kubuxu: Thanks, I didn't know of that
<somebody1234>
hostile environment?
<brendyn>
Yes, everything becomes difficult when one can't assume friendliness.
<Kubuxu>
somebody1234: if anyone can join then it is essentially hostile (I might say that I will host some piece of data and just throw it in to /dev/null).
<brendyn>
I would like to contribute to such projects, but I'm a novice coder. How can I get started?
stoopkid has quit [Read error: Connection reset by peer]
ianopolous has quit [Ping timeout: 258 seconds]
JesseW has quit [Ping timeout: 244 seconds]
zorglub27 has quit [Ping timeout: 252 seconds]
<apiarian>
so i think i've managed to get my node through my firewall.. maybe. can somebody try an ipfs get on Qmc82YwQnYeQ7wNtzcoG21Q1tPTdfj2yhcAiYPtYb143d5 ?
rgrinberg has quit [Ping timeout: 244 seconds]
<apiarian>
oh, wow! it is available through the public gateway! w00t!
Encrypt has joined #ipfs
zorglub27 has joined #ipfs
<Kubuxu>
sol56: I just was rebuilding my nodes from AUR
<Kubuxu>
no problems, IDK what is going on.
<Kubuxu>
are you running custom kernel maybe?
<Kubuxu>
I used to had problems with Go when I was running -ck patchset.
j12t has joined #ipfs
<sol56>
yes
<sol56>
grsec kernel
j12t has quit [Ping timeout: 252 seconds]
<sol56>
go seems to be fine though
<sol56>
it builds and runs everything else i use that is built in go.
<sol56>
coyim and others
dmr has quit [Ping timeout: 250 seconds]
<whyrusleeping>
alu: are you at hope?
<demize>
- running post install for gogo-protobuf: /home/user/build/go-ipfs-git/src/src/gx/ipfs/QmZ4Qi3GaRbjcx28Sme5eMH7RQjGkt8wHxt2a65oLaeFEV
<demize>
fatal error: unexpected signal during runtime execution
<demize>
Hmmm...
<demize>
sol56: Anything in dmesg?
espadrine has quit [Ping timeout: 260 seconds]
<sol56>
nothing
<demize>
It's segfaulting in the post-install at least.
reit has joined #ipfs
<sol56>
waht the heck
<sol56>
it worked this time...
<demize>
Yay, random one-off errors are awesome, aren't they...
rgrinberg has joined #ipfs
<sol56>
yeah..
<sol56>
but the weird thing is i've got it more then once
<sol56>
and it's a random one-off that it compiled
<demize>
Mmmm.
<demize>
clean chroot?
<sol56>
i didn't change where i was compiling it from
<sol56>
but thats a good idea for the future
Combined2857 has joined #ipfs
neuthral has quit [Ping timeout: 240 seconds]
neuthral has joined #ipfs
cwahlers_ has joined #ipfs
cwahlers has quit [Ping timeout: 258 seconds]
cryptix has quit [Ping timeout: 244 seconds]
xenic has quit [Ping timeout: 258 seconds]
zorglub27 has quit [Ping timeout: 240 seconds]
j12t has joined #ipfs
onla has joined #ipfs
<onla>
Is the dht bandwidth still some 10-20kB/s or smth, or has it been optimized in recent months?
<onla>
and I wonder what was the name of this one project that was working on some IM/facebook -like thing for ipfs
<dawuud>
i'll work on this more later when i'm not so tired
<dawuud>
it seems i'll have to write yet another implementation of the net.Addr interface... but it will need to have some extra attributes to handle tor properly etc.
herzmeister has quit [Ping timeout: 244 seconds]
reit has quit [Ping timeout: 252 seconds]
herzmeister has joined #ipfs
Encrypt has quit [Quit: Dinner time!]
nycoliver has quit [Ping timeout: 265 seconds]
dmr has joined #ipfs
stoogenmeyer has quit [Remote host closed the connection]
rendar has quit [Ping timeout: 240 seconds]
Shrike[m] has joined #ipfs
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
stoogenmeyer has joined #ipfs
dmr has quit [Ping timeout: 265 seconds]
wallacoloo has joined #ipfs
j12t has joined #ipfs
dmr has joined #ipfs
stoogenmeyer has quit [Ping timeout: 240 seconds]
lmatteis has joined #ipfs
j12t has quit [Ping timeout: 244 seconds]
rendar has joined #ipfs
Encrypt has joined #ipfs
dmr has quit [Ping timeout: 250 seconds]
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
chriscool has joined #ipfs
nycoliver has joined #ipfs
<alu>
whyrusleeping: yes
<alu>
are you here too? lol
<lgierth>
alu: lisbon
<alu>
yer in lisbon, I'm in new york
<lgierth>
o/
<lgierth>
say hi to the nycmesh and toronto mesh people if you see them :)
<alu>
WHERE HOW?
<alu>
I really want to talk to them
<alu>
They really should meet me
<alu>
I'm starting LA mesh lgierth
<alu>
putting together an elite team
<lgierth>
hehehe
<lgierth>
voxelot: you hear that? ^
<alu>
He already knows
<lgierth>
all the better
<lgierth>
let me find you contacts
<alu>
^.~
<alu>
yes pls, tell them im at overflow room between friedman / lamarr
j12t has joined #ipfs
<lgierth>
so nycmesh's contact is on their website, and they're usually pretty responsive: https://nycmesh.net/
<alu>
do any of them use IRC
<mactrent>
5
<lgierth>
alu: probably, but i'm not sure what their nicks are
stoogenmeyer has joined #ipfs
stoogenmeyer has quit [Ping timeout: 240 seconds]
espadrine has joined #ipfs
<alu>
any ipfs peeps here?
<alu>
@ hope lol
<demize>
Nope, none of the people here has ever actually used ipfs. Why would we?
<alu>
its p cool
rgrinberg has quit [Ping timeout: 264 seconds]
<demize>
It has its fair share of weirdness still, but it's pretty cool, yeah.
jager has quit [Read error: Connection reset by peer]