Nyaa/Sukebei Recovery Thread

This thread is for the discussion of the possible replacement/mirror of the late Nyaa.se (which also hosted Sukebei); taken down for reasons still under speculation.

Previous Thread
Recap of last thread:
>OP looking for tools to create a new site
>An AMAZING user creates a backup site using some old (17 months old) databases of nyaa and sukebei.
>Another user was archiving nyaa and sukebei database, but it is still 7 months behind due to being too confident and stopped making backups.
>Discussion back and forth of what to use to host the new site.

Other urls found in this thread:

strawpoll.com/6eabdyg
sukeibei.pantsu.cat/search/2?q=fakku
nyaa.pantsu.cat
sukeibei.pantsu.cat
srotonpaga.appspot.com/nyaa/
void.cat/d0ab28bc2c406aa88f105c4836f04c97e08a01a9&v
void.cat/7be796c2bcc49b3e816c463a2887d63a38b92581&v
torrent2magnet.com/
0x0.st/3Fs.7z
0x0.st/3h5.torrent
github.com/odangomoe/nyaascrape/
sukeibei.pantsu.cat/
twitter.com/NSFWRedditImage

Assuming nyaa comes back up one day


How do i go about setting up a daily backup?
I have tons of raspberry pis and know some Linux (bash, cron) but don't know web stuff

Poll: strawpoll.com/6eabdyg
Use whatever proxies you got. Hopefully it will all even out.

bumpObump

Would hosting on .onion be feasible if it's just a place to get magnets? Tracker is dead but DHT still works (well enough).

Both nyaa pantsu and sukebei pantsu are down right now. What's going on?

Maybe someone can help me out here. I'm trying to seed what I've got and at first DHT works, however after about half a day, my DHT nodes drop to 0. Restart uTorrent (2.2.1) and the DHT nodes come back. Then they eventually all drop off again. How do I keep my DHT nodes alive? I'm using a Socks5 proxy.

Pretty sure deluge RSS plugin doesn't work over tor

migration done on nyaa.pantsu.cat and sukeibei.pantsu.cat
will correct url issue soon

Nazis from Brussels, user.

So we scrap RSS functionality. Keep it bare bones but at least up-to-date. (unlike TT)

I think the owner is fixing the typo he had on sukebei.

Dht will never die, read how it works. As long as somebody is seeding and you know the hash, you can get it.

I wrote a python dht crawler but its really fucking slow - i get thousands of unique hashes a day, very few actually lead to a working torrent.

The first thing is we need someone with a backup that's not so old the US had a different president when it was made.

Not sure if nyaa had an API, so you'll probably have to scrape the HTML for the metadata you want to store yourselves. Shouldn't be too hard with something like BeautifulSoup, I've used that to great success for a similar purpose. You could try monitoring new torrents through their RSS.

They're both up right now and the typo is indeed fixed so that could've been it.

Also, fuck whoever decided to set several of the torrents to disable DHT. Why would you do that. Now they are just completely dead.

You only need BeautifulSoup if the HTML is messed up. lxml is faster and easier to use.

This.

I know how DHT works.

Fuck. Well. What about original owner? Prepared to share?

not use animetosho you morons.

Right now we can't even get ahold of him. Who knows if we'll ever?

>tfw nyaa was run by Julian Assange

RIP

I'm pretty sure I tried using that at first, but for some reason I got tons of compiler errors when trying to install it through pip.

It's one season back. That's not awful considering almost everyone thought it was untouchable.

The real dream is that the site owner will show up with the full db.

Beautifulsoup can use lxml as the back end for speed and dunno but the bs4 api is well, fucking beautiful

How many nodes are you spawning to crawl the DHT?

Oh shit is the second kristallnacht.

So i guess it's best to wait a while for him to make an appearance.

You need specific tooling to compile and run lxml, its a PIA. a lot easier to just get some nix os, works out of the box

On Windows there's some magery that you have to go through involving VisualStudio to get things to compile. There's a website that has 64 bit MSI installers that are pre-compiled.

No, I think we'll call it M/a/y Day.

Nyaa hasn't even made a public statement himself about why the site went down, I think giving the db is above him.

All you fucking newfags don't even know about tt/at.

Just a few. Thats probably the issue. Also, is there an irc for this shit, id join tommorow, its fucking late.

ITT: butthurt weebs scramble to keep their animated pedophilia piracy alive

I have some of the newer uploads, but no databases backed.

More importantly, what will the logo be?
Will J-List still sponsor?

I got TT, but that still doesn't solve the problem. It just meant (((their))) work isn't done yet.

Really? I had issues with BeautifulSoup even when using lxml as the backend as opposed to using lxml directly. This was last year when I tried to collect a list of links and it never returned the correct amount. I then did the same thing with lxml and got the proper amount of links.

How does it not solve the problem?

I'll do the logo

Is the search function on sukebei pantsu broken for anyone else?

Should be working now if not try clearing cache
sukeibei.pantsu.cat/search/2?q=fakku

Because the problem we have is the same one I've been saying we have had for years: a few single points of failure. We need to create something robust, before the copyright hounds get more maulings in and the bleeding increases.

Dunno, worked fine for me. Used it half a year ago. These days i prefer scrapy, it hides a lot of magic from you, you pretty much only have to write a parser and some rules what to get.

That's already been done. Nyaa existed, and tokyotosho and animetosho are both mirrors of it.

All you're doing is creating yet another point of failure if you're trying to make it anything more than a mirror.

Weird. It works for some phrases, but for some it doesn't. Is the issue here on my side on or the site's side?

Backup site:
nyaa.pantsu.cat
sukeibei.pantsu.cat (yep, it's misspelled)

Well I'll be releasing source of nyaa.pantsu.cat later today and some user was planning on using ipfs for it.
So we could use ipfs.
Probably still buggy. What phrases don't work?

Only if you're talking about a plain HTTP website. These protocols under discussion cannot be taken out by DMCAs or operators pussying out.

you both gotta stop misspelling sukebe

>Probably still buggy. What phrases don't work?
Tried some random fragments of hentai and eroge titles. "Boku no Pico" worked, "Discode" worked but didn't search for the right thing (I was searching for "Discode: Abnormal Sexuality"), "Life with a slave" returned no results.

Sure they can. All that matters is the part where people can actually find them. You take out that, and whatever you set up is worthless as anything more than that super secret thing you and someone else have written down in a notepad.

All you're doing is no more valuable than the magnet links that already exist for nyaa stuff.

oh it's intentional sorry im retarded

2016-02-06 (SQLite), 2016-02-20 (SQLite), 2016-09-19 (JSON) dumps:
srotonpaga.appspot.com/nyaa/

2016-10-11 dump (SQLite):
void.cat/d0ab28bc2c406aa88f105c4836f04c97e08a01a9&v
void.cat/7be796c2bcc49b3e816c463a2887d63a38b92581&v

It's a minor issue but is it only me that page begin with page 3 or 4.

Before we move that to a new infrastructure, we also need to rebuild the latest season. Everyone who still has .torrent files from then, run them through torrent2magnet.com/ and start collecting the magnet links.

Anonymous P2P clients can use a meshed network to find things. Bootstrapping is the only issue, and Sup Forums can probably serve that role, if so needed.

>and some user was planning on using ipfs for it.
That was me but since your site uses server-side programming so I'll do my own thing since IPFS can only serve static content.
But first I'll do something else ( ). Not now though, I've been up since nyaa was shut down.

To add to that: searching for anything on sukebei pantsu only returns one result when it works, usually the wrong one.

Yeah people love still using mule.

All the other super awesome Sup Forums projects are totally still run. haha remember the times Sup Forums tried to just run mirrors of Sup Forums. Those are still going, right?

All I really need are file name and hashes.
Odd I'll look into it for you.
yeh I know I'm just waiting to gen a new ssl cert.

>So we scrap RSS functionality.
At that point, I'd rather use XDCC.

Be sure to read , too. It seems to be a serious bug with the searching.

That works too, I guess.
One step at a time.

Anyway, I got to go. Might be back later, dunno. Keep working though. Ganbatte!

2016-09-19 dump merged into a single valid JSON:
0x0.st/3Fs.7z
0x0.st/3h5.torrent

Spam filter thinks that magnet link is spam for some reason.

I beg you gentoomen, save us, save anything you can. Please. Don't let it end like this.

Nothing's ended. The mirrors still exist.

How big?
I'm on phone for a bit

Whoever posted this in the last thread - how recent is your backup?

142MB

...

Cool, will seed soon™

Reminder to host the new site in the middle of Antarctica

Nope, me as well.

It's not just Nyaa that get this problem either. So many torrent uploaders trip into this basic mistake. One tracker + disabled DHT = single-point failure for your torrent.

>[RAW] Attack on Titan 08
Super old, most likely. Like, at least a year plus some. I'm just guessing off the the actual release date of that episode though.

What about using the gopher protocol, or something else suitably ancient and dis-used that no one knows about?

github.com/odangomoe/nyaascrape/

WHAT THE FUCK? Why did nyaa have to go down of all things?

gopher is fast to set, like put all the links in a folder and start the server, that fast

I still don't know why hasn't this been done in the meantime

That's what I mean. And all the normie (chrome, IE, etc.) plebs will have no way to access it.

>tfw used sukebei shortly before it went offline

Short version, the jews.

Thanks!

Somebody wants a dump from 02 April 2017?

yes please

I've got a big dump from a few minutes ago I can give you, if you want that one too.

Yes link it familia

Please do.
We need any and all nyaa data from the past 8 months.

I will suck your dick.

I need to convert 20gb of torrents into magnets before, but it shouldn't take too much on a meme drive

I was literally shitposting, user. I don't have anything for you.

As well.

Either you're a troll or an absolute madman.

that's mean and rude :(

So where can I just download a big archive of torrents? I have some shit I stopped seeding because autism, maybe I can start it back up again.

please have both nyaa and sukebei

Just read this thread, the links are here.

sukeibei.pantsu.cat/ is the right link, but the on-site link is to sukebei.pantsu.cat and therefore doesn't work.

I started seeding the Json .7z

how do I use it for myself?
just open it in firefox as a page?

Lmao, this spelling confusion is getting out of hand.

>nyaa se is gone
im fucking terrified for the future of the internet