silotis has quit [Remote host closed the connection]
<apiarian>
you could also use a sort of directory object at your IPNS root, with links pointing to whatever various things you'd like to track
silotis has joined #ipfs
zopsi has quit [Quit: About that...]
zopsi has joined #ipfs
Leonardoas26 has quit [Quit: Leonardoas26]
ckwaldon has joined #ipfs
<wes_>
apiarian: can you describe that in more detail? with instructions on actually doing it?
<ipfsbot>
[js-ipfs] diasdavid created greenkeeper-lodash.get-4.4.2 (+1 new commit): https://git.io/v6uya
<ipfsbot>
js-ipfs/greenkeeper-lodash.get-4.4.2 0cefbf4 greenkeeperio-bot: chore(package): update lodash.get to version 4.4.2...
<apiarian>
wes_: here's an example of the process: https://github.com/apiarian/go-ipgs/blob/master/ipgs/state/state.go#L319-L362 . you resolve the current IPNS entry. if there isn't one, you create a new empty object. then you take that object and you do a patch rm-link command on it to remove (if it exists) the current [your-namespace] link. then you add your latest [your-namespace] link to that object. pin it, publish it, and maybe unpin the old
<apiarian>
one if it is actually different
<apiarian>
so that's something like an `ipfs name resolve`, `ipfs object new` (if necessary), `ipfs object patch rm-link` (if [your-namespace] exists), `ipfs object patch add-link`, `ipfs name publish` (with some pins thrown in if you want)
<apiarian>
i'm pretty sure i read somewhere in some doc that this root-object-with-namespaced-subobjects is one of the possible ways to have more than one app share your node. think of it like dot-directories in your home
<ipfsbot>
[js-ipfs] diasdavid created greenkeeper-lodash.set-4.3.2 (+1 new commit): https://git.io/v6uy7
<ipfsbot>
js-ipfs/greenkeeper-lodash.set-4.3.2 6348967 greenkeeperio-bot: chore(package): update lodash.set to version 4.3.2...
<ipfsbot>
[js-ipfs] diasdavid created greenkeeper-lodash.sortby-4.7.0 (+1 new commit): https://git.io/v6uyd
<ipfsbot>
js-ipfs/greenkeeper-lodash.sortby-4.7.0 407083e greenkeeperio-bot: chore(package): update lodash.sortby to version 4.7.0...
<baffo32>
apiarian how does this process differ from publishing a directory of files?
<baffo32>
err a directory of directories
<apiarian>
i suppose ultimately it doesn't. it just isn't specifically tied to a filesystem
<apiarian>
with this process those namespace-root-objects might be "directories" in the filesystem sense. or it might be some fancy arbitrary data object. or a mix of both?
<wes_>
awesome help apiarian, thanks!
<apiarian>
so yes, this turns your IPNS entry into a directory of *
<ipfsbot>
[js-ipfs] diasdavid deleted greenkeeper-lodash.get-4.4.2 at 0cefbf4: https://git.io/v6uSC
zopsi has quit [Quit: Bailing]
<ipfsbot>
[js-ipfs] diasdavid deleted greenkeeper-lodash.set-4.3.2 at 6348967: https://git.io/v6uS6
ckwaldon1 has joined #ipfs
Foxcool has joined #ipfs
ckwaldon has quit [Ping timeout: 265 seconds]
ckwaldon1 is now known as ckwaldon
wes_ has quit [Ping timeout: 265 seconds]
JesseW has joined #ipfs
wes has joined #ipfs
wes is now known as Guest88518
Sonarpulse has joined #ipfs
doesntgolf has quit [Ping timeout: 276 seconds]
goose_ has quit [Quit: Page closed]
Guest88518 has quit [Read error: Connection reset by peer]
wes_ has joined #ipfs
j12t has quit []
gigq has quit [Ping timeout: 250 seconds]
gigq has joined #ipfs
vandemar has quit [Ping timeout: 250 seconds]
vandemar has joined #ipfs
pfrazee has joined #ipfs
russelsteapot has joined #ipfs
j12t has joined #ipfs
russelsteapot has quit [Ping timeout: 258 seconds]
chriscool has quit [Read error: Connection reset by peer]
doesntgolf has joined #ipfs
jaboja has joined #ipfs
Encrypt has quit [Quit: Quitte]
chriscool has joined #ipfs
wes_ has left #ipfs [#ipfs]
Foxcool has quit [Ping timeout: 258 seconds]
pfrazee has quit [Remote host closed the connection]
ralphtheninja has quit [Quit: leaving]
cemerick has joined #ipfs
G-Ray has joined #ipfs
<ipfsbot>
[webui] greenkeeperio-bot opened pull request #458: Update postcss-loader to version 0.10.0
ckwaldon1 has joined #ipfs
rendar has quit [Ping timeout: 258 seconds]
doesntgolf has quit [Ping timeout: 260 seconds]
ckwaldon has quit [Ping timeout: 264 seconds]
ckwaldon1 is now known as ckwaldon
ygrek has joined #ipfs
cemerick has quit [Ping timeout: 250 seconds]
apiarian has quit [Ping timeout: 265 seconds]
chris613 has joined #ipfs
apiarian has joined #ipfs
chriscool has quit [Read error: Connection reset by peer]
vbsd has quit [Ping timeout: 265 seconds]
reit has quit [Ping timeout: 258 seconds]
rendar has joined #ipfs
<ipfsbot>
[js-ipfs] diasdavid created greenkeeper-hapi-14.2.0 (+1 new commit): https://git.io/v6ud7
<ipfsbot>
js-ipfs/greenkeeper-hapi-14.2.0 822fbaf greenkeeperio-bot: chore(package): update hapi to version 14.2.0...
<ipfsbot>
[js-ipfs-api] dignifiedquire created greenkeeper-hapi-14.2.0 (+1 new commit): https://git.io/v6udF
<ipfsbot>
js-ipfs-api/greenkeeper-hapi-14.2.0 dc22a86 greenkeeperio-bot: chore(package): update hapi to version 14.2.0...
<tokyo_jesus>
The strangest part is that the file has a size, but when I try to execute it I'm told it doesn't exist
<cjd>
shared library
<cjd>
>>> ./gx --help
<cjd>
zsh: no such file or directory: ./gx
<cjd>
strace ./gx
anewuser has joined #ipfs
<cjd>
you're missing a shared library, when the loader is missing a shared library execve() fails with an ENOENT just to make people as confused as possible
<codehero>
lol
<tokyo_jesus>
ok. how do I fix that?
<Kubuxu>
tokyo_jesus: what is the result of: `ldd ./gx`?
<Kubuxu>
and can you show: uname -a
<codehero>
how the hell is it doing that and why? D:
<codehero>
i'm so confused right now
<codehero>
why is ipfs making me disconnect
<Kubuxu>
codehero: router with low conntracker size
<tokyo_jesus>
/run/current-system/sw/bin/ldd: line 117: ./gx: No such file or directory
<tokyo_jesus>
Linux home 4.4.16 #1-NixOS SMP Wed Jul 27 16:48:30 UTC 2016 x86_64 GNU/Linux
<codehero>
Kubuxu: hm. how would i check that?
<tokyo_jesus>
I can build gx manually, if that helps
<codehero>
tokyo_jesus: oh, you're using nixos
<codehero>
that makes things more complicated
<codehero>
like a lot
<Kubuxu>
unfortuanatelly we weren't successful in migrating to utp (the lib was buggy) so we still are using TCP which has nasty problem of overflowing conntracker.
<codehero>
tokyo_jesus: you'll need to write a nix expression to build it
<codehero>
or enter a nix shell at least
<Kubuxu>
nix == weird libc?
<tokyo_jesus>
I *can* build it
<codehero>
Kubuxu: it's a fancy package manager
<Kubuxu>
I know.
<codehero>
tokyo_jesus: oh well, yeah execute i mean
<cjd>
umm utp is not going to be better, NATs track "connections" even if they're just exchanges of udp packets
<codehero>
Kubuxu: oh, okay. well, libc isn't really visible to executables outside the nix stores iirc
dignifiedquire has quit [Quit: Connection closed for inactivity]
<Kubuxu>
cjd: utp would allow us to maintain "connected but not active" connections, which wouldn't pollute conntracker.
<codehero>
tokyo_jesus: you should ask this question in #nixos since it's definitely nixos specific
giodamelio has quit [Quit: Bye!]
<Kubuxu>
with TCP we can only have connected and active state which means they are in conntracker.
<cjd>
ahh you mean "connections" which never exchange a packet for a matter of a few minutes so that the NAT evicts them ?
<Kubuxu>
yup
<Kubuxu>
this is how many torrent clients do it.
<cjd>
makes sense
<cjd>
although your port numbers are gone after you're evicted from the nat so I'm not sure how you plan to reactivate that connection
<cjd>
I thought that torrent clients just kept only like 10-50 connections max, TCP or UDP
<Kubuxu>
we do punching either way so we have stable outside ip/port pair.
<cjd>
ok so that proves I don't understand how NATs work, if it gives you 1 port and you can reuse it for talking to n ips then that's super cool
Encrypt has quit [Quit: Sleeping time!]
LumberTech has joined #ipfs
zorglub27 has quit [Ping timeout: 240 seconds]
<Kubuxu>
most NATs don't track the sender address/port, so you can open connection to whomever, ask him for your external IP/port pair and reuse it.
<Kubuxu>
ok, I am off for today.
<codehero>
good night Kubuxu
<codehero>
so, i suppose the problem would go away if i could reduce the number of peers?