In theory, the 1Fichier-encrypted links won't expire, as they are part of a separate premium account set up by admin to try for some longer term storage solutions.  But a similar setup didn't work with other services previously, so we'll see.

Check the blacklist option in your account settings and just put in some of your favorite groups.

Most releases do follow a similar format, but (ignoring outliers) there can be some complications at times.
An example: "Teekyuu - 50" gets labeled as "Teekyuu" whilst "Teekyu S5 - 50" is labeled as "Teekyuu (2015)", even though these should be the same. The script could be a bit more intelligent with detecting these cases, but they're ultimately quite hard to solve.

I suppose most people won't really be put off by this, so probably not really a concern.

Thanks for the reply.

That's an interesting idea actually.
I've never used Usenet unfortunately so don't know much about how it works.
I looked at some guides on how to post files and the process seems a little convoluted, which is a bit worrying.
There definitely won't be RAR support (might replace with .7z), so there probably won't be split files if that has any importance (is it important?). Will also need to see how much load the PAR2 generation has, as well as the posting process - am hoping that it doesn't thrash the disk too much.
From the looks of things, I should be able to just get a block account and get unlimited uploads.

Do you have any recommendations on how uploads should be posted?

Thanks for the suggestion!

can't you enable a filter for fansubs? I've already downloaded anime for color-blind people several times on mistake :/

They could post to the usenet, most servers have 2500+ days of retention.

AT doesn't retain any video files, if that's what you mean.  They're just processed and deleted.  The cost of storage would be huge.

one thing I'm not clear on: in your design how many clicks until download?  Do I have to tunnel down to get to download service choices?  In the current AT page un-split files are one click to the dl service.  One click is good, more than one is bad.

sorry that I can't tell by looking.

Hmm my idea was to use it for the Main page. I agree about the episode numbers being quite hard to get (especially for batches) so maybe it should be removed or not show for torrents with multiple files.

To filter the fansub group and the Anime name from the torrent name shouldnt be so hard for most of the major fansub groups so maybe the design could change and show a white / dark torrent box for torrents which do not fall into the major fansub groups which then shows the torrent name.

For the resolution it could be easier to just use ??? until the torrent is completed or not show the torrent at all until it is completly downloaded (or like you said just make it a seperate page).

I don't think it is a mirror, it is just a link from another source.

At does, or did.  For example:
https://animetosho.org/file/chihiro-_m...aac.229256

Usenet is third row down.

But not in listings since Nyaa database problem a few days ago?

animetosho should mirror to usenet also and post a .nzb because it will have a very good retention compared to free hosts and people could get the release easily.

the 1fichers encrypted files still works.

I believe one or more of the AT dl services used to offer streaming playback, but I don't see that option anymore.  So I was going to direct you to another anime site that specialized in streaming but they've dropped streaming as well.  Don't know why.

You're better off if you download, watch and delete.  Streaming playback has always been low bitrate regardless of user connection speed and used crappy software like flash that relies on hardsub conversions.

which of the downloading services allows streaming ? My internet can handle streaming and I don't have any more space.

There's some pretty interesting ideas there.
The design looks really good for mobile devices, something that AT's current design doesn't really try to accomodate. I presume that desktop users would just get borders on either side?

I don't really want to get too technical yet, but I thought I should point out a few things though:

- Splitting the group name and anime name is nice, but difficult to automate reliably (especially episode numbers), so something based on ugly torrent names may need to stay
    Perhaps this could work better as a separate page rather than the main page? i.e. a page that groups entries together rather than showing a raw dump of all the torrents available

- The resolution idea is pretty cool. Unfortunately the script can't determine the video resolution before a torrent is complete though. There's also some complications with batches containing multiple resolutions, but some heuristics might solve that
    Maybe split incomplete entries off elsewhere?

I'm not particularly good at evaluating designs so can't really comment much unfortunately.
Thanks for the sketch and your interest!

Not saying that HTML parsing is a proper solution (since it can break if the structure is ever changed), but just wanted to mention a tip:
The IDs directly match with AniDB's so perhaps you could use their API if it helps. The anime ID from a series page will be the number after the '.'. You can then use this number to query AniDB for a list of episodes and their episode ID (eid), then just plug it into the following URL structure:

https://animetosho.org/episode/[eid]

(you'll get sent a HTTP redirect to the full URL)

Just a quick sketch to how the design could change: http://i.imgur.com/LB5oBg2.png

Of course I know it is not perfect but the basic idea is that the option button on the top right expands to the bottom which shows the entry options (like 720p and so on) which turn a darker color when selected. Also if you click an entry it expands again to the bottom and reveals all the information which is currently loaded with a new page.

Just wondering what people think :P

There are still some things that are cumbersome from just parsing the HTML, like finding the page for the nth episode given the URL for a series.

You forgot http://chauthanh.info/anime/ .

The motherload is http://nibl.co.uk/bots.php

An index of 162 bots, at the moment.

You can always parse the HTML. I'm doing it to get RSS feeds from search results.

Honestly, I think the current design is perfect. Simple, minimalistic and lightweight, with no "fancy" visual effects.

If a rewrite of the site is happening, one thing I would like to see is tags. Not so much for anime genres, but for the release format - stuff like resolution, 10 bit, audio channels, soft/hardsub, reencode, etc, and then make it possible to filter out stuff you don't want.

If you need to find older releases without torrenting you have a couple of options:
-Anime-Sharing.com forums. Their links stay up a long time.
-XDCC. I reccomend using Hexchat as your client if you don't already have one. Just connect to a channel, and write "/msg [botname] xdcc send [filenumber]" Here's a few packlists:
http://xdcc.utw.me/
http://doki.co/releases/xdcc/
http://intel.haruhichan.com/
http://kyou.x1n.pl:8000/
http://vivid.moe/xdcc.htm

My guess is some database issue on their end. Their admins would have better answers, but my speculation is a database backup.
You may notice that #710169 is dated 2015-07-10, 16:28 UTC, whilst #710170 is dated 2015-07-11, 03:34 UTC (about 11 hours between consecutive IDs).

Data AT has: https://storage.animetosho.org/ul/nyaa...a-loss.png

The public download services we use generally delete the files after 60 days or so anyways.  To create a more permanent archive, we would need expensive private servers instead of free public ones.

We have tried name scrambling and file encryption at times to avoid immediate deletions, but I'm not sure we would say it made much of a difference.

Just sine 2 cents,
for the most part I think this site is good, however there's this problem with links dying especially with old releases.
May I propose that you hide links behind a link protector like ncrypt and maybe rename the files to something less obvious.

It looks like the database at Nyaa has gotten messed up, accidentally or otherwise.  I don't think it has affected our downloads.  For example, right now the last/top listing on Nyaa is [Mori] Wakaba Girl - 02 [60202D17].mkv, which we downloaded 9 hours ago.  And we have had other newer Nyaa downloads within the last hour.

If you find a Nyaa torrent that we didn't get, note it here, please.

What happen on Nyaa? Looks many recently uploaded torrent already deleted.

Nyaa seems lively at the moment.  Are you sure you connected the wires to the detonator correctly?

Oh, I see...  It's a database mess over there, isn't it?

Our last Nyaa-only entry was about 1 hour ago, which post-dates deanzel's note.  So it's hard to say yet if our Nyaa updates are going to be interrupted.

RIP Nyaa.

When are we going to get a new theme for summer season?

I have no intention of culling at this point, as I don't see any need.
Torrent entries take up very little space in the end; storing the actual attachments would take up significantly more space (about 35x more at current estimates).

Just to mention a potential issue on the re-torrenting for subs idea, does admin intend to retain a database entry for every torrent that has ever been downloaded here over the years, or would the db be culled at some point to, say the last 2 years or something?  For performance issues, cost, and such.  Is there any point in extracting subs for db entries that would be gone a few months later?

Thanks for the advice.
At this point, I'm not sure whether I want to get really involved in it, or basically just put up what's needed to integrate with the updates server and see what someone else comes up with. I haven't exactly thought about it too much at this point, and I'm not even sure there's a lot of interest, so trying to stay flexible for now.

AT conceptually has two parts: an updates script (pulls torrents, uploads to hosts etc) and the website (what you see here).
The website is a fairly standard LNMP stack (Linux, nginx, MariaDB, PHP) and Sphinx is used for search.
Regardless of what technology stack is chosen for a rewrite, the database will still need to be MySQL-ish, as updates are sent via MySQL replication. I'm otherwise not too fussed on the technology, although it should be efficient (deal well with low RAM environments with slow disk I/O), not be a pain to administer and probably something that's relatively well known.

I'm not sure what would count as a "mistake". The code isn't well structured, but I knew that when I started - it was something I compromised on to get something working.

Ehh because I had to reload the site I didnt replied to the anonymous this time :c

The website looks quite old i.e like websites in the 90s looked like but with colors. Of course this is better than a flashy website and a simple design is favourable.

The good thing about the Material Design is that it looks simplistic, is not going to be garbage when "flashy" is modern again and has great visual feedback.

Of course I don't mean that the design should change completly, but with the material design the file entries could be a little bigger and when the user click on it instead of opening a new URL, make it appear from the entry click with a smooth transition for example.

Once you've got a script running, you can just leave it running in the background, using spare available bandwidth (does depend on where you run the script though).
Of course, the script would not only have to retrieve the torrents, it would have to extract and package up the attachments.

To get a project like this started, you'll ultimately have to plan it out, inform people what needs to be done, and give them a place to collaborate on it. Creating a design document in an easy to find place, where people can contribute suggestions, would be a good place to start. Do you have any preference for code hosting? If not, Github is a good choice.

I'd also be curious to know more about how the current site works (what programming language is used, how the databases are set up, etc.), and what works well and what does not so we can learn from mistakes.

Nah, the site looks fine the way it is. The current simplistic design has and will stand the test of time, whereas going with whatever appearance is trendy right now will mean the site will look outdated in the future.

Ah, so the issue is taking the time to write a script to retorrent the uploads, not the bandwidth involved in doing so?

I believe a fresh Material Design would fresh up the page :)

Googles Material Design Lite should work fine for this site^^. Sadly I don't have time currently :(

I probably didn't make it clear that I was more interested in a script which would automatically run through a lot of torrents and extract from them (in the end, it would need to match up with AT's database so that I could host it).

Try the comment above for another archive, and I'd recommend uploading subtitles there if you wish to share them (as they accept user submissions).

Thanks for the thorough reply, and I do agree with all your examples (and intend to actually keep it that way).

I should probably expand on some of the things that I was somewhat interested in (although didn't really want to force things towards a strict direction):
- I think the design could be improved a bit. Note that this does NOT mean adding a lot of fancy graphics/banners etc - I want to keep it lean like it is now. I'm not really a designer, so can't really come up with any particularly good ideas, but stuff like the colour coding of items isn't exactly intuitive, or the dump of links when you click into a torrent (source links, article, series information, downloads etc) could probably be organised better
- an API would be nice, as a commenter mentioned
- perhaps better feed (RSS) management, better search capabilities etc
- perhaps make better use of the AniDB metadata that gets attached to torrents (like the episodes page, for example)
- I don't really like to make changes to the current codebase due to it being quite messy
- Performance issues: some parts put an unnecessarily high amount of load on the server.  I'm also somewhat stingy and try to run things on low spec servers to save costs, so performance improvements are always welcome (and there's plenty of scope to improve this)

1. I do want to keep the same feel of Nyaa/TT on the homepage - that is, fairly light and simple.
2. I'm not a fan of chatboxes either. But if it ever happens (quite a bit if), it'll probably be a completely separate page, not something that gets embedded in the sidebar or other distracting variant.
3. I have mused with the idea of a forum, but only as a replacement to this feedback page. I don't intend to actually manage a forum community, so it would be minimal at most. The forum just makes finding stuff a bit easier (it's pretty hard to find a particular post on this feedback page, for example). There are plenty of other anime related forums out there, and I don't see any point in joining in.
4. There's plenty of other streaming sites out there, no need to duplicate functionality. I have been considering compiling a list of external resources for each torrent (for example, other sites which host downloads/streams etc), which might come in handy for someone.
Loss of file hosts: ultimately can't do much about it as I've got no control, so ignorance is the only real option here. From experience though, there's usually something else that pops up.

Hope that addresses your concerns and thanks again for the feedback! :)

Yes there is.  It's pretty much all fansub:
http://bit.ly/1HjgFAW

(I hid it because some don't like the site to be mentioned publicly.)

If the website ever gets a rewrite, I would be interested in putting in an API, so that applications such as streaming clients and auto downloaders could be implemented. Just something to get metadata about series, episodes, and the links to them, nothing too complex.

That's too bad, because there really isn't another website that archives fansub scripts.

I might go through what I have downloaded and extract the attachments.

When a site gets over 4 million pageviews per month and yet gets so few user comments as AT does, it tells you something very important: The users are very satisfied with how things are at your site.  They come to AT, get their anime and go.  Any new features you add to this site might not be the same as progress.  For example:
1. The AT webpage is Spartan (lean, simple) and fast loading.  Adding slower loading fancy rotating banners or graphics = progress?  Maybe not.
2. If you added a homepage chatbox, people will use it.  But progress?  Check out some anime/fansub chatboxes.  
3. Adding forums: It takes a lot of dedicated people with nothing better to do, to keep a forum alive.  Progress?  Maybe it's better if people who want a forum to go to the fansub sites.
4. Adding streaming playback?  Not why I do or would come here.

Based on the user comments I've seen, I think the number one actual user concern would be: What if AT lost Solidfiles or Userfiles?  And with 75GB uploaded to each file service every day, and the long list of fast file services previously lost to AT, that is a worry.

For anyone who happens to see this: is there anyone out there who's willing to help with developing stuff for AT?
With limited spare time, I'm musing with the idea of trying to open this project up a bit more (to contributions), but firstly need to gauge the interest.

Unfortunately AT's codebase is somewhat messy (never really expected AT to stay around for so long, so never initially put much effort), has some odd requirements and, as a free project, there isn't really much in the way of reward (except for personal achievement perhaps), so I've been a little reluctant to reach out. But since we've been around for a while now, and I hope for that to continue, perhaps some of you out there may be interested in implementing stuff you wish AT had (for example, the comment above this one).

Main thing I'm interested in would be improvements to this website, or rather, a complete rewrite of it. I've wanted a rewrite for a while, but have balked at the amount of work required for it. I'd like the rewrite of the website to be open source, as it may help with getting more people involved (and satisfy those who've been pushing for it to be open), although it won't be particularly useful to anyone else without the updates script.
As for the updates script, I intend to continue working on it by myself for now, although I can be flexible.

So is there anyone interested? If so it would be great if you could mention what you'd be interested in helping out with (e.g. website designs, writing the code for a particular feature etc). Again, only trying to see what interest there is out there - there's absolutely no commitment and it may never eventuate.