PseudoNoob has quit [Remote host closed the connection]
<pjz>
well, *someone* has to pin data, or if no one looks at it, it's lost forever
<pjz>
also, currently it's all single-pinned, but in the future I might add multi-pins: pinning by multiple servers in multiple parts of the world, which will help by being 1) more fault tolerant and 2) faster to serve the data
wa7son has joined #ipfs
bigs_ has joined #ipfs
Muis has joined #ipfs
voldyman has joined #ipfs
bret has joined #ipfs
Kane` has quit [Ping timeout: 258 seconds]
Kane` has joined #ipfs
<ansuz>
pjz++
<r0kk3rz>
pjz: ideally there would also be multiple providers
<pjz>
r0kk3rz: well, I'm not stopping anyone from setting up a competing service :)
<r0kk3rz>
yes exactly
<alterego>
Some sweat IPFS AI would be nice. :)
<ion>
Could it implement the smell somehow?
ntninja[m] has quit [Quit: Client limit exceeded: 5000]
lkcl has quit [Read error: Connection timed out]
lkcl has joined #ipfs
anonymuse has joined #ipfs
shizy has joined #ipfs
ashark has joined #ipfs
wlp1s1 has quit [Changing host]
wlp1s1 has joined #ipfs
wlp1s1 has quit [Changing host]
wlp1s1 has joined #ipfs
lkcl has quit [Read error: Connection reset by peer]
jedahan has joined #ipfs
ofdm has quit [Ping timeout: 245 seconds]
lkcl has joined #ipfs
ofdm has joined #ipfs
lkcl has quit [Ping timeout: 265 seconds]
bastianilso_____ has quit [Quit: bastianilso_____]
<richardlitt>
Do we have specific travis tests set up for Swift repositories? Should we have some?
vtomole has joined #ipfs
<vtomole>
Is that what filecoin is going to do? Earn currency for pinning other people's content?
<r0kk3rz>
vtomole: in a nutshell
<vtomole>
I can't find the github page for it, is it going to be developed open source?
AragonTheStrider has quit [Ping timeout: 260 seconds]
<frood>
there's a whitepaper at filecoin.io/filecoin.pdf but it's 2 years out of date, and can't work as written.
retrohacker has joined #ipfs
<r0kk3rz>
vtomole: honestly i expect someone to develop a filecoiny type thing before filecoin itself gets written
pfrazee has joined #ipfs
<vtomole>
I see.. Thank you.
<richardlitt>
Filecoin will be coming. We presented on it at the Ethereum DevCon in September
<richardlitt>
But it's not really related to IPFS, at the moment.
<richardlitt>
I would encourage you to read the white paper.
SuperPhly has joined #ipfs
SuperPhly has quit [Client Quit]
<vtomole>
I definitely will! After I finish my finals of course ;)
s_kunk has quit [Ping timeout: 265 seconds]
<alterego>
whyrusleeping: ping
<alterego>
Kubuxu: you around?
<Kubuxu>
partially
<alterego>
I patched the patch, so dialing actually starts listening rather than trying to connect to a remote service.
galois_d_ has quit [Remote host closed the connection]
anewuser has joined #ipfs
galois_dmz has joined #ipfs
<pjz>
Yeah, once filecoin shows up, pinbits.io becomes potentially a bit redundant. It's meant at least partly as a helpful stopgap to help get the IPFS ecosystem jumpstarted.
<pjz>
sitting here I heard several people say such a thing would be useful, so I built it.
Yatekii has quit [Remote host closed the connection]
nausea has joined #ipfs
nausea has joined #ipfs
nausea has quit [Changing host]
<whyrusleeping>
hrm?
Yatekii has joined #ipfs
<alterego>
whyrusleeping: the corenet conversation last night. Both ends require a service running that the ipfs daemon connects to and pipes data. This is not good, as you want the dialing end to act like a server that a client can connect to.
se3000 has joined #ipfs
anonymuse has joined #ipfs
<retrohacker>
@pjz that is clever
lkcl has joined #ipfs
vtomole has quit [Ping timeout: 260 seconds]
<pjz>
retrohacker: ? what is?
<richardlitt>
sprint-helper: next
<sprint-helper>
There are no more events today.
<richardlitt>
sprint-helper: help
<sprint-helper>
Correct usage: sprint-helper: announce <args> | next | now | tomorrow | help
<richardlitt>
whyrusleeping: I am going to tackle IPFS soon
<richardlitt>
probably tomorrow
<richardlitt>
This will include another round of Readme PRs
<richardlitt>
they should all be fairly easy
<richardlitt>
Things I am doing: checking all links, updating to standardize, checking GitHub descriptions, and making sure that any TODOs in any README have a corresponding issue open
<richardlitt>
I'll do one now so you can see how this looks.
<richardlitt>
It would be 100% more awesome if I can know _now_ if you have any extra stuff, like code coverage, that I should be adding
cemerick has quit [Ping timeout: 252 seconds]
<Kubuxu>
whyrusleeping: I am in favor in leaving the runner for coverage as it is, it is good to have one test done fast, and have the coverage come a bit later.
<richardlitt>
Yesterday I made a module that gets the language of a repo automatically, so I can tell if a repo is go or js related. If they are, I can search for different badges
<Kubuxu>
richardlitt: GH had module for that already, but in Perl or Ruby I think
<Kubuxu>
or do you ask the GH api?
<Kubuxu>
that is probably much easier :p
<richardlitt>
Kubuxu: I made a JS wrapper for Ruby
<Kubuxu>
whyrusleeping: in future when we are sure which CI we settle on then we can have stages, like: gofmt and go build, then if that is ok, go test with coverage in parallel with sharness with coverage, and coverage merges after that so results of tests don't have to wait for the coverage to go through
<whyrusleeping>
Kubuxu: yes, i'd love that
moshe1 has joined #ipfs
moshe1 has left #ipfs [#ipfs]
<alterego>
whyrusleeping, Kubuxu: I'm not trying to get manet to do udp
<alterego>
:)
<alterego>
Erm .. Now.
maxlath has joined #ipfs
Mateon2 has quit [Ping timeout: 240 seconds]
pbondoer has quit [Remote host closed the connection]
<richardlitt>
hsanjuan: you here?
<hsanjuan>
richardlitt: yes
<richardlitt>
hsanjuan: is there a reason you think travis works well cross all repos?
<hsanjuan>
richardlitt: ah, I mean, all repos I have seen have travis configured right?
Kingsquee has joined #ipfs
<hsanjuan>
so it does not hurt to have a default badge in a README template, as it is probable that travis will be used. But it's not a big deal anyway
<kumavis[m]1>
daviddias whyrusleeping jbenet ^ if we want a sponsor
Encrypt_ has quit [Quit: Sleeping time!]
Ronsor has quit [Ping timeout: 258 seconds]
Michael-Dot has quit []
<kumavis[m]1>
me and null_radix identified feb 6-14 as ideal seattle haxx0rs time
ylp has joined #ipfs
Ronsor has joined #ipfs
<jbenet>
nice
rlyshw has joined #ipfs
<rlyshw>
Hey. I was working with the python IPFS api to do some stuff, and maybe I'm misunderstanding how IPFS is supposed to work.
chris613 has joined #ipfs
<rlyshw>
My laptop could retrieve a file I had added to IPFS with a specific hash, but then my desktop could not access it at that same hash
<rlyshw>
it would just timeout. My question is, do files have to be published to be downloaded by different hosts over IPFS?
<Kubuxu>
was your laptop running when you tried to download it with your desktop?
<rlyshw>
Yeah, it was
<rlyshw>
the desktop would just wait forever and not actually retrieve the file
<Kubuxu>
were those in the same network?
<Kubuxu>
and which version of ipfs are you running?
<rlyshw>
I don't think they were on the same network
<rlyshw>
version 0.4.4
<rlyshw>
My original assumption was that when you add something to IPFS and it gives you the hash, you can just access that file via that hash from anywhere. Is that wrong?
<Kubuxu>
and you have to be able to connect to them
<Kubuxu>
s/host/have the file and be on the network
<rlyshw>
okay, is that the same thing as publishing the file?
<Kubuxu>
hmm?
maxlath has quit [Quit: maxlath]
<rlyshw>
The client, or the host with the original file, would have to run "ipfs name publish <file hash>"
<rlyshw>
Or that is unrelated?
slothbag has joined #ipfs
<rlyshw>
By "you have to be able to connect to them" do you mean a direct connection, IE on the same network?
gavinmcdermott has joined #ipfs
<Kubuxu>
it is totally unrelated
<Kubuxu>
we try to penetrate NATs but there are cases in which we fail then we can't connect
<Kubuxu>
you can try connecting those two peers by doing: `ipfs id` and looking for address with external IP and then using this whole address line to do `ipfs swarm connect $ADDRESS` on the other node
<rlyshw>
oh dope, I thought that my understanding of how IPFS was wrong but it looks like the NAT was the issue.
lkcl has joined #ipfs
Taek has quit [Ping timeout: 258 seconds]
ryantm____ has joined #ipfs
UukGoblin has quit [Ping timeout: 252 seconds]
<ryantm____>
Is there some way to encourage my file to get between my two peer computers? I did `ipfs add` on computer that currently has 139 swarm peers, and tried to request it with `ipfs cat` on another that has 189 swarm peers, and `ipfs cat` is just sitting there for many minutes.
<ryantm____>
or some way to debug it
cubemonkey has quit [Ping timeout: 256 seconds]
<rlyshw>
heh I just had the same issue. I think the recommended fix is by trying to force the hosts to recognize each other.
<rlyshw>
" you can try connecting those two peers by doing: `ipfs id` and looking for address with external IP and then using this whole address
<rlyshw>
line to do `ipfs swarm connect $ADDRESS` on the other node"
Taek has joined #ipfs
UukGoblin has joined #ipfs
UukGoblin has quit [Changing host]
UukGoblin has joined #ipfs
<rlyshw>
will published .md files automatically be rendered into HTML by ipfs.io/ipfs links?
<ryantm____>
rlyshw: when I try that from either computer, I get "dial attempt failed" .
<ryantm____>
both machines are using ipfs version 0.4.4
<Kubuxu>
rlyshw: no but I think there is a IPFS website that renders them
ylp has quit [Ping timeout: 260 seconds]
<Kubuxu>
let me find it
<rlyshw>
okay, cool. I was going to write up a guide to solve the issue me and ryantm____ are having.
<rlyshw>
ryantm____: are both machines on the same network, and can communicate with each other?
<rlyshw>
if one is behind a NAT you'll have trouble getting them to see one-another.
<ryantm____>
rlyshw: They are on different WANs.
ylp has joined #ipfs
<ryantm____>
they are both behind different NATs
<rlyshw>
Ah, that is likely your issue. It seems like IPFS can have trouble penetrating the NAT.
<Kubuxu>
I can't find it now, sorry, the approach was easy you had IPFS hash of a HTML that was fetching and rendering markdown that was specified after the fragment in URL
<ryantm____>
Both of them have plenty of peers, but I guess that doesn't help them tunnel through the NAT?
<Kubuxu>
so something like `localhost:8080/ipfs/Qm...AAAA$/ipfs/Qm...BBB