<whyrusleeping>
daviddias: why arent you writing code?
ulrichard has joined #ipfs
<daviddias>
whyrusleeping: why aren't you writing code?
<whyrusleeping>
i am, yo
<whyrusleeping>
daviddias: you should go get me a glass of water
<daviddias>
kumavis[m]: fixed it :)
<daviddias>
whyrusleeping: are you though? :P
kevina has joined #ipfs
cdw1 has joined #ipfs
chriscool has joined #ipfs
Sophrosyne has joined #ipfs
<daviddias>
my computer doesnt lock when i close it
<daviddias>
thats a bad thing
<daviddias>
i have roughly negative opsec
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tiago has quit [Ping timeout: 260 seconds]
Guest66666 has quit [Ping timeout: 244 seconds]
<daviddias>
apparently mac default value is to lock only after 5 minutes
ribasushi has quit [Ping timeout: 248 seconds]
Guest66666 has joined #ipfs
<daviddias>
new default value*
tiago has joined #ipfs
tiago has joined #ipfs
tiago has quit [Changing host]
ribasushi has joined #ipfs
gully-foyle has joined #ipfs
<haad>
morning o/
<gully-foyle>
o/
pfrazee has quit [Remote host closed the connection]
ribasushi has quit [Read error: Connection timed out]
<haad>
whyrusleeping: I see keystore, ooooh! :) ty
<haad>
finally ;)
locusf_ is now known as locusf
ribasushi has joined #ipfs
jchevalay has joined #ipfs
jchevalay has quit [Client Quit]
Guest94493[m] has joined #ipfs
ecloud_ is now known as ecloud
cycad has quit [Ping timeout: 245 seconds]
ylp1 has joined #ipfs
<ianopolous>
keystore!
Kane` has quit [Remote host closed the connection]
bukolay-net has joined #ipfs
s_kunk has quit [Ping timeout: 260 seconds]
rendar has quit [Quit: Leaving]
no-thing_ has joined #ipfs
<dignifiedquire>
good morning
robattila256 has quit [Ping timeout: 248 seconds]
maxlath has joined #ipfs
s_kunk has joined #ipfs
mikolalysenko has quit [Ping timeout: 244 seconds]
mikolalysenko has joined #ipfs
ygrek has quit [Ping timeout: 258 seconds]
wallacoloo___ has quit [Quit: wallacoloo___]
ZaZ has joined #ipfs
Guest62957 has joined #ipfs
jokoon has joined #ipfs
pcre has joined #ipfs
domanic has joined #ipfs
<pcre>
Does anyone know if there is a some p2p tool to make a comment on some url (as hash). Lets say I want to comment a news article without being censored - or i want to comment a youtube video without registering an google+ account. »firefox addon to read comments« ???
<pcre>
Can ipfs be used for this kind of application?
Kane` has joined #ipfs
maxlath has quit [Ping timeout: 268 seconds]
leob has joined #ipfs
<pcre>
Lets say the comment function is turned off on some news article. It would still be possible to do comments on by this p2p software.
<pcre>
The url is the hash to search for on DHT or tracker, then all comments would be displayed.
<pcre>
For privacy a tor proxy could be used.
<pcre>
Of course the comments would only be visable as long as the pc is online, or the nodes containing the search hash could sync each other comments so not all PC need to be online all the time.
<pcre>
Does that make any sense?
<r0kk3rz>
pcre: check out orbit, its an ipfs chat problem
<r0kk3rz>
*program
<pcre>
Looks interesting.
<pcre>
To be honest. I have not jet understood ipfs correctly.
<pcre>
Does a browser need to support ipfs ? So i can use the url like this ipfs://SOMEHASH/folder/file. Where folder/file is somewhere on my local harddrive.
<pcre>
So it can be used like a protocol http some day.?
bukolay-net has quit [K-Lined]
slothbag has joined #ipfs
frmendes has joined #ipfs
maxlath has joined #ipfs
pp7y7 has joined #ipfs
pp99s has quit [Ping timeout: 265 seconds]
<r0kk3rz>
pcre: it would help if they did, but theres other ways
<r0kk3rz>
for instance there is a JS client
<r0kk3rz>
theres also gateways, or you can run a local client which will expose it through your filesystem
screensaver has quit [Ping timeout: 250 seconds]
rendar has joined #ipfs
<sevcsik>
pcre: check out Steem, it's a blockchain-based, reddit-like network. Altough it doesn't have a concept of a content URL - the same URL can be posted multiple times, creating parallel threads, similarly to HN
brodo has joined #ipfs
<sevcsik>
I'm planning to build a Disqus-like comment box on top of it, but haven't got to the implementation yet
pcre has quit [Ping timeout: 260 seconds]
koalalorenzo has joined #ipfs
sametsisartenep has joined #ipfs
koalalorenzo has quit [Quit: This computer has gone to sleep]
bastianilso____ has joined #ipfs
cemerick has joined #ipfs
pcre has joined #ipfs
kenshyx has joined #ipfs
azdle has quit [Remote host closed the connection]
brodo has quit [Read error: Connection reset by peer]
koalalorenzo has quit [Quit: Leaving]
dignifiedquire has quit []
dignifiedquire has joined #ipfs
Akaibu has quit []
Akaibu has joined #ipfs
arpu has joined #ipfs
<pcre>
sevcsik: I have some problems with the blockchain. A node always has to get get the whole content. Ok it is very secure against manipulation, thats a plus.
reit has joined #ipfs
Sophrosyne has quit [Ping timeout: 260 seconds]
espadrine has joined #ipfs
<victorbjelkholm>
dignifiedquire, ah, perfect! I'll use that as reference for what I'm doing as well. I'll continue tomorrow, national holiday today
<victorbjelkholm>
dignifiedquire, btw, the instructions in awesome-docs endeavour are not correct anymore, there is a added build step and pointing to /lib with the theme you did now
jonnycrunch has joined #ipfs
<dignifiedquire>
victorbjelkholm: right, I hope to release the new aegir today or tomorrow then it will be much easier :)
dominic__ has joined #ipfs
<victorbjelkholm>
sweet :)
<dignifiedquire>
have fun on your holiday :)
jonnycrunch1 has joined #ipfs
domanic has quit [Ping timeout: 250 seconds]
<sevcsik>
pcre: true, it's not as decentralised as IPFS
<pcre>
IPFS is more like using bittorrent library, isn't it?
nunofmn has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<pcre>
Using some distributed hash table.
cehteh has joined #ipfs
nunofmn has joined #ipfs
ZaZ has quit [Read error: Connection reset by peer]
anonymuse has joined #ipfs
mrpoopyb1 has quit [Ping timeout: 246 seconds]
<Mateon1>
I've seen some people say that a blockchain somehow makes things less decentralized. While in some sense that's true, pretty much all commonly used blockchain implementations handle network splits badly, in my opinion the worse thing is that blockchain based applications become bloated relatively quickly. Bitcoin's blockchain is over 70 gigabytes at the moment, for example.
jonnycrunch1 has quit [Ping timeout: 244 seconds]
<sevcsik>
Mateon1: less decentralised than P2P, yes, can't argue with that. especially those blockchains which have a relatively small network - like the aforementioned Steem, where I'm pretty sure most of the nodes are run by the creators
mildred has joined #ipfs
mildred has quit [Client Quit]
mrpoopyb1 has joined #ipfs
<pcre>
Mateon1: "become bloated relatively quickly" <- thats the point.
<sevcsik>
about the size, I think small, application-specific blockchains would be more maintainable - the problem with Bitcoin and ETH is that they are used as a general platform
<sevcsik>
so they get bloated
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
cehteh has quit [Ping timeout: 258 seconds]
nunofmn has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Kubuxu>
I am still waiting on a blockchain where you are the one required to keep the proof that your coins are valid
<Kubuxu>
or you can delegate that to someone providing service like that
leob has quit [Quit: leaving]
cehteh has joined #ipfs
maxlath has quit [Quit: maxlath]
briarrose has quit [Quit: bye bye]
frmendes has quit [Ping timeout: 258 seconds]
briarrose has joined #ipfs
briarrose has joined #ipfs
briarrose has quit [Changing host]
nunofmn has joined #ipfs
<fleeky__>
speaking of blockchains , does anyone know of a basic digital contract where you use public/private keys just to validate identity ?
<fleeky__>
i have a scheme on how to do it myself but figured i would ask around also first
jonnycrunch1 has joined #ipfs
pfrazee has joined #ipfs
<r0kk3rz>
fleeky__: you dont even really need a contract, just a signed transaction with a payload
<fleeky__>
can you explain what you mean a bit more
<fleeky__>
not sure what you are referring to as payload
bwerthmann has joined #ipfs
<r0kk3rz>
depending on the blockchain, most of them have a data field on the transaction, where you can put in whatever you want
<fleeky__>
ah im not talking about blockchains
<fleeky__>
ive been thinking a lot lately of just simple digital contracts, that are standalone
<fleeky__>
more like a provable agreement between two entities
<fleeky__>
thats it
Yatekii_ is now known as Yatekii
<r0kk3rz>
smart contracts are usually run on blockchains
cemerick has quit [Ping timeout: 248 seconds]
<fleeky__>
im talking about smart contracts though
<r0kk3rz>
they're not much use otherwise
<haad>
dignifiedquire: I can't uglify orbit with ipfs :( getting "SyntaxError: Unexpected token: name (i) [app.js:11088,6]". works perfectly if I don't import ipfs module. what's the magic trick to make it work?
<fleeky__>
a digital contract is very usefull, being able to sign a contract with a public key is more secure then a signature for instance
<fleeky__>
the problem with smart contracts imo, is there is no real interface to "real" currency
<dignifiedquire>
haad: two options, one transpile ipfs using babel to es2015, two use uglify beta with support for es6
<fleeky__>
unless you use a trusted third party like coinbase for example
<dignifiedquire>
s/es2015/es5
<fleeky__>
you still need to build bridges to where current value is , and that in a way defeats the purpose imo
<fleeky__>
so why not just have cryptographically provable agreements?
<haad>
dignifiedquire: I'm trying to transpile it, this is the include: "include: /node_modules\/(hoek|qs|wreck|boom|promisify-es|logplease|ipfs.+|orbit.*|crdts)/"
<r0kk3rz>
fleeky__: you're conflating two very different ideas. its understandable though, 'smart contracts' arent contracts
<r0kk3rz>
they're code
<haad>
dignifiedquire: something's missing from that I believe
<daviddias>
haad: going to merge and release the go-ipfs dep
<haad>
dignifiedquire: and what's the uglify beta?
<haad>
daviddias: yay, thank you!
<daviddias>
super cool PR! :)
<daviddias>
no, thank you :D
<haad>
daviddias: shall we make official 0.4.5-pre1 or should we wait for 0.4.5-rc1 (it's very soon)?
<dignifiedquire>
haad: you can open the file and look inside at which module it's failing
<daviddias>
that feature of specifying the version in the package.json makes so much sense! :D
<dignifiedquire>
haad: this is the uglify install we are using to minify js-ipfs github:mishoo/UglifyJS2#harmony
<daviddias>
haad: if I understand correctly, it can continue 0.4.4-N to install 0.4.4 by default and then you can do go-ipfs: 0.4.5-pre1 in orbit, right?
<haad>
daviddias: yeah, that's the idea. although, because npm and semver, I suppose if someone includes it as "^0.4.4" their fresh npm install will pick up anything 0.4.5 :/
<haad>
frikkin semver
<haad>
daviddias: not sure if we should do a 0.4.5-pre1 as it's still a bit buggy and if we do, people will get the new version automatically, and... they might get pissed :)
<fleeky__>
r0kk3rz: i think its interesting that you cant really have a truly decentralized smart contract , namely a smart contract without a blockchain
<daviddias>
yeah, going to do 0.4.4 release
<fleeky__>
atleast i dont know of any smart contracts that work without a blockchain
<daviddias>
switching versions for people without them knowing can't bring anything good
ansuz has joined #ipfs
<ansuz>
hi
<r0kk3rz>
fleeky__: a centralised 'smart contract' is a computer program, we have those everywhere :P
<ansuz>
oh yea, programs are the worst
<fleeky__>
the issue for me is i run a business , and i would like to be able to do digital contracts, even smart contracts, but the problem is still how to distribute the money when they send it to you typically in the form of a check.
ashark has joined #ipfs
shizy has joined #ipfs
<fleeky__>
r0kk3rz: your missing the point
<haad>
daviddias: yeah I think holding off releasing pre 0.4.5 versions is prolly a good idea. let's start doing 0.4.5s when we have an rc.
<fleeky__>
all snarkyness aside.
<ansuz>
wait is ipfs doing smart contracts now?
<ansuz>
IPSC
<fleeky__>
no
<fleeky__>
sorry this is offtopic i can stop
<ansuz>
it's not bothering me
<ansuz>
but I'm new here
<fleeky__>
mostly was trying to clarify an idea to r0kk3rz
<fleeky__>
i was however thinking of storing 'digital signed contracts' on ipfs for agreements between people i hire
<fleeky__>
since it would be easy for them to view/sign the file potentially
<fleeky__>
that way
<haad>
dignifiedquire: Multiaddr is where uglify fails. do I need to include that separately for the loader?
<r0kk3rz>
fleeky__: digitalised signed legal contracts are not even nearly related to smart contracts, imho 'smart contract' is a misnomer
<dignifiedquire>
haad: yes because it doesn't start with ipfs
<haad>
dignifiedquire: but it's a dep of ipfs, not mine
<dignifiedquire>
that doesn't matter
<dignifiedquire>
webpack just looks at your node_modules folder
<dignifiedquire>
or rather the absolute path of thing
<dignifiedquire>
and because npm@3 flattens modules multiaddr will not have "ipfs" in its path
<haad>
dignifiedquire: yeah, this is not gonna work as a long term solution. we need to be able to have it so that people can just pick it up for webpack without having to understand which deps come with it. perhaps we can provide a template webpack config or smth?
<dignifiedquire>
not sure we can get that general enough
<haad>
daviddias: wait what, did you change the package name? :O
jonnycrunch1 has quit [Ping timeout: 250 seconds]
<haad>
dignifiedquire: as a user of the module, I shouldn't need to know which *ipfs'* dependencies I need to include, that's just too much work for the end user.
<fleeky__>
r0kk3rz: ok r0kk3rz let me try to continue this line of thought, what i am trying to discuss are two things, if anyone else has thought about interesting ways to do digitally signed agreements, and has anyone thought of truly standalone decentralized smart contracts
<fleeky__>
the main issue of a smart contract in a case like this, is how to do you handle the third party escrow
<daviddias>
Oh, now while updating npm-go-ipfs, I see that I understood the suggestion from Juan wrong
<daviddias>
going to bring it back
<dignifiedquire>
haad: that's not how the tooling works today I'm afraid
<daviddias>
sorry about that haad , ixing it
<fleeky__>
and this actually goes to the heart of 'normal' smart contracts as well
<fleeky__>
in terms of there issues
<fleeky__>
in that being they are still dependent on older institutions
<fleeky__>
and dont really solve a lot , probably in the end will create more centralization
aaaaaaaaa____ has quit []
ploop has quit [Ping timeout: 256 seconds]
<r0kk3rz>
fleeky__: combining IPFS and a blockchain helps things, you can store a digitally signed PDF on ipfs, and put the hash into a blockchain transaction as a public record of when the agreement took place
aaaaaaaaa____ has joined #ipfs
<haad>
dignifiedquire: whether or not they work that way today, we can't put that responsibility on user. we need a way to either make it automatic for the user or provide instructions. we have currently neither.
<dignifiedquire>
we do, we have examples in webpack and browserify setup without transpiling
<dignifiedquire>
we can add docs for using the uglify harmony branch
ploop has joined #ipfs
frmendes has joined #ipfs
<haad>
dignifiedquire: the readme doesn't mention anything about webpack, just browserify and even that says the ipfs code is ES5 (outdated?)
<dignifiedquire>
if you find a good way of documenting the setup to transpile it, please make a PR
jokoon has quit [Read error: Connection reset by peer]
bwerthmann has quit [Ping timeout: 268 seconds]
ploop has quit [Ping timeout: 248 seconds]
<haad>
dignifiedquire: we can't just add docs for using the uglify beta because it doesn't work out of the box. that example is not correct, it's missing bunch of modules that need to go through babel loaders, like multiaddr/peer-id/ipld.+/more.
<dignifiedquire>
haad: the example is fine, it's just not doing transpiliation
<dignifiedquire>
and no minificatin
<dignifiedquire>
adding uglify beta would make it work out of the box for minification as well
<dignifiedquire>
not everyone needs to do the whole babel dance anymore
<fleeky__>
r0kk3rz: ipfs + blockchain .. just use ethereum instead ?
<fleeky__>
also i dont need a blockchain to do what you described on ipfs right now
<fleeky__>
i could easily use ipns at this point to do that
<haad>
dignifiedquire: so, essentially, what you're saying is: js-ipfs doesn't support minification with webpack at the moment?
<pcre>
hey i managed to add a file. »ipfs add somefile.txt«
<pcre>
:)
<pcre>
how to use ls then?
<pcre>
»ipfs ls« ?
<pcre>
»ipfs ls /« returns Error: invalid ipfs ref path
<r0kk3rz>
fleeky__: sure, ethereum + swarm can do the same thing, or ipfs + ethereum or whatever. Without the transaction you dont get an immutable public record of the agreement, but in some cases that might not be needed
<fleeky__>
i would say the immutability of the public record rests between the willingness of the two parties to store the record
<fleeky__>
in this case
<fleeky__>
as in ipfs without a blockchain
<fleeky__>
also there are other ways to store things like time and date into a document like this, also because ipfs is immutable works to our advantage in the case of a contract
<gsf>
pcre: `ipfs pin ls` will list everything locally pinned (which includes files added)
<fleeky__>
if you really wanted to, you could have a third party e.g. a judge or some other mediator pin the contract
<fleeky__>
again the problem with blockchains and smart contracts is they dont solve the problem of how to get value A to value B to anyone unless there is some kind of bridge or other agreement made between yet another middle man ,, so in the end why even use it ?
<fleeky__>
unless the value is the cryptocurrency itself that is
domanic has joined #ipfs
<fleeky__>
maybe with the advent of the iot you could start to programmatically switch ownership but the real world is not completely digitized yet
<pcre>
so to print then i use »ipmfs cat Qusomehash«
dominic__ has quit [Ping timeout: 248 seconds]
<r0kk3rz>
fleeky__: a blockchain is a shared database, its useful for things that databases are useful for and thats about it
<fleeky__>
agreed
<haad>
daviddias: oops! that's definitely missing. will fix it.
<fleeky__>
i really wanted to believe in the goal of blockchains which is to further decentralize things like money and power but i worry that they can easily be subverted to do the exact opposite.. also in terms of code they seem like large cathedrals that are never built properly,, ive started to look for smaller more robust software ideas at this point
<gsf>
pcre: to output file contents, yes `ipfs cat <ipfs-path>`
ulrichard has quit [Remote host closed the connection]
<gsf>
pcre: run `ipfs` without a command to see the short list of commands with help text
<fleeky__>
i think blockchains would be great for an mmo or something though
ploop has joined #ipfs
<pcre>
gsf how come that when adding a file. The filename is lost or not shown with an ls? other see getting-started file
<fleeky__>
r0kk3rz: so the question in the end is, is there some way to do what blockchains do without the centralization of a blockchain ?
<fleeky__>
in the end that would also reduce the crazy overhead of blockchains
<r0kk3rz>
fleeky__: to achieve what exactly
<fleeky__>
or maybe you have mini blockchains
<fleeky__>
for instance you have blockchains between two parties
<fleeky__>
so a non global blockchain
<r0kk3rz>
yeah, of course you can have multiple chains, i always thought using the term 'the blockchain' so much would make people think its like highlander
<fleeky__>
to achieve more future proof privacy and less possibility for exploits , also creates a smaller target, blockchains are a big target that attracts a lot of attacks
<fleeky__>
haha
<fleeky__>
that would actually be kind of funny , highlander blockchains that somehow fight eachother to become the one true blockchain
<r0kk3rz>
fleeky__: look at the competition between bitcoin et al. :)
<fleeky__>
yeah i guess it is kind of a highlander scenario a bit..
<gsf>
pcre: add a directory or use the `-w` option to wrap files in a directory when adding to include filenames
<dignifiedquire>
haad: it does support it, we just don't provide step by step instructions currently
<daviddias>
haad: all good now :)
<pcre>
gsf: Thx for the help.
<gsf>
pcre: you're welcome!
<haad>
daviddias: cool
<daviddias>
thank you :)
<daviddias>
really like the whole feature set
toXel has quit [Remote host closed the connection]
<haad>
dignifiedquire: that's what I mean: if the user can't access the information needed to use it, then it doesn't support it. it's not the user's responsibility, but ours, to make sure we don't break the expected behaviour and provide information to the user how to use our modules.
toXel has joined #ipfs
<dignifiedquire>
if you want to go that route it was never possible...
<haad>
daviddias: with that release, we can now build for any platform on any platform (eg. build orbit's electron app for windows on a linux CI machine)
<dignifiedquire>
but just because docs are missing doesn't mean it's not possible
<daviddias>
\o/ awesome, pfrazee will like that too! :D
<dignifiedquire>
it just means that, docs are missing
<haad>
dignifiedquire: the docs are the key here. something was changed in that process because it used to work. now sometime after, trying to build same code, it doesn't work anymore. that's breaking behaviour and there's no info on how it has changed and how can I make it work again. if it's only about docs, then let's fix that asap!
SoreGums has quit []
SoreGums has joined #ipfs
rugu has joined #ipfs
rugu is now known as waha
apiarian has quit [Ping timeout: 265 seconds]
<waha>
hey anyone familiar with quantum resistant cryptography ??
<waha>
is ecdh using sha512 curve quantum resistant?
apiarian has joined #ipfs
pfrazee has quit [Remote host closed the connection]
ylp1 has quit [Remote host closed the connection]
ogd has quit []
ogd has joined #ipfs
Guest62957 has quit [K-Lined]
isothujone has joined #ipfs
pfrazee has joined #ipfs
f4_ has joined #ipfs
no-thing_ has quit [Remote host closed the connection]
nullstyle has quit []
nullstyle has joined #ipfs
maxlath has joined #ipfs
jedahan has joined #ipfs
jedahan_ has joined #ipfs
jonnycrunch1 has joined #ipfs
jedahan has quit [Ping timeout: 250 seconds]
galois_d_ has joined #ipfs
galois_dmz has quit [Ping timeout: 260 seconds]
flyingzumwalt has quit [Quit: Connection closed for inactivity]
waha has quit [Quit: ChatZilla 0.9.93 [Firefox 50.0.2/20161130094234]]
dominic__ has joined #ipfs
tjade273 has joined #ipfs
domanic has quit [Ping timeout: 258 seconds]
dominic__ has quit [Ping timeout: 250 seconds]
jonnycrunch1 has quit [Read error: Connection reset by peer]
Remramm has quit [Changing host]
Remramm has joined #ipfs
Remramm has joined #ipfs
cemerick has joined #ipfs
s_kunk has quit [Ping timeout: 260 seconds]
bastianilso____ has quit [Quit: bastianilso____]
jedahan_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ashark has quit [Ping timeout: 260 seconds]
robattila256 has joined #ipfs
galois_d_ has quit [Remote host closed the connection]
galois_dmz has joined #ipfs
maxlath has quit [Ping timeout: 250 seconds]
ashark has joined #ipfs
Encrypt_ has joined #ipfs
nunofmn has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
jedahan has joined #ipfs
tjade273 has quit [Ping timeout: 265 seconds]
ashark has quit [Read error: Connection reset by peer]
nunofmn has joined #ipfs
ashark has joined #ipfs
cubemonkey has joined #ipfs
maxlath has joined #ipfs
IlanGodik has joined #ipfs
maxlath has quit [Ping timeout: 250 seconds]
<daviddias>
haad: I want to help finish these pubsub tests
<daviddias>
which branch/fork/PR of js-ipfs-api should they be run against?
<daviddias>
and close the other one, if not used anymore
<haad>
yeah we can, just didn't do it yet
<haad>
daviddias: btw. good news: got orbit running on js-ipfs today.
<daviddias>
\o/ :D
<daviddias>
awesome
<daviddias>
with floodsub?
<haad>
daviddias: pubsub obviously isn't there yet so can't really chat with anyone but I can use it in "offline" mode :)
<daviddias>
roger
<daviddias>
my goal of today would be reviewing and finish interface-ipfs-core and js-ipfs-api floodsub (merged and released) and then help (hopefully still today too) to make that happen in js-ipfs
<haad>
daviddias: here https://github.com/ipfs/js-ipfs-api/pull/470. it's a bit fckd because it depends on the stream refactoring PR. see the indiviual commit for a cleaner view to what pubssub adds/changes.
<haad>
yeah will finish that first, easier to look at them individually
PseudoNoob has joined #ipfs
<haad>
tomorrow
<haad>
maybe
<haad>
:)
<haad>
let's see how everything else goes. if you make progress with pubsub in js-ipfs today, leave me a message somewhere if there's something to check.
<haad>
daviddias: probably the biggest help right now would be to get the interface-ipfs-core pubsub tests running with js-ipfs pubsub. they're not working yet and there's plenty of implementation work to be done to get the tests pass. js-ipfs-api is ok from my perspective atm, I have a working setup so that Orbit works, we're just missing basically official package releases. whereas browser version is not fully working yet (pubsub), so it'd be better to get that sorte
<daviddias>
yeah, that is the first step
<daviddias>
without tests passing in js-ipfs-api
<daviddias>
doing it in js-ipfs would be walking in the dark
<daviddias>
going to work on that
<haad>
daviddias: the tests pass in js-ipfs-api, that PR I just did
<daviddias>
the interface-ipfs-core tests/
<daviddias>
?
<daviddias>
testing it
<haad>
yeah
<daviddias>
sweet!
<daviddias>
so, js-ipfs-api doesn't have a go-ipfs version in its package.json
<daviddias>
shouldn't it be go-ipfs {version v0.4.5} somewhere?
<daviddias>
haad: can we get that custom ipfsd-ctl as a PR too?
<haad>
daviddias: the only difference is that the custom version downloads 0.4.5 and is namaspaced under my username so I can publish it in npm. this has the downside that it needs to be required as '@haad/js-ipfs-api'.
<daviddias>
I see
<daviddias>
another way is to do
<haad>
let me check those tests
<daviddias>
"ipfsd-ctl": "haad/js-ipfsd-ctl"
anonymuse has joined #ipfs
<daviddias>
and this way you don't even have to publish
<haad>
daviddias: yeah, didn't commit the tests to the PR :)
<pinbot>
now pinning /ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz
<pinbot>
[host 2] failed to pin /ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz: cannot store pin state: write /data/ipfs/datastore/381203.log: no space left on device
<pinbot>
now unpinning /ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz
<pinbot>
[host 2] failed to pin /ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz: cannot store pin state: write /data/ipfs/datastore/381203.log: no space left on device
<whyrusleeping>
nvm
<pinbot>
[host 6] failed to grab refs for /ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz: Post http://[fc29:9fda:3b73:c1d2:9302:31e3:964c:144c]:5001/api/v0/refs?arg=/ipfs/QmYRorrvxycsH8qzvVGt3EBrBNB12BoM65spkn2KMkHVNz&encoding=json&stream-channels=true&r=true&: dial tcp [fc29:9fda:3b73:c1d2:9302:31e3:964c:144c]:5001: getsockopt: connection timed out
<alu>
whyrusleeping: yeah I'm still looking for a ticket but im coming :)
<pcre>
single core
<whyrusleeping>
ah yeah... 1GB vps nodes arent happy on 0.4.4. The number of peers in the network has been steadily increasing, and i havent gotten connection closing figured out yet
<whyrusleeping>
0.4.5 should be much better on memory usage
<whyrusleeping>
and then 0.4.6 should have better connection management
<alu>
damn
<alu>
so I should upgrade now
<whyrusleeping>
i'm going to release another 0.4.5 prerelease likely today
<whyrusleeping>
and then cut a release candidate soon
<whyrusleeping>
i think thats part of the identify remote address reporting
<whyrusleeping>
arkadiy: the reason for it is that we use tcp reuseport
<whyrusleeping>
and you could potentially be dialing back on that
<whyrusleeping>
is it not possible for listening on port zero to use a port in that range?
<arkadiy>
ah so it keeps listening on an outgoing port
<whyrusleeping>
yeah, potentially
<whyrusleeping>
it might be the case that even though thats technically possible, it never happens in practice...
palkeo has joined #ipfs
<arkadiy>
whyrusleeping: I don't believe the OS will choose ports in that range
<whyrusleeping>
interesting... okay
<arkadiy>
ah hmm it seems implementation specific
<arkadiy>
so some OSes may actually explicitly give you an ephemeral range port, and others will give you a port in 1024-49151
<arkadiy>
hmm ok probably won't resolve this right now, will open an issue if we find it problematic in practice
gavinmcdermott has joined #ipfs
<whyrusleeping>
arkadiy: yeah, please do open an issue
<whyrusleeping>
better address discovery is really needed
arkadiy has quit [Ping timeout: 260 seconds]
<herzmeister>
when i publish new content on my home server, it takes forever until it's available via at gateway.ipfs.io... to an ipfs instance that runs on a different machine in my home network, but behind a vpn in a far away country, the content is propagated instantly. what is this sorcery?
bwerthma1n has quit [Ping timeout: 248 seconds]
<pcre>
stupid question. Is there a search function to find files on other nodes with ipfs? Similar to torrent search via DHT on bittorrent?
<pcre>
sorry to ask. Im new to ipfs
<pcre>
q[O-O]p
mildred has quit [Ping timeout: 256 seconds]
<whyrusleeping>
pcre: theres some early ipfs search engine sites
<whyrusleeping>
ipfs-search.com i think is one
<whyrusleeping>
herzmeister: connectivity is weird, lol
<whyrusleeping>
herzmeister: your node is probably difficult to connect to from the outside
<whyrusleeping>
so the delay youre noticing is the time it takes for those nodes to find and connect to you
<whyrusleeping>
the nodes youre seeing things propogate to instantly are ones youre likely already connected to
<whyrusleeping>
so that node wants X, it can quickly see that Y (you) has X, and then sees its already connected, makes the request, and boom!
<pcre>
Then search for this hash 01afcd77d88a1a2326a433a7d3fc3034dd953a03f575601ad9ce1efa9e5f0f6a @ http://ipfs-search.com/
<pcre>
could take some time until it is syncronised.
<ruby32_>
no results (yet)
<pcre>
could take some time.
<pcre>
DHT search.
<ruby32_>
commentit is probably a bad name
<pcre>
but for comments on an url not so urgent
<pcre>
yeah.
<pcre>
make a suggestion
<ruby32_>
commented.io is an available domain name
<ruby32_>
so commented.sh would work
<ruby32_>
almost sounds the same, too
<pcre>
probably a firefox extension would also be better than some bash script
<pcre>
commented sound better
bollo has quit [Ping timeout: 258 seconds]
<ruby32_>
i like naming things, it's fun
<ruby32_>
it's really hard though
cemerick has quit [Ping timeout: 260 seconds]
<pcre>
it should contain somethig with "no censorship" and "comment" or "news"
<pcre>
So people can make comments without mass media press censoring it.
<pcre>
another hash 043b301bb5de20b4b54d203c24f4f7e5f8046d1ec89e8730636b944e27b4b849 A comment i made an hour ago. It is available
<ruby32_>
i think a lot of people looking for solutions to censorship are short-sighted
<ruby32_>
let's say 99% of people move to ipfs or some such decentralized, uncensorable system
<ruby32_>
those who want to censor things will simply promote half-truths, untruths, or dissenters in that system
<ruby32_>
effectively the result is the same
chriscool has quit [Ping timeout: 244 seconds]
<ruby32_>
or, they will compromise trusted individuals/groups
<pcre>
Yes.
<ruby32_>
however, it is pretty hard to obfuscate some things
<ruby32_>
like video evidence for example
<ruby32_>
it might be easy to fake a video, but if something is corroborated from many different sources, it's not easy to deny that it's real
<ruby32_>
in that way, systems like this are a big advantage
bwerthmann has joined #ipfs
<pcre>
true.
Encrypt_ has quit [Quit: Quit]
<pcre>
Also im sick of createing hundert acounts just to comment things on different news sites.
mildred has joined #ipfs
<ruby32_>
that's because every news site off-loads the responsibility of identity verification onto the email provider
<pcre>
I guess it is also some data mining.
<deltab>
if it gets popular, it'll be spammed
<pcre>
deltab: yeah. Maybe some machine learning algorithm to sort things.
slothbag has joined #ipfs
maxlath has joined #ipfs
dmr has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
PseudoNoob has quit [Remote host closed the connection]
maxlath has quit [Ping timeout: 246 seconds]
_fil_ has joined #ipfs
maxlath has joined #ipfs
pfrazee has quit [Remote host closed the connection]
maxlath has quit [Ping timeout: 244 seconds]
<kevina>
whyrusleeping: you around?
<ianopolous>
whyrusleeping: re earlier comments about cbor, How can I get an object in cbor format? I've tried "ipfs object get --enc cbor QmVwJaX6m8U8phrwxb3faAYWaqtPPaMfdUVCsLuiHRTfzL" but that only supports protobuf, json and xml.
galois_dmz has quit []
maxlath has joined #ipfs
<whyrusleeping>
ianopolous: huh, good point
<whyrusleeping>
oh, wait
<whyrusleeping>
use the `ipfs dag get` command
galois_dmz has joined #ipfs
<whyrusleeping>
you can also use `ipfs block get` for the raw data
<whyrusleeping>
which will give you serialized CBOR
<whyrusleeping>
kevina: yeah, i'm trying to pin jbenet down to discuss it with him
<whyrusleeping>
i'll have a response today
<ianopolous>
whyrusleeping: fantastic, thank you
<kevina>
whyrusleeping: are you okay with going with the last two/digit characters?
<kevina>
whyrusleeping: at least for now
<kevina>
whyrusleeping: it unconventional, I know but I don't see how it can cause a problem
maxlath has quit [Quit: maxlath]
<ianopolous>
whyrusleeping: "ipfs dag get --enc cbor QmVwJ..." says no marshaller found for cbor (in 0.4.5-dev) and "ipfs block get --enc cbor QmdfT..." appears to return protobuf. Am I doing something wrong?
<whyrusleeping>
ianopolous: 'Qm' objects are all protobuf
<ianopolous>
ah it has to be a cid. ok
<whyrusleeping>
Yeah, if the object wasnt cbor in the first place, you cant pull it out as cbor
lkcl has quit [Ping timeout: 246 seconds]
spilotro has quit [Ping timeout: 268 seconds]
<ianopolous>
is it possible to create cbor objects using object patch addlink and setdata from a new object?
<ianopolous>
or does that always end up as protobuf?
<whyrusleeping>
no, we cant do that yet
<whyrusleeping>
daviddias was working on figuring out how that would work
<whyrusleeping>
right now you just write them (as json) and put them with `ipfs dag put`
<daviddias>
ianopolous: you won't be using object API