<kanzure>
also, you should have a merkle tree of all the hashes published on the site so that you only have to use one giant-ass merkle root
<kanzure>
using ipfs is not a substitute for good design
<kanzure>
(i have been on an opentimestamps timestamping spree.)
<keks>
kanzure: i suggest you ask the author why s/he did it that way and ask why not the other way. maybe they had a reason you didn't see yet. also hash archive doesn't have much to do with ipfs
<kanzure>
no, i think it's a good assumption for me to make that ipfs users are probably in here.
<kanzure>
i don't really care if he sees my messages in particular--- perhaps others will read my messages and consider the ideas for their own purposes in the future.
<keks>
you opened with the question who the operator of that page is so i figured you wanted to discuss ideas for improvement
<keks>
also there was a lengthy discussion in a github issue between btrask and jbenet somewhere about that particular topic...i'll look it up
<keks>
kanzure: both ways have their upsides and downsides. don't assume you are more clever, just talk to people about your ideas. maybe you learn something
<kanzure>
this issue thread is vomit. dags don't matter; only merkle trees.
<kanzure>
(in particular, calling it a dag is silly because it implies some sort of specific content to the preimages)
<kanzure>
*content in the preimages
<kanzure>
i'm a little confused why he would think that ipfs would be the right place to bring up these issues though
<kanzure>
the only reason i'm here is because i assumed he was an ipfs user, not that i thought he was trying to use ipfs in this way
<keks>
I don't see what you problem with the dag is
<keks>
kanzure: i guess he wanted to contribute to the design
<kanzure>
merkle trees are like dags except you don't make assumptions about the leaf nodes
<kanzure>
or any assumptions about the structures of the preimages
<whyrusleeping>
neither of those even mention the word tree though
tooldels has joined #ipfs
<kanzure>
note that i even consider exotic structures like merkle sum trees to be merkle trees
<kanzure>
or even some merkle mountain ranges
<kanzure>
in a merkle dag you can store arbitrary content, my last link seems to indicate this is invalid in ipfs
<kanzure>
well-- arbitrary hash content at least
<keks>
well there is people working on ipld, which is not trying to be a file system but a dag for application data
<whyrusleeping>
unixfs is the filesystem abstraction on top of our current protobuf merkledag, the base 'dag' layer isnt trying to be a filesystem at this point either
<keks>
and the dag (as in directed acyclic graph) is using hashes to refer to other nodes
<kanzure>
could one of you elaborate on why one of you says ipld and one of you says uniffs
<kanzure>
.. unixfs.
<whyrusleeping>
ipld is just the new way of describing and creating objects we're working on
<keks>
because ipld is for making dags and unixfs is a filesystem abstraction
<keks>
not the same
<whyrusleeping>
its going to be more flexible that having to fit inside a protobuf
<whyrusleeping>
^ also what keks said
<kanzure>
the leaf preimages in ipfs are explicitly only protobufs?
<whyrusleeping>
currently, the system doesnt enjoy it if you try to link to raw data.
<whyrusleeping>
but thats being fixed with ipld
<kanzure>
are these irc logs publicly archived and linkable?
<whyrusleeping>
yeap
<whyrusleeping>
its in the topic
* whyrusleeping
realizes how long the topic has gotten
<kanzure>
got it, thanks
<kanzure>
botbot.me has really poor anchorlinks on their default page.
<kanzure>
he is a little tricky to type-- he speaks quickly, but thankfully i type faster.
<whyrusleeping>
reminds me of the merkle mountain range idea
<whyrusleeping>
haha, hes gotten a bit better at speaking slower lately
sametsisartenep has joined #ipfs
reit has joined #ipfs
dignifiedquire has quit [Quit: Connection closed for inactivity]
ppham has quit [Ping timeout: 248 seconds]
lkcl has quit [Ping timeout: 264 seconds]
<A124>
Guys. where can I find how to assemble link objects?
dmr has joined #ipfs
ppham has joined #ipfs
wallacoloo has quit [Quit: wallacoloo]
reit has quit [Ping timeout: 252 seconds]
<apiarian>
what happens to a blockchain when its participants are separated by interplanetary distances? or when the sun gets between the mars and the earth nodes. or when some section of the earth becomes cut off from the rest? i.e. are blockchains compatible with the possibly not-fully-connected world of IPFS?
ygrek_ has quit [Ping timeout: 263 seconds]
<A124>
Fork?
<apiarian>
isn't the smaller blockchain fork destined to die?
ppham has quit [Ping timeout: 264 seconds]
ppham has joined #ipfs
kulelu88 has quit [Quit: Leaving]
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
ygrek_ has joined #ipfs
sametsisartenep has quit [Quit: zzzZZZ]
ppham has quit [Ping timeout: 244 seconds]
ppham has joined #ipfs
anandthakker has quit [Ping timeout: 244 seconds]
ygrek_ has quit [Ping timeout: 272 seconds]
ppham has quit [Remote host closed the connection]
gmcquillan__ has joined #ipfs
<A124>
Yeah, Im not a blockchain guy.
<A124>
But yes, that is what I was pointing to.
<A124>
So you would have to merge.
gmcquillan__ has quit [Ping timeout: 252 seconds]
ygrek has joined #ipfs
<whyrusleeping>
apiarian: you should watch vitaliks talk about sharding at devcon2
<A124>
whyrusleeping .. urgent. js-ipfs can connect to go-ipfs?
<whyrusleeping>
A124: correct, ish
<whyrusleeping>
i'm not sure how the js-side of things is going though
<A124>
To get content, no need to seed.
<whyrusleeping>
yeah, you should be able to do that
<A124>
We are on meeting and they are stuck on js-ipfs as the punchline.
<kanzure>
anyway, if you need interplanetary consensus then you need to decrease the block interval (or you can sprinkle magic sharding dust on the problem)
<kanzure>
the trouble is that if you use million light-year interplanetary consensu is that -- barring any rabbit-out-of-hat tricks from aubrey de grey and the sens foundation -- that most of the transaction participants will be long dead by the time that you have any useful number of confirmations (many hundreds of millions of years)
PrinceOfPeeves has quit [Quit: Leaving]
anewuser_ has joined #ipfs
anewuser has quit [Ping timeout: 248 seconds]
Oatmeal has quit [Ping timeout: 244 seconds]
ygrek_ has joined #ipfs
ygrek has quit [Ping timeout: 272 seconds]
<A124>
whyrusleeping and ipfs passed.
<A124>
If you can throw examples or stuff of how to fix the current js-ipfs we need to do that.
ppham has joined #ipfs
rsync1 has quit [Quit: Leaving]
tmg has joined #ipfs
matoro has quit [Ping timeout: 252 seconds]
mgue has quit [Quit: WeeChat 1.5]
ygrek_ has quit [Ping timeout: 248 seconds]
ralphtheninja has quit [Ping timeout: 265 seconds]
ralphtheninja has joined #ipfs
Oatmeal has joined #ipfs
mgue has joined #ipfs
anewuser_ has quit [Quit: anewuser_]
A124 has quit [Read error: Connection reset by peer]
anonymuse has joined #ipfs
A124 has joined #ipfs
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
tmg has quit [Ping timeout: 248 seconds]
anonymuse has quit [Remote host closed the connection]
Ragnis has quit [Ping timeout: 265 seconds]
ppham has quit [Ping timeout: 244 seconds]
ppham has joined #ipfs
ppham has quit []
ygrek_ has joined #ipfs
matoro has joined #ipfs
trn has joined #ipfs
wallacoloo has joined #ipfs
pfrazee has quit [Remote host closed the connection]
JesseW has quit [Ping timeout: 264 seconds]
tmg has joined #ipfs
<haad>
keks: oh nice, that'd be awesome (portin Orbit to Go)! I'm working on improving the documentation a lot this week, hopefully that'll clear the API and code too. short answer to your question: password is not used (remnants of old code).
<haad>
keks: I started porting the very core module of Orbit (ipfs-log) to Go but haven't had the time to finish it. would be awesome to have a port of that and orbit-db in Go, too.
kudaweb has joined #ipfs
pfrazee has joined #ipfs
pfrazee has quit [Ping timeout: 265 seconds]
ulrichard has joined #ipfs
ulrichard has quit [Ping timeout: 252 seconds]
rgrinberg has quit [Ping timeout: 252 seconds]
chriscool has joined #ipfs
trn has quit [K-Lined]
ulrichard has joined #ipfs
mildred has joined #ipfs
wallacoloo has quit [Quit: wallacoloo]
alu has quit [Quit: WeeChat 1.6-rc1]
rendar has joined #ipfs
pepesza has joined #ipfs
ylp1 has joined #ipfs
konubinix has joined #ipfs
zorglub27 has joined #ipfs
cwahlers has quit [Ping timeout: 265 seconds]
cwahlers has joined #ipfs
lacour has quit [Quit: Leaving]
pfrazee has joined #ipfs
G-Ray has joined #ipfs
pfrazee has quit [Ping timeout: 252 seconds]
Choups has joined #ipfs
Choups has left #ipfs [#ipfs]
<chriscool>
Hi everyone!
zorglub27 has quit [Quit: zorglub27]
corvinux has joined #ipfs
ianopolous has joined #ipfs
s_kunk has joined #ipfs
s_kunk has quit [Max SendQ exceeded]
ianopolous_ has quit [Ping timeout: 244 seconds]
s_kunk has joined #ipfs
lkcl has joined #ipfs
mgue has quit [Ping timeout: 244 seconds]
pepesza has quit [Quit: Terminated with extreme prejudice - dircproxy 1.0.5]
s_kunk has quit [Ping timeout: 272 seconds]
lkcl has quit [Ping timeout: 244 seconds]
mgue has joined #ipfs
kudaweb has quit [Quit: Page closed]
Kamilion has joined #ipfs
s_kunk has joined #ipfs
espadrine has joined #ipfs
dignifiedquire has joined #ipfs
pepesza has joined #ipfs
pfrazee has joined #ipfs
captain_morgan has quit [Ping timeout: 252 seconds]
pepesza has quit [Quit: ZNC 1.6.3+deb1 - http://znc.in]
pfrazee has quit [Ping timeout: 264 seconds]
pepesza has joined #ipfs
Kane` has quit [Ping timeout: 272 seconds]
Kane` has joined #ipfs
pepesza has quit [Quit: ZNC 1.6.3+deb1 - http://znc.in]
pepesza has joined #ipfs
<apiarian>
thanks whyrusleeping and kanzure
<r0kk3rz>
kanzure: videos which can be adequately replaced with text arent fully using the medium :P
ygrek_ has quit [Ping timeout: 248 seconds]
Kane` has quit [Ping timeout: 272 seconds]
Kane` has joined #ipfs
cemerick has joined #ipfs
reit has joined #ipfs
structuralist has joined #ipfs
herzmeister has joined #ipfs
xeed has joined #ipfs
pfrazee has joined #ipfs
pfrazee has quit [Ping timeout: 264 seconds]
tmg has quit [Quit: leaving]
structuralist has quit [Remote host closed the connection]
corvinux has quit [Ping timeout: 265 seconds]
structuralist has joined #ipfs
s_kunk has quit [Remote host closed the connection]
s_kunk has joined #ipfs
structuralist has quit [Remote host closed the connection]
structuralist has joined #ipfs
Oatmeal has quit [Ping timeout: 248 seconds]
Mateon2 has joined #ipfs
<Mateon2>
Hi, I want to ask a quick question about using proxies with IPFS. Currently, I'm behind a NAT that IPFS is unable to punch through, as well as a firewall that blocks all ports apart from few like FTP, HTTP/S SSH, etc.
<Mateon2>
I can SSH into a server of mine and forward the :4001 port as well as open a SOCKS proxy, but I don't know of anything in IPFS that would make it work over such a weird tunnel
Mateon3 has joined #ipfs
Mateon1 has quit [Ping timeout: 272 seconds]
Mateon3 is now known as Mateon1
PseudoNoob has joined #ipfs
anonymuse has joined #ipfs
mildred1 has joined #ipfs
mildred has quit [Ping timeout: 252 seconds]
PrinceOfPeeves has joined #ipfs
<Mateon2>
So, is it possible to use IPFS over SOCKS, and advertise a custom multiaddr? (of the server I reverse forward :4001 from)
cemerick has quit [Ping timeout: 265 seconds]
tomoaki_ has joined #ipfs
tomoaki_ has quit [Client Quit]
structuralist has quit [Remote host closed the connection]
anewuser has joined #ipfs
Boomerang has joined #ipfs
<Mateon2>
Okay, so it's not a quick question. I'll ask in /faq
cemerick has joined #ipfs
<Kubuxu>
Mateon2: I am in similar situation as you NAT behind PPPoE which IPFS is not yet able to punch through.
<daviddias>
I did, going to test it in my machine now :)
kaotisk has joined #ipfs
<chriscool>
great, tell me how it goes
<richardlitt>
daviddias: done :)
<richardlitt>
Thanks for mentioning that!
<dignifiedquire>
Kubuxu: are you doing the stream or should I start my setup?
<Kubuxu>
dignifiedquire: do it
<Kubuxu>
If I did it we would have to figure out how to download VOD from beam and transport it to YT and so on
<daviddias>
chriscool: left you a comment on the PR
mildred1 has quit [Ping timeout: 264 seconds]
<chriscool>
daviddias: yeah I am looking at it
anewuser has quit [Quit: anewuser]
<daviddias>
btw, going to add you to the JS team chriscool , so that you don't have to be working on a fork (makes it easy for collaboration in the team to have branches at the main repo)
<lgierth>
dignifiedquire: hahaha wer hats erfunden
<dignifiedquire>
:D :D
sixeth has joined #ipfs
Encrypt has joined #ipfs
<Guest22313>
When a file is requested from gateway.ipfs.io, is there any way for the file's repo to remain anonymous, e.g., hide the sending IP address?
<Guest22313>
anonymous P2P web...
<Mateon1>
Guest22313: IPFS will allow to host a node over Tor and similar software. Right now it's not anonymous in the slightest
Encrypt has quit [Client Quit]
matoro has joined #ipfs
cemerick has quit [Ping timeout: 264 seconds]
<ianopolous>
hey everyone
<ianopolous>
I'm trying to implement the add command for directories over the http api and just wanted to clarify a few things that aren't obvious from reading the js implementation..
<Kubuxu>
lgierth: what is the status of the rec gx update tooling?
<Kubuxu>
ianopolous: most of us are on a call right now so it will be hard to get a response right now
<ianopolous>
when a dir is added recursively, does it do a bunch of Multipart calls to add each directory, or somehow do the entire subtree with one call?
<ianopolous>
kubuxu: thanks I'll wait
Encrypt has joined #ipfs
<lgierth>
Kubuxu: you can use it by 22 utc
<Kubuxu>
sorry if it sounded pushing it.
<haad>
lgierth: can I search from pinbot's pin labels for a hash?
<Kubuxu>
ianopolous: whyrusleeping might be able to help you as he knows the golang code for that
<whyrusleeping>
ianopolous: we do a single http request and send a multipart file for each file in the directory we are adding
Guest22313 has quit [Quit: Page closed]
<ianopolous>
whyrusleeping: thanks, but what about sub directories, are they separate posts?
sixeth has quit [Ping timeout: 264 seconds]
<haad>
lgierth: excellent! that's an awesome data set :) found exactly what I was looking for
<haad>
pinbot's awesome
<whyrusleeping>
ianopolous: nope, just another multipart
<ianopolous>
whyrysleeping: ok cool. thanks!
<whyrusleeping>
ianopolous: if you want to see what any api request looks like, run `nc -l 6000` in some terminal, and then run your ipfs command with `--api=/ip4/127.0.0.1/tcp/6000`
<ianopolous>
whyrusleeping: excellent
palkeo has joined #ipfs
palkeo has joined #ipfs
palkeo has quit [Changing host]
<lgierth>
whyrusleeping: looks like mars was reboot by DO -- better check everything's up
<lgierth>
whyrusleeping: ah sorry no, it's gonna be rebooted on the 11th
<whyrusleeping>
lgierth: incorrect
<whyrusleeping>
WHAT
<whyrusleeping>
NO
<lgierth>
We have identified a potential issue with the physical machine that your Droplet is hosted on. In order to ensure the best possible performance, and prevent any unexpected disruption, we will be performing a reboot of the underlying physical machine. We expect downtime to your Droplet to be only a few moments, but have scheduled an extended monitoring period to ensure everything progresses smoothly.
<whyrusleeping>
nooooo
<whyrusleeping>
my uptime :(
<lgierth>
if they'd be as serious about uptime as you, they'd just hot-transfer the vm to another host
<A124>
Though in this case, what is running on that server? Isn't it gateway? API has different port. Plus there are some settings in ipfs config about CORS and permissions.
herzmeister has quit [Remote host closed the connection]
herzmeister has joined #ipfs
<A124>
So... gateways. What happens if you throw bunch of requests at it?
<A124>
In my experience it keeps at "connecting". Doing that on public gateway (more peers) it took longer to choke it up then on gateway with 70 peers.
<A124>
Wonder if it is by any means related to the suspected ipfs pin lock issue I experienced last time.
cketti has joined #ipfs
cketti has joined #ipfs
cketti has quit [Changing host]
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
captain_morgan has quit [Ping timeout: 264 seconds]
fleeky_ has joined #ipfs
fleeky__ has quit [Ping timeout: 252 seconds]
konubinix has quit [Ping timeout: 244 seconds]
<lgierth>
whyrusleeping: why does gx add every file individually to ipfs, instead of doing add?r=true ?
<whyrusleeping>
lgierth: so we can respect gitignore and gxignore rules
<lgierth>
mh and something about symlinks
rsync has quit [Quit: Leaving]
quotidien_afk_ has quit []
konubinix has joined #ipfs
<lgierth>
it just makes gx.PublishPackage() really slow, maybe it's faster to make a temp copy of the package with the .gitignore and .gxignore applied, then add -r that
<lgierth>
or have a --ignorefile option for ipfs add, and write a concatenated gitignore+ipfsignore+gxignore file for use with that option
<lgierth>
mmh.
anonymuse has joined #ipfs
herzmeister has quit [Quit: Leaving]
<lgierth>
published go-libp2p-peerstore @ QmYkwVGkwoPbMVQEbf6LonZg4SsCxGP3H7PBEtdNCNRyxD (took 1.446628862s)
herzmeister has joined #ipfs
<lgierth>
(i'm waiting for the go-ipfs publish to finish so that i can post it's duration :))
<lgierth>
published go-ipfs @ QmW1MeBUdz8oiw8MgBbkwt9cYvXjxw1S9259LgBxmFds7X (took 46.410494308s)
<whyrusleeping>
what? it never takes that long for me >.>
<whyrusleeping>
LouisJencka[m]: thats still true for now, we're going to be working on fixing that constraint fairly soon though
<kevina>
so what I want to talk about, is the "Server Side Adds" feature. Right now it is only enabled via a config setting.
<kevina>
If that is still not good enough I could make the user change a line in the code. But than I am not sure how to test the feature.
<LouisJencka[m]>
whyrusleeping: cool, thanks
<whyrusleeping>
kevina: the issue is that to make that feature secure enough, we have to really ensure that you cant change it via the api
<whyrusleeping>
and to do that, we will need a lot of thought put into doing that correctly (look at all the hackery we currently have around the private keys)
<kevina>
Maybe, for now I can get this setting by reading a file on the repo? Rather than using the config API...
<whyrusleeping>
Mostly I just don't think we have the bandwidth to think about this and work on this with the attention it needs to ensure that its done properly.
<whyrusleeping>
And that you shouldnt block filestore work waiting for that issue to be resolved
<kevina>
So would a setting by reading a file (rather than the API Server) be secure enough?
<kevina>
If not I will disable it a comment out the tests for it.
<whyrusleeping>
it might be secure enough, but randomly adding things to the repo is going to lead to things getting messy
<whyrusleeping>
i think we should just disable it for now