<Aranjedeath>
be careful relying on oomkiller to do its job
<Aranjedeath>
I have had nothing but issues (I run with zero swap)
<Aranjedeath>
problem with oomkiller working is that it's on a timer by default, and as soon as you're oom the system can't correctly tell time
<Aranjedeath>
I have to change the oomkiller mode via a tunable to make it the dumb version
<Aranjedeath>
the smart one (the default) tries to intelligently kill something using a lot of ram
<Aranjedeath>
the dumb one just instakills whatever asked for that most recent bit of ram, which I find in practice to be much more effective at saving my ass
structuralist has joined #ipfs
cemerick has joined #ipfs
herzmeister has quit [Quit: Leaving]
dignifiedquire has quit [Quit: Connection closed for inactivity]
apiarian has joined #ipfs
herzmeister has joined #ipfs
ilmu has joined #ipfs
cketti has quit [Quit: Leaving]
gsf has joined #ipfs
clownpriest has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
ilmu has quit [Ping timeout: 265 seconds]
matoro has joined #ipfs
ppham has quit [Remote host closed the connection]
ppham has joined #ipfs
r04r is now known as zz_r04r
kvda has joined #ipfs
ppham has quit [Ping timeout: 240 seconds]
jedahan has joined #ipfs
jedahan has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
jedahan has joined #ipfs
jedahan has quit [Client Quit]
cemerick has quit [Ping timeout: 244 seconds]
Peeves has joined #ipfs
<ploop>
how is ipns going to work once you can have more than one name per node?
stwcx_ is now known as stwcx
ckwaldon has quit [Ping timeout: 272 seconds]
<deltab>
which part of it?
<deltab>
the UI and API will need to change so that you can specify which name
<deltab>
it'll have to store the private keys differently
kulelu88 has quit [Quit: Leaving]
DiCE1904 has joined #ipfs
<ploop>
deltab: but how will the name system actually work? will each node have like a "zone" and they can set names within their zone, like owning a domain and setting subdomains on it?
ppham has joined #ipfs
captain_morgan has quit [Ping timeout: 272 seconds]
<deltab>
that's already possible, because you can point the name at a directory
reit has joined #ipfs
<deltab>
the 'names' are actually hashes of private keys, unless you use dnslink, in which case they're domain names
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
ppham has quit [Remote host closed the connection]
herzmeister has quit [Remote host closed the connection]
herzmeister has joined #ipfs
ppham has joined #ipfs
anonymuse has joined #ipfs
yosafbridge has quit [Quit: Leaving]
nausea has quit [Ping timeout: 250 seconds]
yosafbridge has joined #ipfs
nausea has joined #ipfs
nausea has joined #ipfs
nausea has quit [Changing host]
<ploop>
oh neat
<deltab>
er, hashes of public keys
<ploop>
oh yeah i figured
<ploop>
wait so, I head that ipns is going to be completely changed at some point, but being able to point at a directory solves every issue I thought of with ipns
<ploop>
what's going to change exactly?
<deltab>
I don't know, but I think it's related to the work being done to store data structures
uzyn has joined #ipfs
mgue has quit [Quit: WeeChat 1.5]
<whyrusleeping>
ploop: ipns isnt going to change completely
structuralist has quit [Remote host closed the connection]
<whyrusleeping>
at some point we're going to expand its functionality to better be able to set rules about record validity
mgue has joined #ipfs
<ploop>
whyrusleeping: what do you mean? what more is there to validity than "signed by the key of the node that owns it"?
<whyrusleeping>
ploop: great question, theres a few different scenarios that are rather important. The first is a valid record isnt just one that we've signed, but the most recent one we've published
<whyrusleeping>
You don't want people to be given old out of date records
<ploop>
ah
<whyrusleeping>
think of pointing to a release of a binary with this, you don't want replay attacks forcing people to download code with vulnerabilities
<ploop>
yeah
<whyrusleeping>
You might also want to say that your record expires after a certain time
<whyrusleeping>
You may also want to say that any record along some set of records is valid, so long as it is linked into a hash chain type structure
<whyrusleeping>
something like an ad-hoc blockchain built on ipns should be easily possible given an extensiblie validity system like this
<ploop>
replay attacks and expiration date could be solved by signing a publishing time and/or an expiration time in addition to the record, couldn't they?
chris613 has quit [Quit: Leaving.]
ppham has quit [Remote host closed the connection]
PrinceOfPeeves has quit [Quit: Leaving]
ppham has joined #ipfs
<whyrusleeping>
time is very relative though
<whyrusleeping>
If youre just using a normal unix timestamp, you could mess with anyone whose system clock is off
<whyrusleeping>
Another notion of time could be by referencing globally verifiable events such as a particular block in something like the bitcoin blockchain
buhrmi_ has joined #ipfs
<buhrmi_>
heelllooo u merkeldags
<whyrusleeping>
buhrmi_: sup dag
<buhrmi_>
suh dud
ckwaldon has joined #ipfs
<WardCunningham>
victorbjelkholm: You suggested I echo $? (exit status) when my ipfs daemon was killed.
<WardCunningham>
I did, it was 137. I looked in the source for that code but didn't find it. Do you get any meaning out of it?
<whyrusleeping>
WardCunningham: yeah, that sounds like the system killed it due to out of memory
<whyrusleeping>
ipfs itself will only ever return 1
<whyrusleeping>
how much ram did your digitalocean box have?
<deltab>
WardCunningham: 137 is 128 + 9, which means it was killed with signal 9 (SIGKILL)
<deltab>
the kernel log (/var/log/dmesg) should have a report from the oom killer
<ploop>
when I'm in the webui, under "Files", when I click the x to remove a file, it throws me back into the "Home" tab and doesn't remove the file. I don't suppose this is the intended behavior?
<deltab>
ploop: sounds like a js bug; check the dev tools console for an error message
<deltab>
or the network panel in case it's from the api
<ploop>
"e.preventDefault is not a function"
<ploop>
and the x is a link to # so I can see why
structuralist has joined #ipfs
<deltab>
so e isn't an event, or you have a very old browser
<ploop>
it's current firefox stable, so I doubt it's the latter
<deltab>
yeah, only IE8 lacks it
<deltab>
(if that)
anonymuse has quit [Remote host closed the connection]
<ploop>
is there a way to remove files by hash directly with go-ipfs? I don't see anything in --help
<whyrusleeping>
ploop: if the file isnt pinned, you can run a gc with `ipfs repo gc`
<whyrusleeping>
that will remove all unpinned content
<whyrusleeping>
if you just want to remove a single file, and it is not pinned, you could do: `ipfs refs -r <file hash> | xargs ipfs block rm` and that should work
<whyrusleeping>
but, do note that it only deletes files from your local cache
<whyrusleeping>
if anyone else pulled them from you they will still be available on the network
<buhrmi_>
is there a possibility in ipfs that files replicate onto nodes that never requested them?
<whyrusleeping>
buhrmi_: nope, ipfs is purely pull only at this point
<buhrmi_>
so at this point, if A puts a file into ipfs, and B and C pull it, and then A B C all go offline forever, that file is also gone forever
<whyrusleeping>
sometime in the future, there will be optional bitswap actors you will be able to enable that will help the network seed content given some parameters
<whyrusleeping>
buhrmi_: correct
<buhrmi_>
cool got it
<WardCunningham>
whyrusleeping: I'm running ipfs daemon in 512MB machine, Digital Ocean's smallest.
<buhrmi_>
ah, and in order to incentivize nodes to store some other's files, that's where all that recent filecoin thing comes into play?
<buhrmi_>
or was that something else
<WardCunningham>
I will learn more about oom killer, configuring swap, or getting more memory.
<whyrusleeping>
WardCunningham: yeah... I think 512 shared between ipfs and the rest of the system is a bit tight right now (especially with no swap)
<kvda>
What's the easiest way to play an IPFS video object?
<whyrusleeping>
we definitely want ipfs to be able to run with very small amounts of ram though, so i think running on a 512MB DO droplet is a great goal for us to set
<whyrusleeping>
kvda: localhost:8080/ipfs/<hash> in your browser
<whyrusleeping>
if youre running a daemon that is, otherwise ipfs.io/ipfs/<hash>
<kvda>
Hmm yes thought that just uses the browser video player right, no streaming/buffering
<kvda>
i can still play the files through ipfs.io/ipfs/<hash> without the deamon running
<kvda>
but ipfstube is dependent on it being up
<kvda>
is there anything else like ipfstube?
probitlabs[m] has joined #ipfs
WardCunningham has quit [Quit: WardCunningham]
<A124>
whyrusleeping 633 MB residual for me atm.
<A124>
The context problems not yet re-analyzed foc current stable 4.3, but in general most of the performance problems are gone... thankfuly and finaly.
<A124>
The networking spikes are still troublesoume, it should be able to handle the peaks more gaussian curve that is much wider
<A124>
Also another thing... pinning stuff from commandline seems to go one after one. So pinning a single small resource after doing recursive pin for a resource tree with missing resources blocks even the small resource. Not sure to whom I should eve point this, so please delegate else.
<A124>
Oh.. and the network and CPU spikes go along which is terrible on a shared resource witout using cgroups and/or other limiting.
Tv` has quit [Quit: Connection closed for inactivity]
structuralist has quit [Remote host closed the connection]
warner` has joined #ipfs
WardCunningham has joined #ipfs
Luzifer_ has joined #ipfs
wa7son_ has joined #ipfs
nullstyle_ has joined #ipfs
zielmicha_ has joined #ipfs
mesch__ has joined #ipfs
nskelsey_ has joined #ipfs
risk_ has joined #ipfs
jclay_ has joined #ipfs
prettymuchbryce_ has joined #ipfs
robmyers_ has joined #ipfs
Muis_ has joined #ipfs
sugarpuff_ has joined #ipfs
nicolagreco_ has joined #ipfs
bigbluehat_ has joined #ipfs
tg` has joined #ipfs
gorhgorh has joined #ipfs
rflot_ has joined #ipfs
elimiste1e has joined #ipfs
barnacs_ has joined #ipfs
zmanian___ has joined #ipfs
Stskeepz has joined #ipfs
Guest61185_ has joined #ipfs
pjz_ has joined #ipfs
Guest66676 has joined #ipfs
martinBrown_ has joined #ipfs
zignig_ has joined #ipfs
Mateon1 has joined #ipfs
area_ has joined #ipfs
Mateon1 is now known as Guest17282
WardCunningham has quit [Quit: WardCunningham]
ppham has quit [Remote host closed the connection]
frabrune- has joined #ipfs
gozala_ has joined #ipfs
teaso_ has joined #ipfs
Ragnis_ has joined #ipfs
oleavr_ has joined #ipfs
sickill_ has joined #ipfs
cmeik_ has joined #ipfs
manveru_ has joined #ipfs
M|tar has joined #ipfs
MarkOtaris1 has joined #ipfs
bmpvieira_ has joined #ipfs
Aaron101- has joined #ipfs
jhulten_ has joined #ipfs
Magik6k_ has joined #ipfs
Nukien_ has joined #ipfs
kants_ has joined #ipfs
zombu2- has joined #ipfs
lachenmayer_ has joined #ipfs
Guest963391 has joined #ipfs
Ralith1 has joined #ipfs
irx[m]1 has joined #ipfs
Guest96339 has quit [*.net *.split]
irx[m] has quit [*.net *.split]
LugariusMtrx has quit [*.net *.split]
kegan[m] has quit [*.net *.split]
M-mistake has quit [*.net *.split]
frabrunelle has quit [*.net *.split]
warner has quit [*.net *.split]
MarkOtaris has quit [*.net *.split]
Ralith has quit [*.net *.split]
Ragnis has quit [*.net *.split]
sugarpuff has quit [*.net *.split]
sickill has quit [*.net *.split]
bmpvieira has quit [*.net *.split]
kyledrake has quit [*.net *.split]
robmyers has quit [*.net *.split]
nicolagreco has quit [*.net *.split]
wa7son has quit [*.net *.split]
prettymuchbryce has quit [*.net *.split]
NeoTeo has quit [*.net *.split]
nskelsey has quit [*.net *.split]
zielmicha has quit [*.net *.split]
nullstyle has quit [*.net *.split]
Luzifer has quit [*.net *.split]
risk has quit [*.net *.split]
bigbluehat has quit [*.net *.split]
Guest61185 has quit [*.net *.split]
jclay has quit [*.net *.split]
jhulten has quit [*.net *.split]
cmeik has quit [*.net *.split]
manveru has quit [*.net *.split]
retrohacker has quit [*.net *.split]
whyrusleeping has quit [*.net *.split]
kevina has quit [*.net *.split]
codebam has quit [*.net *.split]
deltab has quit [*.net *.split]
metaf5 has quit [*.net *.split]
sega01 has quit [*.net *.split]
Magik6k has quit [*.net *.split]
zmanian__ has quit [*.net *.split]
barnacs has quit [*.net *.split]
Muis has quit [*.net *.split]
rflot has quit [*.net *.split]
gorhgorh_ has quit [*.net *.split]
elimisteve has quit [*.net *.split]
teaso has quit [*.net *.split]
bsm1175321 has quit [*.net *.split]
edubai__ has quit [*.net *.split]
danemacmillan has quit [*.net *.split]
thufir has quit [*.net *.split]
mesch_ has quit [*.net *.split]
zignig has quit [*.net *.split]
alu has quit [*.net *.split]
oleavr has quit [*.net *.split]
Mateon1_ has quit [*.net *.split]
Kubuxu has quit [*.net *.split]
dvn has quit [*.net *.split]
Mitar has quit [*.net *.split]
Tugger has quit [*.net *.split]
kragniz has quit [*.net *.split]
kants has quit [*.net *.split]
gozala has quit [*.net *.split]
Kane` has quit [*.net *.split]
martinBrown has quit [*.net *.split]
Xe has quit [*.net *.split]
tg has quit [*.net *.split]
Nukien has quit [*.net *.split]
zombu2 has quit [*.net *.split]
Aaron1011 has quit [*.net *.split]
pjz has quit [*.net *.split]
Arw_ has quit [*.net *.split]
Guest66666 has quit [*.net *.split]
area has quit [*.net *.split]
Stskeeps has quit [*.net *.split]
lachenmayer has quit [*.net *.split]
rabbitface3 has quit [*.net *.split]
zombu2- is now known as zombu2
jhulten_ is now known as jhulten
martinBrown_ is now known as martinBrown
Ragnis_ is now known as Ragnis
tg` is now known as tg
M|tar is now known as Mitar
Arw has joined #ipfs
gozala_ is now known as gozala
jclay_ is now known as jclay
nullstyle_ is now known as nullstyle
deltab has joined #ipfs
zielmicha_ is now known as zielmicha
sickill_ is now known as sickill
nicolagreco_ is now known as nicolagreco
Luzifer_ is now known as Luzifer
kegan[m] has joined #ipfs
nskelsey_ is now known as nskelsey
dvn has joined #ipfs
M-mistake1 has joined #ipfs
Xe has joined #ipfs
LugariusMtrx has joined #ipfs
sugarpuff_ is now known as sugarpuff
robmyers_ is now known as robmyers
bmpvieira_ is now known as bmpvieira
kumavis has quit [Ping timeout: 255 seconds]
zrl has quit [Ping timeout: 255 seconds]
cmeik_ is now known as cmeik
risk_ has quit [Ping timeout: 265 seconds]
codebam has joined #ipfs
prosody has quit [Ping timeout: 255 seconds]
Guest61185_ is now known as Guest61185
sega01 has joined #ipfs
risk_ has joined #ipfs
Muis_ is now known as Muis
risk_ is now known as risk
wa7son_ is now known as wa7son
Foxcool has joined #ipfs
kevina has joined #ipfs
Tugger has joined #ipfs
whyrusleeping has joined #ipfs
Kane` has joined #ipfs
oleavr_ is now known as oleavr
rabbitface3 has joined #ipfs
danemacmillan has joined #ipfs
retrohacker has joined #ipfs
alu has joined #ipfs
kragniz has joined #ipfs
rflot_ is now known as rflot
Kubuxu has joined #ipfs
bsm1175321 has joined #ipfs
metaf5 has joined #ipfs
edubai__ has joined #ipfs
prettymuchbryce_ is now known as prettymuchbryce
manveru_ is now known as manveru
NeoTeo has joined #ipfs
kyledrake has joined #ipfs
bigbluehat_ is now known as bigbluehat
apiarian has quit [Ping timeout: 255 seconds]
prosody has joined #ipfs
apiarian has joined #ipfs
zrl has joined #ipfs
kumavis has joined #ipfs
Stskeepz is now known as Stskeeps
Stskeeps has quit [Quit: Reconnecting]
Stskeeps has joined #ipfs
warner` is now known as warner
<whyrusleeping>
dude, whats with the netsplits?
whyrusleeping has quit [Changing host]
whyrusleeping has joined #ipfs
<whyrusleeping>
kvda: ipfstube is really neat, i hadnt seen that yet
alu has quit [Changing host]
alu has joined #ipfs
rgrinberg has quit [Ping timeout: 265 seconds]
<Aranjedeath>
the entirety of freenode split yesterday I think
cosmos has joined #ipfs
thufir has joined #ipfs
<Aranjedeath>
there was some really nasty traffic weather across cogent's network at the time
<Aranjedeath>
I suppose if I really wanted to know I could go fetch the nanog archives, they'd say
ecloud is now known as ecloud_wfh
<Aranjedeath>
yeah, looks like it's still the ddos storms
* Aranjedeath
likes weather analogies for network traffic
<whyrusleeping>
Its a pretty good analogy
<whyrusleeping>
and also why i would love if something like cjdns got some major adoption
<whyrusleeping>
your traffic would dynamically reroute around these problems
<A124>
whyrusleeping BGP re routing can do the same, but in real world the traffic still has to get somewhere, so not sure if that simple implications as one wuold like.
<Aranjedeath>
yeah, but bgp will keep sending data over bad links until the bgp flaps
<Aranjedeath>
which is both awesome and terrible
<A124>
Well cjdns would be a better thing for peers, the server side is more complicated.
<A124>
Also I did not found how it can bridge to internet etc, so practically I did not try myself yet.
Guest7581 has quit [Changing host]
Guest7581 has joined #ipfs
<A124>
More docs needed.
Guest7581 is now known as DarkFox
herzmeister has quit [Ping timeout: 272 seconds]
<Aranjedeath>
there's a docs page about precisely that thing
<noffle>
okay. just wanted to share that I've been putting time into ipget lately and it's on its way to being ready for a nice stable 1.0.0 release: https://github.com/ipfs/ipget/issues/18
<lgierth>
yay :)
<keks>
nice
ygrek has joined #ipfs
<A124>
The pinning in case of missing things in a tree is awful and blocking following pins.
<A124>
Umm.. even sigterm on those pins seems not to solve it.
soloojos has joined #ipfs
WardCunningham has quit [Quit: WardCunningham]
s_kunk has quit [Ping timeout: 265 seconds]
<dignifiedquire>
richardlitt: can you post the announcment please?
cyberwolf has quit [Quit: Konversation terminated!]
<LugariusMtrx>
i have a question
<LugariusMtrx>
are hashes limeted?
<richardlitt>
Limeted?
<LugariusMtrx>
i mean .... this hashes are only 46 chars long with numbers and letters a and A so .... there is a file limit right?
<LugariusMtrx>
lets say there are endless files
<LugariusMtrx>
endless different files
<LugariusMtrx>
and this hashes for example QmWHyrPWQnsz1wxHR219ooJDYTvxJPyZuDUPSDpdsAovN5 are not so long
MarkOtaris1 has quit [Changing host]
MarkOtaris1 has joined #ipfs
Anduck has quit [Quit: Reconnecting]
Anduck has joined #ipfs
<keks>
I think datagrams have the additional benefit that you don't need to keep an overview of open connections
<keks>
oh, github issue
<richardlitt>
Kubuxu: can you add the all-hands notes to ipfs/pm/meeting_notes?
<richardlitt>
LugariusMtrx: technically, there is a file limit; but it is an astronomically high number
MarkOtaris1 is now known as MarkOtaris
<richardlitt>
The odds of collision - using the same hash for two things - is basically null. And the chance of hashes running out, in our lifetime, isn't particularly high.
<richardlitt>
think of it this way: 0-9 is 10 digits, and there are 26 letters in the alphabet - that's 36 possible options.
<LugariusMtrx>
but what when?
<richardlitt>
36^48 is 5.04103876e74.
<lgierth>
keks: yeah! :) and there are usecases where reliability is just a bad idea
<richardlitt>
That will basically never, ever happen.
<richardlitt>
Assuming that hashes were just five letters long, there are still 60466176 variations.
<richardlitt>
but there aren't five - there are loads more.
<richardlitt>
Make sense?
<LugariusMtrx>
i mean theoratically, what happens?
<ansuz>
upgrade to a new hash type
<ansuz>
:D
<ansuz>
via multihash
<LugariusMtrx>
yes this number is high... but that interests me
<richardlitt>
No reason, just, you know, it is cold in Poland, things happen.
<ansuz>
this is why everything should be done via commandline
<richardlitt>
brb going to buy a flight
<ansuz>
without tab-complete
<richardlitt>
LugariusMtrx: theoretically, little happens. it depends on which hashes collide.
<richardlitt>
The system assumes that it works.
<richardlitt>
If the hash is in a merkledag, than it is possible that that section of the merkledag will be malformed. But I'm not sure.
<LugariusMtrx>
malformed?
<richardlitt>
And, to be honest, 504103876000000000000000000000000000000000000000000000000000000000000000000 is a lot of 0s
<richardlitt>
Well, it may point to the wrong subset of data. But don't quote me on that.
<LugariusMtrx>
some guys sure said the same with ipv4
<richardlitt>
:D
<ansuz>
and ipv4 is fiiine
* ansuz
handwaves
ygrek has joined #ipfs
<richardlitt>
LugariusMtrx: a 46 character hash is stronger than ipv6.
<richardlitt>
By quite a bit.
<LugariusMtrx>
no really, if we use IPFS for years.... and in million of years we have to rewrite IPFS becouse this
<ansuz>
the hashes can upgrade
<ansuz>
like I said, multihash
<richardlitt>
That's why we use multihash. :)
<LugariusMtrx>
how?
<richardlitt>
github.com/multiformats/multihash
<LugariusMtrx>
ok
<richardlitt>
Go take a look.
<ansuz>
multi-all-the-things
<ansuz>
crypto-agility
<LugariusMtrx>
i will :)
<richardlitt>
basically, the first few characters of a multihash tell us which function created it; so we can just update the protocol to use different hashes, but still recognize old ones
<LugariusMtrx>
OHH
<richardlitt>
And we can switch to 49-char+ length hashes
<LugariusMtrx>
""It is useful to write applications that future-proof their use of hashes""
<richardlitt>
Exactly.
<LugariusMtrx>
perfect
<richardlitt>
Sorry, should have gotten to that point earlier.
<richardlitt>
jbenet basically came up with the idea of multihash, I believe.
<richardlitt>
He also, you know, runs IPFS.
* richardlitt
looks at the channel name with a quizzical look on his face
<ansuz>
richardlitt: irc moves fast :p
<LugariusMtrx>
like he cared about this already
<richardlitt>
yep!
<LugariusMtrx>
i know
Akaibu has quit [Quit: Connection closed for inactivity]
<richardlitt>
:D
<LugariusMtrx>
Juan Benet is a f *n genious
<richardlitt>
How do you know him? (Who are you? (Welcome!!)
rgrinberg has quit [Ping timeout: 276 seconds]
ulrichard has joined #ipfs
<ansuz>
richardlitt: lemme get that last paren for you
<ansuz>
)
<LugariusMtrx>
I know him from the Decentrelized Web summit
<richardlitt>
ah! Awesome. Glad you were there man.
<LugariusMtrx>
same for ipfs
<richardlitt>
Yes: hashes are pretty important to IPFS, and while collision rates are so far off as to be ridiculous, they still need to be thought of.
<richardlitt>
There's more than just hashes; having self-describing protocols is super important.
<richardlitt>
Go check out github.com/multiformats/multiformats for moe on others.
<LugariusMtrx>
a good developer solves problems before they happen, he already made IPFS future proof
<Stebalien>
nicolagreco: I'm listening in on the call. Please see my comments on the etherpad.
jedahan has joined #ipfs
rendar has quit [Ping timeout: 265 seconds]
taw00 has quit [Read error: Connection reset by peer]
atrapado has joined #ipfs
<dignifiedquire>
Thank you everyone for a great sync!
<ansuz>
np dignifiedquire
gavinmcdermott has joined #ipfs
gavinmcdermott has quit [Client Quit]
<dignifiedquire>
ansuz: no thank you
incog has left #ipfs [#ipfs]
<lgierth>
Kubuxu: when do and don't we use `gx-go rewrite`
<lgierth>
Kubuxu: i imagine something like, libs (e.g. go-multihash) don't rewrite, apps (e.g. go-ipfs) do write?
<lgierth>
whyrusleeping: ^
rendar has joined #ipfs
matoro has joined #ipfs
Stebalien has quit [Remote host closed the connection]
<ReactorScram>
I noticed if I pin the file from Computer A, then request it from Computer B, Computer B will hang until I request it from the gateway, at which point it suddenly gets it
<lgierth>
ReactorScram: sounds like NAT issues
<lgierth>
you can try to directly connect your nodes with `ipfs swarm connect <maddr>`, where maddr is one of the addresses from `ipfs id`
PseudoNoob has quit [Quit: Leaving]
<lgierth>
and you can debug content routing using `ipfs dht findprovs <hash>` to get a list of provider peer ids, and `ipfs dht findpeer <peerid>` to get maddrs
palkeo has joined #ipfs
<ReactorScram>
thanks, I'll look into that
palkeo has joined #ipfs
<ReactorScram>
Computer A's firewall is very old and has not had a software update in years, maybe that's it?
chungy_ has quit [Read error: Connection reset by peer]
atrapado has quit [Quit: Leaving]
fredthomsen has joined #ipfs
ulrichard has quit [Remote host closed the connection]
chungy has joined #ipfs
mildred has joined #ipfs
Guest39382 is now known as wking
WardCunningham has joined #ipfs
WardCunningham has quit [Client Quit]
taw00 has joined #ipfs
clownpriest has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<Kubuxu>
lgierth: we rewrite only go-ipfs right now
<Kubuxu>
as we don't want people to build it with unpinned deps