<doublec>
Is it safe to expose the 8080 port from ipfs as a gateway or do they have special settings for it?
<alexandria-devon>
oh yowza, thats a zippy one
<alexandria-devon>
getting 500kb...
<alexandria-devon>
we'll see if it holds up
<alexandria-devon>
yup, thats what i figured...
<alexandria-devon>
got up to ~700kbs for a bit, hit the limit that the other nodes had reached and slowed down to my paltry upload speed ;)
<alexandria-devon>
coolness, nice to see how much it multiplies from a few other nodes
<tperson>
Can't build ipfs on my 512mb DO node. Doesn't appear to have enough memory anymore.
pfraze has joined #ipfs
pfraze_ has joined #ipfs
* zignig
has managed to squidge ipfs into a rkt container
<zignig>
needs work on config management.
pfraze has quit [Ping timeout: 264 seconds]
<zignig>
oh and logging chews up the tmpfs pretty quickly
Wallacoloo has quit [Ping timeout: 244 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
anshukla has quit [Remote host closed the connection]
<headbite>
whyrusleeping: are the 8 nodes in some kind of a round robin or something?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<alexandria-devon>
does IPFS not yet support some form of local peer discovery?
<headbite>
I heard local discovery was added but I haven't tried it out myself.
<headbite>
I did end up trying "offline" mode last weekend. Basically starting the daemon while to connected to wifi still lets you view your pinned content.
<alexandria-devon>
cant find any specific commands in the ipfs commands list
<headbite>
*while not connected to wifi
<alexandria-devon>
for searching for a file from local peers
<tperson>
Peers on local networks are automatically connected.
<alexandria-devon>
got two computers here in LAN, the host of the file and another one - the other seems to be downloading just as quickly as the gateways are, which seems mostly limited by my outbound upload speed
<tperson>
(though may of been disabled, I haven't been complete kept up with status)
<alexandria-devon>
if it was peering directly i would expect it to be much faster
<tperson>
You can check your peers with `ipfs swarm peers`
<alexandria-devon>
can i check which peers have an IPFS file?
<alexandria-devon>
pinned, or local already
<tperson>
You can find providers of a hash
<alexandria-devon>
i see the other local computer in my swarm peers list
<alexandria-devon>
what is the command to do that, do you know of the top of your head?
<alexandria-devon>
tperson
<tperson>
`ipfs dht findprovs <hash>
<alexandria-devon>
gracias
<alexandria-devon>
fabulous, ty ;)
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
EricJ2190 has quit [Ping timeout: 264 seconds]
<whyrusleeping>
the local peer discovery works :)
<whyrusleeping>
also, for faster speeds, try latest master
<whyrusleeping>
just merged some good code today
pfraze_ has quit [Remote host closed the connection]
<alexandria-devon>
local peer discovery works on its own or do i need to do anything special?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
works as is
alexandria-devon has quit [Ping timeout: 246 seconds]
hellertime has quit [Quit: Leaving.]
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed refactor/importer from 89d7625 to 8dae027: http://git.io/vTrp7
<ipfsbot>
go-ipfs/refactor/importer 8dae027 Jeromy: change pinning to happen in a callback
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<zignig>
whyrusleeping: so the bitswap double-awesome-mega-mega patch has landed ?
inconshr_ has joined #ipfs
alexandria-devon has joined #ipfs
inconshreveable has quit [Ping timeout: 256 seconds]
<zignig>
whyrusleeping: wow, that is heaps faster , the resolving feels snappier as well.
<zignig>
whyrusleeping: good work :)
chriscool has joined #ipfs
inconshr_ has quit [Ping timeout: 265 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] chriscool created t0060-stdin-works (+1 new commit): http://git.io/vTivY
<ipfsbot>
go-ipfs/t0060-stdin-works e7732ba Christian Couder: t0060: expect success with stdin...
<whyrusleeping>
zignig: glad you like it :)
* krl
on latest master as well
<whyrusleeping>
krl: i think your nodes stopped serving the geoip stuff after you left last night
<krl>
oh
<whyrusleeping>
could you give me the hash again?
<krl>
one sec
<ipfsbot>
[go-ipfs] chriscool opened pull request #1277: t0060: expect success with stdin (master...t0060-stdin-works) http://git.io/vTivw
<jbenet>
I fear perf is going to suck. If so bug whyrusleeping
<jbenet>
(He's good at that)
<whyrusleeping>
that should be my job title
<kyledrake>
Nope, I've got it under control so far. Dealing with the current output was easy: output.split("\n").last.split(' ')[1]
<jbenet>
whyrusleeping: could you check that krls geoip _really is there_? Count the refs
<whyrusleeping>
i trust pinbot
<jbenet>
(On pin bot)
<jbenet>
Yeah I don't
<jbenet>
It's flaked multiple times before, remember 1150 exists too
<kyledrake>
Phase 1 is local, phase 2 is I need to have it fire up an IPFS server somewhere, and have it send the command remotely.
<kyledrake>
Well, I need to run an ipfs server, and then send the commands to it remotely, rather.
<whyrusleeping>
1150 is solved i thought
<jbenet>
Yeah maybe :)
<jbenet>
(We hope!)
<krl>
i'm writing tests for the api btw
<whyrusleeping>
krl: awesome!
<kyledrake>
It would be easier to just run ipfs straight off our replication server, but I'm a little concerned about running into another netscan thing and having Hetzner freak out on our main backups
<jbenet>
krl: yes yes yes
<krl>
it's unfortunately quite broken it seems
<krl>
having http requests that hang, for ipfs.cat for example
<jbenet>
kyledrake: let's get that fixed on our end first
<kyledrake>
Err not Hetzner, that's on a different host. But you get the idea.
<whyrusleeping>
jbenet: did you hear my intermediary idea?
<whyrusleeping>
i wanted a config setting for ip ranges not to dial
<jbenet>
Not yet haven't seen backlog
<jbenet>
About to fly again
<jbenet>
Took a break in hotdoglandia
<kyledrake>
jbenet How long are you there again?
<whyrusleeping>
jbenet: youre coming to seattle after berlin right? :D
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
<whyrusleeping>
jbenet: also, pinbot does in fact have the entire thing
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
chriscool has quit [Ping timeout: 256 seconds]
<jbenet>
kyledrake: till tuesday
<jbenet>
whyrusleeping: I can probably be there june-ish
<jbenet>
whyrusleeping: that's great! All 65k objects?
<jbenet>
pinbot: good bot!
<jbenet>
krl: request hanging? What's the example?
<krl>
ipfs.cat('hash', ...
<whyrusleeping>
jbenet: yeap 65217
<jbenet>
Hmmm :/. But does ipfs cat work in the terminal with daemon on? (It uses the api)
<krl>
it might be a local issue for me, or weird mocha interactions.
<jbenet>
zignig: awesome getting it into rkt!
<jbenet>
zignig would be awesome to make a "rkt launcher" with ipfs support
<krl>
jbenet: just started looking at this, i'll ask for help if i dont figure it out
<krl>
ok, i think it's mocha acting up, not the actual command.
<krl>
however, i should really get some sleep. see you tomorrow!
<jbenet>
See you in a few!
<whyrusleeping>
jbenet: if you get a sec could you look at 1274?
<zignig>
jbenet: I'm m rolling rocket support into astral boot. but a rkt of ipfs will aid testing @scalle and d easy ipfs access for coreos users.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
whyrusleeping I would but my computer's battery ran out. And then Lufthansa doesn't believe in electrical plugs at the seats. (I expected more from German engineering?) oh I guess that's design. Yeah this isn't Denmark.
<jbenet>
That PR is complex. Will need to look carefully
<whyrusleeping>
1276 is less so
<kyledrake>
jbenet Lufthansa is.. a bit spartan I noticed.
<jbenet>
✈ ✈
sharky has quit [Ping timeout: 244 seconds]
patcon has quit [Ping timeout: 258 seconds]
tilgovi has joined #ipfs
<cryptix>
jbenet: good flight Ö=
<cryptix>
:)*
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
sharky has joined #ipfs
Tv` has quit [Quit: Connection closed for inactivity]
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] whyrusleeping created fix/bs-cmd-offline (+1 new commit): http://git.io/vTi8x
<ipfsbot>
go-ipfs/fix/bs-cmd-offline 648526d Jeromy: make bitswap commands error out properly offline
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1278: make bitswap commands error out properly offline (master...fix/bs-cmd-offline) http://git.io/vTi4k
pfraze has joined #ipfs
pfraze has quit [Ping timeout: 276 seconds]
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]
<cryptix>
sometimes "ipfs get" times out with error code 0.. which is really bad
<kyledrake>
phase 2 is getting this to work on a remote server. Right now it only works locally. I'm not running ipfs on the main web server, so it needs to do a remote shellout.
<cryptix>
i wonder if we can get the writable http interface feature complete - i guess that would be less overhead
<kyledrake>
the thing is, even if the API supported multiple-file, it would still be more efficient to do an ipfs add -r via a mounted filesystem.
<kyledrake>
Otherwise I would have to dump the entire site into a REST api every time it updated. Not good.
mildred has joined #ipfs
<tperson>
kyledrake: The API does support multi file, just not the `node-ipfs-api` package.
<cryptix>
the shell in online mode also does a rest-ish multipat wrapping for you afaik
<tperson>
Since you're writting it in ruby anyways, you could totally roll your own. Multi files are just multiparts inside of multiparts
<Darksecond>
yeah, but you need the ipfs daemon running, otherwise you see no real advantages (and you would still be pulling from the internet)
<zignig>
correct.
<Darksecond>
ok :)
<zignig>
there are gateway machines that the team is running , they cleanse themselves after 12 hours.
<zignig>
so you can get material off ipfs via the internet.
inconshreveable has quit [Remote host closed the connection]
kyledrake has joined #ipfs
<zignig>
Darksecond: how did you come across ipfs ?
<Darksecond>
I saw that playback had ipfs support (https://github.com/mafintosh/playback). (Was looking at example electron apps) and didn't know what ipfs was :)
okket has quit [Max SendQ exceeded]
atomotic has joined #ipfs
okket has joined #ipfs
<zignig>
looks like a cool project for ipfs.
<Darksecond>
uhuh
<Darksecond>
Is there a way to see how many copies of a file are floating about?
<Darksecond>
sorry about all the questions, just curious ^^
<zignig>
yep , on the command line ; ipfs dht findprovs <hash>
<zignig>
keep asking question are gooooooood. ;)
<Darksecond>
:D
rht_ has joined #ipfs
<Darksecond>
I like how the ipfs command/daemon is one executable :)
<zignig>
that golang , if you are a programmer check it out.
<dbolser>
if you were to add up all the torrents currently indexed, how many Pt do you get?
<dbolser>
is that the nominal / potential capacity of IPFS?
<zignig>
dbolser: the name space size and 256k block ( at the moment ) , means that is lots of EXA bytes.
<zignig>
dbolser: your entire porn collection might fit. ;P
<dbolser>
pff
<dbolser>
but that's the theoretical limit right?
<dbolser>
what's the 'target' capacity? if that makes sense ... like ... OK, what is the current capacity?
<zignig>
dbolser: you can't tell, the swarm won't tell you of block , it will only tell you about blocks that are near your node id.
<zignig>
unless you search for them.
<dbolser>
also (sorry for showing my noobs), I'm assuming once I add a file, at some point it gets seeded by others, what's the mechanism for that?
<dbolser>
zignig: you mean 'search for them explicitly by id'?
<zignig>
dbolser: it will only get seeded by others _if_ they download it.
<dbolser>
ah
<zignig>
your node tells other nodes they you have it, the first id resolution is by proxy.
<dbolser>
so there (currently?) is no concept of automatic redundancy?
<dbolser>
from the paper (I'm only part way through), nobody is forced to host a block
<zignig>
dbolser: I have a cron job and a webservice that grabs and id and replicates it onto another node ( and pins it ) , but in ipfs ; not yet.
<dbolser>
pins it?
<zignig>
stop the garbage collecter from deleting it.
domanic__ has joined #ipfs
<dbolser>
you grab ids from the gateway? Else, how do you guess IDs?
<dbolser>
ahh... whats the gc?
patcon has quit [Ping timeout: 276 seconds]
<zignig>
these are ids of files that I have added.
<dbolser>
oh right
<zignig>
gc: ipfs repo gc , will cleasne the files that you have stored locally
<dbolser>
and 'another node' in this case is just someone outthere? or you need to control that node?
* dbolser
feels no wiser
* dbolser
clenses his brains
<zignig>
I have control over it, ( its the machines that is running this irc session )
<zignig>
dbolser: what vexes you ?
domanic_ has quit [Ping timeout: 265 seconds]
<dbolser>
cleasne the files?
<dbolser>
what do you mean by cleasne the files?
<zignig>
bork ! typo ! , clean
<dbolser>
so I can't randomly put file x onto host z?
<dbolser>
What do you mean by clean the files?
<zignig>
nope , host z must ask for them.
<dbolser>
ok
<dbolser>
so if you're a google host, you could go looking for all files?
<zignig>
sort of , you have to have the refs first.
<zignig>
the important thing is that the same file is also cached, if there are 500 web application in ipfs and they are using the same version of jquery , it uses the same block.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
lgierth has quit [Quit: Ex-Chat]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
domanic__ has quit [Ping timeout: 264 seconds]
<dbolser>
nice
m0ns00n has quit [Remote host closed the connection]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
domanic has joined #ipfs
pfraze has joined #ipfs
kyledrake has quit [Ping timeout: 265 seconds]
<jbenet>
hey dbolser thanks for the interest. Make sure you read the code of conduct of this channel and the IPFS community https://github.com/ipfs/community/blob/master/code-of-conduct.md -- we do not discuss piracy or illegal or questionable content here.
<krl>
jbenet: where you at?
<jbenet>
krl: i'm over by montbijou park
<jbenet>
krl: about to get some coffee-- want to join?
<jbenet>
meetup at 18:00
<krl>
i'd like to grab some food before, maybe around there?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
domanic has quit [Ping timeout: 256 seconds]
<whyrusleeping>
i got conned into catsitting for a friend today...
<whyrusleeping>
i may not get anything done until saturday
<substack>
caturday
Darksecond has quit [Quit: leaving]
<whyrusleeping>
and its not just a cat, its a very ferocious manx kitten
<whyrusleeping>
who is currently destroying his squeaky mouse toy
<jbenet>
video (over ipfs) or didn't happen
* whyrusleeping
thinks about how to get video off of his phone
patcon has joined #ipfs
hellertime has joined #ipfs
bmcorser has joined #ipfs
Tv` has joined #ipfs
<whyrusleeping>
jbenet: are the gateways running latest master?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
notduncansmith has quit [Read error: Connection reset by peer]
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
flugsio has joined #ipfs
patcon has joined #ipfs
<tperson>
I have no idea why that test is failing where it does right now.
<whyrusleeping>
the code in core resolve got changed so that only /ipns/ paths trigger it
<tperson>
aaah
<tperson>
Would it be appropriate at this time to pull that error out and name it inside of path, and use it for both of those errors (as they are basically the same thing).
<whyrusleeping>
yeah, i'd do that
<ipfsbot>
[go-ipfs] whyrusleeping created fix/bs-race (+1 new commit): http://git.io/vT1AM
<ipfsbot>
go-ipfs/fix/bs-race 934aaca Jeromy: fix minor data race in bitswap
cryPhone has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
joeyh has quit [Read error: Connection reset by peer]
cryPhone has quit [Client Quit]
<dbolser>
what's the plan to publicise files?
<whyrusleeping>
publicise?
joeyh has joined #ipfs
<ipfsbot>
[go-ipfs] travisperson opened pull request #1282: Bug/panic on empty path (master...bug/panic-on-empty-path) http://git.io/vTMJP
<tperson>
Like an index?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
lgierth has joined #ipfs
<Tv`>
maybe dbolser is asking about ipns
<dbolser>
Tv`: I think i may be... (need to read more)
<dbolser>
yeah, like an index or a search engine
<whyrusleeping>
ah, we're planning on building a search engine type service
<dbolser>
anyone working on a distributed database? i.e. where you could register your files
<dbolser>
trying to get my head round how you keep spam out of a distributed index
<cryptix>
what.. having this strange thing on my linux node right now.. in online mode i only get 'Error: can't Lock file "/home/cryptix/.ipfs/repo.lock": has non-zero size'when running commands
<jbenet>
woah
<jbenet>
what
<jbenet>
why did that post _now_
<jbenet>
i sent that htis morning
<whyrusleeping>
cryptix: interesting... that could be my fault
<jbenet>
m3s that's about as crowded as i've seen anything in berlin
<cryptix>
whyrusleeping: works on osx - trying to repro on an 2nd box
<jbenet>
m3s go live in berlin and/or zurich. you'd love both.
<whyrusleeping>
cryptix: linux what arch/os/version/etc?
<cryptix>
whyrusleeping: debian
<whyrusleeping>
cryptix: krl hows berlin?
<cryptix>
whyrusleeping: awesome as always :)
<cryptix>
whyrusleeping: i have no idea about the lock.. works on the 2nd box..
<whyrusleeping>
cryptix: i switched the lock checker to explicit error checking
<whyrusleeping>
should the lock have been locked?
<kyledrake>
A bus in Berlin. The subways are (usually) less busy FYI.
<kyledrake>
Good luck figuring out the crazy ticket system though
<kyledrake>
If it looks clean, you're in the wrong place. There should be graffiti on pretty much everything if you're doing it right.
<whyrusleeping>
kyledrake: lol, i think ill let you show me around berlin :P
<lgierth>
jbenet: can i have the id of that video you showed earlier?
<kyledrake>
My main advice is watch where you step. Dogs are not.. cleaned up after very well.
<kyledrake>
Anyways, day 2 I need to get an IPFS node server running and handling async messages, so that's my day
<jbenet>
<3
<cryptix>
whyrusleeping: the daemon is already running and the pid inside the lock file is correct
<cryptix>
whyrusleeping: resetting to c561c09 makes it work again
<kyledrake>
Now's a good time to let me know if there's a way to do replication of nodes
<cryptix>
whyrusleeping: might not be right before that one, that was just where my repo was before
<ipfsbot>
[go-ipfs] jbenet pushed 5 new commits to master: http://git.io/vTyvZ
<ipfsbot>
go-ipfs/master 42289d4 Travis Person: Daemon panics if no path is given...
<ipfsbot>
go-ipfs/master dcc4da0 Travis Person: Replaced old logic to check for valid path...
<ipfsbot>
go-ipfs/master d96246c Travis Person: Fixed tests to actually test for the error we are seeking...
<kyledrake>
My assumption is that an rsync won't work for the running node instance
<nell>
whyrusleeping yo
<whyrusleeping>
nell: are you working on this metaverse project full time?
<nell>
yes
<kyledrake>
That's an interesting question actually. How do you back up an IPFS node?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<nell>
I'm playing with an android implementation
<whyrusleeping>
kyledrake: zip up the .ipfs folder
<kyledrake>
whyrusleeping is it safe to do that hot?
<whyrusleeping>
nell: thats awesome!
<whyrusleeping>
kyledrake: i wouldnt do it hot
<kyledrake>
ok.
<whyrusleeping>
unless you want to manually unlock leveldb
faceface has quit [Quit: Lost terminal]
<jbenet>
kyledrake: soon you'll be able to back the whole thing with ... ipfs. (and s3)
<jbenet>
kyledrake repo state will be boiled down (close to) just a singel hash. (single hash + couple files, maybe)
<kyledrake>
yeah
<kyledrake>
true
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/bs-race from 934aaca to a3356b0: http://git.io/vTyU7
<ipfsbot>
go-ipfs/fix/bs-race a3356b0 Jeromy: fix minor data race in bitswap
<lgierth>
is there any port that is considered standard? i see 4001 a lot. and: if i just throw a bunch of peer ips at it, will it eventually discard the ones that aren't available, e.g. don't run an ipfs daemon?
<whyrusleeping>
lgierth: yeah, 4001 is our current 'standard'
<lgierth>
i wanna start it without peer discovery and then feed cjdns' peers into it with a standard port
<whyrusleeping>
thats a neat idea
chriscool has joined #ipfs
<lgierth>
yes :P
<lgierth>
i have a little golang+ipfs bootstrap/update script that i just used on three of my cjdns nodes
<lgierth>
so let's say i throw 20 ipv6 addresses at it for peering, and only 2 of them are actual ipfs nodes on port 4001
<whyrusleeping>
yeah, it will try them and chuck ones that dont work
<lgierth>
anything problematic with that?
<whyrusleeping>
sounds fine to me
<lgierth>
coolio
<whyrusleeping>
the upside of our current dialer system is that its really good at sifting through a ton of addresses to get a good connection
<whyrusleeping>
(the downside is we dial everybody and their uncles grandmother in the process)
<lgierth>
have fun dialing all of fc00::/8
<lgierth>
;)
ei-slackbot-ipfs has quit [Remote host closed the connection]
<lgierth>
we don't have anything like multicast at the moment
ei-slackbot-ipfs has joined #ipfs
<whyrusleeping>
lol
<lgierth>
you can only discover other nodes using cjdns' peers and routing table tools
<lgierth>
it's all one big LAN, except for, well, multicast
therealplato has quit [Ping timeout: 252 seconds]
mildred has quit [Ping timeout: 256 seconds]
<whyrusleeping>
lol
void has joined #ipfs
therealplato has joined #ipfs
<lgierth>
multicast is doable but noone has bothered so far
<Tv`>
that 4+ year laptop has keys swapped out because i literally typed through the keycaps
<Tv`>
and now some of the keys bounce off their brackets randomly, too
<whyrusleeping>
ive had the same mechanical keyboard for four years now
<whyrusleeping>
pretty sure its going to outlast everything its connected to
<Tv`>
i mourned the passing of the PS/2 connector, needed to donate a bunch of model m's
atrapado has quit [Quit: Leaving]
Darksecond has quit [Quit: leaving]
<whyrusleeping>
i'd take a model m
pfraze has joined #ipfs
<whyrusleeping>
i have a few computers laying around that only take PS/2
<Tv`>
i used to buy them for $3 a pop
<whyrusleeping>
and my openwrt router doesnt know what a usb keyboard is
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
Wallacoloo has quit [Quit: Leaving.]
Wallacoloo1 has joined #ipfs
<whyrusleeping>
apparently my laptop either lied about 70% battery life
<lgierth>
so it looks like i need to know node's id before i can peer with it
<whyrusleeping>
lgierth: this is a true fact
<lgierth>
meh
<whyrusleeping>
i forgot about that
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
<lgierth>
so i'll fire up `echo $nodeid | nc -ukp 4002`
Wallacoloo1 has quit [Ping timeout: 264 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
nell has quit [Quit: WeeChat 0.4.2]
<whyrusleeping>
Tv`: you use linux on a macbook air?
<Tv`>
whyrusleeping: yes
<whyrusleeping>
what distro?
<www1>
just an idea i had: if people start spamming IPFS there could be mining, too. so that hashes start with Q000... or something before they get accepted
<Tv`>
whyrusleeping: ubuntu
<Tv`>
whyrusleeping: i used to run debian on a powerpc apple ibook you know ;)
<Tv`>
i like small laptops
<Tv`>
except the new apples -- that keyboard is just too much
mildred has quit [Quit: Leaving.]
<whyrusleeping>
www1: take a look at section 3.1 of the paper
<whyrusleeping>
Tv`: and everything runs nicely?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<www1>
whyrusleeping: 3.1 Identities ?
<whyrusleeping>
yeap!
<www1>
how does this relate to mining or spam?
<whyrusleeping>
read it :)
<Tv`>
whyrusleeping: well, the firmware looks different from a typical PC (it's an early, bastard, version of EFI), but once you're past the bootloader it's just an intel box.. broadcom wifi though, and they hate open source
<whyrusleeping>
okay, yeah
<whyrusleeping>
the broadcom is what im having issues with
<whyrusleeping>
but i just turn those off instead of dealing with them
<Tv`>
i wouldn't try that with newer apples, though
<Tv`>
they've gone more and more proprietary over the years
<Tv`>
remember, this was 2011 model
<whyrusleeping>
hmmm, yeah, ive got a late 2014
<Tv`>
i had good sources saying it works when i bought it
<www1>
whyrusleeping: so?
<whyrusleeping>
www1: it discusses mining
<www1>
don't get it
<whyrusleeping>
Tv`: yeah, i wish they were replaceable
<www1>
i must be stupid
<www1>
my idea was about the content and not about identities
<whyrusleeping>
oooooh, okay. i didnt catch that distinction
<www1>
was anyway pretty raw and not thought through
<whyrusleeping>
mining for ID's is going to be important
<whyrusleeping>
the thing about mining for content is that ipfs is a pull-not-push type of data network
<www1>
i am not sleeping. why do you ask me whether I am sleeping all the time? :D
<whyrusleeping>
so i can generate all the content i want, and it doesnt go anywhere else until someone else grabs it
<whyrusleeping>
www1: lol
<www1>
but still, somewhere needs to be stored where the content ist
<www1>
ok, hashes can be faked an there might be no content
<whyrusleeping>
yeah, this is a problem
<www1>
at least I understood now how mining for identities works in IPFS
<whyrusleeping>
you can spam the DHT, and it will bog things down
lgierth has quit [Quit: Ex-Chat]
zootella has joined #ipfs
<www1>
any idea how to solve this? a fee structure maybe
<whyrusleeping>
no... that will make us lose legitimate users
<whyrusleeping>
dont want to impede the freedom of the network
www has joined #ipfs
<www>
so spamming the dht is actually not an issue? whyrusleeping
www1 has quit [Ping timeout: 240 seconds]
<whyrusleeping>
no, its definitely an issue
<whyrusleeping>
i'm worried about how provider records scale even without factoring in people spamming it on purpose
<whyrusleeping>
if i add a large file, and it has thousands of child blocks
<whyrusleeping>
i need to announce each of those blocks to the network
<whyrusleeping>
thats difficult
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
every solution ive come up with has rather large drawbacks
<www>
people will use this service to share things which other people don't like to be shared, so the latter group might have an incentive to 'spam'. a fee structure would help, i think. better fees than a broken network.
<whyrusleeping>
s/broken/slow/
inconshreveable has quit [Remote host closed the connection]
<whyrusleeping>
provider records do have to be signed by the provider
<whyrusleeping>
and thats a semi expensive operation
<whyrusleeping>
it also makes it easy to identify a spammer
<whyrusleeping>
and if peer ID generation is made difficult by the ID mining process
<whyrusleeping>
then a system to 'ban' spammers from the network could address the purposeful spam issue
<whyrusleeping>
the only problem then becomes identifying who is a spammer and who is not
<whyrusleeping>
(spammer vs someone adding a lot of files)
EricJ2190 has joined #ipfs
<www>
this also would involve centralization....
<www>
this banning
<whyrusleeping>
no
<whyrusleeping>
it would just require each node to make a decision on whether or not a given person is a spammer
<whyrusleeping>
you can also ask nodes that you trust (by some trust metric) who they think are spammers
<www>
so a WoT
<www>
possible, yes
<whyrusleeping>
yeah, a web of trust is a good idea
ei-slackbot-ipfs has quit [Remote host closed the connection]
ei-slackbot-ipfs has joined #ipfs
headbite has quit [Ping timeout: 258 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
flugsio has quit [Quit: WeeChat 1.2]
anshukla has joined #ipfs
anshukla has quit [Remote host closed the connection]
mappum has quit [Ping timeout: 252 seconds]
insanity54 has quit [Ping timeout: 272 seconds]
insanity54 has joined #ipfs
vonzipper has quit [Ping timeout: 252 seconds]
mappum has joined #ipfs
lgierth has joined #ipfs
headbite has joined #ipfs
<kyledrake>
whyrusleeping www Neocities uses a shared list of known spammers, it's pretty common (and not a new problem) http://www.stopforumspam.com
<whyrusleeping>
kyledrake: okay, i figured someones probably encountered the issue before
vonzipper has joined #ipfs
<kyledrake>
Things like DDoS are a huge problem with the HTTP web too, and that one's never really been solved either. People will try to mess with the network, it's just a thing that will happen.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<kyledrake>
Plenty of solutions available of varying complexity and strategy
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to pin: http://git.io/vTyhZ
<ipfsbot>
go-ipfs/pin faecf92 Jeromy: bump fsrepo version to 3
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]