<jbenet>
if we have no peers, we use the local records.
<jbenet>
(if they're valid).
<noffle>
jbenet: it looks like we actually return an error if there are no peers in the routing table; not use local value
<noffle>
(looking at routing/dht/routing.go:170
<noffle>
)
<jbenet>
like, the idea is this: "use the 'most valid' record you can find" -- if you're offline or have no peers, you can still find one. and that's why we search the DHT hard before using them.
<noffle>
jbenet: if we do an online query even if the EOL is valid, what is the semantic meaning of EOL? it *doesn't* mean we trust it to be up to date, as per what you've written, yes?
<jbenet>
noffle: I think when it comes to _records_, that's a bug. (DHT operation could very well error when it isn't "in connected / stable configuration", but then we should push up the use of the local records somewhere higher up in the call stack.
<jbenet>
it depends on the semantics we want the DHT to have. i think in general it is a good idea to get an error from the DHT when it's not connected to anything else. however, when getting _a record from the record system_ we should be able to use local records even when totally offline / disconnected.
<jbenet>
the distinction is subtle, but basically, the DHT is used to implement a record system. -- this is not very well specified in the current go-ipfs codebase because the spec moved forward.
<jbenet>
this is what the IPRS doc talks about -- https://github.com/ipfs/specs/tree/master/records -- @daviddias and I have been calling "the thing that takes care of this" -- the "Record Store". right now it is totally implicit in go-ipfs (because it was implemented as just the DHT), but it should become a package on its own.
<noffle>
jbenet: okay yes, the distinction between dht and record system makes sense -- agreed
<noffle>
I've been conflating these for days in my discussions -- shoot
Matoro has joined #ipfs
<jbenet>
im sorry, it's my fault. it took us months to realize there was a very important distinction there.
<jbenet>
one of the upshots is that the record store (or "iprs" as a thing) can then be layered upon things like DHTs, pub/sub, etc. very cleanly.
<noffle>
yes <3
<noffle>
okay, this helps. I'm going to do some reading on those links and think about this some more.
<noffle>
thanks for that
<voxelot>
daviddias: can you see if i'm doing something wrong with the protobuf, the one I created decodes fine but the encoding looks slightly different then config file
<noffle>
but it sounds like you agree with me that the end behaviour from the record system's PoV is: "if I can't find my key on the network but I have it locally, I'll use this local unexpired value"?
vanila has joined #ipfs
<vanila>
hello
<vanila>
is there anything ilke youtube, but ipfs backed?
<daviddias>
voxelot you can use the same file and fs.readFileSync it, this way you are sure to use the same and also in the future, if there are changes, we just need to update de proto file and not go into the code
<voxelot>
ohh good idea, so do a fs read on the cryto.proto so we are sure, thanks let me try that
<jbenet>
noffle: that's right. as long as it's valid. (validity may expire)
<daviddias>
noffle: in an IPRS world, the IPNS record is a type of IPRS record that points to the signature of the actual name record, this signature has the hash of the public key which is capable of validating the signature of the name record, this key will be stored and made available in the network as a MerkleDAG object, so that when you look up for the key, you
<daviddias>
will use something like the DAG service which will get you the key ( locally or network )
<jbenet>
validity may depend on the node's own position in spacetime.
<jbenet>
aside from the obvious time notion, it is possible in the future people may do things like geofencing records, like "when in Rome, resolve to A. otherwise, B".
pfraze has quit [Remote host closed the connection]
<jbenet>
this of course is up to the node to decide + opt to follow, but a bunch of other interesting things can fall out of app-defined validity constraints.
<daviddias>
Once you have the signature merkle DAG node, the public key merkle DAG node and the MerkleDAG node that has actually the name, you can execute the validity function ( see IPRS specs ) and check if it is still valid
<haad>
good morning
<alu>
sup fam
<voxelot>
morning gents
<voxelot>
err afternoon to alu
<daviddias>
jbenet we should add the geofencing idea to the IPRS doc as well :)
<alu>
*tip*
kaiza has quit [Quit: Leaving]
<jbenet>
is there a non-technical word for something like web-chain, like chain mail, or chain link fabrics. trying to use a name without the words "web", "merkle dag", "dag", or "secure"
<Kubuxu>
It is 2h, over 200MB, quite possibly streamed over IPFS from my server to the gateway.
<cojy>
so the gateway is doing something special to request it sequentially?
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to feat/mfs-flush-cmd: https://git.io/vgL1Z
<ipfsbot>
go-ipfs/feat/mfs-flush-cmd cd491ec Jeromy: Add a lock...
<cojy>
Kubuxu: well that's a problem
<cojy>
that's just 1 server with high bandwidth
<cojy>
i meant from many peers
<Kubuxu>
If this file was distributed in network it would use many peers
<Kubuxu>
It is not so it uses only few.
<cojy>
but would it sitll be able to stream
<Kubuxu>
yes
<cojy>
if each individual peer was below the bitrate of the media
simonv3 has quit [Quit: Connection closed for inactivity]
voxelot has quit [Ping timeout: 250 seconds]
<cojy>
how does ipfs schedule the downloads?
redfish has quit [Ping timeout: 250 seconds]
<Kubuxu>
it is bitswap, meaning: I have no idea.
redfish has joined #ipfs
r04r is now known as zz_r04r
<cojy>
that's surprising because vanilla bitorrent isn't like this at all so i didn't expect it to work
<cojy>
thanks
<Kubuxu>
cojy: you can request segments in bittorrent sequentially.
<cojy>
yes b ut it's not by default and doing it right is really hard
<cojy>
some nodes will be really slow and break it
<Kubuxu>
You get worst over all download speed (as you choose peers selectively) but it is possible.
<cojy>
you need to use a specially tuned client which as far as i know was proprietary
<cojy>
i havent looked recently though
<cojy>
yea that's what i mean that's a big issue
<cojy>
if you h it a block from a really slow peer that ruins the idea
Matoro has quit [Ping timeout: 245 seconds]
<noffle>
jbenet: what piece of go-ipfs qualifies as the "record system" you talked about above?
kaiza has quit [Ping timeout: 272 seconds]
<jbenet>
noffle: the DHT's storing of records and implementing the "Routing" interface.
patcon_ has joined #ipfs
Matoro has joined #ipfs
<noffle>
jbenet: how can you then distinguish between "getting a value from the dht" and "getting a value from the record system"? as per your comments above, they can have different semantics
<noffle>
e.g. dht failing when there are no peers, but the record system handing you the cached copy
<jbenet>
users of the record system should not have to care about that, as long as the records are valid, they're good.
ygrek_ has joined #ipfs
<jbenet>
can implement this by making a "record store" abstraction like i started in that go-iprs repo
<noffle>
jbenet: right, so this abstraction does NOT yet exist in go-ipfs :)
<jbenet>
which wraps the DHT, and makes its own decisions. OR can also implement it the other way around, make the DHT use a record store, and _not_ return the error.
<jbenet>
(we get to chose the semantics we want here, and am persnally not settled on which is better)
<jbenet>
noffle: that's correct.
null_rad- is now known as null_radix
SWingedSeraph has joined #ipfs
blarp1 has joined #ipfs
<noffle>
hm. the scope of this keeps on increasing
yellowsir1 has joined #ipfs
<noffle>
I should've realized "let's just make ipns keys resolve fast" wasn't a quick fix
HostFat has quit [Read error: Connection reset by peer]
<tmg>
[file] This is a directory - adding to playlist.
<jbenet>
noffle: yeah not easy. but it's also very important stuff to work on though.
<noffle>
yes
<noffle>
jbenet: thanks for the notes on the 'ipfs mount' issue
jhulten_ has quit [Ping timeout: 250 seconds]
jamie_k has quit [Ping timeout: 248 seconds]
jamie_k has joined #ipfs
reit has quit [Quit: Leaving]
vanila has quit [Quit: Leaving]
dignifiedquire has quit [Quit: Connection closed for inactivity]
<The_8472>
<jbenet> is there a non-technical word for something like web-chain, like chain mail, or chain link fabrics. <- a mesh?
jamie_k has quit [Ping timeout: 240 seconds]
<jbenet>
The_8472 yeah i've been using "mesh", like "the net" -> "the web" -> "the mesh"
<The_8472>
latticework
<jbenet>
but it's also confusing because mesh network protocols
simonv3 has joined #ipfs
lmatteis has quit [Quit: Connection closed for inactivity]
<The_8472>
<jbenet> requirements are right. can start validating the records beforehand. the point is to get 16 _valid_ records. or Exhaust the dht query
kvda has joined #ipfs
pfraze has quit [Remote host closed the connection]
<The_8472>
in my experience it's much better to optimistically return the first result as soon as you get it with the option to invalidate it later
<The_8472>
that way the next phase of your lookup can start speculatively
<The_8472>
DHT time-to-first-result is (or can be) much much lower than time-to-lookup-completion
jamie_k has joined #ipfs
<jbenet>
The_8472 yes, indeed, but right now we're talking about the result externalized to the client. there is no way to "return a invalidate-promise but keep resolving" that is checked before externalizing to the app.
<jbenet>
The_8472 entirely agreed that we should do that :) -- just not there yet.
<jbenet>
cc noffle for consideration o/
<jbenet>
Kubuxu: if you get a chance to make one of the 3D logos with the white IPFS in front, we can make stickers with it
<jbenet>
(i can go do that later this weekend too)
parkan has joined #ipfs
jhulten_ has joined #ipfs
pfraze has joined #ipfs
computerfreak has quit [Quit: Leaving.]
yellowsir1 has quit [Remote host closed the connection]
disgusting_wall has quit [Quit: Connection closed for inactivity]
nuun has joined #ipfs
kaiza has joined #ipfs
patcon_ has quit [Ping timeout: 272 seconds]
voxelot has joined #ipfs
jhulten_ has quit [Ping timeout: 240 seconds]
ygrek_ has quit [Ping timeout: 260 seconds]
gaboose has quit [Ping timeout: 245 seconds]
hellertime has joined #ipfs
jamie_k has quit [Quit: jamie_k]
tmg has quit [Ping timeout: 240 seconds]
pfraze has quit [Remote host closed the connection]
pfraze has joined #ipfs
<nuun>
what similarites are between the OSD standard and how IPFS stores?
<voxelot>
nuun: i don
<nuun>
I recently encountered documentation in the linux kernel documentation about exofs and OSD. It seemed like something similar to how IPFS intends to have data presented.
<voxelot>
*don't know much about OSD but most of the IPFS datatype storage is something like JSON, but OSD seems to be more of file storage, which i think ipfs does more of a regular block storage
<voxelot>
alu: sounds trippy, show me at nullspace
aquilax has quit [Ping timeout: 276 seconds]
<alu>
aite
tymat has joined #ipfs
tymat has quit [Client Quit]
tymat has joined #ipfs
tymat has quit [Client Quit]
tymat has joined #ipfs
alu has quit [Quit: WeeChat 1.4-dev]
tymat has quit [Client Quit]
tymat has joined #ipfs
alu has joined #ipfs
tymat has quit [Quit: Leaving]
pfraze has quit [Remote host closed the connection]
Not_ has quit [Remote host closed the connection]
tymat has joined #ipfs
tymat has quit [Client Quit]
tymat has joined #ipfs
tymat has quit [Client Quit]
tymat has joined #ipfs
tymat has quit [Client Quit]
<daviddias>
morning everyone :)
tymat has joined #ipfs
<alu>
Hey
jhulten_ has quit [Ping timeout: 245 seconds]
<haad>
hello!
<haad>
daviddias: got some cool pics for you from (underwater) Raja Ampat later this week, as soon as I sort them out
<daviddias>
hey :)
<daviddias>
niiiiicee! :D
<daviddias>
I so want to see them
<haad>
I tell you, it was mind blowing! you need to go there one day
<haad>
absoutely incredible
<daviddias>
my friends have been pitching me the idea of one of those dive holidays on a boat with 3 dives a day, but I'm sure if I talk about the trip you organized, they will be blown away
<haad>
daviddias: I also went ona liveaboard, 4 dives a day :)
<daviddias>
ok, that is really the one to do :D
<haad>
there's not much else to do ;)
<daviddias>
well, do you needed more? :)
<haad>
nope, I was totally happy to just dive, eat, sleep, and repeat :)
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to feat/mfs-flush-cmd: https://git.io/vgtR3
<ipfsbot>
go-ipfs/feat/mfs-flush-cmd ccb5ed4 Jeromy: cleanup a bit...
<daviddias>
.tell vijayee that I need some kind of async way to reach out to him when he is not in IRC, since he doesn't have a bouncer on IRC :)
<daviddias>
oh, we don't have multivac anymore? I loved multivac
flounders has quit [Ping timeout: 240 seconds]
tymat has quit [Ping timeout: 276 seconds]
<ipfsbot>
[go-ipfs] whyrusleeping pushed 2 new commits to master: https://git.io/vgt0m
<ipfsbot>
go-ipfs/master aaa6569 Thomas Gardner: trivial: various superficial fixes...
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feature/object: https://git.io/vgt6x
<ipfsbot>
js-ipfs/feature/object f28a0f3 David Dias: links and stat
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feature/object: https://git.io/vgtPI
<ipfsbot>
js-ipfs/feature/object 80d098f David Dias: data
jhulten_ has joined #ipfs
s_kunk has joined #ipfs
hashcore has quit [Ping timeout: 264 seconds]
jhulten_ has quit [Ping timeout: 260 seconds]
hashcore has joined #ipfs
mildred has quit [Ping timeout: 276 seconds]
ygrek_ has quit [Ping timeout: 245 seconds]
<Kubuxu>
jbenet: no problem, I just needed to know the font as it was too few letters for font detector to work.
Tv` has quit [Quit: Connection closed for inactivity]
sivachandran has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
<sivachandran>
Can anyone tell how do I know the download status of pinned object? Say I've pinned a 100GB file and it is getting downloaded. Now how do I tell whether the object is fully downloaded or not?
<computerfreak>
is there any Idea to have decentral user-authentication over ipfs-files as databases?
<computerfreak>
so users can have an webapp as ipfs-folder (like webui) and ''login'' to get their keys etc... by filling in username/password and search for the resulting hash in an (couch?)Database from an ipfs/ipns Link(hash) ...
joshbuddy has joined #ipfs
<xelra>
computerfreak: AFAIK, extensive access control features are planned with ipfs-cluster. You can find 2 issues regarding that on GitHub.
<tmg>
dignifiedquire: can you gen and add an arbitrary hash for me?
<dignifiedquire>
hpk: who wrote "no api is the best API" in js?
<dignifiedquire>
lol nvm read the link
<dignifiedquire>
and how exactly has power-assert no api -.- it has loads of functions
<hpk>
because you can use the assert statement/function
<hpk>
the thing is that usage requires no api, of course its implementation has functions :)
<dignifiedquire>
hmmm
<dignifiedquire>
sounds fishy to me
<dignifiedquire>
everything related to testing in js is lightyears behind what pytest does anyway
<tmg>
dignifiedquire: are you still holding QmahuxqczQ1aAmpVt3MjEVg93L64vnro1ubKGnociBNFT9? can you /msg me your peerid :^)
pfraze has joined #ipfs
joshbuddy has quit [Quit: joshbuddy]
voxelot has joined #ipfs
<ipfsbot>
[go-ipfs] RichardLitt force-pushed docs/cleanup-id-mans from babd707 to d5153db: https://git.io/vgqDK
<ipfsbot>
go-ipfs/docs/cleanup-id-mans d5153db Richard Littauer: Cleaned up `ipfs id` mans...
reit has quit [Ping timeout: 250 seconds]
hashcor has joined #ipfs
The_8472 has joined #ipfs
voxelot has quit [Ping timeout: 240 seconds]
sivachandran has quit [Quit: Leaving...]
jhulten_ has joined #ipfs
jhulten_ has quit [Ping timeout: 240 seconds]
Matoro has quit [Ping timeout: 272 seconds]
lmatteis has joined #ipfs
kayaelle has joined #ipfs
elima has quit [Ping timeout: 250 seconds]
<kayaelle>
Hello all - New to IPFS. Really amped about it. Have a question. I added some image files while running the daemon on a server. (I have ipfs on my computer as well). If I stop the daemon, the only files I can access are the ones that I already clicked on from my peer id's gateway. Assuming that's because my computer's node is hosting those files. Does it take some time for files to get picked up by other peers? Thanks
<dignifiedquire>
kayaelle: if nobody on another peer requests the file, they will not load the file
<kayaelle>
dignifiedquire: ok - that's what I thought. I was hoping to serve these images from a web app but unless my daemon is running or someone else using ipfs requests them, they won't be accesible.
hashcor has quit [Ping timeout: 252 seconds]
TheWhisper has quit [Quit: Leaving]
TheWhisper has joined #ipfs
<kayaelle>
dignifiedquire: thanks :)
elima has joined #ipfs
ecloud_ is now known as ecloud
ashark has joined #ipfs
jamie_k has joined #ipfs
mildred has quit [Ping timeout: 245 seconds]
elima has quit [Ping timeout: 276 seconds]
<misalias>
autopeer?
* misalias
autopeer/text
zootella has joined #ipfs
Tv` has joined #ipfs
zootella has quit [Client Quit]
tmg has quit [Ping timeout: 240 seconds]
aquilax has joined #ipfs
reit has joined #ipfs
elima has joined #ipfs
Hory_ has joined #ipfs
dlight has joined #ipfs
flounders has quit [Quit: leaving]
voxelot has joined #ipfs
voxelot has quit [Changing host]
voxelot has joined #ipfs
dlight has quit [Client Quit]
kerozene has joined #ipfs
ylp has quit [Quit: Leaving.]
vijayee has joined #ipfs
m0ns00nfup has quit [Quit: undefined]
jhulten_ has joined #ipfs
jhulten_ has quit [Ping timeout: 276 seconds]
reit has quit [Quit: Leaving]
<dignifiedquire>
whyrusleeping: jbenet and anybody else who knows things can you explain to me how ethereum storage is supposed to be reliable? according to their whitepaper I'm supposed to split my data into as many blocks as possible and distribute them, but without replication + abillity to restore missing blocks how is that even comparable witg storing on my local disk
<dignifiedquire>
in terms of reliability
Matoro has joined #ipfs
jamie_k has quit [Ping timeout: 240 seconds]
jamie_k has joined #ipfs
m0ns00nfup has joined #ipfs
m0ns00nfup has quit [Client Quit]
Encrypt has joined #ipfs
reit has joined #ipfs
takinbo has quit [Ping timeout: 276 seconds]
shyamsk has quit [Ping timeout: 276 seconds]
vakla has quit [Ping timeout: 276 seconds]
reit has quit [Client Quit]
shyamsk has joined #ipfs
vakla has joined #ipfs
Ragnis has joined #ipfs
rtlong has quit [Ping timeout: 276 seconds]
takinbo has joined #ipfs
rtlong has joined #ipfs
disgusting_wall has joined #ipfs
randomguy has joined #ipfs
lgierth_ is now known as lgierth
<voxelot>
dignifiedquire: it seems the key piece to that is making sure nodes don't choose to forget your data
<dignifiedquire>
that sounds like gambling
<voxelot>
and you can verify that by watching the contract, if it is paying out, that means someone is verifying that they have your data
<dignifiedquire>
okay so I will be notified of my data is lost, but I still can't recover it
<voxelot>
that would seem to be the case
<voxelot>
unless i'm missing something
<dignifiedquire>
it just needs a single hard drive of a single node to be corrupted
<voxelot>
well hopefully more than one node would have that piece of data that got randomly selected
<dignifiedquire>
so if I want to use this storage I need to build raid myself on top of it :/
<voxelot>
just the first to reply gets paid so
<dignifiedquire>
but if the data is not part of the chain how will it get replicated?
<voxelot>
i think it maybe it gives incentives to people to keep the data and try to get htat payment though
<voxelot>
via ipfs :)
<dignifiedquire>
hmmm there is no mention of such things in the "whitepaper"
<dignifiedquire>
but ipfs does not replicate if noone requests
<voxelot>
yeah, that was written awhile back, i wonder if anyone has thought out the decentralized file storage idea more
<voxelot>
maybe filecoin haha
<dignifiedquire>
and if this is my personal encrypted data nobody will
<voxelot>
well they might if you are paying them to
ulrichard has quit [Read error: Connection reset by peer]
<dignifiedquire>
sounds fishy to me all in all
<lgierth>
whyrusleeping: Kubuxu: i'm pretty sure the listening on individual addresses has something to with how swarm handles addresses per peer and so on
<voxelot>
yeah i just trust whyrusleeping is storing my data somewhere
<dignifiedquire>
lol
<lgierth>
whyrusleeping: finishing touches on the docker stuff right now -- i didn't make it to the computer yesterday :/
<whyrusleeping>
voxelot: about that...
<voxelot>
noooooo
<whyrusleeping>
lgierth: no worries, i got sidetracked and worked on something else
<dignifiedquire>
whyrusleeping: don't go around loosing our data :P
<voxelot>
daviddias: reading crypto.proto into fs has the same affect as using the schema I wrote, I can't get the output protobufs encodings to exactly match go's, fairly certain i'm pulling the right cryptographic info as well
jamie_k has quit [Quit: jamie_k]
jamie_k_ has joined #ipfs
<dignifiedquire>
woot "Memory, an infinitely expandable byte array" looks like ethereum can replace my computer
<dignifiedquire>
whyrusleeping: lets just store everything in this magical infintely expandable byte array :)
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to feat/debug-events: https://git.io/vgmay
<ipfsbot>
go-ipfs/feat/debug-events e0fb690 Jeromy: better connect to provider printing...
<dignifiedquire>
no need for you building you supernode anymore
<xicombd_>
hey all, I’ve just published a new version of my chrome extension, if you can, please take it for a spin and let me know your thoughts :)
jaboja has quit [Remote host closed the connection]
wiedi has quit [Read error: Connection reset by peer]
jamie_k_ has quit [Quit: jamie_k_]
jamie_k has joined #ipfs
jhulten_ has joined #ipfs
mildred has quit [Ping timeout: 240 seconds]
<lidel>
Kubuxu, thank you :) by the way, do you think I could use your shaded SVG? The version I've seen yesterday was really nice.
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feature/object: https://git.io/vgmAC
<ipfsbot>
js-ipfs/feature/object 8e4b058 David Dias: apply CR comments
<Kubuxu>
lidel: it will be becoming official soon, so of course, I don't know how licensing will work but it is CC either way.
mildred has joined #ipfs
<Kubuxu>
(working on nodejs script to generate all those rasters).
m0ns00nfup has joined #ipfs
jhulten_ has quit [Ping timeout: 264 seconds]
<lidel>
Kubuxu, we discussed licensing in https://github.com/ipfs/logo/issues/1 If you will make a PR to ipfs/logo, then it automatically be CC BY-SA :)
ianopolous has joined #ipfs
<lidel>
xicombd_, I am not using Chromium as my daily driver, but it looks really good! Number of peers on a badge is neat, I will reuse this idea :)
<ipfsbot>
[js-ipfs] diasdavid deleted feature/object at 8e4b058: https://git.io/vgmhW
<xicombd_>
lidel awesome :)
pepesza has quit [Excess Flood]
wiedi has joined #ipfs
pepesza has joined #ipfs
m0ns00nfup has quit [Quit: undefined]
pepesza has quit [Client Quit]
pepesza has joined #ipfs
<xicombd_>
lidel dignifiedquire: was thinking that we could try to extract the station UI as a separate module and re-use it on the different projects
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to master: https://git.io/vgYeY
<ipfsbot>
js-ipfs/master 70ed97b David Dias: Release v0.3.0.
<dignifiedquire>
xicombd_: that would be nice for sure, I just haven't done that yet because I wasn't sure if I would settle on this UI (and I didn't expect people to actually reuse it tbh)
ygrek_ has joined #ipfs
<alu>
Is it fair to say IPFS distributes infrastructure?
jamie_k has quit [Quit: jamie_k]
<lgierth>
i'm power-cycling 3 gateway/bootstrap nodes whose network connection had somehow locked up -- if you noticed slow loading times or timeouts, that should be fixed now
Matoro has quit [Ping timeout: 272 seconds]
<whyrusleeping>
alu: that would be fair to say
m0ns00nfup has joined #ipfs
Peer3Peer has joined #ipfs
<ipfsbot>
[go-ipfs] whyrusleeping pushed 1 new commit to feat/mfs-flush-cmd: https://git.io/vgYk3
elima has quit [Remote host closed the connection]
* bigbluehat
stumbles in to let the world know that copy.com is shutting down--on May 1st
<bigbluehat>
yet another fabulous opportunity for a sanity-inducing "land grab" in the decentralized storage space world ^_^
<bigbluehat>
just sayin' ;)
Encrypt has joined #ipfs
HostFat has quit [Read error: Connection reset by peer]
<whyrusleeping>
mmmm
<whyrusleeping>
if only i had the money
mildred has joined #ipfs
<richardlitt>
whyrusleeping: ipfs object links should return what object types?
<richardlitt>
Of, what content type
<richardlitt>
JSON, right?
rendar has quit [Ping timeout: 250 seconds]
HostFat has joined #ipfs
<jbenet>
richardlitt: depends on the encoding you request
<jbenet>
richardlitt: this is api, right?
<richardlitt>
jbenet: exactly! Great. What encodings **should** I be able to request?
<jbenet>
bigbluehat: thanks! yeah that sucks.
<richardlitt>
jbenet: yes.
<jbenet>
richardlitt: for now, JSON, Protobuf, XML. in the future we'll also have {CBOR, YML, CSON}
<richardlitt>
I've found, experimentally, that I can get XML back. Protobuf works for some subcommands, but not for `ipfs object links`, strangely.
<jbenet>
yeah we should have a table somewhere showing what encodings are available where
<richardlitt>
OK. So that is definitely two bugs then: 1) not documenting encoding in the ipfs man for ipfs object links 2) not marshalling protobuf proerly
<jbenet>
ideally this should be doable with everything.
<jbenet>
well, ipfs object links likely wont return protobuf as we dont have a protobuf for just ipfs object links.
<richardlitt>
It doesn't.
<bigbluehat>
jbenet: yeah...a "house-to-house" encrypted, private IPFS would pretty much make my year!...and it's only February!!1! ;)
<richardlitt>
But `ipfs object get` does return protobuf
<richardlitt>
Just, interesting.
<jbenet>
bigbluehat :) \o/ let's do it
<richardlitt>
ok. Will note this.
<jbenet>
richardlitt: yes, the object itself is encoded in protobuf, that's the native format, too
<richardlitt>
back in a bit, just realised sun is setting and I need to go for a run now if I don't want to be running in the dark.
<jbenet>
richardlitt: extracting only the links and then outputting them would require another protobuf (that's how protobuf works)
<jbenet>
richardlitt: sounds good
rendar has joined #ipfs
cemerick has quit [Ping timeout: 264 seconds]
Matoro has joined #ipfs
m0ns00nfup has joined #ipfs
ianopolous has quit [Ping timeout: 245 seconds]
montagsoup has joined #ipfs
lmatteis has joined #ipfs
mildred has quit [Ping timeout: 252 seconds]
ugjka has quit [Ping timeout: 250 seconds]
<whyrusleeping>
lgierth: docker things????
m0ns00nfup has quit [Quit: undefined]
afdudley has joined #ipfs
ianopolous has joined #ipfs
joshbuddy has joined #ipfs
Edur has joined #ipfs
ashark has quit [Ping timeout: 240 seconds]
pfraze has quit [Remote host closed the connection]