computerfreak has quit [Remote host closed the connection]
arpu has quit [Ping timeout: 240 seconds]
arpu has joined #ipfs
igork[m] has joined #ipfs
ploop has joined #ipfs
arpu has quit [Ping timeout: 260 seconds]
arpu has joined #ipfs
draget[m] has joined #ipfs
arpu has quit [Ping timeout: 260 seconds]
arpu has joined #ipfs
arpu has quit [Ping timeout: 276 seconds]
arpu has joined #ipfs
espadrine has joined #ipfs
j12t has joined #ipfs
j12t has quit [Ping timeout: 252 seconds]
herzmeister has quit [Quit: Leaving]
A124 has quit [Read error: No route to host]
herzmeister has joined #ipfs
A124 has joined #ipfs
computerfreak has joined #ipfs
ashark has joined #ipfs
stoogenm_ has joined #ipfs
ashark has quit [Ping timeout: 258 seconds]
stoogenm_ has quit [Ping timeout: 272 seconds]
MahaDev has quit [Read error: Connection reset by peer]
corvinux has quit [Quit: Leaving]
dignifiedquire has joined #ipfs
cketti has quit [Quit: Leaving]
curiouscraig has joined #ipfs
curiouscraig has quit [Client Quit]
yuryk[m] has joined #ipfs
zorglub27 has quit [Ping timeout: 240 seconds]
nessence has quit []
dd[m] has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
j12t has joined #ipfs
j12t has quit [Ping timeout: 252 seconds]
zz_r04r is now known as r04r
zorglub27 has joined #ipfs
reit has quit [Ping timeout: 240 seconds]
Guest21903[m] has joined #ipfs
<Guest21903[m]>
pretty cool
<Shibe>
yeah
stoogenmeyer has joined #ipfs
computerfreak has quit [Quit: Leaving.]
stoogenmeyer has quit [Ping timeout: 258 seconds]
Guest21923[m] has joined #ipfs
<Guest21923[m]>
hi
<haad>
hello everyone! back online...
<Guest21923[m]>
hi this is cool man..!!
ashark has joined #ipfs
chevdor[m] has joined #ipfs
lmatteis has joined #ipfs
vijayee has joined #ipfs
<cwahlers>
haad: welcome back :)
zorglub27 has quit [Ping timeout: 276 seconds]
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
monzoon[m] has joined #ipfs
<haad>
cwahlers: o/
<haad>
cwahlers: you on this side of the atlantic already?
<cwahlers>
yep :)
<haad>
cool
zorglub27 has joined #ipfs
kus has joined #ipfs
mildred1 has joined #ipfs
Avinash[m] has joined #ipfs
mildred has quit [Ping timeout: 250 seconds]
ashark has quit [Ping timeout: 244 seconds]
OutBackDingo has quit [Ping timeout: 240 seconds]
computerfreak has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
<lgierth>
o/
<monzoon[m]>
hi guys!
<monzoon[m]>
I'm new to ipfs, just finished watching the demo video and I've got a question - since several files may have the same hash, they will be de-duplicated
<monzoon[m]>
but what do you do with the creation/modification date?
<monzoon[m]>
is it saved? and if it is, which timestamp is chosen? the first file with the hash or the last one?
<lgierth>
filesystem metadata is not saved at the moment
<lgierth>
modification/access dates, permissions, etc.
<monzoon[m]>
ok, got it. But is there a design decision on what to do with the metadata?
stoogenmeyer has quit [Remote host closed the connection]
j12t has quit [Ping timeout: 252 seconds]
stoogenmeyer has joined #ipfs
Tv` has joined #ipfs
<ipfsbot>
[webui] ystyle opened pull request #421: Add a new Chinese Translation (master...patch-1) https://git.io/vKJhH
stoogenmeyer has quit [Ping timeout: 250 seconds]
Mitar has joined #ipfs
<kpcyrd>
does it make sense to start working on something that is able to follow /ipns/, so 1) resolve 2) pin hash 3) unpin old hash (if any) 4) gc? or should I wait for ipfs-cluster?
ashark has joined #ipfs
<Kubuxu>
ipfs-cluster will take a while
zorglub27 has joined #ipfs
<lgierth>
ipfs-cluster has not so much to do with that
<lgierth>
IPRS does, the record systems
<richardlitt>
Hey, so the stream died -- join the hangout if you want to join in.
<richardlitt>
Will try to not have kernel panic next time
<lgierth>
IPNS records will be first class ipfs objects and can be pinned
<dignifiedquire>
whyrusleeping: Kubuxu what's the pr documenting the changes in the repo?
<whyrusleeping>
Kubuxu: if i run the code with a race detector and i see a report with your code in it. I'm going to reach all the way over to poland and turn your computer off
<Kubuxu>
It might be detected as race, it is just not meaningfull race.
<whyrusleeping>
doesnt matter
<Kubuxu>
Then I should add 400ns latency to code that runs 900ns (of which there is already 400ns lock latency)?
<whyrusleeping>
check out sync/atomic
vijayee has quit [Ping timeout: 240 seconds]
Boomerang has quit [Quit: Lost terminal]
<Kubuxu>
Race detector detects race there ...
<Kubuxu>
but this is false positive as it is synced on channel :p
Guest22220[m] has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
<lgierth>
kpcyrd Kubuxu: IPRS after IPLD and keystore
<Kubuxu>
yup
<ipfsbot>
[js-ipfs] nginnever created interface/block from master (+0 new commits): https://git.io/vKULg
palkeo has joined #ipfs
palkeo has quit [Changing host]
palkeo has joined #ipfs
MadcapJake has quit [Remote host closed the connection]
MadcapJake has joined #ipfs
zorglub27 has quit [Ping timeout: 264 seconds]
cketti has joined #ipfs
cketti has quit [Changing host]
cketti has joined #ipfs
OutBackDingo has joined #ipfs
<Kubuxu>
whyrusleeping: used atomic
<Kubuxu>
:w
<Kubuxu>
(not this window)
<whyrusleeping>
Kubuxu: thank you!
<Kubuxu>
whyrusleeping: should I remove the `k == ""`?
<whyrusleeping>
can you think of a reason why that would be there?
<Kubuxu>
If we are using keys correctly it shouldn't happen
<whyrusleeping>
i dont see it in the original write_cache code
<whyrusleeping>
ahhh, i remember now
reit has quit [Ping timeout: 258 seconds]
<whyrusleeping>
its in the normal blockstore code
<Kubuxu>
right
<whyrusleeping>
if we accidentally pass in a bad key we want to report it as 'not found'
<Kubuxu>
nah, so I will just pass it with no caching
<whyrusleeping>
so keeping what you have? hard coded to return 'true' ?
<whyrusleeping>
or hard code it to return false
<Kubuxu>
oh, right. So I should say that this record has invalid cache, to pass it to blockstorage
<whyrusleeping>
either that or have your cache report it as not found
zorglub27 has quit [Remote host closed the connection]
zorglub27 has joined #ipfs
vijayee has joined #ipfs
<kpcyrd>
lgierth: so nothing I can use right now :) ?
<apiarian>
kpcyrd: i'm working on a system which uses IPLD entries to communicate. so i publish something to an IPFS tree, stick the node in the current /ipns/[node]/[namespace] link (creating or replacing whatever was on that link before), and then unpin the old ipns entry and pin the new one. over the go-ipfs-api
<kpcyrd>
apiarian: is IPLD already a thing?
<apiarian>
there's a publish step in there somewhere
stoogenmeyer has joined #ipfs
<apiarian>
err, IPFS, not IPLD (TOO MANY IP** ACRONYMS!)
<apiarian>
(sorry)
<apiarian>
err, IPNS entries to communicate *
<kpcyrd>
btw, I'm getting some of these on ipfs daemon: 20:48:49.640 ERROR dht: Error correcting DHT entry: dial to self attempted routing.go:126
<whyrusleeping>
kpcyrd: ah, i should drop that to a warning
<kpcyrd>
whyrusleeping: how does that happen? I've finished the mirror for now and next step is trying to fix performance :)
<whyrusleeping>
what do you mean how does that happen?
<kpcyrd>
whyrusleeping: I've tried running `ipfs pin $hash_of_whole_mirror` once, then canceled after a while and the errors sounds like it thinks it already has the object
aluu has joined #ipfs
<kpcyrd>
whyrusleeping: also, pulling arch packages from 127.0.0.1 almost never works, so I was wondering if there's something wrong with my local .ipfs/
<whyrusleeping>
doesnt work as in it hangs?
<kpcyrd>
whyrusleeping: pacman tries: 127.0.0.1:8080, h.ipfs.io, ipfs.io, h.cdn.telekollektiv.org, cdn.telekollektiv.org, h.mirror.rxv.cc, mirror.rxv.cc (requests are canceled by pacman if there it doesn't receive any bytes for 10s)
<whyrusleeping>
hrm... no errors from the daemon?
<whyrusleeping>
other than the dht thing
<ipfsbot>
[go-ipfs] whyrusleeping closed pull request #2885: Add ARC caching and bloom filter for blockstorage (master...feature/bloom-cache) https://git.io/voKhW
ianopolous has quit [Ping timeout: 272 seconds]
ianopolous has joined #ipfs
ashark has quit [Ping timeout: 276 seconds]
<Kubuxu>
whyrusleeping, kpcyrd is using IPNS, there is still quite high initial latency of IPNS resolution.
<Kubuxu>
<rant> and cache is valid for only 60s "because consistency" </rant>
<whyrusleeping>
does the ipns entry being used for the arch mirror belong to the node youre getting those errors on?
frankyr has quit [Quit: Connection closed for inactivity]
<Kubuxu>
it is slow also on other nodes
<whyrusleeping>
i know
<Kubuxu>
also it gives error that it couldn't resolve that name quite qucily
<whyrusleeping>
but i'm trying to debug one particular issue
<whyrusleeping>
having to deal with that error message he posted
<Kubuxu>
right
cemerick has quit [Ping timeout: 260 seconds]
<kpcyrd>
whyrusleeping: no, different node
<whyrusleeping>
kpcyrd: okay, cool. that helps
<kpcyrd>
whyrusleeping: the no link ones are fine, pacman seems to trigger 404s
<whyrusleeping>
Kubuxu: i've got latest master with your bloom filter code now running on mars, any info i can provide you from it?
<Kubuxu>
whyrusleeping: :5001/debug/vars after it runs some time.
vijayee_ has joined #ipfs
<whyrusleeping>
cool, will do
<Kubuxu>
or better after about 15min after start (bloom should be active for sure then) and then after some time (hour or few hours),
<Kubuxu>
or not, as I think I included those vars only in custom build ....
<Kubuxu>
I plan having metrics done this week, so we can start tweaking it after that.
vijayee_ has quit [Ping timeout: 258 seconds]
Foxcool has quit [Read error: Connection reset by peer]
rendar has joined #ipfs
ygrek has quit [Ping timeout: 244 seconds]
<kpcyrd>
hey. somebody mentioned that 400 might be a bad response code when ipfs/ipns fails :)
<kpcyrd>
some people ran into issues with the gateway
<lgierth>
yes there are quite a couple bad response codes in the gateway
<lgierth>
it's on the list -- help very very welcome :)
<Kubuxu>
yes it does but adding different response code is quite hard, lgierth is reworking gateway though
<lgierth>
i'm not reworking the gateway -- just extracting it
<lgierth>
so it's easier to rework
<lgierth>
btw Kubuxu when are you coming to berlin?
<lgierth>
i'm off to lisbon on the 11th? are you coming there too?
Encrypt has joined #ipfs
<Kubuxu>
I planned to come for start of July but after all I will come only on 10th to fly on 11th morning to Lisbon.
<lgierth>
heh ok
<Kubuxu>
I need to find place to sleep for 10th to 11th night.
ashark has joined #ipfs
Entroacceptor has joined #ipfs
vijayee has joined #ipfs
ashark has quit [Ping timeout: 240 seconds]
zorglub27 has quit [Quit: zorglub27]
vijayee has quit [Ping timeout: 276 seconds]
<kpcyrd>
is there a way to figure out a rough estimation which/how many nodes have a specific object?
<apiarian>
don't nodes publish their ownership of a specific object?
<kpcyrd>
yes, I'm trying to access it
kus has quit [Ping timeout: 240 seconds]
<kpcyrd>
ah, ipfs dht findprov
vijayee has joined #ipfs
arajakul1 has quit [Ping timeout: 272 seconds]
socon has joined #ipfs
erde74 has joined #ipfs
erde74 has quit [Client Quit]
Foxcool has joined #ipfs
neurosis12 has joined #ipfs
neurosis12 has quit [Changing host]
neurosis12 has joined #ipfs
<lgierth>
yeah findprovs
<lgierth>
the upper bound of many providers there are is always unknown
<lgierth>
*how many
<lgierth>
just as with the network size of cjdns :)
<ipfsbot>
[go-ipfs] Kubuxu created feature/blocks-cache-arc (+2 new commits): https://git.io/vKUCd
<ipfsbot>
go-ipfs/feature/blocks-cache-arc 985d177 Jakub Sztandera: blocks/blockstore: add CacheOpts - structure of cache config...
<ipfsbot>
go-ipfs/feature/blocks-cache-arc 8407cad Jakub Sztandera: blocks/blockstore: introduce context passing to blockstore...
<Kubuxu>
I am off for today, have nice rest of a day everyone.
<whyrusleeping>
Kubuxu: gnite!
vijayee has quit [Ping timeout: 250 seconds]
<lgierth>
o/
ashark has joined #ipfs
leotreasure has quit [Read error: Connection reset by peer]
ashark has quit [Ping timeout: 258 seconds]
leotreasure has joined #ipfs
vijayee has joined #ipfs
taaem has quit [Read error: Connection reset by peer]
Boomerang has joined #ipfs
taaem has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
Foxcool has quit [Remote host closed the connection]
ygrek has joined #ipfs
Foxcool has joined #ipfs
ralphtheninja has quit [Ping timeout: 264 seconds]
ralphtheninja has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
herzmeister has quit [Quit: Leaving]
herzmeister has joined #ipfs
ashark has joined #ipfs
ashark has quit [Ping timeout: 272 seconds]
vijayee has quit [Ping timeout: 244 seconds]
Boomerang has quit [Quit: Leaving]
jaboja has quit [Ping timeout: 264 seconds]
<lgierth>
whyrusleeping: we can move go-libp2p to the libp2p org
jaboja has joined #ipfs
mildred has quit [Ping timeout: 272 seconds]
vijayee has joined #ipfs
<deltab>
monzoon[m]: note that deduplication is done at the block level, not just for whole files; so there could be one file with timestamp 123 and content ABC, and another with timestamp 456 and content ABC, and the content would be deduplicated
<deltab>
so storing both takes up only a few dozen more bytes than just one
ofdm has quit [Ping timeout: 244 seconds]
<lgierth>
do you update the access timestamp every time someone fetches the file from ipfs or the gateway? ;)
<lgierth>
and what if they only fetch a few blocks, but not the whole file?
vijayee has quit [Ping timeout: 272 seconds]
pguth has quit [Remote host closed the connection]
pguth has joined #ipfs
jaboja has quit [Ping timeout: 264 seconds]
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]