[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vp / vr / w / wg / wsg / wsr / x / y ] [Home]
4chanarchives logo
/ipfsg/
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.

You are currently reading a thread in /g/ - Technology

Thread replies: 183
Thread images: 5
File: ipfs.png (70 KB, 1024x1024) Image search: [Google]
ipfs.png
70 KB, 1024x1024
Old thread >>51477984

>What is IPFS?
It's basically BitTorrent on steroids.

>why would one use it
* Distributed, decentralized filesharing (for now - ipfs is merely the communication protocol, and additional applications can exist on top of it).
* You can have a mutable address (i.e. always points to the latest version of a site), or a static address (points to a specific file). Yes, you can host sites over IPFS.
* Peers are found fast for new downloads. You don't need to wait that much to start a download.
* You can watch your animu while it downloads, I watched few episodes that way and it didn't even buffer.

>how to upload a single file
$ ipfs add ./$file
Access it at localhost:8080/ipfs/$outputted-hash
>how to upload a dir
$ ipfs add -r ./$dir
Access it at localhost:8080/ipfs/$last-outputted-hash
>how to make the thing mutable
$ ipfs name publish ./$file-or-dir-hash
Access it at localhost:8080/ipns/$output-hash-aka-peerid (it's ipNs not ipFs)
To update, publish another hash and it will be available at the same IPNS address.

>gateways (how to access IPFS if you don't have it installed)
https://gateway.ipfs.io/

>most recent talk about it by the dev
https://www.youtube.com/watch?v=HUVmypx9HGI

>I2P and Tor support coming soon™. We need that thing anonymous so pls halp.
https://github.com/ipfs/go-ipfs/issues/1118
https://github.com/ipfs/notes/issues/37
https://github.com/jbenet/go-multiaddr-net/issues/7

Daily reminder to pin files that you care about.
ipfs.io/ipns/gindex.dynu.com/
( aka ipfs.io/ipns/QmaGks9KKzu2WykHQjJFJkcUAN4ZoF7ok9h2hXj1WQn47U/ ) has a non-comprehensive list of content.
>>
Only big guys will pin this QmQWwypyp4Gfv4NPd2cM
cLsi4gGHEzpNPGj1Tr1La5bULT
>>
>>51492926
If you want me to include your content on the site, just reply to this post with your link.
>>
I can't seem to pin anything from other people. I do:
ipfs pin add -r <hash>
And it gets stuck for hours and does seemingly nothing.

Tips?
>>
>>51493086
What are you trying to pin? Normally it should be done almost instantly unless you're downloading something massive as well.
>>
>>51493086
Sometimes happens to me too, I'm guessing it might just be a bug?
>>
>>51493100
https://ipfs.io/ipfs/QmTmMhRv2nh889JfYBWXdxSvNS6zWnh4QFo4Q2knV7Ei2B/

In retrospect, it does seem to be a bit big, so I'm guessing upload bottleneck.
But then I tried pinning just the networking directory, and it seems to stall too.
>>
>>51493148
The gentoomen library is around 50GB now, it's pretty massive. Better to pin just the sections you care about probably.
>>
>>51493148
When you pin the directory it will recursively pin everything it contains.
Yeah, speeds will be slow if there's only one person seeding it.
Use ipfs get to at least have an idea about how much progress it has done and pin it once it has finished downloading.
You should also make sure you have enough space on your drive and in your ipfs repository (StorageMax in the Datastore section of the config file in ~/.ipfs/).
>>
>>51493285
You don't need to adjust StorageMax, that's only for the passive cache I think.
>>
>>51493022
Add this under 18+ if you want:
>>51491510
>>
>>51493337
Done. Publishing new version. Should be up soon.

Of note to whoever wants to try making dynamic sites over ipfs: there's now ipfsbin and an ethereum demo for simple storage over ipfs which might be useful references.

Additionally, I'm trying to get the content from ipfs.io/ipfs/QmTNsb6JHeG1k2UfxF1tz
4wL8YsJwcrFGiEnhj88LitXA2/hypertext
and ipfs.io/ipfs/QmTYdo9E59qWSuqes
n4E8VfXxHCUx59hRhjkgaSbPPJRyo/
If someone can access it and tell me what it is exactly, that would be great.
>>
>no gnunet thread whatsoever
WE DID IT! WE KILLED THE GNU!
>>
>>51493503
Could I get a link to that ehereum demo please?
>>
>>51493635
Open wide: Aaaaah~
ipfs.io/ipfs/QmPnH8thybK
yU8ySEHgqo2AMS2Bb7y87
WnsgM5vxvktJHT
>>
>>51493705
Thanks senpai
>>
>>51493705
Do I have to have Ethereum installed for this to work?
>>
>>51494085
Yes, obviously.
>>
I swear pinning looks broken.

I tried to ipfs pin add -r the Pepe folder, and nothing's happening. That doesn't look like that big of a folder.
>>
>>51494245
How is this going to help web dev then. You cant expect everyone browsing your site to have Ethereum installed.
>>
>>51494281
Restart the daemon and try again. Its a fair size of folder. If restarting doesnt work, try downloading it first without pinning, and then pin after its downloaded.
>>
>>51494295
In the same way that you can't expect everyone to have ipfs am i rite? In the same way that you can't expect everyone to have a javascript-capable browser, in the same way that you can't expect everyone to use http?
>>
>>51494342
Eventually ipfs is going to be a very widespread thing, most everyone will have the capabilities to use it. Same with http and jabbascript.

Is Ethereum a service lightweight enough to come packaged with every browser in the future?
>>
>>51494391
Yes and no. Ethereum uses a currency (eth) to encourage nodes to process requests, and one has to store the blockchain to know the state of the network. However, thanks to the pandacoin devs, it's now possible to sync blockchains without downloading the full thing, so storage of the blockchain is actually not an issue in 2015. The ethereum software itself is very lean. There are several currency-free platforms out there, though, such as eris.
>>
>>51494471
What are the benefits of currency free vs currency not-free platforms?
>>
>>51494512
Incentive to do processing work.
>>
>>51494540
What are the consequences for not doing processing work?
>>
>>51494512
Currency-free reduces the barrier of entry. There's no need to either perform possibly intensive work or use real money. On the other hand, nodes have a strong incentive to help the network, so you expect network health to stay consistently higher.

Also I'm not sure what the payment model for eth is anymore, I actually think it's original uploader pays the network to verify contracts for a set amount of time, not the users.
>>
>>51494588
Network dies if nobody does work because there's nobody to verify that requests are valid (so websites can be attacked) and everything goes super-slow (because nobody's there to process shit).
>>
>>51493503
Wow, that index is really starting to get meaty, good job
>>
>>51493630
once tpp comes around and the copyright agencies go on their inquisition people will come crawling back to GNUnet/Freenet/i2p for it's anonymity.
>>
so how's it different from freenet/ i2p/ mega/ whatever
>>
>>51495259
i2p can't be used alone because it's centralized content hosting.
Freenet has limited capacity and thus isn't viable for large content archival.
Gnunet is slow as balls, buggy as shit, has no 3rd party interests, and very little development efforts or plans.

However, ipfs + i2p is what it's all going to be about.
>>
It would be good if we scraped libgen.
Other things to upload include: anime, music, tvshows, movies, vidya. Someone's already collecting mame roms, and we have the full atari collection up.
>>
>>51495702
Theres an anime tracker running already. Doesnt look like many people are uploading animes to ipfs yet though. Makes sense I suppose. Considering uploading them copies them in your disk.

/ipns/QmUqBf56JeGUvuf2SiJN
JahAqaVhFSHS6r9gYk5FbS4TAn
>>
Someone just upload a wiki website software so we can all add our content to it without OP needing to do it.
>>
>>51495774
wait what... copies them in ur disk?

it dosnt just use the location of the file i already have?
>>
>>51495808
>>>/reddit/
>>
>>51495790
Do you mean the ipfs wiki or something else?
>>
>>51495808
Nope. It copies the files you add to your cache. Its planned to be able to not copy them, but its low priority apparently.
>>
>>51495808
Not yet, it's being looked at I think

>>51495774

>S4TAn
SPOOKED
>>
>>51493705
This isnt working for me. I have ethereum installed and running. Page loads fine too. But setting and getting the variable doesnt work.
>>
>>51495836
im mean the equivalent of the tor darknet wikis

somewhere everyone can just go and add links, without needing to post them here and have OP add them.
>>
>>51495851
Maybe it should be renamed to the satan anime tracker.
>>
>>51495867
Did you run geth --rpc --rpcport 8101 --rpccorsdomain "*" ?
>>
>>51495897
And also, are you on the latest eth network? Olympic is the really old one, then there's another I don't recall the name of, and the current one might be one version later even.
>>
>>51495897
No, I cant. After running the install, it decided I was to hold the eth command instead of geth. Followed directions here: ethereum .org/cli

But I ran the command given at the bottom and I successfully connected to the network.

>>51495951
Yeah. Im connected to Frontier.

>>51495871
There is a site like that. It was posted in the OP

/ipns/QmaGks9KKzu2WykHQjJFJkcUAN4ZoF7ok9h2hXj1WQn47U/
>>
>>51495976
>There is a site like that. It was posted in the OP
but no one can edit it but the OP
>>
>>51495976
You should use the geth client, it's the most developed one. Most of the examples are incompatible with the non-geth clients.
>>
>>51496006
Oh you mean like that.

Currently a site like that is impossible with IPFS. Dynamic content is very limited. Static sites are pretty much all we can host.
>>
>>51496018
How can I install the geth client? I went through the install instructions that ethereum gave me. It installed eth in place of geth.
>>
>>51496032
i thought ipfs was versioning? why could a system not be constructed where users make new versions of html files and submit them. and some sort of tree showing version history.
>>
>>51496043
Oh wait nevermind. I made a mistake. Ill install it now.
>>
>>51496071
because otherwise a random person could come along and thrash a random site by posting disruptive content or just throwing the site away. hence, signing is required.
>>
>>51496087
but there's be version history. it's git like right?
>>
>>51496082
Also make sure you have a non-0 eth balance.
>>
>>51496071
When files are edited to make new versions, they generate an entirely different hash. Which means an entirely different link to get there.

This is partially averted by using ipns, but only the original publisher can publish to their ipns.
>>
>>51496104
No. Git and this are entirely different.
>>
>>51496113
so really we should just have a non ipfs wiki with links to ipfs files
>>
>>51496136
Rather, an ipfs wiki with ipfs files, but a non-ipfs interface that allows modifying the ipfs files, causing a republish, basically.
>>
>>51496136
That would be optimal until ipfs can support dynamic pages. But that aint much fun is it.
>>
>>51496149
thats a neat idea that would hopefully require minimal work

i hope someone does it
>>
>>51496151
E R I S
R
I
S
https://db.erisindustries.com/distributed%20business/2015/11/01/eris-and-ipfs/
https://eng.erisindustries.com/tutorials/2015/08/05/ipfs-as-a-service/
>>
>>51496182
>https://eng.erisindustries.com/tutorials/2015/08/05/ipfs-as-a-service/
This is literally just the same commands ipfs has. But start with eris instead of ipfs.
>>
>>51496209
It's ipfs integration inside eris. It allows eris contracts to talk natively with ipfs objects.
>>
>>51495841
Can't I just use a hardlink?
>>
>>51496476
Oh I misunderstood, I apologise
>>
Is it possible to change the location of the cache? I have only 500GB free on the drive my home folder is on but I have a couple TB on another drive.
>>
>>51496151
Maybe this can be adapted?
https://github.com/eris-ltd/toadserver
>>
>>51496982
> ipfs config datastore.path ~/.ipfs/datastore
>>
I keep getting errors, and not receiving data. is it my connection, or is it because I am using the windows version?
>>
>>51499817
Is your connection stable?
>>
So is this like glorified dropbox for loonix users or what?
>>
>>51500800
You could do your own research instead of making dumb comments.
>>
>>51495702
>It would be good if we scraped libgen.
To avoid anyone having to download the whole collection, everyone should pick a random number N between 1 and 1403 and download http://libgen.io/repository_torrent/r_${N}000.torrent, add that to ipfs and post the hash. As I described in http://127.0.0.1:8080/ipfs/QmQUTjk7VG8qGg66rZWhDSBREga MFdyW8zVFJWfWnTLsxh/index.html#p51486484 someone could then combine the hashes into a single directory and pin _only_ that directory (i.e. non-recursively). This would create the entire database without any single person having to download all of it.
>>
>>51501288
>random number
Or we could do it sequentially, but that's harder to coordinate.
>>
I have a pretty decent collection of Korean movies and series. My internet is utter shit, and I'm not online all day. Should I even bother with hashing all this shit?
>>
dumping scraped music from jetsetrad.io
/ipfs/QmbQbpaJLm9jCqfKUsEEwab7EHNnFhz3QPfb9DfTyToz82
>>
>>51500800
It's more like bittorrent for the web.
>>
Is the absolute windows path (including drive letter) in published filenames a bug?
>>
>>51502328
yea, it was designed around unix paths

"\" is a valid filename character in unix systems, infact, any full windows path is a valid unix filename
>>
>>51502374
That's weird because Go's path package has functions that go to and from OS-specific paths.
>>
seems like the version available in /ipfs/QmZiSAYkU7g ZtqYeZWL21yuw gFtRnJu1JjDzR6 Qd2qdDBr/ is slightly outdated, even if it says to install from there.
>>
>>51492926
whats the difference between this and demonsaw
>>
>>51501840
thanks
>>
>>51500800
dropbox is a centralized file syncronization service, not even close

ipfs is a web implemented as a distributed filesystem, it's flexible enough that it can be used like dropbox, or like bittorrent, or like a CDN, or like a (static) http server
>>
>>51501798
No, don't bother. IPFS relies heavily on 24/7 availability and good network connection. Only items of high demand might be propagated properly. The goal of IPFS is different and is to speed up loading of data, by caching it on the receiving machines in congested local networks.
>>
>>51503848
The goal is also to make the network censorship resistant. It is fully decentralized and distributed after all. One has to have people willing to keep the content up, though.
>>
>>51502454
Yeah, it's an ipfs link. If someone can find the ipns link that points to the dist site it would be great.
Really, if anyone finds the ipns sites corresponding to any of the ipfs sites I list, please tell me.
>>
Where is the ipfs cache on windows?
>>
>>51504747
%AppData% probably?
You could look for a file containing the string "datastore"
>>
>>51504747
X:/Users/you/.ipfs/
>>
>>51504830
That's the first place I looked.
>>51504882
Thanks. :3
>>
So it's just DOS file that flashes some text and then disappears?
>>
>>51504961
>can't into cl
>>
>>51504961
yes, that is literally all it is/does
>>
>>51504961
Use ipfs in cmd.
>>
>>51501288
Is the dbbackup_last_all not a full dump?
>>
>>51495976
>>51492926
>Course Technology Cplusplus Programming From Problem Analysis to Program Design 7th.Edition[A4].pdf 268MB
>268MB
wat
>>
>>51505101
Doubt it. It's just 3 GB. The Gentoomen library is 15 times that size already.

Once when they were in danger of being shut down a few years ago, I remember them asking users to do exactly the "download one random torrent" thing. They wouldn't exactly have to do that if it was that small.

dbbackup is probably just metadata.
>>
Could someone make a client-side content search program? Make a search box, when we use it we give it a db to search from and the key, and it generates a table with the returned columns?
This way it should be much easier to manage dumps of large websites - just build a name - meta - ipfs link table and serve a simple webpage with the search program.
>>
>>51505286
I can't even get the torrents to resolve anyway so whatever.
>>
>>51505286
The .rar are extremely highly compressed. standarts is 37mb compressed, 365mb uncompressed. mirknig is 182mb compressed, 1.8gb uncompressed. If it's the same ratio for the dbbackup_last_all, it should be around 25GB uncompressed.
>>
>>51505363
Well, I just downloaded one random torrent to check, and it ended up on 2.8GB uncompressed. There are 1400 of them, so a rough estimate gives a total size of around 4TB.
>>
>>51505394
Wonder how the db got that big then.
>>
>>51505408
It might contain the cover images for every book or something. We should probably download it and check.
>>
>>51505317
Me neither. I see one peer in the swarm, but he refuses to connect to me. Probably the KGB.
>>
>>51505455
I'm downloading basically everything not part of the torrent, only the 2.7GB file is left.
>>
localhost:8080/ipns/gindex.dynu.com should be live for most people now.
>>
>>51505780
Nice. Looks like it's come along quite a bit since I last checked.
>>
>>51505853
>>51505780
This is starting to remind me of webrings, by the way.
>>
>>51505867
The next couple of years on ipfs are going to be glorious until the normies invade.
>>
>>51506029
>>51505867
If you want to help out, help us archive libgen ( >>51501288 ), submit anime to the anime tracker ( /ipns/QmUqBf56JeGUvuf2SiJNJahAqaVhFSHS6r9gYk5FbS4TAn ), start your own tracker for other content, play with smart contract frameworks to try to get a fully dynamic site up and running (tendermint, eris, ethereum, or something completely different), start an http portal to upload content for an ipfs site, or write us a generic client-side 'dumb' search engine.
>>
>>51506107
I'm going to host the archive.moe db dump soon-ish (without the images) but I need to buy a couple of new HDDs first.

I'm also thinking about creating a place where people could add their hashes and tag them that would mostly run on IPFS.
>>
>>51506166
>without the images
;_;
>>
>>51506329
If you want them you can download them from the internet archive. It's only a few terabytes.
>>
>download speed to libgen is now down 90%
Wew!
>>
Why does it take so long to pin /ipns/gindex.dynu.com?
>>
>>51507444
Because you have to do a dns lookup, then resolve the ipns name, and finally download and pin the site itself.
>>
>>51507497
Should this take over 10 minutes?
>>
>>51507497
That shouldn't take long at all, anon. The real answer is that there is a bug with long-running daemon sessions that makes pin hang.
>>
>>51507526
Restarted the daemon and pin still seems to be hanging.
>>
>>51507610
Strange, I just tried it myself and I get the same problem. Not even ls works, and that should only have to pull the directory listing file.
>>
>>51507663
Looks like I can't get ls to work either.

Weird.
>>
>>51507511
No, it should take around 1 minute tops. Try restarting your daemon maybe.
>>
>>51507663
>>51507685
>>51507610
What happens if you ipfs get the site?
>>
>>51507708
Looks like that's not working either.

Yet somehow it works fine in the browser?
>>
>>51507764
are you accessing it through ipfs.io or through localhost:8080?
>>
>>51507791
http://127.0.0.1:8080/ipns/gindex.dynu.com/
>>
>>51507791
Also try direct access/direct pin perhaps. ipns name: QmaGks9KKzu2WykH
QjJFJkcUAN4ZoF7ok9h2hXj1WQn47U
ipfs name: QmRhW3Ed73zpftdG1
nUdyezGASDpaix2hHd5wdFgB6wP6q
>>
>>51494342
In the same way that you can't expect people to browse, then? You won't expect to reach everyone, but you must plan yourself to reach as many people as possible that may be interested in what you're providing. And because there are STANDARDS, by those you are assured that everyone who browses WILL have a JavaScript-capable browser using HTTP.

>also, the average Doe today still doesn't know how to torrent, even some who use FTP for work.
>>
>>51507821
>mfw he's serious
>nfw
>>
>>51507817
Works in browser but not with cli.

Spooky.
>>
>>51507997
It's just a simple index.html page, so you can just download it (ctrl+s in your browser) and re-add it anyway. Content-addressing stronk.
>>
>>51508026
>Content-addressing stronk.
Yep, pretty cool stuff.

Still confused why some files work through browser but not cli though.
>>
>>51508109
I think it's because the cli tool uses ipfs dns to resolve names, and the daemon uses its own method. In that case, dnslinks that point to ipns addresses won't work in cli (but plain ipns or ipfs addresses should work). Someone should test it though.
>>
>>51508219
I can't even get some plain ipfs ones to work. Seems pretty random.

For example: QmWZtn3ahqqpGBBRZqPd thcWz2n1rxc1UuiDoWXrgrHKzZ and QmVbCbGE39kSpgwX33j3
s5XiidgFaY8SdE23VBqtbf8rgc both work in browser for me, but the first doesn't in cli.
>>
I haven't managed to pin shit. Even trying freezes down the daemon and slows my network to a crawl. Hours pass and nothing happens.
>>
I don't know why, and if this will even cause a problem, but I have not yet seen a file with an extension online. For example, I knew what extension my files were so there were not probs there. But getting an unknown file, how am I supposed to file the filetype corresponding with it?
(So if you're posting links in here, do you need to add the extension too?)

See >>51493021
>>
>>51508778
>still using windows
just use file nigga
>>
>>51508700
Dunno what happened, but I finished pinning the touhou animus successfully.

I had started angrily double clicking everything as if I was using windows. Damn frustrating.
>>
>>51508778
That's what you get for using an operating system that doesn't care about magic bytes, senpai. Look for a windows version of ffprobe, or rely on files wrapped inside a directory. You can also use ipfs ls and ipfs cat to obtain more information about the file.
>>
>>51509265
Pinning files means you're downloading them. If you're downloading large files, this can take a while.
>>
File: a.png (31 KB, 414x303) Image search: [Google]
a.png
31 KB, 414x303
>>51508778
real operating systems don't have this problem
>>
Of interest:
>>>/t/662698
Get dat shit on ipfs ASAP
>>
>>51508778
>>51509657
ps. but seriously, if you want/need to preserve the filename, pin the file in a folder (you can do this automatically with "ipfs add -w file"), this way you get an ipfs address which is folder with the file in it, with its original name
>>
File: head.jpg (149 KB, 1680x1740) Image search: [Google]
head.jpg
149 KB, 1680x1740
>>51492926
This is the most autistic idea I've come across on this board. Literally worse than the faggot who made a chandelier computer.
>>
>>51509743
>torrents are autistic
Top kek
>>
I remember that all YIFY movies combined in '1080p' are like 6TB

Someone could pretty easy create a popcorn time based on this technology.
>>
>>51509722
for example;
QmRX6sKftFAS4QMeMBpvN
EQbc4noNhj1Lj3i55xGMSjJ4r
(made up a name, not necessarily the original)
>>
>>51509755
>Torrents
>This piece of shit
Pick one
>>
>>51509779
sounds like someone doesn't understand ipfs and/or bittorrent

file transfers ipfs work very similarly to how a trackerless torrent works
>>
I finally got libgen_dbbackup-last-all-2015-11-21.rar downloaded, after many times the download broke. Now when I go to extract it I reach an unexpected end of archive error near the end. Rip that.
>>
I'm confused. So i'm having problems making my dir mutable. I have the hash for the dir I'm updating but it never gets updated unless I upload the entire directory again?? (with all the stuff Ive already uploaded)
>>
>>51510235
A folder is just a list of links. When you upload a new folder, none of the content that was already there gets uploaded, only the new content (due to recursion) and the list of links.
>>
>>51507708
Sorry for the late reply, but I get the same result. I can't ipfs resolve or ipfs dns the name either, but when I run
$ dig txt gindex.dynu.com
, I get
gindex.dynu.com.    2427    IN    TXT    "dnslink=/ipns/QmaGks9KKzu2WykHQjJFJkc UAN4ZoF7ok9h2hXj1WQn47U"
.

If I try to resolve that name, I do get the path /ipfs/QmRHRW4capXNBziDHZZAzQ VTMyzYeux2bHBFF5MRWvF2G8, but any attempt to access it (get, ls, pin, cat) stalls.
>>
>>51510235
You mean you add things to the actual OS directory that you originally added? IPFS stores its data in a separate location. You have to add the directory "again" to get the new hash.
>>
>>51510435
ls and get works from the web UI however. Not sure how to pin stuff from there, though, so I haven't tried it.
>>
>>51510610
ipfs dns only resolves ipfs dnslink entries, and this is an ipns dnslink entry as you can see. Not sure if that's the issue. Try downloading the file manually and then ipfs add it back.
>>
>>51510483
>>51510334

I just want to add new files to the existing ipns without uploading everything again

example https://ipfs.io/ipfs/QmfXH9XtP7xmoTH8WAp4HNSduqWMwLTH8B8TvbTkdgzNAa

surely if i wanted to upload a file to that directory i wouldn't have to upload all the files already in there?
>>
>>51510712
No, because ipfs is content-addressed. This means that if a file hasn't changed, then the hash will remain the same and you aren't uploading it again.

Directories are always just special files which point to other files on disk. In ipfs, they're special files which point to other files in ipfs.
>>
>>51510637
>ipfs dns only resolves ipfs dnslink entries, and this is an ipns dnslink entry
That's a long standing bug due to some inconsistencies between the different cli commands, IIRC.

>Try downloading the file manually and then ipfs add it back.
Yes, that would work, obviously. Strange that the cli tools don't work for this. Or that it seems to do for everyone but me and the other guy. What version is everyone running? I'm on 0.3.10-dev, built from git revision g93c2309 (if I read the Arch package version correctly).

Strangely enough, ipfs get works on my own domain, which uses the exact same kind of dnslink record pointing to an IPNS name. The IPFS path hasn't been visited on this computer (at least not since the last garbage-collect), so that shouldn't have to do with it.
>>
>>51510785
I'm using 0.3.10-dev, the git version isn't registered in the versionstring of my build. It's from gentoo using the overlay that pops up on google.
>>
>>51510749
so when i do ipfs add -r why is it uploading + generating a new hash for files already uploaded?
>>
>>51510712
>surely if i wanted to upload a file to that directory i wouldn't have to upload all the files already in there?
You don't. ipfs add doesn't actually upload anything. It just computes a hash or two and stores a record in your local database saying what the hashes points to (and stores the files themselves in the database to prevent accidental modifications). You only upload something when someone else downloads it.

>add new files to the existing ipns without uploading everything again
IPNS pointers are not updated automatically. You need to publish the new hash of your directory. Note that the hash always changes after you add a file to it (because it's no longer the same directory, content-wise).
>>
>>51510843
It's generating the hash but it's not uploading them. You have to generate the hash to know whether or not you have to upload it again.
>>
Nk simple db with links and commends
>>
>>51510843
>generating a new hash for files already uploaded
Not true. Example:
$ mkdir test
$ echo a > test/a
$ ipfs add -r test
added Qmbvkmk9LFsGneteXk3G7Y LqtLVME566ho6ibaQZZVHaC9 test/a
added QmP8P6d8yUwEPLXRG9ejjS mGvE7ktvjybgU8MEtaSKxAnS test
$ echo b > test/b
$ ipfs add -r test
added Qmbvkmk9LFsGneteXk3G7Y LqtLVME566ho6ibaQZZVHaC9 test/a
added QmR9pC5uCF3UExca8RSrCV L8eKv7nHMpATzbEQkAHpXmVM test/b
added QmXJZu5sfdRZZWE5jAUKBW yybLezLkqxfmKDgngTHF3aNK test


Note that the hash for test/a is the same in both commands. All that changes is the hash for test and test/b (which is the files and directories that were modified).
>>
File: txrx.png (23 KB, 858x182) Image search: [Google]
txrx.png
23 KB, 858x182
>>51510873
Yeah it does generate the same hash back actually, still dunno why it uses a lot of TX when calculating the hash
>>
>>51511033
Nodes are probably connecting to you. When you publish to your key, you have to push the result out to nearby nodes, but not when you generate the hashes.
>>
File: _Cover.jpg (420 KB, 1339x1024) Image search: [Google]
_Cover.jpg
420 KB, 1339x1024
Here's my contribution.

/ipfs/QmaPpuqeCeeq2LyazGLXw3moGr9C3LCzGoH1zHpYYmX5Cz
>>
Could yacy be adapted to work over ipfs?
>>
some tasks /g/ could undertake with this:

1) put pepe collection up (done)
2) libgen (is this even at all feasible right now?)
3) get a comprehensive collection of reaction pictures (there's already a good smug anime girl collection)
4) back up 4chan archives

What else can we do?
>>
>>51513058
>2) libgen (is this even at all feasible right now?)
Yes, ipfs is really fast already. The issue is downloading the data (the library is massive!) more than uploading it or having other users download it. Of course, that's assuming we can publicize it well enough that we get pins, or that we can keep hosting for a long time.
>4) back up 4chan archives
Someone seems to be working on that.
Actually I don't agree with this at all, 4chan was always meant to be ephemeral in nature. Back when the archive only saved content that was voted on, it made more sense - good threads that people want to remember are saved in much the same way one could just save an image, but in a more convenient format. But I digress.

>what else
>>51509700
>>51506107
>>
>>51511306
>shirow
>ow
>w
This rustles the jimmies
>>
So how anonymous is IPFS compared to lets say, TOR?
>>
>>51513058
>get a comprehensive collection of reaction pictures (there's already a good smug anime girl collection)
Is there a way we could thumbnail preview these?
>>
>>51514164
It's tor, not TOR. And ipfs is not anonymous, it provides a decentralized, distributed, encrypted means of data transfer compatible with other technology, including tor, i2p, anonet, etc. For anonymity, you want to run ipfs over one of these technologies. That is on purpose because this way they don't have to introduce a new, experimental, non-vetted, unaudited solution, and since the other networks are already large, a 51% or sybil attack is not probable.
>>
>>51514232
Got it
>>
>>51514225
Yes, download the image and generate a thumbnail, then upload both.
>>
>>51511306
>/ipfs/QmaPpuqeCeeq2LyazGLXw3moGr9C3LCzGoH1zHpYYmX5Cz

How do I get this? Single files I can do, but i don't get it otherwise.
>>
>>51515141
ipfs get the-link
You can also use -o "output-name"
>>
>>51515141
Do
ipfs get -r QmaPpuqeCeeq2LyazGLXw3moGr9C3LCzGoH1zHpYYmX5Cz
The r stands for recursive. It will download everything.
Thread replies: 183
Thread images: 5

banner
banner
[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vp / vr / w / wg / wsg / wsr / x / y] [Home]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
If a post contains personal/copyrighted/illegal content you can contact me at [email protected] with that post and thread number and it will be removed as soon as possible.
DMCA Content Takedown via dmca.com
All images are hosted on imgur.com, send takedown notices to them.
This is a 4chan archive - all of the content originated from them. If you need IP information for a Poster - you need to contact them. This website shows only archived content.