<pjz>
whyrusleeping: minor feedback: what keeps me from DOSing you with a bunch of keys?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<whyrusleeping>
pjz: i have to run 'ipfs key recv'
<whyrusleeping>
with your peer ID
<whyrusleeping>
to receive a key from you
<pjz>
whyrusleeping: ah, okay, that's not listed in the spec
<whyrusleeping>
pjz: ooooh, huh
<pjz>
whyrusleeping: also, I kind of feel like ipfs crypt decrypt <> might not be necessary if the data is on a node and you have key(s) from that node
<whyrusleeping>
pjz: but what if you want other nodes to assist in the distribution of that data?
<whyrusleeping>
plus, if you want data to be readable only by a certain set of peers
<whyrusleeping>
the only 'safe' way to do that is to encrypt the data and share the key with those peers
<pjz>
whyrusleeping: right
<pjz>
whyrusleeping: but my point is that those peers shouldn't have to do an explicit decrypt
<whyrusleeping>
pjz: very much agreed
<pjz>
whyrusleeping: if you've shared encrypted with me, and sent me your key, when I go to look at that data it should automatically try any keys you sent me against it and see if they work
<whyrusleeping>
'ipfs cat <encrypted hash>' should output the unencrypted data for objects i have the keys to
<whyrusleeping>
my idea was to write the keys hash in the header of the encrypted object
<whyrusleeping>
sure thats leaking a little bit of data, but i think its fine
<pjz>
whyrusleeping: oh, that would work too :)
<pjz>
whyrusleeping: and is actually a much better idea than mine :)
voxelot has quit []
<whyrusleeping>
pjz: haha, its a little easier i think
<pjz>
whyrusleeping: and it will autodecrypt if you do an 'ipfs cat' or an 'ipfs get', yes?
<pjz>
I'm not sure what 'ipfs crypt decrypt' is for
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to get-fix: http://git.io/vYX3d
<ipfsbot>
go-ipfs/get-fix 8176766 Jeromy: naming and cleanup...
<lgierth>
whyrusleeping: or create /ipfs/<hash-of-encrypted-hash>
<lgierth>
where encrypted hash is encrypted with your public key
<lgierth>
that'd be an additional roundtrip though
<whyrusleeping>
pjz: well, the encrypt and decrypt functions dont necessarily add into ipfs
<lgierth>
you wouldn't leak the public key that it was encrypted with into the header
<whyrusleeping>
you could do 'echo "hello" | ipfs encrypt >afile'
<whyrusleeping>
and the use decrypt to read that back
notduncansmith has joined #ipfs
<whyrusleeping>
its important to have as a porcelain command
notduncansmith has quit [Read error: Connection reset by peer]
<pjz>
I see
tilgovi has joined #ipfs
compleatang has quit [Quit: Leaving.]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to master: http://git.io/vYXWR
<ipfsbot>
go-ipfs/master aaaf884 Juan Benet: Merge pull request #1519 from ipfs/get-fix...
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<daviddias>
jbenet: lgierth I left my Driver's license at home, but if we do rent a Tesla, I would be for it anyways :D (I think I would have 24 hours to show up with the driver's license before I would get a ticket)
<daviddias>
I was just informed that in Germany, there are road without any speed limit
<daviddias>
We could make the 24 hours
<daviddias>
:P
<whyrusleeping>
autobahn!
<whyrusleeping>
i beleive you have to be a german citizen though and have a german drivers license
<jbenet>
whyrusleeping can you rebase 1510 on master? errs were notif failures
<whyrusleeping>
jbenet: sure thing
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed pinfail-debug from 7c91e63 to 138198e: http://git.io/vYXEM
<ipfsbot>
go-ipfs/pinfail-debug 138198e Jeromy: print output on test failure for t0081...
<ipfsbot>
[go-ipfs] jbenet pushed 2 new commits to master: http://git.io/vYXuS
<ipfsbot>
go-ipfs/master 818d3af rht: Clear progress bar on `ipfs cat` exit...
<ipfsbot>
go-ipfs/master 870d850 Juan Benet: Merge pull request #1516 from rht/progressbar...
rschulman__ has quit [Quit: rschulman__]
<ipfsbot>
[go-ipfs] jbenet pushed 2 new commits to master: http://git.io/vYXzl
<ipfsbot>
go-ipfs/master f417cb5 Juan Benet: Merge pull request #1510 from ipfs/pinfail-debug...
<ipfsbot>
go-ipfs/master 138198e Jeromy: print output on test failure for t0081...
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman__ has joined #ipfs
<whyrusleeping>
random thought
<whyrusleeping>
if your bookmarks bar was an ipfs object
<whyrusleeping>
you could cd into webpages
<whyrusleeping>
and just browse them
<reit>
if i'm reading https://github.com/ipfs/go-ipfs/issues/1522 correctly, it appears there is a method of dumping files directly into ipns at which point Magic happens and converts everything into ipfs entries on the fly, is there any more information on how to use this/how it works?
domanic has quit [Ping timeout: 256 seconds]
<whyrusleeping>
reit: you run 'ipfs mount' and cd into /ipns/local
<whyrusleeping>
if you want details see fuse/ipns in the codebase
<reit>
if i cp a file/directory into it, am i correct in assuming the system
<reit>
1) recursively converts the file/directory into ipfs blockstore entries (and then deleting the originals)
<reit>
2) calculates the new hash of the root node and
<reit>
3) publishes this new hash to your ipns name?
<whyrusleeping>
jbenet: thoughts on 0.4.0 soon? it will solve a good number of perf issues, and also make things easier on daviddias for testing multistream stuff
<reit>
if so that's really cool
<whyrusleeping>
reit: that is correct
<whyrusleeping>
although it might be broken right now
<whyrusleeping>
we havent touched it in two months...
<reit>
yeah, i can't seem to get it working
<whyrusleeping>
and i just fixed my fuse install i think
<whyrusleeping>
so i'll see about getting it back in action
notduncansmith has joined #ipfs
<reit>
cool, looking forward to it
notduncansmith has quit [Read error: Connection reset by peer]
cdata has joined #ipfs
reit has quit [Quit: Leaving]
reit has joined #ipfs
AndChat|77184 has joined #ipfs
<whyrusleeping>
huh, the ipns tests pass on linux on my machine
<whyrusleeping>
oh wait nvm
reit has quit [Ping timeout: 256 seconds]
AndChat|77184 has quit [Quit: Bye]
reit has joined #ipfs
<whyrusleeping>
the tests were all disabled
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman__ has quit [Quit: rschulman__]
qqqueue is now known as qqueue
cdata has quit [Ping timeout: 240 seconds]
cdata has joined #ipfs
<ipfsbot>
[go-ipfs] rht opened pull request #1523: Add reversible gc test (master...reversible-gc) http://git.io/vYX1c
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
rschulman__ has joined #ipfs
hellertime has quit [Quit: Leaving.]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
cdata has quit [Read error: Connection reset by peer]
cdata1 has joined #ipfs
rschulman__ has quit [Quit: rschulman__]
simonv3 has quit [Quit: Connection closed for inactivity]
<ipfsbot>
[go-ipfs] whyrusleeping created fix/coremock (+1 new commit): http://git.io/vYXQy
<ipfsbot>
go-ipfs/fix/coremock 0adad66 Jeromy: set identity on core mock to fix ipns tests...
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1524: set identity on core mock to fix ipns tests (master...fix/coremock) http://git.io/vYX7v
tilgovi has quit [Ping timeout: 272 seconds]
<ipfsbot>
[go-ipfs] whyrusleeping force-pushed fix/coremock from 0adad66 to c64f743: http://git.io/vYX7W
<ipfsbot>
go-ipfs/fix/coremock c64f743 Jeromy: set identity on core mock to fix ipns tests...
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
cdata2 has joined #ipfs
cdata1 has quit [Ping timeout: 240 seconds]
cdata2 has quit [Ping timeout: 244 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
substack: it seems that webcomponents -- with its rel="import" -- just dumps all the tags into the global page namespace, instead of doing qualified imports ("sweet as!"). am i right? if so, we have to fix this before it's too late.
<eternaleye>
I was under the impression that it was more polymer = backcompat-polyfill(webcomponents)
<jbenet>
substack: i dont think you can supply names... maybe we can ask for a "namespace='foo'" attribute to be added that allows the importing page to refer to things as <foo.X>
<jbenet>
eternaleye: i think polymer came first.
<jbenet>
eternaleye like http2 = w3c(spdy)
<jbenet>
eternaleye but i might be off.
<substack>
I thought people had soured on html imports
<substack>
well anyways just like all the other stupid ideas on the web platform, you can ignore this one
tilgovi has joined #ipfs
<ipfsbot>
[go-ipfs] gatesvp closed pull request #1448: Add ignore and hidden file support to add (master...master) http://git.io/vqZlT
<gatesvp>
Some of them seem legit, like the panic in `peer_request_queue_test.go`
alu has quit [Ping timeout: 246 seconds]
<gatesvp>
But problems like argument parsing from stdin (where we use a file to proxy stdin) or like "embedded data not equal" seem like the type of things that just may not work
besenwesen has quit [Quit: ☠]
besenwesen has joined #ipfs
besenwesen has quit [Changing host]
besenwesen has joined #ipfs
besenwesen_ has joined #ipfs
<jbenet>
gatesvp makes sense
besenwesen has quit [Ping timeout: 246 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
lidel has quit [Ping timeout: 256 seconds]
reit has joined #ipfs
besenwesen_ has quit [Ping timeout: 272 seconds]
reit has quit [Ping timeout: 244 seconds]
lidel has joined #ipfs
<gatesvp>
@jbenet: thanks... I'm for tonight... will start looking at those Windows test issues tomorrow
<jbenet>
gatesvp: sounds good, thanks so much!
besenwesen has joined #ipfs
besenwesen has joined #ipfs
besenwesen has quit [Remote host closed the connection]
besenwesen has joined #ipfs
besenwesen has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
dignifiedquire has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
gatesvp has quit [Ping timeout: 246 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<mafintosh>
if i want to share some vm images using ipfs should i just encrypt them myself and there the encrypted images? or does ipfs have something for that
<mafintosh>
*and share
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
atomotic has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
fleeky has quit [Remote host closed the connection]
fleeky has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<pguth2>
I would be interested in knowing that too.
<cryptix>
afaik this is planed but depends on the keystore stuff - im not sure ont he status of that
<cryptix>
whyrusleeping: do you have standalone tests for the metrics stuff?
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
silotis_ has quit [Ping timeout: 246 seconds]
silotis has joined #ipfs
<jbenet>
mafintosh: encrypt them yourself first. we'll have encryption soon (there's a branch with it somewhere) but not in master yet
<jbenet>
mafintosh i think that's right, double check. i dont use gpg much.
<mafintosh>
jbenet: to guarantee that something is seeded on ipfs do i need to have my own node running or does the gateway currently replicate stuff?
<mafintosh>
(its too much data for my raspberry pi ipfs node)
<mafintosh>
but i can spin up on of digital ocean if thats better than relying on the gateway
<jbenet>
mafintosh: you can curl things on the gateway to make it fetch them, but they'll get expunged in a few hours. but-- just throw me a hash and i'll pin it
<mafintosh>
jbenet: sweet thanks
<jbenet>
mafintosh: i wanna get a "1 click spin up ipfs node on digital ocean" button.
<mafintosh>
jbenet: the node leveldb community thanks you
<jbenet>
domanic has quit [Ping timeout: 264 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<jbenet>
but the fact that "id?stream-channels=true" finished ok, means its not a full break. something is buggy with one of the commands and interrupting webui flow.
<jbenet>
the webui hasn't changed so it must be something in the api
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<slothbag>
I just rolled back to the commit before http get fix and it works
<Luzifer>
jbenet: refs gobuilder#55: where should this special cli tool be running? is it specially for you to mirror from S3 to IPFS? How do users get the IPFS hash to load via IPFS?
<jbenet>
slothbag: thanks-- would you mind filing an issue and tagging whyrusleeping?
<jbenet>
Luzifer: i meant, a build of gobuilder-cli could use ipfs to download things instead.
<jbenet>
so could make a separate bin if dont want to change the main one
<Luzifer>
jbenet: yeah possible but currently there is nothing pushing into ipfs
<Luzifer>
and it also would need a central "registry" of ipfs hashes
<Luzifer>
both currently doesnt exist
<Luzifer>
because if you just have the filename you are not able to determine which hash the file has
<Luzifer>
and for the push-to-ipfs topic: I currently don't have a good solution for the mirroring… would need a node with plenty of storage getting the files pushed and then adds them to ipfs
<Luzifer>
or: an ephemeral node on the build server which adds and then another node pins and pulls the file using IPFS…
<Luzifer>
currently my builder units don't have enough disk space to hold all the assets
<Luzifer>
(besides I didn't want them to hold all the assets. they are builders, not a stateful storage node)
<jbenet>
Luzifer: why not store the ipfs hash with the page you render?
<Luzifer>
yeah. that would be the plan as soon as the push-to-ipfs "problem" is solved
pfraze has joined #ipfs
notduncansmith has joined #ipfs
pfraze has quit [Remote host closed the connection]
notduncansmith has quit [Read error: Connection reset by peer]
alu has joined #ipfs
rschulman__ has joined #ipfs
<ipfsbot>
[go-ipfs] jbenet pushed 3 new commits to master: http://git.io/vYDue
<ipfsbot>
go-ipfs/master a5521fc Juan Batiz-Benet: sharness/repo: fix pinning test...
<ipfsbot>
go-ipfs/master 8fe7d2f gatesvp: Fix refs -r -u for #1211...
<ipfsbot>
go-ipfs/master 490ed41 Juan Batiz-Benet: sharness/ipfs refs bigger tests...
<jbenet>
wtf thats not what i did.
<Luzifer>
THE BOTS ARE AGAINST US!!!
<Luzifer>
(scnr)
<jbenet>
sorry people. that was weird.
<ipfsbot>
[go-ipfs] jbenet force-pushed master from 490ed41 to 29f67b4: http://git.io/vfG15
<ipfsbot>
[go-ipfs] jbenet created fix-refs-u (+3 new commits): http://git.io/vYDz9
<ipfsbot>
go-ipfs/fix-refs-u 8fe7d2f gatesvp: Fix refs -r -u for #1211...
<ipfsbot>
go-ipfs/fix-refs-u a5521fc Juan Batiz-Benet: sharness/repo: fix pinning test...
<ipfsbot>
go-ipfs/fix-refs-u 490ed41 Juan Batiz-Benet: sharness/ipfs refs bigger tests...
<ipfsbot>
[go-ipfs] jbenet opened pull request #1527: Fix refs u (master...fix-refs-u) http://git.io/vYDg6
slothbag has quit [Remote host closed the connection]
<ipfsbot>
[go-ipfs] jbenet force-pushed fix-gobuilder-dist from 8e588d3 to 470de92: http://git.io/vYDLp
<ipfsbot>
go-ipfs/fix-gobuilder-dist 470de92 Juan Batiz-Benet: updated gobuilder dist...
notduncansmith has joined #ipfs
notduncansmith has quit [Ping timeout: 244 seconds]
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<Luzifer>
lol. found a bug in go-vet… "constant 0 not a string in call to LogF" - yeah, right. first parameter expects to be uint. first parameter is uint. -_-
<ipfsbot>
[go-ipfs] jbenet opened pull request #1529: Fix CORS Support (master...cors-fix) http://git.io/vYyea
therealplato has joined #ipfs
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<rschulman__>
jbenet: Nice. Thanks for working on the CORS stuff
<rschulman__>
alu: Thanks, that was an nice read. :)
<alu>
mhm
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
tilgovi has joined #ipfs
<Luzifer>
yeah. I hate that unmarshal-fuck… "well no, I can't cast map[interface{}]interface{} to your struct" - so why the fuck do you create such a map?…
<Luzifer>
currently 8/8 build threads are hanging in an infinite for-loop as they are unable to read the build config… xD
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to cors-fix: http://git.io/vYy3Z
<ipfsbot>
go-ipfs/cors-fix 395b120 Juan Batiz-Benet: implement arbitrary HTTP header support...
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to master: http://git.io/vYy3g
<ipfsbot>
go-ipfs/master 075e687 Juan Benet: Merge pull request #1526 from ipfs/fix-gobuilder-dist...
Leer10 has quit [Ping timeout: 240 seconds]
tilgovi has quit [Ping timeout: 272 seconds]
<ipfsbot>
[go-ipfs] jbenet force-pushed cors-fix from 395b120 to e126194: http://git.io/vYyee
<ipfsbot>
go-ipfs/cors-fix a78fc4e Juan Batiz-Benet: Added API + Gateway support for arbitrary HTTP headers...
<ipfsbot>
go-ipfs/cors-fix fcc536d Juan Batiz-Benet: cmds/http: remove referrer check...
<ipfsbot>
go-ipfs/cors-fix e126194 Juan Batiz-Benet: implement arbitrary HTTP header support...
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<ipfsbot>
[go-ipfs] jbenet closed pull request #1444: Allow CORS for gateway handler. (master...gateway-cors) http://git.io/vqkzt
<whyrusleeping>
jbenet: its been a long time coming, but green tests are more frequent now :)
<jbenet>
:)
<ipfsbot>
[go-ipfs] jbenet pushed 1 new commit to dev0.4.0: http://git.io/vYyRE
<ipfsbot>
go-ipfs/dev0.4.0 25fd6b9 Juan Benet: Merge pull request #1488 from ipfs/s3-0.4.0...
<ipfsbot>
[go-ipfs] whyrusleeping created fix/diag (+1 new commit): http://git.io/vYyEQ
<ipfsbot>
go-ipfs/fix/diag f2b44e3 Jeromy: rmeove duplicate timeout option from diag...
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1530: rmeove duplicate timeout option from diag (master...fix/diag) http://git.io/vYyEN
<ipfsbot>
[go-ipfs] whyrusleeping created fix/stream-header (+1 new commit): http://git.io/vYyu1
<ipfsbot>
go-ipfs/fix/stream-header 2cad99d Jeromy: only set stream header on streamed output...
<ipfsbot>
[go-ipfs] whyrusleeping opened pull request #1531: only set stream header on streamed output (master...fix/stream-header) http://git.io/vYyuj
* whyrusleeping
gets off the couch to go make some coffee
Tv` has joined #ipfs
<rschulman__>
whyrusleeping: how’s that coffee looking?
<Luzifer>
woah… some people… the F5 key must be worn down at their keyboards… "oh my build is not yet ready, better hit F5 20 more times" *ragemode*
<whyrusleeping>
Luzifer: holding down ctrl+r works too :D
<whyrusleeping>
rschulman__: i have yet to grind it, waiting for the water to get a little hotter
<rschulman__>
I actually don’t drink coffee but my wife does and seems incapable of making her own, so I make coffee all the time.
<rschulman__>
all of which is incredibly pertinant to your life, obviously
<rschulman__>
which is why I shared it.
notduncansmith has joined #ipfs
notduncansmith has quit [Read error: Connection reset by peer]
<rschulman__>
jbenet: Is it possible to return a json reply when the web RPC gets a Path Resolve Error? Would just make it easier to parse if the returns were always json. Unless there’s a particular reason to have done it the way it is?
<jbenet>
api?
<jbenet>
or gateway
<jbenet>
the gateway ones will be nice error pages later on
<rschulman__>
api in this case, haven’t tried on the gateway
<jbenet>
api should be all json usually, when isn't it?
<jbenet>
(i mean ... the api has a gateway in it -- yeah thats the api's gateway)
<rschulman__>
oh, gotcha
<rschulman__>
oh, I think I see what you mean
<jbenet>
"api" is a bad name for it. "priviledged http server with a gateway and an api" is a more accurate description
<rschulman__>
as opposed to localhost:5001/cat/<hash>
<jbenet>
/api/v0/cat/<hash> -- but yeah
<rschulman__>
er, /api/v0/cat/<hash>
<rschulman__>
yeah
<rschulman__>
got it
<voxelot>
so the "api" connects to the daemon on localhost right? rschulman__ are you playing with ways to get that on a server so clients don't have to DL anything to use ipfs?
<voxelot>
been kinda confused by that work flow
<rschulman__>
voxelot: No, just trying to use the api from a javascript in browser
<rschulman__>
but users would still hav eto have ipfs installed
<rschulman__>
otherwise, you could call the public gateways
<voxelot>
think Eris did something like that if you want to check that out
<rschulman__>
but then you’re just recentralizing
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to fix/stream-header: http://git.io/vYyXY
<ipfsbot>
go-ipfs/fix/stream-header 98559d0 Jeromy Johnson: put comment back
<voxelot>
yeah, we just want to make things feel easy for the user, would there be a way for us to snoop traffic if we centralized the server
<jbenet>
yo mappum: can you provide QmYeRGr8PY3yigx2aTvK8B9e4HVUy3h4b2r3tsZHaYBduD ? am trying to see what's up with the CSRF problem. (several sites claim correct CORS should address this, but im starting to think only a session token would do it)
<jbenet>
gordonb hello \
<gordonb>
jbenet: howdy
<jbenet>
gordonb welcome! -- just saying hello. if want to discuss anything at any point lmk-- but otherwise dont mean to distract
<gordonb>
just hanging out
<mappum>
jbenet: I readded it just now, but I'm on a bad connection
<jpinbot>
now pinning /ipfs/QmRuhVrY9JXVwgATPjAPN9wY3qAj8RdDGyuYtu4nZv85iL/success.jpegthisshouldfail
<jpinbot>
[host 0] failed to grab refs for /ipfs/QmRuhVrY9JXVwgATPjAPN9wY3qAj8RdDGyuYtu4nZv85iL/success.jpegthisshouldfail: no link named "success.jpegthisshouldfail" under QmRuhVrY9JXVwgATPjAPN9wY3qAj8RdDGyuYtu4nZv85iL