<tanizaki>
I reckon if GNS had been implemented in ipfs, it would stand for Galactic Name System, or be changed for IGNS
<ispeedtoo>
Is the location information an attribute of the config file or is it NAT derived?
The_8472 has joined #ipfs
zorglub27 has quit [Quit: zorglub27]
<brimstone>
Tanizaki: can you set TXT records with GNS?
<tanizaki>
I guess so, brimstone.
<tanizaki>
GNS zones are similar to those of DNS zones, but instead of a hierarchy of authorities to governing their use, GNS zones are controlled by a private key. When you create a record in a DNS zone, that information stored in your nameserver. Anyone trying to resolve your domain then gets pointed (hopefully) by the centralised authority to your nameserver. Whereas GNS, being decentralised by design, stores that information in DHT. The validity of
<tanizaki>
…records is assured cryptographically, by signing them with the private key of the respective zone.
<tanizaki>
Anyone trying to resolve records in a zone your domain can then verify the signature on the records they get from the DHT and be assured that they are indeed from the respective zone. To make this work, there is a 1:1 correspondence between zones and their public-private key pairs. So when we talk about the owner of a GNS zone, that's really the owner of the private key. And a user accessing a zone needs to somehow specify the corresponding p
<tanizaki>
…c key first.
The_8472 has quit [Ping timeout: 264 seconds]
tilgovi has quit [Ping timeout: 240 seconds]
<ion>
What is its difference compared to IPNS?
<Ronsor>
yeah
<Ronsor>
but named records i think
<Ronsor>
from what i read
<ion>
so /ipns/foobar.baz.<hash of pubkey>/blah. But wouldn't /ipns/<hash of pubkey>/foobar/baz/blah cover that?
<ion>
well, baz/foobar/blah to have the same hierarchy
tilgovi has joined #ipfs
The_8472 has joined #ipfs
<tanizaki>
ion, I suppose GNS can take care of the longevity and memorability of ipns records
tanizaki has quit [Quit: AtomicIRC: The nuclear option.]
joshbuddy has quit [Quit: joshbuddy]
coeern has joined #ipfs
<coeern>
Hello. Could ipfs use i2p or Tor? Or would it use something of it's own for anonymous usage? If it comes to that.
The_8472 has quit [Ping timeout: 240 seconds]
reit has quit [Quit: Leaving]
simpbrain has quit [Ping timeout: 265 seconds]
joshbuddy has joined #ipfs
<brimstone>
coeern: you can hide ipfs behind Tor
<brimstone>
it's possible
<brimstone>
i'd imagine i2p worked the same
<coeern>
Sweet.
<coeern>
I want to help with ipfs but I am still just learning Golang.
The_8472 has joined #ipfs
tilgovi has quit [Ping timeout: 250 seconds]
joshbuddy has quit [Quit: joshbuddy]
chrisdickinson has quit [Ping timeout: 240 seconds]
joshbuddy has joined #ipfs
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
chrisdickinson has joined #ipfs
voxelot has quit [Ping timeout: 250 seconds]
The_8472 has quit [Ping timeout: 264 seconds]
prf has quit [Remote host closed the connection]
The_8472 has joined #ipfs
The_8472 has quit [Ping timeout: 240 seconds]
dignifiedquire has quit [Quit: Connection closed for inactivity]
The_8472 has joined #ipfs
simpbrain has joined #ipfs
IlanGodik has quit [Quit: Connection closed for inactivity]
disgusting_wall has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
The_8472 has joined #ipfs
prf has joined #ipfs
The_8472 has quit [Ping timeout: 240 seconds]
The_8472 has joined #ipfs
libman has quit [Remote host closed the connection]
The_8472 has quit [Ping timeout: 264 seconds]
yellowsir1 has quit [Quit: Leaving.]
reit has joined #ipfs
coeern has quit [Remote host closed the connection]
notdaniel has quit [Remote host closed the connection]
notdaniel has joined #ipfs
<jbenet>
thanks.
<jbenet>
that picture keeps on giving
<vakla>
I keep slipping IPFS into conversations with people and hope they'll look it up
<jbenet>
vakla, have you tried super-liminal approaches? like "Here, you HAVE to see this. ---> 5min demo"
<vakla>
I mentioned it at our research administrators group yesterday
<vakla>
After someone was talking about the Hitachi Content whatever thing they have (super vendor-locked)
<jbenet>
hey whyrusleeping i've been collecting a lot of feedback for improvements. we should discuss them when you're back to normalcy
<vakla>
jbenet: Not overtly as I'm not really in the "decision-making" ranks here yet
The_8472 has quit [Ping timeout: 264 seconds]
<vakla>
However I've gotten some people to test it in interesting contexts thus far
<vakla>
Have you heard of the OSG?
<whyrusleeping>
jbenet: sure thing!
<whyrusleeping>
normalcy ETA 4 days
<jbenet>
vakla which OSG?
<vakla>
Open Science Grid
<jbenet>
vakla yeah tangentially, dont know much
<vakla>
I work with it and IPFS is SUPER interesting to ship out to the OSG with jobs since then you can have something like your own filesystem out on the grid
<vakla>
since you don't really choose where you're going to flock your jobs to, so some jobs could run in, say, Indiana, and others could be running in California
<vakla>
but if you run an ipfs daemon within your job then you can share your intermediate files without any hard work
<vakla>
so people are interested in this
<jbenet>
vakla: yep absolutely.
<jbenet>
vakla: depending on the job duration, it will be important to pre-fetch some of the data.
<jbenet>
vakla, but yes lots of interesting applications
<vakla>
Yeah, I had a few discussions with people about that today
The_8472 has joined #ipfs
<jbenet>
vakla: if you guys are interested in a pilot, we can help setup a deployment that will perform better
<vakla>
pre-staging data in locations isn't an easily solvable issue
<vakla>
Here, we use gluster
<whyrusleeping>
i just wrote some code to run *just* a dht node
<vakla>
gluster is VERY slow
<jbenet>
(i.e. dont connect to the rest of the network, etc)
<jbenet>
whyrusleeping: yay libp2
<jbenet>
2
<jbenet>
p
<whyrusleeping>
jbenet: well, it doesnt use that yet sadly
<jbenet>
:c
<whyrusleeping>
i need to PR using gx into go-ipfs
<vakla>
r/w performance has been clocked at 40% slower than local disk, without network bottlenecking
<whyrusleeping>
i'm just worried about making everyone use it, having there be *one more tool*
<achin>
p2p2p2p2p
<jbenet>
vakla: ipfs will be slow atm. we will get the perf up there, but lmk if you want to try some test deployment stuff, we can help.
<vakla>
I'd definitely be interested
<vakla>
Most of our pre-staging is super giant files (genetics data and the like)
<vakla>
And I work with Icecube as well, an Antarctic research team who handles petabytes of data and doesn't really know how to best store/handle their files
<vakla>
They re-evaluate every year, basically, because they need absolute best performance/reliability since the data is precious
<jbenet>
vakla: the best target range for current perf of ipfs is <1TB. we'll get orders of magnitude better but not yet there.
<vakla>
right now they're using lustre
<jbenet>
vakla: we're probably pretty good for archival.
<jbenet>
vakla: but not for high perf archival et
<jbenet>
yet*
<vakla>
Yeah, it sure seems like it
<jbenet>
so much to optimize
<vakla>
I'm very excited to see where it's going
* whyrusleeping
wishes he had time to sit and optimize
brixen has quit [Excess Flood]
<vakla>
jbenet: re:target range do you mean single file size?
user24 has joined #ipfs
user24 has quit [Client Quit]
brixen has joined #ipfs
<vakla>
I did give a presentation to a grad student research group on using IPFS as a personal CDN, that was an interesting Q&A session.
cow_2001 has quit [Quit: Offense is taken, not given. No need to disarm the world. Just make yourself bullet proof. Sticks & stones people. Block. Unfollow. Laugh.]
r04r is now known as zz_r04r
The_8472 has quit [Ping timeout: 240 seconds]
The_8472 has joined #ipfs
jfntn has quit [Remote host closed the connection]
tilgovi has joined #ipfs
chriscool has quit [Quit: Leaving.]
The_8472 has quit [Ping timeout: 264 seconds]
chriscool has joined #ipfs
The_8472 has joined #ipfs
tilgovi has quit [Ping timeout: 272 seconds]
bicorne has joined #ipfs
The_8472 has quit [Ping timeout: 240 seconds]
prf has quit [Remote host closed the connection]
The_8472 has joined #ipfs
prf has joined #ipfs
<SWingedSeraph>
About multiaddress, what do 0 and V mean in the size field for various protocols? I was interpreting 0 as meaning variable, but in that case, what is V for?
<SWingedSeraph>
Also, I assume that it uses network endianness?
computerfreak has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
prf has quit [Remote host closed the connection]
The_8472 has joined #ipfs
jedahan has joined #ipfs
The_8472 has quit [Ping timeout: 240 seconds]
<warner>
jbenet: I am! let's try for lunch, maybe next week or the one after
<jbenet>
warner: let's do next week, may not be here the following.
<jbenet>
warner: i'm staying in south mission so should be easy to find a time :) lunch or coffee or dinner.
kvda has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<jbenet>
hey warner, Dignifiedquire and daviddias (both likely asleep now) know a lot about electron stuff, feel free to ask here if you run into things
<warner>
cool, thanks
The_8472 has quit [Ping timeout: 240 seconds]
The_8472 has joined #ipfs
nii236 has joined #ipfs
<nii236>
Hey all, I'm new to ipfs and have run through the quickstart just last night. How viable is video streaming over the swarm? Does it gather chunks for a single file from multiple peers or does it pull it from a single machine?
<jbenet>
nii236 can do either
<jbenet>
and welcome
<nii236>
Cool thanks
<nii236>
Is there a reason filecoin progress has been kept behind closed doors as opposed to open source?
nii2361 has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
nii236 has quit [Ping timeout: 245 seconds]
grahamperrin has joined #ipfs
The_8472 has joined #ipfs
computerfreak has quit [Remote host closed the connection]
<jbenet>
nil2361 it will be open soon enough. good things take time to brew :)
<nii2361>
Alright =)
ulrichard has joined #ipfs
wiedi has quit [Ping timeout: 240 seconds]
<ipfsbot>
[go-ipfs] jbenet deleted fix/corsconfig at f080158: https://git.io/vz20O
The_8472 has quit [Ping timeout: 240 seconds]
computerfreak has joined #ipfs
The_8472 has joined #ipfs
disgusting_wall has quit [Quit: Connection closed for inactivity]
The_8472 has quit [Ping timeout: 264 seconds]
mildred has joined #ipfs
<jbenet>
hey mildred i made some examples with ipld to play around
<jbenet>
feel free to make more, too. or try out different pathing schemes
<ipfsbot>
go-ipfs/master 981b389 Jakub Sztandera: Add Server field in HTTP API...
<ipfsbot>
go-ipfs/master a2b0287 Juan Benet: Merge pull request #2141 from Kubuxu/patch-1...
computerfreak has quit [Remote host closed the connection]
The_8472 has joined #ipfs
Tv` has quit [Quit: Connection closed for inactivity]
dignifiedquire has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
m0ns00nfup has joined #ipfs
The_8472 has joined #ipfs
m0ns00nfup has quit [Client Quit]
m0ns00nfup has joined #ipfs
nii2361 has quit [Ping timeout: 265 seconds]
The_8472 has quit [Ping timeout: 240 seconds]
felixn has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
The_8472 has joined #ipfs
tlevine has quit [Ping timeout: 245 seconds]
The_8472 has quit [Ping timeout: 264 seconds]
_fe_ has joined #ipfs
The_8472 has joined #ipfs
m0ns00nfup has quit [Quit: undefined]
<vakla>
jbenet: I'd be really interested in a test deployment if you're willing, I talked to a colleague and he's also interested; we've got over 50,000 cores betwixt us both so that would be ultra cool.
The_8472 has quit [Ping timeout: 240 seconds]
<jbenet>
vakla yes, it would. what does the network look like?
<jbenet>
vakla: want to discuss details of what this might look like in a note over in github.com/ipfs/notes ?
<vakla>
Sure, could do
<jbenet>
(or email me if you want it to be a private discussion)
<vakla>
We run academic clusters
<vakla>
here we've got three main server rooms, we run HTCondor as a job scheduler
<vakla>
(they run condor there as well)
nii2361 has joined #ipfs
<vakla>
currently we use gluster and AFS as auxiliary filesystems where users can pull input data from
<vakla>
(in the distributed realm)
<vakla>
They can pull data normally as well but we like to keep big files from going through the submit nodes since that's a bottleneck
<vakla>
As I mentioned earlier, gluster is good for these large files but...VERY slow (it's kind of like a distributed RAID 10)
grahamperrin has left #ipfs ["Leaving"]
<vakla>
and OSG is a separate beast completely, colleague has been considering doing a talk on IPFS @ Condor week in May about it if he can get something concrete together
The_8472 has joined #ipfs
<vakla>
His boss is pretty excited about it so that's cool. Just need to convince the important people it's worth the time (and corresponding grant money ;) )
The_8472 has quit [Ping timeout: 264 seconds]
<jbenet>
vakla: would be great-- we're happy to help however we can. in particular making sure you have good perf.
kvda has joined #ipfs
<vakla>
Yeah -- this stuff is all firewalled so it's not really accessible from the plain ol' web. It'd be super awesome to run IPFS on it though, and I'd be way more than willing to make some cool graphs and get them back to you
<jbenet>
we really need a good way to launch a cluster, run some tests, and make graphs. all of this repeatably.
<vakla>
definitely
<jbenet>
i think this will make us improve perf orders of magnitude.
<jbenet>
so the sooner we get it the better-- i dont have bw to do this, but if someone can help would be huge
<vakla>
We're rolling out grafana here, it seems like a good way to conduct monitoring (my colleague uses it heavily as well)
<vakla>
I'm interested in working on it, my go-fu is constantly improving w/ the help of personal projects so...
nii2361 has quit [Ping timeout: 260 seconds]
<Kubuxu>
It was mentioned on latest sprint that it would be worth monitoring possible performance regressions.
<jbenet>
vakla well it would be a great contribution to the project
<jbenet>
Kubuxu: indeeed
<vakla>
I'll have to take a look at the grafana stuff, I'll be sitting in a few meetings about it in the near future
<vakla>
I'd really like to get some performance graphs and perhaps write a condor plugin for transferring input files w/ IPFS
Myagui has quit [Ping timeout: 256 seconds]
Myagui has joined #ipfs
The_8472 has joined #ipfs
bicorne has quit [Ping timeout: 256 seconds]
ylp1 has joined #ipfs
unvoluptuous has joined #ipfs
IlanGodik has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
The_8472 has joined #ipfs
M-trashrabbit has quit [Ping timeout: 265 seconds]
M-trashrabbit has joined #ipfs
rendar has joined #ipfs
nii2361 has joined #ipfs
The_8472 has quit [Ping timeout: 240 seconds]
ecloud is now known as ecloud_wfh
<Kubuxu>
jbenet: will be there format to write IPLD schemes?
The_8472 has joined #ipfs
nii2361 has quit [Ping timeout: 240 seconds]
Myagui has quit [Ping timeout: 250 seconds]
Myagui has joined #ipfs
<dignifiedquire>
good morning everyone
<vakla>
good morning dignifiedquire
_fe_ has quit []
m0ns00nfup has joined #ipfs
m0ns00nfup has quit [Quit: undefined]
m0ns00nfup has joined #ipfs
The_8472 has quit [Ping timeout: 264 seconds]
The_8472 has joined #ipfs
m0ns00nfup has quit [Quit: undefined]
<ipfsbot>
[js-ipfs-api] Dignifiedquire created test/ls (+1 new commit): https://git.io/vz2dM
Akaibu has quit [Quit: Connection closed for inactivity]
ulrichard has quit [Remote host closed the connection]
prf has quit [Quit: pfraze has joined]
edubai__ has quit [Quit: Connection closed for inactivity]
<Quiark>
hi, I think ipfs is pretty cool, but what are some real world applications? how can it be used today? #lackofimagination
Tv` has joined #ipfs
mildred has quit [Quit: Leaving.]
voxelot has joined #ipfs
<xiaomorph>
Quiark: try disconnect yourself from internet and try use your local mesh network :)
<Quiark>
xiaomorph, ok that's more of a long-term goal :)
corvinux has quit [Ping timeout: 265 seconds]
<Kubuxu>
Quiark: no, that is something that already can be used only if we had some substantial applications on IPFS but those are WIP.
Not_ has joined #ipfs
<Kubuxu>
Berlin mesh can function w/o internet, with IPFS it would be actually useful.
<Quiark>
I'm just saying that I know exactly 2 sites that uses ipfs (ipfs.pics, your homepage)
<Kubuxu>
There is in works forum on IPFS. Derp from Hyperboira plans to employ IPFS in urlcloud (you won't find a lot about it as it is Hyperboira only service).
corvinux has joined #ipfs
yellowsir1 has joined #ipfs
<Kubuxu>
I've also set up bin.kubuxu.ovh, syntax highlight viewer for now.
<Kubuxu>
It is all WIP but you have to start with something.
<SWingedSeraph>
jbenet, I have a few questions about multiaddr; what endianness does the binary representation use, and what is the canonical way to handle transports with arbitrary length addresses? Can I assume that they are null-terminated strings?
corvinux has quit [Client Quit]
<Quiark>
Kubuxu, good, I actually was having trouble googling up those WIP projects
<Kubuxu>
It isn't good idea to publish: I want to do something. It is much better to do it and then say: look on what have I created.
<Kubuxu>
Also positioning of sites on IPFS isn't great as it is like site links to itself.
e-lima has quit [Ping timeout: 260 seconds]
<SWingedSeraph>
Kubuxu, are there any good resources on how to architect applications for IPFS? I.e., how to handle server side interaction, etc
pfraze has joined #ipfs
<Quiark>
Kubuxu, yeah you can't make circular links :)
<Quiark>
Kubuxu, do you know what is the forum using for mutability?
<Kubuxu>
SWingedSeraph: i am on a phone but read through ipfs/FAQ and ipfs/noyrs
<SWingedSeraph>
OK, will do. I was sort of assuming that all applications would be mostly client side but you mentioned a forum which absolutely requires server side interaction so I was curious
<Kubuxu>
Quiark: for now it uses own protocol and users' IPNS, users reference each other. In future I will use POST
<Kubuxu>
SWingedSeraph: it is client side only, there are small problems with aggregation but it is also worked on.
<Kubuxu>
If you want to create hybrid site it is really easy.
<Kubuxu>
Just make location /ipfs/ of your site pass to your gateway and post links/images to it.
<Kubuxu>
It is exactly what ipfs.pics does.
Encrypt has joined #ipfs
<ipfsbot>
[go-ipfs] Luzifer closed pull request #2170: Switched to amd64 version of ipfs (master...docker_amd64) https://git.io/vuKQf
<xiaomorph>
Quiark: there are lot of mesh networks which are content-less ;( but there are projects like Hyperboria (cjdns) :)
<xiaomorph>
but it still forming thing
ilyaigpetrov has joined #ipfs
j0hnsmith has joined #ipfs
<ipfsbot>
[go-ipfs] chriscool created pin-ls-count (+1 new commit): https://git.io/vzVfc
<ipfsbot>
go-ipfs/pin-ls-count 636c45d Christian Couder: core/commands/pin: implement --count for 'ipfs pin ls'...
Guest8342 has quit [Remote host closed the connection]
fe has joined #ipfs
fe is now known as Guest89221
arpu__ has joined #ipfs
arpu_ has quit [Ping timeout: 265 seconds]
<raxrb>
can someone guide me, regarding contributing to ipfs. I am new to this
disgusting_wall has quit [Quit: Connection closed for inactivity]
parkan has joined #ipfs
arpu__ has quit [Quit: Ex-Chat]
<parkan>
is the ipfs.io gateway acting up? I can't get it to load hashes that I see fine on cubepin and my own nodes
ygrek has joined #ipfs
<parkan>
QmXx283TyuUPo8Wt8A5aCzX5hfxyexaSyceSXjmeNBmnGA loads one cubepin and locally but not through gateway, for ex
<ipfsbot>
[js-ipfs-api] diasdavid deleted test/ls at ed298ac: https://git.io/vzV3U
Quiark has quit [Ping timeout: 272 seconds]
devbug has quit [Read error: Connection reset by peer]
devbug has joined #ipfs
nonaTure has quit [Ping timeout: 264 seconds]
j0hnsmith has joined #ipfs
maxlath has quit [Ping timeout: 260 seconds]
Guest89221 has quit [Remote host closed the connection]
j0hnsmith has quit [Quit: j0hnsmith]
nonaTure has joined #ipfs
Matoro has quit [Ping timeout: 256 seconds]
Matoro has joined #ipfs
Akaibu has joined #ipfs
brixen has quit [Excess Flood]
brixen has joined #ipfs
The_8472 has joined #ipfs
tilgovi has quit [Ping timeout: 272 seconds]
parkan has quit [Ping timeout: 276 seconds]
maxlath has joined #ipfs
ilyaigpetrov has quit [Quit: Connection closed for inactivity]
Matoro has quit [Ping timeout: 276 seconds]
dissoluble has quit [Remote host closed the connection]
Matoro has joined #ipfs
joshbuddy has quit [Quit: joshbuddy]
devbug has quit [Ping timeout: 276 seconds]
tilgovi has joined #ipfs
<achin>
richardlitt: i'm ACKing weekly/pull/14. will add a list of code contribs as soon as i'm home from work
Encrypt has quit [Quit: Quitte]
The_8472 has quit [Ping timeout: 240 seconds]
libman has joined #ipfs
hashcore has quit [Ping timeout: 256 seconds]
fiatjaf_ has joined #ipfs
joshbuddy has joined #ipfs
computerfreak has joined #ipfs
joshbuddy has quit [Quit: joshbuddy]
Not_ has quit [Ping timeout: 250 seconds]
devbug has joined #ipfs
tanizaki has joined #ipfs
Matoro has quit [Ping timeout: 256 seconds]
hashcore has joined #ipfs
<whyrusleeping>
i hate everything
<whyrusleeping>
airports are the worst
<vakla>
agree
Matoro has joined #ipfs
<tanizaki>
Dear Community, could you please tell me why go has been chosen as a primary language for the ipfs implementation?
<vakla>
because go is awesome
<whyrusleeping>
^
<vakla>
It's fast and it's not a pain to write like C can be
<whyrusleeping>
Tanizaki: why do you ask?
<vakla>
it's got great standard libs
The_8472 has joined #ipfs
<vakla>
Rob Pike developed it who is absolutely brilliant
<whyrusleeping>
and that one ken guy
<whyrusleeping>
i heard he was pretty neato too
<vakla>
oh yeah that ken guy
<vakla>
also russ cox
<tanizaki>
whyrusleeping, I am wondering whether ipfs would continue to use go in the long term. Is go used primarily for prototyping ipfs, or are there any plans for alternative implementations? What are the advantages of go before C other than its non-pain of writing and its history, and what are its disadvantages??
fe has joined #ipfs
<vakla>
Tanizaki: there's WIP implementations in other langs already
fe is now known as Guest66844
<tanizaki>
A related question: What languages and in what sequence would you recommend to learn? What is the ideal language for ipfs?
<vakla>
There's not an objective answer to that
<tanizaki>
Vakla, I do not ask for objectivity
<vakla>
golang is very nice for a lot of reasons, especially when it comes to networking
<tanizaki>
Vakla, even though there are other implementations, its first one is in go
<vakla>
It is different than other languages though; I personally learned to program with java, and from there it's quite easy to pick up other langs
<tanizaki>
I was wondering whether it was a random decision and what the plans were
<whyrusleeping>
Tanizaki: i think go is the ideal language for ipfs
<whyrusleeping>
its fast, is great for networking, has awesome support for concurrency, and has a lot of good stuff in the std lib
<whyrusleeping>
it was designed for building networked systems
<tanizaki>
Isn't go lower than java?
<whyrusleeping>
lower?
Guest66844 has quit [Ping timeout: 264 seconds]
<tanizaki>
of lower level?
<vakla>
well, java uses the jvm
<vakla>
and java bytecode
Matoro has quit [Ping timeout: 256 seconds]
<whyrusleeping>
the primary implementation of ipfs will be in go for the foreseeable future
<vakla>
java is also terrible to write, though, so...
<tanizaki>
I see, whyrusleeping. Do you think go will spread as c has?
<whyrusleeping>
Tanizaki: C has its place, as does go
<vakla>
There is no one perfect language for everything
<vakla>
use the best tool for the job
Matoro has joined #ipfs
neurosis12 has joined #ipfs
<whyrusleeping>
Tanizaki: if you want, you can start an implementation in C
<vakla>
it's definitely not going away anytime soon since google uses it internally and google is everything
jasaarim has joined #ipfs
<tanizaki>
whyrusleeping, just like perl in shells. Whyrusleeping, Vakla, do you think go will be used more widely in the future?
<vakla>
I believe it's "catching on"
<tanizaki>
whyrusleeping, thank you for the encouragement!
<vakla>
but it's used in a lot of places already people don't normally see
<vakla>
for example, twitch uses it for their backend
<vakla>
digitalocean uses it
<vakla>
it's a systems lang
<vakla>
generally
<tanizaki>
Aren't the goals of ipfs opposite of the aims of google as a centralised and insecure service?
<vakla>
google encompasses many many services
<tanizaki>
Vakla, but most of them have the same concepts of centralisation and snooping-proness in the core: Maps, Hangouts, Search, Drive. Regarding go, what about the pc and not the server environments? Do you see potential of go in the context of client-server coalescence of ipfs?
<tanizaki>
*proneness
<tanizaki>
Do you see any potential in the c implementation of go, or would it stagnate as gnunet does? What are the downsides of c in the context of ipfs?
jaboja has joined #ipfs
<tanizaki>
I am sorry, c implementation of ipfs is meant
neurosis12 has quit [Remote host closed the connection]
edubai__ has joined #ipfs
parkan has joined #ipfs
tlevine has joined #ipfs
voxelot has quit [Ping timeout: 240 seconds]
parkan has quit [Ping timeout: 256 seconds]
Matoro has quit [Ping timeout: 256 seconds]
<tanizaki>
Is golang stable enough for ipfs, considering it is not standardised and its integrity is arguably based on the existence of Google?
<achin>
a few months ago, there was 1 bug in IPFS that was tracked down to a bug in golang itself
<achin>
but that's only a small datapoint
<tanizaki>
How can we be sure that the backbone of ipfs is not rotten in its core?
<tanizaki>
Thank you, Achin, for the info!
tilgovi has quit [Remote host closed the connection]
<Kubuxu>
Tanizaki: problem with C implementations is that you have to create whole code base around them. There aren't really universal libraries for C (even if you think there are).
<achin>
i am not an IPFS developer, and so i don't want to speak for them. but my impression is that, in general, using golang has been a success for ipfs
Encrypt has joined #ipfs
<Kubuxu>
See: cjdns where I had to implement qsort as glibc was doing stupid shit (glibc's qsort was calling `open` which is blocked syscall in cjdns).
<tanizaki>
Thank you, Kubuxu. Isn't it the goal of ipfs to grant the full control to the users and developers?
ryepdx has joined #ipfs
<Kubuxu>
go is great for quickly developing projects that want to be fast and native. Also golang can live without Google's support.
<Kubuxu>
What do you mean by that?
<tanizaki>
Achin, why do you think so?
<tanizaki>
Is creating libraries safer in c or golang?
<Kubuxu>
From my experience: IPFS would not get 30% of contributions it got by being written in Go, it would require much greater codebase as everything would have to be implemented just for IPFS
<Kubuxu>
golang
<tanizaki>
Well, the licensing of golang dooms the future developers to everlasting remembrance of google
<tanizaki>
I see, Kubuxu. My question is whether ipfs will transition at some point to the implementation in another language
<Kubuxu>
It is already transitioning to JS.
<Kubuxu>
maybe not transitioning but being reimplmented.
<achin>
my guess is that the "reference implemetation" of ipfs will always be in golang
<tanizaki>
Thank you for the info, Kubuxu! Why JS?
<achin>
other implementations may be written, though, perhapns in other languages
<tanizaki>
Why, Achin, do you guess so?
<achin>
but i also guess there will not likely be an implemention in C
<achin>
it seems more likely to me that other implementations would be in higher level languages. for example, i'd like to see an implementation in rust
<Kubuxu>
You can embed it everywhere. Browser can embed it and use it if system does not provide IPFS, site can embed it if browser does not provide IPFS.
alexandria-devon has joined #ipfs
<alexandria-devon>
hey @whyrusleeping, you around homie?
jaboja has quit [Ping timeout: 276 seconds]
<alexandria-devon>
if you are, would you mind popping into http://dloa.slack.com for a few? - one of our devs had an IPFS question i couldn't answer
rendar has quit [Ping timeout: 272 seconds]
<achin>
Tanizaki: there is significant cost to restarting a large project in another language. i'm not sure the costs would be worth the benefit, so i guess that the go impl of ipfs will be always be around
<Kubuxu>
Also there is little reason for coding something in pure C nowadays.
Matoro has joined #ipfs
<achin>
there are good reasons for coding something in C. but those reasons also apply to other languages
<achin>
(reasons like: "really good interoperability with other systems")
<tanizaki>
But isn't JS transport-dependent, i.e. it depends on the current infrastructure of the internet? How does JS go with mesh networking infrastructures for instance?
<Kubuxu>
It isn't the best part of C, if you are doing anything bigger than printing and simple processing, interops start to be a mess (again cjdns).
<achin>
doesn't everything depend on the current infrastructure of the internet in some way?
<Kubuxu>
Tanizaki: it isn't a problem, nodejs implementation could work exactly as go's, browser's would use different transport but it is still possible.
<tanizaki>
Kubuxu, and what about the environments without browsers?
<Kubuxu>
nodejs
<achin>
a better question might be: "how well could a project/technology adapt to a changing environment"
<tanizaki>
achin, isn't it one of the goals of ipfs to change how the internet works?
<Kubuxu>
yes but before that we have to adapt to current one.
<achin>
i think they are trying to change how the "web" works. it's a subtle distinction, but i think it's important
rendar has joined #ipfs
<tanizaki>
I've thought this is what jbenet has meant talking about the infrastructure engineers engaging with ipfs
<jbenet>
alexandria-devon: can it be asked here? Other people may help.
<jbenet>
well we must do both: improve and connect a smooth ramp to current tech
<Kubuxu>
IPFS uses existing infrastructure and enables meshes to be useful.
<jbenet>
Else hard to switch
<jbenet>
Adoption cost needs to be as close as we can get to zero
<tanizaki>
Kubuxu, it means building a distributed web on centralise infrastructure, doesnt it? What about changing the dichotomy of offline and online?
<Kubuxu>
IPFS can work on current infrastructure and adoption of IPFS would influence how infrastructure is the build/modified.
<Kubuxu>
s/the/later
The_8472 has quit [Ping timeout: 240 seconds]
<Kubuxu>
If ISP finds out that big part of traffic is from one his user to another he might think about modifying his network so burden of this internal traffic is lower.
<tanizaki>
Jbenet, what about guifi.net and other examples of mesh networking in not-as-affluent-as-US locales? Their difficulties seem to make the transition easier, isn't it?
<tanizaki>
Kubuxu, might. Might not. Might block the direct connection via ipfs and that is it.
ispeedtoo has quit [Ping timeout: 252 seconds]
<jbenet>
Tanizaki: not sure what you ask. Integration would them already would be nice and easy and an improvement
voxelot has joined #ipfs
<tanizaki>
jbenet, I am questioning the paradigm and the implementation of ipfs. There is a huge pressure on the bittorent protocol in such countries as Russia. Ipfs may face the same obstruction. Some firewalls are blocking ipfs.io already. Are there any plans of implementing ipfs in another language?
<Kubuxu>
You can't block just one protocol, you can keep trying but as long as you allow encrypted streams through ...
<tanizaki>
What about dpi?
<Kubuxu>
(ipfs.io might be blocked but it is only transition strategy, you should run node and use local gateway if you want to assist IPFS).
<Kubuxu>
Encryption screws it over.
<Kubuxu>
As long as you allow users to create encrypted channels you can't filter it out really, especially in case of protocol that creates mesh.
<jbenet>
Tanizaki while TLS streams exist no protocol is blocked
<jbenet>
Yes other languages. But Your concerns about go are unfounded
<jbenet>
Read the language spec yourself
<vakla>
yeah, it's open, so...
<jbenet>
Write your own compiler or use gc
e_low is now known as prosody
<tanizaki>
Jbenet, I am not sure if focus on integration with the centralised infrastructure may be productive in the long term, as well as stress on such architectures as nodejs with the implied security, implementability and stability issues. Am I wrong, and why?
<tanizaki>
Jbenet, I am sorry for the wrong word -- not blocked, but persecuted.
<redfish>
Tanizaki: I don't follow your concern. There's nothing tieing IPFS to the IP infrastructure. You can route IPFS packets over anything (possibly some transports will need implementation work, but that's all).
<Kubuxu>
Tanizaki: I would love the hear your solution.
<tanizaki>
redfish, my concern is letting ipfs develop to the point of no return contradicting the initial paradigm described by jbenet.
<tanizaki>
Kubuxu, I would love too. Thank you for inspiration!
<Kubuxu>
IPFS already works on cjdns (which is mesh network). Possibilities are limitless.
<redfish>
Tanizaki: that's pretty abstract... you have some concrete anchors? like examples of decisions made by other projects that were good and bad?
libman has quit [Remote host closed the connection]
lothar_m has joined #ipfs
<tanizaki>
What do you think must ideally be changed in the tech in the context of ipfs?
<tanizaki>
redfish, from the link:"Fixing security issues as an afterthought is much harder."
<redfish>
Tanizaki: I don't see what design decision(s) in IPFS preclude any of the major use-cases of GNUnet
joshbuddy has joined #ipfs
<redfish>
a lot of the responsibilities can (and often should) be delegated to lower levels
<tanizaki>
Kubuxu, thank you for the info!
montagsoup has joined #ipfs
<Kubuxu>
also IPFS deployed official gateways in cjdns (Hyperboria network) and there is WIP discovery using cjdns' DHT routing.
<redfish>
Kubuxu: interesting, so there's belief in Hyperboria, here. I ran a CJDNS node some time ago, but then lost faith in the CJDNS project. It seems to solve to nebulous a problem. Perhaps, I'm too stuck in the present.
<zjohnson>
I just want to see popular browser support and a reddit/imgur/youtube implementation entirely in ipfs :)
<tanizaki>
redfish, so is satisfaction of the goals of ipfs such as the nonexistence of central point of failure dependent on the lower levels?
<redfish>
Tanizaki: if you refer to failure of *communication links*, then yes. A node can be connected to its peers via many different links concurrently (Cable ISP, 4G network, neighbor's wifi, etc). Failure of any one of these does not disconnect the node.
<redfish>
Tanizaki: if you refer to failure of *a peer*, then no, that's solved at higher protocol level (IPFS CDN level): content important enough will be pinned by multiple nodes. Failure of one node does not erase the content.
<tanizaki>
I see, redfish, thank you for your detailed answer. Thank you all!
disgusting_wall has joined #ipfs
ralphtheninja has joined #ipfs
luigiplr has joined #ipfs
fe has joined #ipfs
fe is now known as Guest44189
e-lima has joined #ipfs
computerfreak has quit [Remote host closed the connection]