Yeah. If you pay for Amazon S3 Glacier or Gsuit unlimited. So thanks for your generosity.

Mega is no, but what is it about Mega you are looking for?

Is it possible to upload on mega.nz ? I know they got a 50 gb free cap, but maybe there are other hoster like it?

:)

Should be fixed now, thanks.

(issue solved, nevermind)

So it should be related to the issue I have.

Some entries that were available yesterday disappeared today.

You can compare https://animetosho.org/search?q=HorribleSubs 1080p and https://nyaa.si/user/HorribleSubs

I'm actually more confused over what you mean there.
Most entries only have one MAL link. For the few that have more than one, they look like "MAL (2)", where the underlines indicate the links.
I don't understand what you mean by "the link could be anywhere".

For your specific search example, you can just invert the sort order (oldest -> newest), though I suppose that isn't obvious, and won't work if there's 3 seasons and you want the middle one.
You make a good point there - I'll have a think about it.

Though unintuitive, searching on AniDB and copying the ID does also work.

Thanks for the explanation!

Security measures?
Erm, don't do anything silly I guess?

Thanks for pointing that out!
It looks like there's problems with building the search index at the moment, which I'm trying to fix.
I've rebuilt the index for now, so the above should be fixed. There may be more similar search issues in the coming few days.

@MAL links:
In my experience the behavior is inconsistent.
There was a time, where I thought, I simply had to click the highest number. That the link was moving away from "MAL", as new ones where added.
But over time, this wasn't always the case. The link could be anywhere. So I just started clicking them all.


@search:
- If you search for series A season 1. But season 2 is ongoing. There will be lots of releases for episodes of season 2. The latest release for season 1 might be (several pages deep) in "Older Entries". And other releases for season 1 might be buried even deeper. What you need, is to see only the releases for season 1. A direct search for that link, would be more convenient.

- The naming of releases is not uniform. One simply cannot expect to find all releases, using a narrow search pattern. "Season 1" vs "S1" vs "1st Season" ...  title in english vs. japanese.
Thus.. the search usually delivers more or less mixed results - different shows, different seasons. That makes it harder to find and compare releases. Plus you might still be missing releases, which do not fit your chosen search pattern. So again, what you actually want, is to look at the list of releases for the series in question.

It is hard for me to keep this short. The way I see it, opening the specific page/list for a show, is the most common reason for using the search. The most direct path to your goal.
So while the search works as it is. It would simply be more convenient/efficient to somehow have that link reliably served up on the first page.

But I agree the priority should be to keep things simple. Otherwise its value and likelihood of use diminishes.

Just go to hell, Eric ;)
We don't need your crap here.

Thanks for that feedback!

You're right. My experience is limited to small scale stuff hosted on heroku and aws. I think to start I'll just focus on the file processing part to start with. There are plenty of indexer websites out there already. So to start I need file sources (torrents I guess), a monthly estimate of bandwidth needed, a cost estimate for a server(s) to handle that bandwidth and a usenet provider unless I use one of my personal accounts. Anything missing? Advice on security measures?

[spam]

It doesn't seem to be possible to search for torrents from 20200411:

- the torrents exist: https://animetosho.org/view/horriblesu...kv.1449709
- they appear in the series page: https://animetosho.org/series/gal-kyouryuu.15072
- but you can't search for them: https://animetosho.org/search?q=horrib...ryuu&aids=
- and you can search for torrents from the day before and the day after: https://animetosho.org/search?q=horrib...080p&aids=

Apologies if someone's already mentioned this.

Thanks for the suggestion.

Archive creation here is done via stream manipulation, so resource-wise, it's basically free on AT's servers.
The main downsides are that users have to extract the contents, and impedes those that use the DDL services for streaming purposes.

As others have noted, it hasn't noticeably made an effect here. I think the problem is that, at the scale AT operates at, regular obfuscation techniques don't really work.
Unfortunately it's a problem that's likely to get worse with time, and I don't really know of any good solutions that I particularly like.

Having said that, with the new EU regulations coming into play at some point, it may make more sense to switch uploads to use archive processing in the future.

What URL are you using?
What does it look like for you / in what way is it odd?

Thanks for all the feedback!

One thing to note is that I don't really want AT being an anime information service - there's much better websites available for that (hence the links) and I don't intend to duplicate their functionality.

Having said that, I do think linking to related series does make sense, so I've added that in.
I don't quite understand your MAL link comment - isn't that always the case now?

I'm hoping the search is enough to let you find what you want without adding too many filters or complexity. Do you have an example where your suggestions would be beneficial?

Alternatively, you can see related anime from anidb, then use the id to open it on animetosho. For example, one piece anidb site is https://anidb.net/anime/69 . You can open it on animetosho by going to https://animetosho.org/series/69 .

Dead Links:

Over the years we have tried many things and I can't say that we found anything that noticeably extended the life of links for targeted shows.  Last season the only thing that worked more or less reliably was for users to plan for less than a 48 hour window to download their favorite shows from their favorite ddl-service.  I think it is unreasonable to expect links to last long anymore no matter what we do, if a show has been targeted.

It's also true that last season the targeted shows were quickly removed from certain ddl-services but remained available at other ddl-services.  I don't know if those other sites ignored take-down requests or were just not contacted.  I don't want to mention names because I've seen discussions in our Feedback that resulted in changes elsewhere.

I'll be collecting info for this season to see if anything has changed.  I already see they have expanded the ddl-services at AT which are targeted and/or compliant.  They will probably also look for quicker ways to enforce.  It also looks like dub has become more of prime target.

The general answer will probably continue to be for users to plan for downloads of their favorite shows as soon as they are available, and flexibility in which ddls and  "brands" of shows they use when necessary.

Sendspace has file names like

file_705325.7z

They still get taken down on targeted shows.

i wont be surprised if crapimation is behind all those links going down fast. quite a few crapimation licensed shows dont last even a week after being uploaded.

Is it just me or does the RSS feeds look kinda odd in qBtorrent? I don't think it can parse the names of any feeds from this site.

Reply to FAQ: Why so many dead links? (for a recent file)

You are probably aware of this, and it's not possible to implement (resources). Nevertheless I thought I'd mention it, just to make sure.

Over the years I have noticed .rar/.zip/.7z files often stay up longer. Especially when the filename is unrelated to the content. I may stumble into an old thread in a forum (several years old), all the links to .avi/.mp4/.mkv are dead. But that one link to a "22392873.rar" file is still working. Obviously not always, but I've seen it quite a few times. Likely depends on policies of the host.
If processing power was available to create archives (uncompressed) of each episode. An automated script could assign each show a specific number + number for release (+ number for file, if multiple). And upload it like that.

Recently uploads have been going down very fast. In the past you could still find working links for files uploaded 1-2 years ago. Now links from Jan./Feb. are already down.
It might be worth it, to do an experiment (manually?), just renaming files. Maybe that would already help a bit. Since files go down quickly, the results would be available fairly soon.

Once links are dead, it's over. So the hassle of having to rename files back after dl, is worth it (imo).
But of course, if someone is taking the files down by following their links from AT. Nothing will help, I guess?

it would be nice, if the header/info for a "series" would contain links to related "series" on AT.
this would help explore/search and get it all quicker.
also, it would be more convenient. if the main MAL link of the "series" in question, was always accessible via the letters "MAL". even when there are multiple links (which is great).

for the same purpose as above, it would be great (to have an option) to search for "series" instead of releases. it would deliver cleaner results. a more direct way to find the "series", and then the release you are looking for.

since some releases seem to end up unsorted. it would then be necessary to also offer a search within unsorted files only.

-------------------
otherwise I believe AT is pretty much as good as it can get, which quite frankly is pretty damn awesome. thank you.

Thanks.
Shows with many/complex seasons are difficult to deal with unfortunately. I think I've managed to fix most of those.

Thanks for reporting that.

ClickNUpload's uploads are done by posting <80MB chunks. Unfortunately, their upload handler doesn't give a reliable way of knowing whether a chunk was uploaded successfully (I've only got the HTTP response code to rely on).
To put it another way, the script doesn't have any way of easily knowing whether the file was fully received by ClickNUpload - it can only guess. This means that if there's a failure, the file could be missing a chunk or more. The example you gave follows this pattern.

ClickNUpload doesn't seem to have an API to query received sizes with, or a documented way of sending uploads.
This makes it hard to solve this problem here unfortunately. Usually the script tries to detect faults and retry, but there's no reliable way to do it with ClickNUpload.

Note that I consider AT to be in two main parts: file processing (downloads/uploads) and the website (what's shown to users). I run these on separate servers, and I'll assume you'll do the same; you can run everything on the same server (may be handy for testing/development), but I'd advise splitting up the website and processing components unless you want the complexity of managing QoS and the like.

I generally wouldn't recommend doing anything off a home connection. File processing may be feasible, but you've indicated that you couldn't do the volume with that. For testing/development, you could always just run with a smaller sample.
I strongly would discourage running a website off a home connection, though, like above, you could use it for testing/development purposes.

This means that you'll need to rent one or more servers. What you get exactly will depend on many factors, which you'll have to work out. Bandwidth will likely be a concern for mass uploading so you'll need to ensure that there's enough, though if you've only got one target (usenet), it won't matter as much. I/O (disk speed) might also be a concern, if you're like me and stick with cheap single disk servers.
For the website server, you'll probably need to experiment with what works for you, depending on how you manage/develop it. I generally wouldn't go with anything less than an SSD (referring to disk speed that is), though you could do fine with a HDD if you have sufficient RAM.

My opinion on tech stacks is that they really don't matter. Use whatever you're comfortable with, unless you actually want to learn something new (which is also fine, but it'll be more effort on your part).

I'm guessing you don't have much experience with running websites yourself? This definitely isn't the simplest thing to start off with, but if you do continue, there's a lot to learn from it.
One piece of advice I'd give you is to think beyond the technical side of things. Newbie programmers like to think that's the only thing that matters, but you should consider aspects like your goals, funding, whether or not you'll still be running it in 10 years time, etc.

I noticed that, and on AnonFiles the file sizes were larger than they were supposed to be. When I tried a force recheck on the ClickNUpload files with torrent to finish them off it would go to 23% or 77% and then drop to 0% after a few seconds and start over. Other torrents working fine so the file must be corrupted.

Episodes of the Pokemon TV series from SRS are identified as being releases of Mewtwo Strikes Back: Evolution

https://animetosho.org/search?q=pokemon srs

i think its more a case of failed uploads than CNU altering it. before there used to be like 3 uploads to CNU - one from animetosho, one from MultiUp and one from Jheberg so if even one of those uploads were broken, you still had two other CNU links to get from. now it's just one, if that is a failed upload then i go to DropAPK but it's so slow :( i have also seen some CNU files being way over the actual size.

It seems ClickNUpload alter the files in some cases. I hadn't encountered this problem before on other sites and when I upload files to iClickNUpload myself.
Also AnonFiles is OK. It just shows the size of files wrongly.

Recently, I have found that some uploads on ClickNUpload are broken because the files are smaller than they're supposed to be.

https://animetosho.org/view/horriblesu...kv.1448909
Here, the actual file size is supposed to be 579.8 MB but on ClickNUpload it's only 507.3 MB, which makes the file unusable.

Other examples:
https://animetosho.org/view/lostyears-...it.1448733
https://animetosho.org/view/golumpa-so...ma.d321182

Any idea why this happens?

Basically tech stack advice. Uploading to usenet isn't a technical problem but I couldn't do that volume from my home connection. Advice for that? Thanks to my day job I can do basic sites with python but I assume that's not optimal. Advice on what to learn?

I see, makes sense now. Oh well!

How disappointing.

That sounds exciting - hope it goes well!
What sort of advice were you looking for?

Thanks for the explanation.

Unfortunately, the naming scheme confuses the classifier and they don't even consider it to be One Piece, so One Pace releases will get mapped to random series, based on what appears to be in the title.
I basically throw my hands up in the air with those...

I'm thinking of doing a javtosho. Upload to usenet only. Maybe ddl also at some point. Would you be willing to advise admin?

[spam]

I think the whole Anime world is also flat because of the Corona virus.

These are all One Piece releases (fan re-edits to remove filler and make them closer to manga pacing), but they get tossed in random series EXCEPT One Piece. :D

This reminds me of Sega games logos.

Looks interesting!

Thanks for pointing that out.
Unfortunately, it looks to be the same issue with another YakuboEncodes' release, which is being rejected due to slashes in the name. It might be best if they could fix their torrents not to do that, as it appears a number of torrent clients don't handle it well.

Thanks for the feedback.
As mentioned above, if you want to supply a new design, I can look at replacing it.

???

Ossan to Kyouryuu? WTF? XD (おっさん, ossan: uncle; middle-aged man).

Japan is as weird as ever <3 :D

Can someone please get this one for me: https://nyaa.si/view/1230771 the site didn't upload it.

[spam]