<nicolagreco>
can someone help me set up my ipns & dns?
<nicolagreco>
or pointers to online tutorial if any
captain_morgan has quit [Ping timeout: 240 seconds]
<nicolagreco>
I am planning to write a blog post on how to do so
<achin>
you can probably look at em32.net or ipfs.io as two examples
<achin>
create a TXT record with "dnslink=/ipfs/QmHash" or "dnslink=/ipns/QmHash"
<nicolagreco>
and how to I get the /ipns/nicola.io ?
<nicolagreco>
s/to/do
<achin>
with a TXT record
<achin>
(do you know what dns TXT records are?)
<nicolagreco>
yes
<nicolagreco>
sorry, I think I misunderstood the ipns then
<nicolagreco>
I thought I had to have a mutable hash
<achin>
if you have dnslink=/ipns/QmHash then the QmHash needs to be a peerid
<nicolagreco>
ok this is probably what I meant
<nicolagreco>
that is my default peerid or shall I create a new one?
<achin>
that's up to you
<achin>
most people i think will only be running 1 node
fingertoe has joined #ipfs
<achin>
i think i'm missing something obivous, but: if i'm loading a page via /ipfs/QmHash, how can i an an ajax call to /api/v0/object/get.... doing a simple jQuery.getJSON() just returns a 403 Forbidden
qgnox has quit [Quit: WeeChat 1.3]
patcon has quit [Ping timeout: 246 seconds]
<nicolagreco>
maybe they did not allow CORS
<achin>
i see a "csrftoken" cookie being set, but i'm not sure what to do with it
hoboprimate has quit [Quit: hoboprimate]
sonatagreen has joined #ipfs
voxelot has joined #ipfs
voxelot has joined #ipfs
<deltab>
nicolagreco: I think at the moment it makes more sense to use dns pointing to an ipfs hash, not an ipns peerid
<deltab>
you can set the dns record's ttl, update it easily, and you can have more than one
Guest73396 has joined #ipfs
<sonatagreen>
https://ipfs.io/ipns/sonatagreen.com/ works fine, but I can't for the life of me figure out how to resolve it on the command line -- if I want to find the ipfs hash /or/ the ipns node id, I have to do 'ipfs name resolve' (with no further arguments) on the machine running that node.
Guest73396 has quit [Ping timeout: 246 seconds]
<achin>
dig -t TXT sonatagreen.com.
<sonatagreen>
hah, thanks
mungojelly has joined #ipfs
Oatmeal has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] tilgovi opened pull request #1910: Leave overflow logic to go-msgio (master...cleanup-secio-reader-overflow) http://git.io/vWbHa
simonv3 has quit [Quit: Connection closed for inactivity]
captain_morgan has joined #ipfs
Oatmeal has joined #ipfs
tilgovi has quit [Ping timeout: 264 seconds]
Oatmeal has quit [Max SendQ exceeded]
rongladney has quit [Quit: Connection closed for inactivity]
ygrek_ has quit [Ping timeout: 260 seconds]
<ipfsbot>
[go-ipfs] whyrusleeping created fix/api-content-type (+1 new commit): http://git.io/vWb5m
<ipfsbot>
go-ipfs/fix/api-content-type 82e5b49 Jeromy: content type on command responses default to text...
Oatmeal has joined #ipfs
Oatmeal has quit [Ping timeout: 240 seconds]
captain_morgan has quit [Ping timeout: 264 seconds]
sseagull has quit [Quit: Lost terminal]
amstocker_ has joined #ipfs
captain_morgan has joined #ipfs
reit has quit [Quit: Leaving]
<spikebike>
victorbjelkholm: nice ipfsbin is yours?
Oatmeal has joined #ipfs
Oatmeal has quit [Ping timeout: 240 seconds]
reit has joined #ipfs
Oatmeal has joined #ipfs
Oatmeal has quit [Max SendQ exceeded]
hellertime has quit [Quit: Leaving.]
captain_morgan has quit [Read error: No route to host]
<nicolagreco>
I wonder how one could do a log database on top of ipfs
<nicolagreco>
so to give you an example, I want to have all the updates of my log change by change
<nicolagreco>
so that if I roll back to 6 months ago I could see how it was - essentially git
<spikebike>
yeah, not sure if IPFS exposes merkledag widgets at all to let you do the git like commits
<ion>
IPFS will definitely have commit objects in the future.
edrex has quit [Read error: Connection reset by peer]
edrex has joined #ipfs
<nicolagreco>
is there a way for me to do `ipfs add . --exclude .git` ?
cvik has quit [Ping timeout: 240 seconds]
<ion>
Does ipfs add have a parameter to take a tarball as input? (I don't remember.) If it does, you could tar c the directory with the exclusion and pipe that to ipfs add.
<nicolagreco>
uhm
<nicolagreco>
since `ipfs add -w a b` won't work
<nicolagreco>
I want to have the same base address for a and b
Guest73396 has joined #ipfs
<davidar>
nicolagreco (IRC): hidden files are excluded by default
<nicolagreco>
davidar, well I did `ipfs add -r -w .`
<nicolagreco>
and it added all the files
<davidar>
Including .git?
cvik has joined #ipfs
<nicolagreco>
hold on a second, I stumbled upon something I should not have seen
<nicolagreco>
I had a node http-server opened on my mac
<nicolagreco>
and I just checked the logs
<nicolagreco>
and I found out that for a week there was someone making far too many request
<AlexPGP>
Hmmm. Just tried embedding that link on a page at neocities. Am seeing 'Waiting for gateway.ipfs.io' on a machine other than the one being used as a peer.
<AlexPGP>
And 'Connected to gateway.ipfs.io' on the peer.
<AlexPGP>
No image, though.
<mappum>
AlexPGP: you shouldn't have to publish a name for the file unless you want to be able to change that link later on ipfs (e.g. to point to a newer version of the image)
<mappum>
application should be just an optional label, so ipfs is good. both for protocol
<AlexPGP>
I figure IP address is the address of my peer machine.
<AlexPGP>
source net, maybe 0.0.0.0?
<mappum>
IP address should be your LAN address (my psychic powers tell me it should be 192.168.10.98)
<mappum>
yeah, 0.0.0.0 should accept connections from everywhere
<AlexPGP>
can you tell me how your psychic powers work? (Where is this info being leaked?) :^)
<AlexPGP>
Ok. Done. Let me try this again.
<mappum>
i figured out your node's peer id by doing `ipfs dht findprovs QmRBiU3wJwrWUJKsueHT3GjWcvX9sfFhawNzTW2hydWfTL`, and since you were the only person with that file there was only one provider
<mappum>
then, i did `ipfs dht findpeer QmbBNFAbS8SkExTVjfzsFzy2hyMpjjAFEHUX2hNPUJYzDQ` to find your node, and it returned all the addresses to try to reach you on, including your LAN address
<AlexPGP>
Aha, Thanks for the education. :^)
<mappum>
hm, still can't ping your node
<AlexPGP>
It helps if I *enable* the port forwarding entry. (Done.)
<AlexPGP>
Any joy?
chriscool has quit [Read error: No route to host]
<mappum>
no :(
chriscool has joined #ipfs
<AlexPGP>
I should not have to reboot the router, or should I?
<mappum>
i don't believe so, but i'm not sure what else to try now
chriscool has quit [Client Quit]
<AlexPGP>
Might my antivirus/internet protection be the culprit?
<AlexPGP>
Kaspersky is shut down.
<mappum>
i don't think so. what version of ipfs are you on? `ipfs version`
<AlexPGP>
0.3.9-dev
<AlexPGP>
Yeah, that didn't do anything. Shields back up.
chriscool has joined #ipfs
<AlexPGP>
mappum: Thanks for your efforts. I've got to finish off some other stuff (work) before hitting ths sack. I'll return to this issue later.
<mappum>
AlexPGP: sorry i couldn't solve it for you :(
<AlexPGP>
Can't win 'em all.
chriscool has quit [Read error: No route to host]
<mappum>
i'll let you know if i figure out what it could be
<AlexPGP>
'k. Thanks.!
chriscool has joined #ipfs
AlexPGP has quit [Remote host closed the connection]
chriscool has quit [Ping timeout: 256 seconds]
chriscool has joined #ipfs
voxelot has quit [Ping timeout: 240 seconds]
jatb has quit [Quit: oops]
jatb has joined #ipfs
<nicolagreco>
I am not sure I am doing something wrong
<nicolagreco>
but I can't get gandi.net to allow me set up my dns with
<amstocker_>
its possible now to run a simple webapp that uses ipfs as a database, but a few people are investigating way to make larger distributed webapps on ipfs
<victorbjelkholm>
Good morning!
edrex has joined #ipfs
domanic has joined #ipfs
Syun has quit [Killed (Sigyn (Spam is off topic on freenode.))]
Johnnie has joined #ipfs
svetter has quit [Remote host closed the connection]
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to master: http://git.io/vWNuf
<ipfsbot>
go-ipfs/master 3c8c729 Juan Benet: Merge pull request #1875 from ipfs/fix/windows-builds...
r1k0 has quit [Remote host closed the connection]
r1k0 has joined #ipfs
svetter has joined #ipfs
NeoTeo has quit [Quit: ZZZzzz…]
mildred has joined #ipfs
patcon has joined #ipfs
patcon has quit [Ping timeout: 250 seconds]
warptangent has quit [Quit: bah bah quit quit]
warptangent has joined #ipfs
amstocker_ has quit [Ping timeout: 246 seconds]
domanic has quit [Ping timeout: 272 seconds]
stoopkid has left #ipfs [#ipfs]
stoopkid has joined #ipfs
Guest73396 has joined #ipfs
border0464 has quit [Ping timeout: 252 seconds]
<stoopkid>
ok so, if i want to deliver a wiki-page off of IPFS to a user's browser, this is no big deal. now let's say that user wants to make an update to the page. let's take two alternatives a) where the user is running a node, b) where the user is not running a node and is solely in the browser
<stoopkid>
what mechanisms are available to send this update out into the network?
<stoopkid>
i was reading the links amstocker pasted but didn't get a very clear picture from this
border0464 has joined #ipfs
koo7 has joined #ipfs
<stoopkid>
or, let's say for simplicity's sake that i'm the only node that's hosting, everybody's just retrieving the wiki pages from my node, and if they can send me an update, i'll be a good CA and make sure to update the content appropriately, so now it's changed for the world to see next time they get the page from me
<stoopkid>
so, no concerns about maintaining consistency/consensus across a decentralized/distributed network yet
<stoopkid>
just, how does the user send me something
<victorbjelkholm>
stoopkid, in the future, the user could be running a node directly in the browser, hence, publishing his new version and you'll just fetch it via IPFS
<victorbjelkholm>
you could also have a centralized service in between, in the mean time, for the users to be able to send you things. Once your service disappear, the content will still be avaible, but someone else have to start acting as a CA
<stoopkid>
victorbjelkholm: ah, thank you. so, in that first version, with the user running a node in the browser, how would my node be alerted of the update in order to fetch?
rendar has joined #ipfs
lgarron has joined #ipfs
<victorbjelkholm>
stoopkid, I'm not saying this is the best or only way but in my tests with having users discovering each other, I've added a directory with directory/id file, that contains 'ipfs-chat' for example, so I know the hash. Then I poll the DHT with findprovs to find nodes with that thing, and then I know they are using the service I'm building
<victorbjelkholm>
probably exists better ways of doing this though
<stoopkid>
i see, thanks, i'll see if i can get that working
<mungojelly>
wow, reusing providing hashes as a way to provide metadata about a node, that's a cool idea, what hash do i share to say i think that's a cool idea :D
<victorbjelkholm>
mungojelly, hah, thanks I guess!
<victorbjelkholm>
but then once I hit the streaming json in findprovs, that was invalid streaming json, I kind of had to stop
<victorbjelkholm>
so... Probably should create a POC about that, would be cool
<mungojelly>
QmUzp857Cmeu8kYFSh3mzTx1PQ76LCXESeNoJRH72RKv98 is the hash of "i think using hashes as node metadata is a cool idea!" so then (theoretically) anyone sharing that thinks that's a cool idea, perhaps
lgarron has left #ipfs [#ipfs]
zz_r04r is now known as r04r
s_kunk has quit [Ping timeout: 244 seconds]
ygrek_ has quit [Ping timeout: 250 seconds]
dignifiedquire has joined #ipfs
dignifiedquire_ has joined #ipfs
Guest73396 has quit [Ping timeout: 246 seconds]
mungojelly has quit [Remote host closed the connection]
s_kunk has joined #ipfs
s_kunk has quit [Changing host]
s_kunk has joined #ipfs
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/api-content-type from 82e5b49 to 2201cfd: http://git.io/vWNdh
<ipfsbot>
go-ipfs/fix/api-content-type 2201cfd Jeromy: content type on command responses default to text...
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/api-content-type from 2201cfd to b994038: http://git.io/vWNdh
<ipfsbot>
go-ipfs/fix/api-content-type b994038 Jeromy: content type on command responses default to text...
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1911: content type on command responses default to text (master...fix/api-content-type) http://git.io/vWNbY
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/api-content-type from b994038 to bb9493a: http://git.io/vWNdh
<ipfsbot>
go-ipfs/fix/api-content-type bb9493a Jeromy: content type on command responses default to text...
<whyrusleeping>
whoa, people are using that service discovery method? awesome :D
NeoTeo has joined #ipfs
lgarron1 has joined #ipfs
bitemyapp has quit [Ping timeout: 255 seconds]
bitemyapp has joined #ipfs
okket_ has quit [Remote host closed the connection]
<stoopkid>
ok, let's say a user makes an update, and hosts the "i want to push an update" file who's hash is <update-hash>, so i can find them using 'dht findprovs <update-hash>', and then they can set their peer-id to resolve to a file which will tell me what they wanted to push, and i can fetch that from them
okket has joined #ipfs
<stoopkid>
so, now the user will need to be informed when i've successfully fetched their update and updated the wiki
<stoopkid>
so that they can remove the "i want to push an update" file (because they're friendly like that)
<stoopkid>
it would be preferable if that would happen automatically, but i don't see how that would work, so it seems the users would have to check to see if i actually fetched and used their update
<ipfsbot>
[go-ipfs] jbenet closed pull request #1911: content type on command responses default to text (master...fix/api-content-type) http://git.io/vWNbY
martinkl_ has joined #ipfs
martinkl_ has quit [Client Quit]
martinkl_ has joined #ipfs
mungojelly has joined #ipfs
<cryptix>
gmorning
<dignifiedquire>
good morning
<mungojelly>
mornin
<victorbjelkholm>
whyrusleeping, yeah! findprovs in the js-ipfs-api (yeah, feels good to say js instead of node) currently have some issues, but other than that it's a fiiinee method for finding users using a service
<cryptix>
whyrusleeping: can we start splitting out packages out of the main repo after 0.4? i dont want to vendor all of go-ipfs to use path.Parse()
Whispery has joined #ipfs
TheWhisper has quit [Ping timeout: 246 seconds]
felixn has quit [Ping timeout: 265 seconds]
felixn has joined #ipfs
tilgovi has joined #ipfs
lgarron1 has quit [Quit: Leaving.]
edcryptickiller has joined #ipfs
edcryptickiller has left #ipfs [#ipfs]
snormett has joined #ipfs
hellertime has joined #ipfs
Guest73396 has joined #ipfs
tilgovi has quit [Ping timeout: 260 seconds]
snormett has quit [Quit: leaving]
<cryptix>
gawd why is this vendoring thing such a _MESS_
gamemanj has joined #ipfs
<ipfsbot>
[go-ipfs] cryptix force-pushed fix/ThatPut from cfc0824 to 0b1572a: http://git.io/vW84K
<ipfsbot>
go-ipfs/fix/ThatPut 0b1572a Henry: fixing putHandler for --writable http gateway...
dinamic has quit [Ping timeout: 272 seconds]
fingertoe has joined #ipfs
koo7 has quit [Ping timeout: 250 seconds]
AndroUser has joined #ipfs
AndroUser has quit [Read error: Connection reset by peer]
AndroUser has quit [Read error: Connection reset by peer]
<achin>
hi there. one of the github repositories is the best place to make new proposals for ipfs-related ideas. maybe the "notes" repo. this allows others to find/search/read/comment on them
<cryptix>
mostly want to look in ipfs/notes and ipfs/ipfs but /archives and /apps also had some amazing discussions
<achin>
cryptix: neat, i didn't know you could search an entire orginization like that
<cryptix>
daviddias showed me after my 30th-ish rant about not finding stuff i knew was there ^^
<achin>
:D
<test982726>
:)
stoopkid has quit [Ping timeout: 240 seconds]
scobiehague has joined #ipfs
<scobiehague>
hi. Would IPFS be vulnerable to someone polluting it with fake files?
<Stskeeps>
scobiehague: files are retrieved through asking for hashes
<Stskeeps>
so it's limited how fake they can be, since you know the hash of the right one you wanted .. :)
<Guest73396>
Yes, but what about all the storage spam? If someone just keeps uploading big spam files
<wscott>
@scobiehague nodes only get what they request
<cryptix>
scobiehague: also, content isnt distributed until other nodes request transfers
<scobiehague>
right. so only by demand could they stay around I suppose.
<scobiehague>
I wonder if someone could generate artificial demand though. Like have access to a few different boxes around the next and generate useless fakes
<scobiehague>
around the *net, I meant to say
<gamemanj>
Then the intermediate nodes might have your data in their caches, once. But:
<gamemanj>
If you generate many pieces of data, then there won't be much demand for an individual.
<victorbjelkholm>
scobiehague, well, if the rogue agent put fake files on his own machines, no one else is gonna get hurt by that
<scobiehague>
yes, true
<gamemanj>
Assuming the caching algorithm kept demand in mind, it wouldn't really matter.
<victorbjelkholm>
What is your definition of fake files by the way?
<gamemanj>
If you reused the same piece of data, then caching would occur at the nodes directly "next" to your nodes in the graph.
<scobiehague>
I meant in geographically different locations, but as gamemanj says, they would just stay in intermediate nodes for a while
<scobiehague>
victorbjelkhoml: just randomly generated files of different sizes
<gamemanj>
And also, another thing: In order to retrieve all this data...
<scobiehague>
filled with garbage
<fingertoe>
I assume other than the lookup, IPFS uses traditional file transfer technologies? So it really isn't any different than any other internet traffic?
cemerick has joined #ipfs
<scobiehague>
if anyone remembers DC++, you may remember they had problems with "share fakers"
<fingertoe>
You aren't routing over DHT or anything? Just straight IP to IP transfer, correct?
<scobiehague>
is there a way to profile encrypted files and say that they are probably a file of a certain type? Or file A is similar to file B?
<scobiehague>
might not be the right place to ask, but it leads me to some other thoughts involving IPFS
<gamemanj>
File size, maybe. In theory. But that won't really help, since that's on the "entire file" level...
<gamemanj>
stuff gets exchanged at the "block of data" level - a big file would be broken into chunks on the unixfs layer, if the words "chunking algorithum" mean what I think they mean.
<scobiehague>
I was thinking if it was possible to implement a deniable encryption over IPFS you see
<scobiehague>
generate randonly sized garabage files, and encrypt each individually with throwaway passphrases, using a simple script
<Stskeeps>
scobiehague: it's pretty easy to figure out who's participating in the swarm around a ipfs file though
<scobiehague>
the idea I had was of a "needle in a haystack" kind of effect
<scobiehague>
say you create an encrypted virtual machine image, but hide it with a load of garbage encrypted files
<scobiehague>
so if your computer gets seized, the agent with ill intent does not know where to look for your files
<scobiehague>
but distribute it over IPFS, so they couldn't use coldboot profiling and look at atimes and such
<scobiehague>
to see which file (e.g., the virtual machine) was being accessed the most
<achin>
recall that ipfs files are cached locally in ~/.ipfs/blocks
<achin>
unless you scrubbed those, they would give hints about what IPFS files you had been looking at
<scobiehague>
true
<scobiehague>
that is why my idea might work better on a network of computers you have physical control over
<scobiehague>
so I can e.g., delete 'Recent Files' and access logs, atimes and such
bergie has joined #ipfs
<achin>
if your computers were seized, would you really have time to delete files and change access logs?
* Stskeeps
waves to bergie
<gamemanj>
achin: if you had an automated script...
<achin>
what would trigger the script?
<scobiehague>
achin: would probably need a killswitch kind of script to delete them
<victorbjelkholm>
Dead mans script
<gamemanj>
in more light-hearted news: ideas for increasing the amount of IPFS nodes on the network... a casual game
<victorbjelkholm>
actually called "dead mans switch" I see now
<scobiehague>
achin: inactivity for a certain period, perhaps
<scobiehague>
there would be many ways to trigger it
<victorbjelkholm>
gamemanj, tell us more! You have any ideas for the game?
<gamemanj>
a casual game called "bitswap", where people act as human BitSwap algorithms
<gamemanj>
ofc this would be connected to the real IPFS network
<gamemanj>
they'd be playing the game with other IPFS nodes (Which might potentially be other instances of the game!)
<victorbjelkholm>
just added ability to use local daemon in ipfsbin.xyz! Try it out :) Next step is to make it offline-ready in case of the network dying
<gamemanj>
they'd only be told about this after uploading 1TB of data ofc
<gamemanj>
ok, I may be joking there :)
<scobiehague>
nice game over message for them
<gamemanj>
"game over" message? I suppose it would be game over if they leeched enough useless data...
<scobiehague>
"game over: thanks for the 1TB of bandwidth"
fiatjaf has joined #ipfs
<ipfsbot>
[go-ipfs] chriscool created t0220-bitswap-stat (+1 new commit): http://git.io/vWx8P
<scobiehague>
gaemanj: I was joking too. tricking people into sharing their bandwidth, unwittingly. They wouldn't be too happy with that.
<gamemanj>
Hmm, true. Though it would be kind of funny, since the game wouldn't be automated. They'd actually be the ones setting their wantlist and giving blocks.
<scobiehague>
anyone know of a static comments system for static sites? I found pooleapp.com, but it seems to be down now
<dignifiedquire>
daviddias: hey are you around, got a question
<ipfsbot>
[js-ipfs-api] Dignifiedquire closed pull request #85: [discuss] Switch to superagent instead of raw http.request (solid-tests...superagent) http://git.io/vWxBx
border0464 has quit [Ping timeout: 256 seconds]
border0464 has joined #ipfs
<ipfsbot>
[js-ipfs-api] Dignifiedquire opened pull request #92: [discuss] Switch to superagent instead of raw http.request (master...superagent) http://git.io/vWxEq
<dignifiedquire>
I have looked a little bit at fetch and polyfills and it victorbjelkholm does make a good point that it is being implemented by browser vendors already
<gamemanj>
Or maybe something like GTK's UI builder language, with Markdown panes.
<gamemanj>
There are all sorts of options, and they're all a lot simpler than HTML nowadays...
<vanila>
test982726 brought up a really good point about markdown - I've been thinking s-expression based markup language could be nice and simple
<vanila>
regular grammar for tokenization, just need a single stack to parse them, easy to work with in code
<daviddias>
dignifiedquire: shared my feedback :) the gist of it is, as long as we don't break things that already work, I'm good to whatever, we should definitely keep experimenting and moving forward as the language evolves
therealplato has joined #ipfs
<dignifiedquire_>
daviddias: thanks, reading now, one other not circleci seems to be running on an outdated node version
barnacs has quit [Read error: Connection reset by peer]
<daviddias>
dignifiedquire_: quick question, what would babel-eslint give us? doesn't standard cover new Node.js features?
<dignifiedquire_>
standard = config for eslint + cli
<dignifiedquire_>
eslint-standard-config = config for eslint used in standard
<dignifiedquire_>
eslint-parser = default parser from eslint
<dignifiedquire_>
babel-eslint = parser for eslint supporting all new language features that babel supports
<dignifiedquire_>
the standard eslint parser does not yet understand all new ES2015 features, that’s why babel-eslint is used for parsing, the same rules are still applied from standard
therealplato has quit [Read error: Connection reset by peer]
cemerick has joined #ipfs
gamemanj has quit [Ping timeout: 260 seconds]
<daviddias>
dignifiedquire_: sounds good :)
AndroUser has quit [Read error: Connection reset by peer]
AndroUser has joined #ipfs
Tv` has joined #ipfs
fingertoe has joined #ipfs
simonv3 has joined #ipfs
therealplato has joined #ipfs
dignifiedquire_ has quit [Quit: dignifiedquire_]
nessence has joined #ipfs
<Zuardi>
standard is a terrible name :)
Guest73396 has quit [Ping timeout: 265 seconds]
martinkl_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<dignifiedquire>
daviddias: recursive uploads are only available from node right?
<daviddias>
I believe they work in the browser (well, that is why vinyl was used in the first place)
fiatjaf has left #ipfs ["undefined"]
sonatagreen has joined #ipfs
voxelot has joined #ipfs
voxelot has joined #ipfs
<ipfsbot>
[go-ipfs] chriscool created test-ipfs-config (+1 new commit): http://git.io/vWpnG
<ipfsbot>
go-ipfs/test-ipfs-config bbe95b8 Christian Couder: t0021: test 'ipfs config show'...
<dignifiedquire>
daviddias: but how? vinyl does not have access to the file system from the browser
<ipfsbot>
[go-ipfs] chriscool opened pull request #1913: t0021: test 'ipfs config show' (master...test-ipfs-config) http://git.io/vWpCt
<daviddias>
jbenet and mappum worked a lot on that, they will know best to answer you
<dignifiedquire>
okay thanks
<dignifiedquire>
but why is the recursive upload test then failing at the moment in the browser?
mcelrath has quit [Remote host closed the connection]
<dignifiedquire>
from the looks of that package and the changes that were done it only allows to use vinyl-fs
<dignifiedquire>
(delete last message)
bsm1175321 has joined #ipfs
bsm1175321 is now known as mcelrath
chriscool has quit [Ping timeout: 250 seconds]
<victorbjelkholm>
does it make sense to save recursively from the browser?
<daviddias>
victorbjelkholm: well, we do want users to be able to drop a folder and expect the same behaviour as it was a native app
<dignifiedquire>
daviddias: loool dropping a folder
<victorbjelkholm>
that makes sense
<dignifiedquire>
daviddias: only chrome supports that though
<dignifiedquire>
and it’s not clear if that will go away again
<daviddias>
dignifiedquire: it might be broken, not 100% sure, that is why I was advocating for us to have tests all this long, I understand how frustrating it is
<dignifiedquire>
let’s just drop multipart and find a decent way to upload files ;)
<daviddias>
multipart is how ipfs HTTP api works and how ipfs cli works
<dignifiedquire>
I’m sure jbenet would be happy xD
<daviddias>
that is why we went with it
<dignifiedquire>
I know..
<dignifiedquire>
just a joke
<daviddias>
but yeah, always happy for easier and clean ways, just making sure we keep in our minds that it is more than just js-ipfs-api
edcryptickiller has joined #ipfs
<victorbjelkholm>
we could just convert the files into json files, upload that and get do the same but reverse
<victorbjelkholm>
+ we should be able to just use the file api without vinyl, no?
<dignifiedquire>
the issue is that go-ipfs does not understand that
<daviddias>
a user can add a bunch of files and then we wrap it it in a folder
<daviddias>
this should be doable in every browser
kode54 has quit [Ping timeout: 260 seconds]
Encrypt has joined #ipfs
domanic has joined #ipfs
<dignifiedquire>
daviddias: they can drop many files, and that’s easy to support to upload via simple multipart uploads (we already do that), the thing is that all these solutions to not support things like nested directories, or a top level directory
hoboprimate has joined #ipfs
cemerick has quit [Ping timeout: 240 seconds]
mildred has quit [Ping timeout: 264 seconds]
NeoTeo has quit [Quit: ZZZzzz…]
<dignifiedquire>
daviddias: victorbjelkholm I think I’m an idiot and there is a simple solution that actually works with the current way go-ipfs is implemented
<dignifiedquire>
if that’s the case I will go and bang my head against the wall for a while
gaboose has joined #ipfs
kode54 has joined #ipfs
<gaboose>
what is this context variable in go-ipfs all about? it gets passed around everywhere
<victorbjelkholm>
hm, not sure I understand what vinyl is for
<victorbjelkholm>
Gaboose, a context in go is the request context, so you can pass things around to other functions
<victorbjelkholm>
dignifiedquire, interested to see what you mean :)
Whispy has joined #ipfs
Whispy is now known as TheWhisper
hoboprimate has quit [Remote host closed the connection]
Whispery has quit [Ping timeout: 265 seconds]
nessence has quit [Remote host closed the connection]
captain_morgan has quit [Ping timeout: 240 seconds]
HoboPrimate has quit [Read error: Connection reset by peer]
HoboPrimate has joined #ipfs
JimmyTheWhale has joined #ipfs
HoboPrimate has quit [Read error: Connection reset by peer]
HoboPrimate has joined #ipfs
domanic has joined #ipfs
cemerick has quit [Ping timeout: 268 seconds]
<JimmyTheWhale>
Hey ipfs. I read your botbot sometimes and it seems pretty clear that ipfs is explicitly a political project. I thought some of you might be interested in www.platformcoop.net, a conference coming up in NYC, Nov 13-14. It will be livestreamed.
<ansuz>
whoa a talking whale
<JimmyTheWhale>
o/
<ansuz>
\o
<vanila>
can anyone elaborate more on the policital side of IPFS?
<JimmyTheWhale>
or at least an instance of it, this was the one i went to.
<ansuz>
tl;dr ISPs break end to end principle, I think
<ansuz>
&& end to end principle is intrinsicly related to human rights
<ansuz>
ish
<ansuz>
that's my take on it
<JimmyTheWhale>
It also seems to have implications for offering a different monetization model for the web? I know it wouldn't preclude ads, but it implicitly reframes the web as an information-commons -- once it is shared, it becomes public property of sorts.
lgarron has quit [Quit: Leaving.]
<test982726>
I'd think commons, security (through decentralization) and optimization
HoboPrimate has quit [Read error: Connection reset by peer]
<achin>
JimmyTheWhale: i wouldn't discount the IPFS technology. it borrows a lot of existing technologies, but can stand on its own as a novel creation
<JimmyTheWhale>
?
<achin>
you said above that "ipfs is explicity a politican project"
<vanila>
political
<vanila>
like
<JimmyTheWhale>
political project -- that is, built to achieve certain political goals
<JimmyTheWhale>
like, "freedom"
<JimmyTheWhale>
(& such :P )
<achin>
i know you were not ignoring the other goals, but i thought i'd point it out anyway :) (where 'other goals' means things like 'producing a neat bit of technology)
<ansuz>
achin reserves the right to be novel && unfree
HoboPrimate has quit [Ping timeout: 240 seconds]
<demize>
I'm quite expensive and novel.
<JimmyTheWhale>
:^) Yeah, it's totally a novel creation. I think that's why it might be successful at making an impact.
<ion>
If I have “ttl := …” followed by an if block that includes “respath, ttl, err := …”, which value will ttl have after the if block if the branch was executed?
<ion>
Trying to patch go-ipfs without having learned go. :-P
<ion>
go doesn’t seem to come with a REPL in which i could test these things easily.
<ion>
It seems I’ll have to use another name in the if block and overwrite the outer ttl value in a separate statement.
<codehero>
does this work for you? ipfs.io/ipfs/Qmb6izJfuhw5GmrJRNsfrB4JgPL3PPLGed4Facm2wbTJHq
<ion>
yes
<codehero>
huh
<ion>
At least with https:// it does
<codehero>
oh. hrm
<codehero>
so i need to use https
<codehero>
good to know
<ion>
No, http:// will also work but just redirect to https://
<codehero>
ah
nessence has quit [Remote host closed the connection]
nessence has joined #ipfs
tilgovi has quit [Ping timeout: 255 seconds]
<kyledrake>
whyrusleeping I'm doing a quick blog post on setting up IPFS on a VPS. Anything I should mention? How much RAM do you typically recommend for people?
rendar has joined #ipfs
<codehero>
hmm
<codehero>
using https:// doesn't make it faster either
domanic has quit [Ping timeout: 268 seconds]
<codehero>
the gateway is discriminating me. lol
HoboPrimate has joined #ipfs
<achin>
i'm not sure a RAM recommendation can be made just yet. there are still a few memory-use issues that are still open
<codehero>
it works for other people just fine
NeoTeo has quit [Quit: ZZZzzz…]
HoboPrimate has quit [Remote host closed the connection]
OutBackDingo has joined #ipfs
<codehero>
ipfs just uses 2.2% of my ram, and i only have 4GB
<codehero>
of course, it's not a server that provides lots of data to lots of clients
Vaul has joined #ipfs
forthqc has quit [Quit: Lämnar]
JimmyTheWhale has quit [Remote host closed the connection]
JimmyTheWhale has joined #ipfs
domanic has joined #ipfs
HoboPrimate has joined #ipfs
<victorbjelkholm>
codehero, try running ipfs get Qmb6izJfuhw5GmrJRNsfrB4JgPL3PPLGed4Facm2wbTJHq from another machine, still slow?
<victorbjelkholm>
runs in 1.6 seconds for me
<codehero>
i only have one machine, unfortunately
<codehero>
victorbjelkholm: hm. yeah. then it's definitely the gateway
HoboPrimate has quit [Read error: Connection reset by peer]
NeoTeo has joined #ipfs
HoboPrimate has joined #ipfs
HoboPrimate has quit [Remote host closed the connection]
patcon has joined #ipfs
pierron has joined #ipfs
m0ns00n has joined #ipfs
<m0ns00n>
Hoi.
HoboPrimate has joined #ipfs
<achin>
ohi
HoboPrimate has quit [Ping timeout: 272 seconds]
martinkl_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
barnacs has joined #ipfs
martinkl_ has joined #ipfs
Encrypt has joined #ipfs
<m0ns00n>
achin: How's the project moving along?
Matoro has joined #ipfs
tilgovi has joined #ipfs
Oatmeal has joined #ipfs
<achin>
pretty good! it's got a nice community, a dedicated team of developers working on it, and a working alpha release
<m0ns00n>
That I know :)
<m0ns00n>
I want to use your fs once it's at least beta quality.
<m0ns00n>
I don't know which state it is in now.
<m0ns00n>
I'm part of a company building a new OSS OS.
<achin>
the best thing to do, then, is give it a try, and provide feedback to irc, or to the github issue tracker!
<victorbjelkholm>
m0ns00n, try it out and see how it's working. My view on it is that it's stable enough for usage
<victorbjelkholm>
had minor issues but been more about the js-ipfs-api library than actual ipfs
<m0ns00n>
Ah cool!
<m0ns00n>
We're releasing the first version of our OS on saturday.
<m0ns00n>
Only for invited now.
<m0ns00n>
But a milestone for us.
<test982726>
achin: is there a timeline for ipns, because much of the dynamic functionality depends on that?
<achin>
i don't know. check github, or check with whyrusleeping or jbenet
<test982726>
k, thx
ygrek_ has quit [Ping timeout: 250 seconds]
<ion>
whyrusleeping: Progress! I am getting useful Cache-Control and ETag headers for IPNS requests.
<m0ns00n>
Will be changed and made a bit more commercial :)
<m0ns00n>
"From the horse's mouth" so to speak.
<ogd>
cool fjords
<m0ns00n>
:)
<m0ns00n>
Norway rocks!
<m0ns00n>
(literally)
<codehero>
oh. people are still making js "operating systems" huh
<codehero>
however, the last time i saw something like that was before html5
<codehero>
so they pretty much always sucked
<pierron>
Hi, I watched some of the video, and I was wondering about the anonimity of the network. Is there a way to have peers which are not reachable by locations, or only seeding through hops?
<codehero>
m0ns00n: what is this friendUP? is it just an ipfs gateway thingy?
<m0ns00n>
codehero: It's a web server with resource management (like hardware, libraries etc).
<achin>
ipfs does not quite support relaying at the moment
<codehero>
m0ns00n: ah. okay
<m0ns00n>
codehero: It has a built in desktop environment for web browsers.
<codehero>
and i suppose you can use desktop applications via emscripten?
<m0ns00n>
codehero: If you visit our facebook page (if you have an account) we added lots of screenshots etc.
<pierron>
Where a user can potentially opt-in in a peer-to-peer network for sharing binary software, but not having annonimity might imply that your computer is potentially vulnerable to some attack.
<achin>
if A connects to B, and B connects to C, there is no way for a file that exists only on C to be transferred to A via B. the only think you can do at the moment is manually ask B "please download this hash for me"
<m0ns00n>
codehero: We haven't gotten to that yet - but it's the idea
<codehero>
pierron: oh. nixos. cool. i just hope they could get their sh* together in terms of security updates :/
<codehero>
but yeah. ipfs is perfect for nixos
<m0ns00n>
codehero: We're being very open in the project, anyone of you who want early access, send me an email to titlestad@friendos.com
<codehero>
especially since it also uses hashes, like nixos
<codehero>
m0ns00n: cool :)
<m0ns00n>
codehero: The OS will be available as AGPLv3 in late december.
<m0ns00n>
codehero: That's our "patent". :)
<codehero>
heh
<pierron>
codehero: I am working on it, and presenting at the NixCon a way to ship security updates quickly
<codehero>
oh. sweet
<codehero>
pierron: do you also have a solution for the mass rebuilds that delay security updates to over a month?
<pierron>
codehero: yes ;)
<codehero>
THANK YOU
<codehero>
you're doing god's work
<codehero>
i'm very excited already
<codehero>
so i'm probably going to switch back when you're finished
<codehero>
and if nixos can also get ipfs support *_*
ianopolous has joined #ipfs
<pierron>
codehero: I don't think this would land before December, the time to convince everybody, as usual
<codehero>
yeah
<codehero>
no problem. i'm pretty comfortable with arch right nw
<codehero>
so i can wait
<pierron>
Do you know if ipfs plan to have anonimity between peers?
<codehero>
also about the lack of anonymity. windows also supplies updates p2p and bittorrent isn't anonymous either
<codehero>
i don't really see a problem, especially if you're just distributing packages
<pierron>
codehero: This is at least the playground that Tribler is trying to address.
<codehero>
well. yeah, but onion routing makes things slow
<codehero>
and it still has a problem with vulnerability
<codehero>
you can easily get the ip adresses of your peers
<m0ns00n>
must code more :D
<codehero>
onion routing just prevents outsiders from seeing who is actually downloading what
<pierron>
codehero: the problem is that if you know that openssl-AAAAAA is vulnerable, and opensll-BBBBBB is fixed, and that you can find the forner and not the later on a peer, you know that he is likely running a vulnerable version of openssl.
<m0ns00n>
But I will definately look into making a drive for ipfs when we're closer to v1 beta1.
<codehero>
hm
<codehero>
true true
<m0ns00n>
Now it's coding my ass off for saturday
<m0ns00n>
hehe
<pierron>
codehero: the other way around, if somebody ask you for openssl-BBBBBBB you know that he does not have the fix running yet.
mildred has quit [Ping timeout: 260 seconds]
<codehero>
yeah
<codehero>
hm. well
<codehero>
but onion routing isn't going to fix that either
<pierron>
codehero: that's the claim of Tribler.
<codehero>
hm
<ion>
pierron: You can use Tor for anonymity. go-ipfs will support it in the future.
<pierron>
ion: oh … I didn't thought about that.
<pierron>
ion: good call.
<codehero>
the only problem is that tor makes everything dial up modem slow
<codehero>
which defeats the purpose of distributed package managing
<codehero>
imo
nessence has quit [Remote host closed the connection]
nessence has joined #ipfs
<codehero>
although i suppose having the seeder encrypt the file, then route it through one other node that directs it to the recipient should be enough
<pierron>
oh, sorry I miss the reference to the onion network before, yes, this will only protect the person who is fetching, not the person which is serving.
<pierron>
codehero: as long as the encrypted hash is predictible?
<codehero>
well. you could have the receiver encrypt the request with the sender's public key
<codehero>
send it through a random close node
<codehero>
and the sender encrypts the file with the receiver's public key and sends it back
<codehero>
the middle node will know the ip's but not the data
<codehero>
i think
<codehero>
onion routing is definitely overkill
nessence has quit [Remote host closed the connection]
nessence has joined #ipfs
<pierron>
codehero: The other option is to only seed one version of each package.
<pierron>
codehero: this way, you only have a likelyhood, but this is enough for attackers :/
<ion>
codehero: Are you running 0.3.8 or 0.3.9-dev?
<OutBackDingo>
wants 0.4
<codehero>
umm. 0.3.9-dev, i think
juul is now known as MegaJoule
MegaJoule is now known as zuul
<codehero>
wait a sec
<codehero>
ipfs version 0.3.9-dev
<codehero>
there we go
<ion>
codehero: Ok, it's a different bug than the one <0.3.8 had then.
<codehero>
i'll try to reproduce and make a dump
<pierron>
codehero: I guess this does not only apply to NixOS, but to any form of code. So one could base an attack on the fact that you are running an old version of a web-application.
<codehero>
yeah. although i'm not exactly sure how to fix that without making stuff at least a bit slower
ygrek_ has joined #ipfs
<codehero>
you got to be kidding me
<codehero>
now it works
<codehero>
btw. ion, what's the type column in the pinned files list of the webui
<codehero>
is it supposed to work?
<codehero>
well. it works for now
Not_ has quit [Remote host closed the connection]
<ion>
If it's anything like the pin type in the command line UI, “recursive” means that the object and all its transitive links are pinned and “direct” means that only the object is pinned, not its links AFAIU.
<codehero>
ohh. okay
<codehero>
i wish systemd had better ways to manage which libraries a service can use
martinkl_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
chriscool has quit [Quit: Leaving.]
OutBackDingo_ has quit [Remote host closed the connection]
chriscool has joined #ipfs
Eudaimonstro has quit [Remote host closed the connection]
OutBackDingo_ has joined #ipfs
HoboPrimate has joined #ipfs
chriscool has quit [Client Quit]
chriscool has joined #ipfs
e-lima has joined #ipfs
domanic has joined #ipfs
martinkl_ has joined #ipfs
pfraze has joined #ipfs
elima has quit [Ping timeout: 240 seconds]
lgarron has joined #ipfs
chriscool has quit [Ping timeout: 250 seconds]
OutBackDingo has quit [Quit: Leaving]
pfraze has quit [Remote host closed the connection]
OutBackDingo has joined #ipfs
master8787 has joined #ipfs
neoteo_ has joined #ipfs
captain_morgan has joined #ipfs
pfraze has joined #ipfs
m0ns00n has quit [Quit: Leaving]
amade has quit [Quit: leaving]
ianopolous2 has joined #ipfs
neoteo_ has quit [Ping timeout: 240 seconds]
ianopolous has quit [Ping timeout: 252 seconds]
lgarron has quit [Quit: Leaving.]
CarlWeathers has quit [Remote host closed the connection]
martinkl_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
simonv3 has quit [Quit: Connection closed for inactivity]
Vaul has quit [Quit: Bye]
CarlWeathers has joined #ipfs
Matoro has quit [Ping timeout: 260 seconds]
Encrypt has quit [Quit: Quitte]
<spikebike>
hrmpf, ipfsbin.xyz is acting upo for me
NeoTeo has quit [Quit: ZZZzzz…]
master8787 has quit [Quit: Leaving]
simonv3 has joined #ipfs
<spikebike>
can anyone else get QmWgPhfQ5ixgLUwaAH7k7zUxGc7HFARGWc7oXcpB8NZ4Hb
<spikebike>
I added it with ipfsbin.xyz
<spikebike>
(it's a small harmless script)
test982726 has quit [Ping timeout: 246 seconds]
<achin>
yep
<spikebike>
ah, wonder if you fixed it
Matoro has joined #ipfs
<spikebike>
my ipfs cat stopped for minutes, but just finished
<spikebike>
weird
<achin>
mrow
<spikebike>
it's oddly formatted, ah, \n instead of line breaks.
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
<spikebike>
no idea
<spikebike>
maybe for a simpler apache rewrite rule or somthing
<whyrusleeping>
hashrouting, makes it easy for apps to have a path without having that file structure
<whyrusleeping>
so instead of setting up routes on the server, a static page can have 'routes'
<whyrusleeping>
and process them however it likes
Matoro has quit [Ping timeout: 260 seconds]
<cryptix>
i find it bewildering that js/dom has no api for query params
<ion>
jbenet: I started working on making Resolve et al. return a TTL value. The gateway can take advantage of that for the Cache-Control/max-age header, letting IPNS sites loaded through the gateway take advantage of HTTP caching. Do you have comments about the suggested API change in <https://github.com/ipfs/go-ipfs/issues/1818#issuecomment-152281965>?
Matoro has joined #ipfs
<ion>
I already have the gateway serving the right headers but i haven’t made the test suite work yet.
e-lima has quit [Ping timeout: 255 seconds]
Teeed has joined #ipfs
<Teeed>
hi
<achin>
hi
pfraze has quit [Remote host closed the connection]
pfraze has joined #ipfs
Matoro has quit [Ping timeout: 240 seconds]
Eudaimonstro has joined #ipfs
Johnnie has quit [Killed (Sigyn (Spam is off topic on freenode.))]