<lvernschrock>
(I'm referring to the link in the topic for this channel.)
<whyrusleeping>
lvernschrock: ah, thanks
hoboprimate has quit [Quit: hoboprimate]
lvernschrock has quit [Ping timeout: 240 seconds]
libman has quit [Quit: Connection closed for inactivity]
jamiew has joined #ipfs
infinitesum has joined #ipfs
ashark has joined #ipfs
zuck05 has quit [Read error: Connection reset by peer]
ashark has quit [Ping timeout: 255 seconds]
skeuomorf has quit [Ping timeout: 260 seconds]
palkeo has quit [Ping timeout: 255 seconds]
shizy has joined #ipfs
zuck05 has joined #ipfs
shizy has quit [Ping timeout: 240 seconds]
clavi has joined #ipfs
infinitesum has quit [Quit: infinitesum]
talonz has quit [Remote host closed the connection]
infinitesum has joined #ipfs
appa has quit [Ping timeout: 240 seconds]
appa has joined #ipfs
infinitesum has quit [Quit: infinitesum]
infinitesum has joined #ipfs
ashark has joined #ipfs
zuck05 has quit [Ping timeout: 240 seconds]
ashark has quit [Ping timeout: 240 seconds]
owlet has quit [Ping timeout: 240 seconds]
kaotisk has joined #ipfs
akt has joined #ipfs
shizy has joined #ipfs
talonz has joined #ipfs
lvernschrock has joined #ipfs
zuck05 has joined #ipfs
infinitesum has quit [Quit: infinitesum]
infinitesum has joined #ipfs
lvernschrock has quit [Quit: Leaving.]
jamiew has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
infinitesum has quit [Quit: infinitesum]
ashark has joined #ipfs
ashark has quit [Ping timeout: 255 seconds]
ashark has joined #ipfs
ashark has quit [Ping timeout: 240 seconds]
olivernyc has joined #ipfs
jaboja has quit [Ping timeout: 255 seconds]
The_8472 has quit [Ping timeout: 264 seconds]
The_8472 has joined #ipfs
<olivernyc>
Hey, I added a bunch to static.network, anyone around for some feedback?
<whyrusleeping>
i'm getting a weird error on static.network
<whyrusleeping>
'cannot read property _currentElement on null'
<olivernyc>
Hm is this your first visit?
<whyrusleeping>
no, i posted something and then refreshed the page
<whyrusleeping>
i actually cant seem to use the app now
<whyrusleeping>
it throws that error every time i open it
<olivernyc>
Maybe delete the site data and try again?
<olivernyc>
You're using a current chrome or firefox?
chris613 has joined #ipfs
<whyrusleeping>
current chrome
<olivernyc>
Can you go to application > clear site data in dev tools?
<olivernyc>
application > clear storage > clear site data
<whyrusleeping>
olivernyc: that fixed it
<olivernyc>
Great :)
infinity0 has joined #ipfs
<whyrusleeping>
:D
dimitarvp` has quit [Read error: Connection reset by peer]
ashark has joined #ipfs
<whyrusleeping>
olivernyc: hrm... can i not change my profile picture?
ashark has quit [Ping timeout: 260 seconds]
alu has quit [Quit: WeeChat 1.0.1]
<olivernyc>
Fixed, pushing update now
wking has joined #ipfs
shizy has quit [Ping timeout: 268 seconds]
ashark has joined #ipfs
chris613 has quit [Quit: Leaving.]
olivernyc has quit [Ping timeout: 240 seconds]
ashark has quit [Ping timeout: 246 seconds]
olivernyc has joined #ipfs
olivernyc has quit [Ping timeout: 240 seconds]
cyanobacteria has joined #ipfs
cyanobacteria has quit [Client Quit]
olivernyc has joined #ipfs
cyanobacteria has joined #ipfs
cyanobacteria has quit [Client Quit]
cyanobacteria has joined #ipfs
cyanobacteria has quit [Client Quit]
_whitelogger has joined #ipfs
kaotisk has quit [Ping timeout: 240 seconds]
kaotisk has joined #ipfs
pekman has joined #ipfs
espadrine_ has joined #ipfs
olivernyc has quit [Ping timeout: 255 seconds]
olivernyc has joined #ipfs
oliverny1 has joined #ipfs
Mateon1 has quit [Ping timeout: 240 seconds]
Mateon1 has joined #ipfs
maxlath has joined #ipfs
rendar has joined #ipfs
ShalokShalom has joined #ipfs
Caterpillar has joined #ipfs
gde33 has quit [Ping timeout: 246 seconds]
ashark has joined #ipfs
gde33 has joined #ipfs
maxlath has quit [Ping timeout: 245 seconds]
ashark has quit [Ping timeout: 255 seconds]
ilyaigpetrov has joined #ipfs
oliverny1 has quit [Read error: Connection reset by peer]
oliverny1 has joined #ipfs
_whitelogger has joined #ipfs
m0ns00n_ has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 255 seconds]
maxlath has joined #ipfs
f_ has quit [Read error: Connection reset by peer]
sirdancealot has quit [Ping timeout: 246 seconds]
olivernyc has quit [Ping timeout: 260 seconds]
oliverny1 has quit [Ping timeout: 255 seconds]
pekman has quit [Quit: WeeChat 1.0.1]
olivernyc has joined #ipfs
oliverny1 has joined #ipfs
olivernyc has quit [Ping timeout: 240 seconds]
olivernyc has joined #ipfs
ShalokShalom has quit [Remote host closed the connection]
olivernyc has quit [Ping timeout: 260 seconds]
oliverny1 has quit [Ping timeout: 260 seconds]
olivernyc has joined #ipfs
oliverny1 has joined #ipfs
espadrine_ is now known as espadrine
olivernyc has quit [Ping timeout: 240 seconds]
oliverny1 has quit [Ping timeout: 240 seconds]
appa has quit [Ping timeout: 246 seconds]
appa has joined #ipfs
A124 has quit [Read error: No route to host]
A124 has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 260 seconds]
m0ns00n_ has quit [Quit: quit]
appa has quit [Ping timeout: 260 seconds]
appa has joined #ipfs
Guest259688[m] has joined #ipfs
<Guest259688[m]>
hello
appa has quit [Ping timeout: 240 seconds]
droman has joined #ipfs
btmsn has joined #ipfs
maxlath has quit [Ping timeout: 240 seconds]
Mateon3 has joined #ipfs
Mateon1 has quit [Ping timeout: 240 seconds]
Mateon3 is now known as Mateon1
appa has joined #ipfs
maxlath has joined #ipfs
m0ns00n_ has joined #ipfs
jamiew has joined #ipfs
ashark has joined #ipfs
dimitarvp has joined #ipfs
mahloun has joined #ipfs
ashark has quit [Ping timeout: 255 seconds]
Boomerang has joined #ipfs
Encrypt has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 260 seconds]
chris613 has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 240 seconds]
tglman has joined #ipfs
Yatekii has left #ipfs [#ipfs]
appa has quit [Ping timeout: 240 seconds]
reit has quit [Ping timeout: 260 seconds]
appa has joined #ipfs
_rht has joined #ipfs
jaboja has joined #ipfs
m0ns00n_ has quit [Quit: quit]
jaboja has quit [Ping timeout: 255 seconds]
jaboja has joined #ipfs
m3lt has joined #ipfs
Encrypt has quit [Quit: Quit]
btmsn has quit [Read error: Connection reset by peer]
btmsn has joined #ipfs
appa has quit [Ping timeout: 240 seconds]
maxlath has quit [Ping timeout: 246 seconds]
owlet has joined #ipfs
libman has joined #ipfs
jaboja has quit [Ping timeout: 268 seconds]
m3lt has quit [Ping timeout: 246 seconds]
jmill has joined #ipfs
reit has joined #ipfs
appa has joined #ipfs
infinitesum has joined #ipfs
jaboja has joined #ipfs
Boomerang has quit [Quit: Lost terminal]
Encrypt has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 246 seconds]
maxlath has joined #ipfs
tglman has quit [Ping timeout: 240 seconds]
bbence[m] has joined #ipfs
akt has quit [Ping timeout: 255 seconds]
athan_ has joined #ipfs
infinitesum has quit [Quit: infinitesum]
infinitesum has joined #ipfs
baffo32 has joined #ipfs
m3lt has joined #ipfs
dconroy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
m3lt has quit [Ping timeout: 240 seconds]
m3lt has joined #ipfs
palkeo has joined #ipfs
palkeo has joined #ipfs
palkeo has quit [Changing host]
_rht has quit [Quit: Connection closed for inactivity]
palkeo has quit [Ping timeout: 240 seconds]
_shizy has quit [Ping timeout: 240 seconds]
jmill has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
screensaver has joined #ipfs
chris613 has quit [Quit: Leaving.]
guybrush has quit [Excess Flood]
guybrush has joined #ipfs
guybrush has joined #ipfs
guybrush has quit [Changing host]
JayCarpenter has joined #ipfs
m3lt has quit [Quit: leaving]
atrapado_ has joined #ipfs
droman has quit []
palkeo has joined #ipfs
screensaver1 has joined #ipfs
Jesin has quit [Quit: Leaving]
Encrypt has quit [Quit: Quit]
screensaver1 has quit [Remote host closed the connection]
Jesin has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
jaboja has quit [Ping timeout: 240 seconds]
ygrek_ has joined #ipfs
MrSparkle has quit [Ping timeout: 255 seconds]
MrSparkle has joined #ipfs
espadrine has quit [Ping timeout: 240 seconds]
DiCE1904 has quit [Ping timeout: 260 seconds]
DiCE1904 has joined #ipfs
<pinkieval>
is there an application that provides an S3-like API and stores files in IPFS?
espadrine has joined #ipfs
Encrypt has joined #ipfs
rcat has joined #ipfs
oliverny1 has joined #ipfs
olivernyc has joined #ipfs
skeuomorf has joined #ipfs
<libman>
Seems like you're adding a horse to a motorcar. Why do you need a "S3-like API"?
<kythyria[m]>
S3 assumes the user will pick the bucket and object names, top
pcre has joined #ipfs
<kythyria[m]>
IPFS isn't really S3-shaped
mahloun has quit [Read error: Connection reset by peer]
<libman>
In the beginning there was rsync, and everyone was happy. Then S3 came along with more limited functionality, for Amazon's convenience and no one else's.
<libman>
(Now there's also `rclone`, which does everything that `rsync` does but with various kinds of cloud API support, including S3.)
<pinkieval>
libman: make it work with any software with an S3 backend
<kythyria[m]>
pinkieval: Can't. The bucket name or object path will change as you add files.
<pinkieval>
good point
<kythyria[m]>
Well.... I guess if there was a way to use an IPNS identifier and update it each time you do that sort of thing.
<pinkieval>
what about storing the bucket name in ipns?
<kythyria[m]>
But that would be terribly racy.
<pinkieval>
or just store the list of objects in the bucket outside ipfs
oliverny1 has quit [Read error: Connection reset by peer]
oliverny1 has joined #ipfs
dconroy has joined #ipfs
<pinkieval>
libman: I don't know S3's API much, but it looks like pretty standard use of HTTP's verbs
<pinkieval>
better than most APIs nowadays that use GET and POST to do everything
<libman>
I didn't like the way IPFS stored files prior to --nocopy. Now I'm tinkering with a shell script that adds modified files in place and keeps a .ipfs.ls in each public directory remembering the hashes.
<libman>
There should be one file access API: the file system. You can access everything else through FUSE.
<pinkieval>
Yeah, --nocopy is great. It absence was the biggest point preventing me from using IPFS
<libman>
Operating Systems should have always had a kernel / FS API for accessing a file based on checksum (multihash-to-inode instead of path-to-inode lookup). You'd then be able to expose that to servers, and that'd be more than half of IPFS functionality right there.
<libman>
Modern advanced filesystems (ZFS, btrfs, hammer, etc) store the checksum of every file already. Much duplication of effort between the layers.
oliverny1 has quit [Ping timeout: 240 seconds]
olivernyc has quit [Ping timeout: 260 seconds]
<lemmi>
the checksums are only fast checksums, that almost cost nothing to compute, but they arent necessary hashes you'd call cryptographic. their use is to detect whether a file is still ok or not. collisions are likely.
<lemmi>
but an on-disk filesystems for ipfs sounds interesting, yes
olivernyc has joined #ipfs
<libman>
Ideally it should let you set which hash func to use for which files.
oliverny1 has joined #ipfs
<libman>
I think AWS S3 was always a bad idea. At every point you could get normal web hosting that was cheaper, didn't lock you in, and less prone to censorship.
infinitesum has quit [Quit: infinitesum]
<pinkieval>
libman: that's not a protocol issue
<pinkieval>
only politics/mony
<pinkieval>
only politics/money *
<kythyria[m]>
> There should be one file access API: the file system. You can access everything else through FUSE.
<kythyria[m]>
This works great until you try and mount something that isn't the widely used subset of POSIX filesystems.
<libman>
RamNode accepts Bitcoin, and cheap HDD storage at 120gb costs 2.225 cgm (cents per GB per month; that's $2.67 a month or $8 per quarter). That's about same as S3, but you get a real server which can do anything and serve your files over any protocol.
<kythyria[m]>
That's... not what S3 is about. It's what EC2 is about, though.
<libman>
AWS forces you to separate storage from computation. That's in Amazon's interest, not yours.
<pinkieval>
libman: at Online.net, you can get 10€/TB/month storage
<kythyria[m]>
libman: I fail to see your argument here
<pinkieval>
you need a server that is at least 30€/month to access it, though
<libman>
Yup, it's a competitive free market of servers you can access with normal Unix tools like rsync and run nginx, ipfs, bittorrent, and anything else.
athan_ has quit [Quit: Lost terminal]
<sknebel>
regarding "that'd be terribly racy", would be interesting to see if it would the that different from the consistency guerantees of S3
<kythyria[m]>
libman: At the cost of having to set it up and manage it yourself, including all the stuff S3 does for you, which is the point of S3 in the first place. And you're still comparing different kinds of product.
<sknebel>
I guess the read-after-write consistency for new objects would require some kind of local cache
<libman>
Proprietary cloud protocols were the Microsoft of the last decade, inferior and overpriced. There are now F/L/OSS implementations, but they're still not worth using.
<libman>
Actually the prices I've quoted include bandwidth, for which Amazon charges you extra.
<kythyria[m]>
The only time you've actually compared comparable products is B2 vs S3
m0ns00n_ has joined #ipfs
<libman>
I know I'm comparing different kinds of fruit, that doesn't invalidate my argument.
<kyledrake>
The best way to really get crazy with capabilities is to put together a ceph cluster in a colo with unmetered transit. The cloud junk really isn't designed for lots of small chunks of data.
<sknebel>
libman: S3-style protocols have multiple open-source implementations, both client- and server-side, multiple vendors ship storage devices offering them, ... They are not an AWS-only thing
<libman>
sknebel: I already said that. Read more carefully.
<kyledrake>
Also they charge way, way, way, way, way, way, way, way too much for bandwidth, all of them. I'm paying $0.00125/GB for transit and if you're paying more than that you're not paying market rates.
<kythyria[m]>
libman: You've basically said "S3-esque protocols aren't worth using because VPSes are cheaper"
<kythyria[m]>
So far as I can tell, anyway
<kythyria[m]>
And that doesn't even make sense
<libman>
cheaper but also better and more flexible.
<sknebel>
and you totally can use S3-esque protocols in your cluster of VPSes
<kythyria[m]>
So... by that standard Apache mod_dav doesn't make sense?
<libman>
A good example (given the channel we're in) is IPFS.
<kythyria[m]>
Good luck using IPFS for a store you can mutate one file at a time.
<kyledrake>
With IPFS using S3 you're essentially making an HTTP call for each 256k chunk or whatever it is, and AWS also charges per-request. It's dog slow for pulling together a lot of those chunks at once
<libman>
Um, you can link to one file at a time too.
<kythyria[m]>
Why would you run IPFS on top of S3 unless you completely ignored the bit where you can attach huge volumes to an EC2 instance?
<sknebel>
kyledrake: I think the question was about building an S3 API over IPFS, not about using S3 as the backend for IPFS
<libman>
So you have a server storing some files. You can access those files via HTTP, IPFS, SCP (incl rsync), FTP, etc, etc, etc. Why do you need to access them with a "S3-esque protocol"?
<kythyria[m]>
Well, because said protocol is a layer on top of HTTP, for a start.
<kythyria[m]>
If you like HTTP, you're probably going to prefer using it to write as well as read.
<kythyria[m]>
It's not even all that thick a layer IIRC
<libman>
In my zillion years of rsync'ing updates to a web server I never ever thought, "gee, I wish HTTP was involved in this"
hoboprimate has joined #ipfs
<kythyria[m]>
You didn't.
mahloun has joined #ipfs
<kythyria[m]>
I'd be more surprised at the existence of rsyncfs than a webdavfs too.
<libman>
IPFS has its own web api.
<kythyria[m]>
And? What makes that inherently better than S3's?
<kythyria[m]>
Also, what makes it inherently better than rsync, if you're going to use that as a benchmark
<kythyria[m]>
(rsync isn't SCP, by the way)
<libman>
It wasn't engineered to be "defective by design" to make you pay more for cloud services?
<kythyria[m]>
What exactly is defective about the S3 protocol, then?
<libman>
Does it support partial transfers / binary diffs like rsync? Does it work over SSH which you run anyway for admin remote access?
<kythyria[m]>
rsync is the only protocol either of us have named that can do partial transfers like rsync.
<kythyria[m]>
And anything that runs over TCP can be run over SSH.
sirdancealot has joined #ipfs
<kythyria[m]>
(and yes, you can do resumable uploads with S3 if you think to start the transfer that way.)
<lemmi>
well you could even look it that way: if an s3 api is what it takes to move people from amazon to ipfs, wouldn't that be a fair price to pay :)?
<kythyria[m]>
Except that what would actually be more likely is amazon doing IPFS pinning as a service :)
oliverny1 has quit [Ping timeout: 246 seconds]
olivernyc has quit [Ping timeout: 268 seconds]
Edur has quit [Write error: Broken pipe]
Edur has joined #ipfs
olivernyc has joined #ipfs
<lemmi>
and i wouldn't even be mad about it
<Anchakor>
do really a lot of apps/libaries support S3 API for storage?
atrapado_ has quit [Quit: Leaving]
<lemmi>
there are libraries for almost every usable programming language, and you can almost bet, that most of the stuff that touches ec2, can somehow talk s3
<lemmi>
there are also a bunch of other s3 compatible storage provides
dconroy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Anchakor>
hmm, so maybe S3 API on track to be what webdav was meant to be
<lemmi>
hell i know what webdav was meant to be
<pinkieval>
Anchakor: first, any ruby app using paperclip for file storage
<pinkieval>
s/ruby/ror
<pinkieval>
(which is actually why I asked this question in the first place)
jaboja has joined #ipfs
m0ns00n_ has quit [Quit: quit]
hoboprimate has quit [Quit: hoboprimate]
jsgrant_ has joined #ipfs
dconroy has joined #ipfs
oliverny1 has joined #ipfs
olivernyc has quit [Ping timeout: 260 seconds]
btmsn has quit [Ping timeout: 240 seconds]
Encrypt has quit [Quit: Quit]
<keks[m]>
is there an eta for 0.4.10?
jmill has joined #ipfs
jaboja has quit [Quit: Leaving]
Caterpillar has quit [Quit: You were not made to live as brutes, but to follow virtue and knowledge.]
oliverny1 has quit [Ping timeout: 260 seconds]
Wikipedia[m] has joined #ipfs
<MrSparkle>
is there a roadmap? preferably including past milestones
<lemmi>
on the github repo there is a page literally called "milestones"
<thekyriarchy>
does ipfs build upon or incorporate anything from information centric networking approaches like CCNx or NDN?
<thekyriarchy>
!wikipedia named data networking
<Wikipedia[m]>
*Named Data Networking (NDN)* (related to Content-Centric Networking (CCN), content-based networking, data-oriented networking or information-centric networking) is a Future Internet architecture inspired by years of empirical research into network usage and a growing awareness of unsolved problems in contemporary internet architectures like IP. NDN has its roots in an earlier project, Content-Centric Networking (CCN),
<Wikipedia[m]>
which Van Jacobson first publicly presented in 2006. The NDN project is investigating Jacobson’s proposed evolution from today’s host-centric network architecture IP to a data-centric network architecture (NDN). The belief is that this conceptually simple shift will have far-reaching implications for how people design, develop, deploy, and use networks and applications.
<Wikipedia[m]>
Its premise is that the Internet is primarily used as an information distribution network, which is not a good match for IP, and that the future Internet's "thin waist" should be based on named data rather than numerically ...
* kythyria[m]
is never sure how any of these proposals will handle live, low-latency, streaming effectively
<MrSparkle>
heh don't kill routers milestone
<kythyria[m]>
(or for that matter, it seems to be handwaved how you actually resolve a name down to a link, or how to stop it from being trivial to fill up a router's pending request table)
jaboja has joined #ipfs
maxlath has quit [Quit: maxlath]
shoragan has quit [Ping timeout: 255 seconds]
pcre has quit [Ping timeout: 240 seconds]
shoragan has joined #ipfs
shoragan is now known as Guest44050
<deltab>
thekyriarchy: obviously there's some overlap in concepts and architecture, but I'm not aware of any specific shared technology beyond things such as hash functions
ygrek has joined #ipfs
ygrek_ has quit [Ping timeout: 260 seconds]
<deltab>
however I think they're open enough that there can be interoperability in the future, e.g. a ccnx router could hand off /ipfs/ to an ipfs node
<deltab>
kythyria[m]: from my limited knowledge of CCNx, by removing layers such as IP and TCP, and by implicitly integrating multicast routing, it could handle streaming more effectively than our present systems
<kythyria[m]>
How? Systems like that seem to primarily work by having to request each chunk, rather than subscribing to the whole stream
chris613 has joined #ipfs
ilyaigpetrov has quit [Quit: Connection closed for inactivity]
<deltab>
requests can be sent in advance, e.g. /somewhere/stream/chunks/{1..100}
<deltab>
(or 10 rather than 100)
<deltab>
that builds up the routes in the routers ahead of time
<deltab>
when the data is available, it flows back along the established routes
<deltab>
routers can merge requests they receive from downstream, so they only need to send one request for each chunk upstream; the returned content is then sent to each requesting branch downstream
<deltab>
re filling up a router's pending request table, the router could limit how many requests are accepted at a time from each downstream
espadrine has quit [Ping timeout: 260 seconds]
<deltab>
not sure what you mean by "actually resolve a name down to a link"
<deltab>
unless you mean choosing the network link to send a request on, which would be done be prefix matching in a routing table
pawn has joined #ipfs
<kythyria[m]>
Bloody big table though.
Spferical has left #ipfs ["User left"]
<pawn>
kythyria[m]: What's a bloody big table?
* pawn
just logged on
dominic_ has joined #ipfs
<kythyria[m]>
The routing table for a content-based networking thingy
<pawn>
I just signed on to talk about this kythyria[m] haha
<pawn>
Basically I was wondering how content-address look ups work. Is it just a bit DHT?
<pawn>
big*
jonnycrunch has quit [Ping timeout: 245 seconds]
<deltab>
in IPFS, the DHT is used to map content hash to peers that can provide it
<deltab>
then your node connects to them and requests the content
<baggatea_>
hi all. got a couple of questions about ipfs
<baggatea_>
first up, I'd like to run it long-term for a project, but don't want to maintain it. is there a PPA or similar for Debian/Ubuntu so I don't have to care about updates?
<baggatea_>
secondly, when I pin some files from someone's library, where is the main space kept? are the files split into chunks or stored in a database, or are they just stored like normal files on disk?
<pawn>
deltab: How does a node know when to store a key in a DHT or when to leave it in the hands of peers?