<stevenaleach>
Ah... nm... works fine from the command line: mkdir foo; echo "hello" > foo/test.txt; touch foo/empty.file; ipfs add -r foo... But if I try it with this directory: /ipns/QmY6F81SgJQCAPw1zKtKR6v2LeoATt9JaR5ps8Ai9Z8Jw5/PUB then adding an empty file breaks things it seems.
reit has joined #ipfs
<whyrusleeping>
i just pulled that dir down
<whyrusleeping>
and did 'touch empty && ipfs add -r .'
<whyrusleeping>
it appears to have worked fine, what problems are you seeing?
<stevenaleach>
Not sure what I'm doing wrong here /ipns/QmY6F81SgJQCAPw1zKtKR6v2LeoATt9JaR5ps8Ai9Z8Jw5/PUB/gnosi.py If PUB contains an empty file of any kind, I get null for the pub link - not the directory full of files.
MDude has quit [Ping timeout: 260 seconds]
<kythyria[m]>
What sort of chunk sizes are efficient for IPFS?
bastianilso has quit [Quit: bastianilso]
<stevenaleach>
with gnos an existing directory hash: l = ipfs.add(PUB_DIR,recursive=True)[-1]['Hash']; n = PUB_DIR.split("/")[-1]; gnos = ipfs.object_patch_add_link(gnos,n,l)['Hash'], if PUB_DIR contains no empty files, I get a PUB_DIR named directory in the main directory (what I want). If I put an empty file in PUB_DIR, then I get a link to nothing instead in the main directory..
<stevenaleach>
Argg.. works fine from the commandline but fails with the code above.... here's what I get when the script publishes itself with an empty in ~/PUB: /ipfs/QmcCT9rukfrQ3DdegNBwFKCGnSttKcxTmCyJ1SauszymTo ... when I delete the empty and rerun, /ipfs/QmZ5GPnvhyQ5L8QdrhKHtWYdvmFgGSmbbN5ejGpTindztm...
<whyrusleeping>
stevenaleach: whats wrong there?
<whyrusleeping>
oh, i see. the pub directory
<whyrusleeping>
got it
<Kubuxu>
I would say files api is much simple in this case
<Kubuxu>
and it will properly work when sharding comes into play
<stevenaleach>
but if I add from the command line with "ipfs add -r PUB", I get a good link. If I do it from the python api I get a null directory.
<stevenaleach>
files api?
<whyrusleeping>
stevenaleach: hrm... did you write the python api?
<whyrusleeping>
i wouldnt put much faith in it being correct (i don't beleive its well tested)
<Kubuxu>
`ipfs files --help`/.
<stevenaleach>
Thought that was for copying to mfs from ipfs?
<Kubuxu>
yeah and then you can take it back
akkad has joined #ipfs
<Kubuxu>
ipfs files stat --hash /some/dir
<Kubuxu>
to get hash of that dir
matoro has joined #ipfs
<stevenaleach>
so forget ipfsapi and use the commandline tool directly?
dryajov1 has joined #ipfs
<Kubuxu>
it also works through http api
<stevenaleach>
Ok... so what does that mean in terms of writing python code to talk to it?
dryajov1 has quit [Client Quit]
<Kubuxu>
you would have to have similar bindings you have for `ipfs add` `ipfs object ...` but for `ipfs files ...` which probably wasn't written by anyone
MDude has joined #ipfs
<stevenaleach>
but access those bindings how?
<mguentner>
Kubuxu: it is, only api.files_flush() is missing (will add that soonish)
<Kubuxu>
flush isn't criticla (as long as you are not using --flush=false`
<stevenaleach>
Wow... any speed advantage over python subprocesses calling the CLI tool? If not, I'm guessing that's a whole lot easier than learning how to do tcip connections just to talk to the machine you're currently on :-/
Boomerang has quit [Quit: leaving]
ygrek has joined #ipfs
fleeky_ has quit [Read error: Connection reset by peer]
fleeky__ has joined #ipfs
<stevenaleach>
Ack... I don't get it... I have no idea what the files api does.... what does "ipfs files mkdir" do? does it just create a local files system directory or does it add something to ipfs 'cause I just tried "ipfs files mkdir /foo" and it ran just fine... and didn't give me a hash.. what was created? what does this command, or any of the rest of the "files" do?
stevenaleach has quit [Remote host closed the connection]
jkilpatr has quit [Ping timeout: 255 seconds]
infinity0 has quit [Remote host closed the connection]
stevenaleach has joined #ipfs
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
dryajov1 has joined #ipfs
infinity0 has joined #ipfs
ianopolous has quit [Quit: Leaving]
kvda has joined #ipfs
<stevenaleach>
Whew... Thanks, mguentner! I just ran an update on all my pip installed packages and though I could have sworn I saw "requirement already up to date" for ipfs, things suddenly work - it's happy with both empty directories and empty files. Yay. Thanks!
<mguentner>
stevenaleach: awesome
ygrek has quit [Ping timeout: 240 seconds]
tmg has quit [Ping timeout: 252 seconds]
HostFat_ has joined #ipfs
<kvda>
just manually installing go-ipfs form source for the first time on alpine-linux
<stevenaleach>
Arg... ok... it seems to be random whether it works or not... how do I install the latest git version?
<kvda>
go install ./cmd/ipfs
<kvda>
doesn't seem to make a bin file available in /go/bin?
<stevenaleach>
version in pip isn't working yet.
<kvda>
where does the 'ipfs' bin end up after go install ./cmd/ipfs?
<Kubuxu>
kvda: do you have GOBIN set?
HostFat__ has quit [Ping timeout: 240 seconds]
bmoriarty has left #ipfs [#ipfs]
<kvda>
@Kubuxu no...
<kvda>
it's just on the path
<Kubuxu>
yeah, it was quite recent change in golang iirc
dignifiedquire has quit [Quit: Connection closed for inactivity]
kvda has quit [Remote host closed the connection]
<stevenaleach>
Well.. dern. I was mistaken. Empty files still break things. I'm not going to worry about it for now, I just won't have empty files.
<stevenaleach>
If I want to avoid library issues and talk to ipfs directly (the command line tool adds directories with empty files or empty directories just fine), and if my code will always be running on the same node as the daemon, is there a reason to use the local http gateway versus forking subprocesses running the command line tool?
kvda has joined #ipfs
<kvda>
Kubuxu exporting GOBIN fixed the install
<kvda>
thanks again
<gde33>
is it/will it be possible to remove copyright violations from the network?
<gde33>
or any other kind of not-nice materials
<SchrodingersScat>
gde33: is the existing hashblocking not sufficient?
<gde33>
SchrodingersScat: just curious about the paradox of not nice things vs censorship/centralized control
<gde33>
I'm not entirely familiar with the way ipfs works atm
<SchrodingersScat>
gde33: seems like they can block certain hashes, like there are some you can't reach from ipfs.io
<stevenaleach>
Oh... Crap. It's something fundamentally broken with py-ipfsapi I'm thinking. It sometimes works and sometimes doesn't when adding a directory (at least the way I've been doing it?) Sometimes, the directory and all its contents (to include empty files) is published by my system and sometimes a null link is published in its place. If I run my (kludgy and ugly so far) script again, the dice are re-rolled. Definitely going to
<stevenaleach>
have to rewrite to call go-ipfs from the commandline instead and see if I can get some consistent plumbing here.
<gde33>
SchrodingersScat: seems a can-o-wurms, I cant imagine a nice formula. Will try read up, thanks
<whyrusleeping>
stevenaleach: yeah, the python ipfs api isnt complete or tested yet
<stevenaleach>
Guessing that now :-) Reworking to just call go-ipfs in a subprocess as needed.
tclass has joined #ipfs
tclass has quit [Ping timeout: 240 seconds]
<SchrodingersScat>
gde33: what are your concerns?
<gde33>
SchrodingersScat: just curious
matoro has quit [Ping timeout: 260 seconds]
kulelu88 has quit [Quit: Leaving]
chris613 has quit [Quit: Leaving.]
aeternity has joined #ipfs
stevenaleach has quit [Remote host closed the connection]
<kythyria[m]>
It's nonetheless equivalent to asking if there's some mechanism to remove copyvio from the web.
<SchrodingersScat>
like a hammer?
dryajov2 has joined #ipfs
dryajov has quit [Ping timeout: 255 seconds]
zombu2 has quit [Ping timeout: 255 seconds]
mguentner has quit [Quit: WeeChat 1.6]
Mateon3 has joined #ipfs
kevina_ has joined #ipfs
Mateon1 has quit [Ping timeout: 256 seconds]
Mateon3 is now known as Mateon1
mguentner has joined #ipfs
matoro has joined #ipfs
kevina_ has quit [Quit: Leaving]
Mateon3 has joined #ipfs
Mateon1 has quit [Ping timeout: 240 seconds]
Mateon3 is now known as Mateon1
Mateon2 has joined #ipfs
Mateon1 has quit [Ping timeout: 245 seconds]
Mateon2 is now known as Mateon1
qgnox has joined #ipfs
Foxcool has joined #ipfs
mguentner2 has joined #ipfs
mguentner has quit [Ping timeout: 248 seconds]
qgnox has quit [Quit: Leaving]
aeternity has quit [Ping timeout: 240 seconds]
stevenaleach has joined #ipfs
dignifiedquire has joined #ipfs
<mekel>
hi deltab
<mekel>
successfully moved it to usr/local/bin
<mekel>
now i am stuck on ipfs --help
<mekel>
i get bash: /usr/local/bin/ipfs: cannot execute binary file: Exec format error
<deltab>
run the command file /usr/local/bin/ipfs
<deltab>
to find out what it is
<deltab>
what does it say?
galois_dmz has joined #ipfs
<mekel>
ran file /usr/local/bin/ipfs
<mekel>
got /usr/local/bin/ipfs: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, not stripped
<mekel>
can you decypher that for me deltab? thanks for the help!!
<deltab>
that looks okay: it's an executable for 64-bit linux
<deltab>
my guess is that you're not running 64-bit linux
<deltab>
what does uname -a output?
<mekel>
Linux mekel-VirtualBox 4.4.0-59-generic #80-Ubuntu SMP Fri Jan 6 17:36:54 UTC 2017 i686 i686 i686 GNU/Linux
<deltab>
i686 means 32-bit
<deltab>
you'll need to download the 32-bit version
<mekel>
ah! got it! thanks so much!
lachenmayer has quit [Ping timeout: 260 seconds]
zopsi has quit [Ping timeout: 264 seconds]
xelra has quit [Remote host closed the connection]
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tclass has quit [Ping timeout: 240 seconds]
kvda has joined #ipfs
cemerick has quit [Ping timeout: 248 seconds]
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
aeternity has quit [Ping timeout: 255 seconds]
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
wmoh has quit [Ping timeout: 256 seconds]
Boomerang has joined #ipfs
cemerick has joined #ipfs
pfrazee has quit [Remote host closed the connection]
Oatmeal has quit [Ping timeout: 240 seconds]
Boomerang has quit [Ping timeout: 276 seconds]
<phorse>
does anybody know what's going on with pubsub?
pfrazee has joined #ipfs
Boomerang has joined #ipfs
Boomeran1 has joined #ipfs
Boomerang has quit [Ping timeout: 255 seconds]
Oatmeal has joined #ipfs
shizy has joined #ipfs
aeternity has joined #ipfs
<keks>
phorse: what do you mean?
<phorse>
keks: last night I was reading about various suggested pubsub implementations for ipfs, and I'm curious if there is one under active development
<lapinot>
i was wondering what is happening with ipld: is it useable (or still in early pre-release)? Will it fully replace the current ipfs object model (and thus introduce a break in the api)?
<lgierth>
yeah it's still in pre-release, but usable
<lgierth>
and the old object model (merkledag, unixfs) will continue to work
<lapinot>
i'm asking because i'm eager to see an ipfs+ipld backed git repo (i mean really with the object structure, not just dumping the .git directory in ipfs) to see how it works
<lgierth>
same here :) the cool thing is we can address any content-addressed system's data with that system's hashes, using CID: github.com/ipld/cid
<lgierth>
e.g. kumavis is dealing with ethereum data structures in ipfs, iirc
<lgierth>
and whyrusleeping has built a zcash block explorer on ipfs
<lapinot>
nice then... is there any roadmap? I can't find any! (sure it's great to have such a tiddily organised set of repos but the downside is the learning curve)
pfrazee_ has joined #ipfs
pfrazee has quit [Read error: Connection reset by peer]
<Kubuxu>
keks: so what is the plan for it right now?
<Kubuxu>
as there are a lot of long comments there :)
<lapinot>
as a side question: has there been any high load test of the ipfs network? If ipld is going to be more or less stable, i'm going to work on some fully distributed data store (kinda to replace that awefully centralized torrent community). (note: didn't see this topic really debated here, is this something i should avoid?)
<keks>
Kubuxu: first, rip it out of the go-ipfs tree
<keks>
then, don't pass a response to cmd.Call, but a responseemitter
<keks>
that writes e.g. to the http connection and has full control over it
<keks>
and the http client returns a response
<keks>
that has a smaller api surface than it currently has
<keks>
also i remove the Output() madness
<keks>
now you can ResponseEmitter.Emit(value)
koalalorenzo has joined #ipfs
<keks>
if it is an io.Reader() stream it to the connection
<lapinot>
edit: just read the topic line with link to the 300TB challenge, so it's partially answered
<keks>
else, send it to the response
<keks>
the response type now has a Next() (interface{}, error} method
<Kubuxu>
yeah
<keks>
if you sent an io.Reader, that reader will be returned there
vapid is now known as Redd1t
<keks>
imho that api is just as powerful but has much less weird corner cases
<Kubuxu>
what about allowing to specif the encoder in the API call?
G-Ray_ has quit [Quit: G-Ray_]
<kevina>
Kubuxu: I see I have voice now, cool.
<kevina>
Now what does that mean. :)
* kevina
shows his irc ignorance
<Kubuxu>
almost nothing, higher flood limit
<achin>
it means you have a very attractive hat!
<kevina>
"higher flood limit" ?
<richardlitt>
Note: We are going to take an extra 15 minutes for this All Hands call to do demos about data.gov
<kevina>
never mind, looked it up
tclass has joined #ipfs
<keks>
Kubuxu: yeah that is an option i just want to get legacy compatibility done first
<Kubuxu>
it is already in the PAI
<Kubuxu>
API
<victorbjelkholm>
FrankPetrilli: awesome work on the kubernetes-ipfs stuff! :)
tclass has quit [Ping timeout: 240 seconds]
espadrine has quit [Ping timeout: 255 seconds]
ygrek has joined #ipfs
galois_d_ has quit [Remote host closed the connection]
galois_dmz has joined #ipfs
koalalorenzo has quit [Quit: Sto andando via]
Encrypt has joined #ipfs
<Kubuxu>
sorry that the stream couldn't see what I was talking about
galois_dmz has quit [Ping timeout: 245 seconds]
espadrine has joined #ipfs
arpu has joined #ipfs
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
s_kunk has quit [Ping timeout: 240 seconds]
aquentson1 has joined #ipfs
<keks>
Kubuxu: currently we use Marshalers which I want to deprecate in favor of Encoders, which directly operate on an io.Writer
<dignifiedquire>
is there an irc channel for parity?
john has joined #ipfs
john is now known as Guest87261
rendar has quit [Ping timeout: 255 seconds]
shamb0t has quit [Remote host closed the connection]
<kumavis>
i dont think so
wallacoloo____ has joined #ipfs
<kumavis>
they are on gitter
shamb0t has joined #ipfs
<kumavis>
i invited him to this irc channel
Guest27309 has quit [Ping timeout: 260 seconds]
<kumavis>
seems he is in berlin, maybe an ipfs berliner could meet him at a cafe to help ( just an idea )
<kumavis>
i think the requirements are pretty straightforward
<kumavis>
i just really want to have this somewhat useable by the end of the week, for the eth-ipfs hack week
<kumavis>
things seem to be moving forward but so is the clock
<lgierth>
i don't think it'd be helpful for any random person to meet him
<dignifiedquire>
none of the ones in Berlin know rust so not sure that helps, but I'll try and make myself available, means he is in the same timezone as me at least
<lgierth>
i have 0 context about your work
<lgierth>
yup cool
<dignifiedquire>
lgierth: try to keep up ;)
<kumavis>
that would be awesome, i will be less available during his timezone
<kumavis>
ive been listening in on the meetings, they just overlap with my commute
<kumavis>
so hard to join properly
apiarian has quit [Ping timeout: 258 seconds]
apiarian has joined #ipfs
galois_dmz has joined #ipfs
<kumavis>
daviddias: if i add my name to js impl of eth-objects
<kumavis>
what do i need to do
<kumavis>
to fulfill my duty as sprint participant
tilgovi has joined #ipfs
tclass has quit [Remote host closed the connection]
emerson has quit [Remote host closed the connection]
emerson has joined #ipfs
<daviddias>
A lot of this sprint is focus on getting all the stuff we need in order to support eth dev on top of IPLD, the tasks listed represent our understanding of what is needed, but getting a status update from you and a list of things you expect to be able to do next week would be awesome to understand what we are missing
<daviddias>
then we can divide and conquer in this list of tasks :)
tmg has joined #ipfs
<kumavis>
cool - daviddias i assume i already missed the relevant call but im available if you want me to jump into one
neurrowcat has quit [Quit: Deebidappidoodah!]
<kumavis>
added to the doodle
<daviddias>
that call was last week, but happy to chat for a bit
<daviddias>
wanna do it now?
cemerick has quit [Ping timeout: 248 seconds]
<kumavis>
sure
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
Encrypt has quit [Quit: Quit]
SuperPhly has joined #ipfs
jonnycrunch has quit [Quit: jonnycrunch]
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
wallacoloo____ has quit [Quit: wallacoloo____]
bastianilso has quit [Read error: Connection reset by peer]
bastianilso has joined #ipfs
frmendes has joined #ipfs
Boomerang has joined #ipfs
dryajov1 has joined #ipfs
dryajov1 has quit [Client Quit]
bastianilso has quit [Quit: bastianilso]
bastianilso has joined #ipfs
galois_dmz has quit [Remote host closed the connection]
aquentson1 has joined #ipfs
<kumavis>
wooo so excited for eth-ipfs hack week
<kumavis>
head my explode
<daviddias>
:D :D
aquentson has quit [Ping timeout: 255 seconds]
rendar has quit [Quit: std::lower_bound + std::less_equal *works* with a vector without duplicates!]
aquentson has joined #ipfs
aquentson1 has quit [Ping timeout: 255 seconds]
jkilpatr has quit [Ping timeout: 276 seconds]
chris613 has joined #ipfs
aquentson has quit [Ping timeout: 248 seconds]
tclass has joined #ipfs
aquentson has joined #ipfs
SuperPhly has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
bastianilso has quit [Quit: bastianilso]
wallacoloo____ has joined #ipfs
tclass has quit [Ping timeout: 255 seconds]
<elopio>
I've just pushed v0.4.5-rc1 to the Ubuntu candidate channel. To get it: $ sudo snap install ipfs --candidate