There's no stated rate limit. I request that clients be "reasonable" with the number and rate of requests sent.
For "reasonable", consider the page browsing activity of an average user.

Hopefully that helps.

(I don't manage or check the Discord server, so you probably won't get responses for these sorts of questions there)

I opened a ticket on discord for this as well, but that didn't receive a response so I'm trying here too :D
What's the current rate limiting with regards to API queries? I'd like to implement the limit client-side to avoid requests being blocked.

Thanks in advance!

This site is new, I suspect the thread starter is the owner. Because only rent a domain for 1 year. https://who.is/whois/vikingfile.com

I see, thanks. I realize it's not that hard to code, although I suppose it's somewhat more work than just slapping some remote embed on your page, and can get more involved when you also need to manage users or implement moderation (depending on how much spam they get). I think many do just slap together some wordpress comments or something of that sort, but then you also need to keep watching for vulnerabilities in those large packages (which happen quite often), so it's still not that hands off either.

Unfortunately the code isn't open source, and it's not really its own component either, meaning that others wouldn't be able to utilise it even if they had the code.
The system here is really quite simple and bare bones, so I can't imagine it to be difficult to implement by someone knowledgeable enough.

I suspect the appeal of Disqus is that it does everything with minimal setup, from user registration/management to comment storage and database maintenance, without the website admins needing to care about any of that. So depending on how much effort the admins want to put in, a self managed open source comment system may not be what they're after.

Looks like that's a 'no'.

So I've reluctantly added an exception for VARYG to favour matching via name over the info they supply. This should improve classification accuracy.

Thanks for the suggestion. Is there any particular reason for the recommendation?

Thanks for your concerns. Sorry if I implied that your searches are having a significant performance impact - it's not really what I meant.
Terms like 'a*' showed up as queries taking the most time, indicating that it potentially could be something I'd want to restrict in the future.
Whilst the searches are relatively expensive, there isn't too many of them.
(also part of the problem is the search server not using the correct index)

When positive terms are specified, they must be present in the results.
So if your 2160p entries don't contain "1080p", they won't show up.

For negative terms, they must be absent in the results. So if !(*720*) is specified, the following examples will not appear in the results:
- [Group] Title - 05 (Web 1080p) [A1E720F8].mkv
- Title.S01E720.1080p-Group.mkv

You can just specify all those cases, e.g. (*x1080|*x1080p|1080p).

admin, is this minimal no-javascript comment system open source or available in some form?
Asking because recently some anime releasing sites (and not only) recently lost their cloud comment sections (banned for piracy by disqus), and I'm sure some of them are in need of something simple, easy, fast and secure. I'm aware of many implementations of such things, but they're always really large and consume a lot of resources and tend to get some 0day every few months.

after a lot of sanity loss of hours later, i settled on:
q=!("DTS"|"TrueHD"|"[EMBER]")!(*720*|*540*|*480*)

sorry for the.... flood of users spaming expanded queries, i didnt realise it had such a perf impact

yeah, I'm fairly sure something is wrong
?aid=18576&qx=1&q=("1080p")|(NOT("720p"|"480p"|"360p"))
does not return any 2160p entries, however
?aid=18576&qx=1&q=!("720p"|"480p"|"360p")
returns 2160p entries, this is woeful

sorry for the spam, but i'm actually at my wits end trying to get this working, while making sure i dont tank your servers with my users, i didnt expect my app to actually impact you this much :/

hm, actually this isnt correct, I think I'd be looking for ?eid=293094&qx=1&q="1080p"|-("720p"|"540p"|"480p") but i cannot get this to work correctly....

There are requirements for that, read them first https://animetosho.org/about

Now, now. You can at least ask nicely. 😉

Add VikingFile as a new DDL provider - vikingfile.com

I'm considering

?eid=293094&qx=1&q=!("DTS"|"TrueHD"|"[EMBER]"|"*720*"|"*540*"|"*480*")"*1080*"

tell me if you'd prefer if that wasn't the query used, i kinda want to use it because some shows are tagged as 1920x1080 or even 1920x1080p which is awful, so *1080* does x1080 x1080p and 1080p

great, my queries now look like this:

?eid=293094&qx=1&q=!("DTS"|"TrueHD"|"[EMBER]"|"720p"|"540p"|"480p")"1080p"

is this still too specific? too laggy for you, should i somehow simplify it??

i tried this and left out the brackets, i'll redo my rss feed with the brackets

just was wondering if there was a way to get results for specific torrent/file quality, i.e search for only 1080p torrents, etc. thanks for clarifying admin.

Unfortunately I don't have any suggestions - it's likely not what you want to hear, but pure exclusion filtering is just hard.
If you don't expect to exclude much, filtering client side might be a solution.

You should now be able to use negative-only terms in searches when an aid/eid is supplied. Hopefully that works out for you.

> and I'm not particularly fond of attempts to bypass this

what do you suggest then, what would be the best way to exclude a keyword

negative only with aid or eid supplied would be absolute peak, and would likely help perf on ur end in the long run too

If you're using Sphinx syntax, try replacing "1080" with "(1080p|*x1080)".
Automatic keyword expansion has been disabled to conserve CPU - you'll need to explicitly specify the terms you want expanded now.

Style "2021q3 - Rio" works for me. Although this style is 2nd oldest now, so it may be removed soon. Can this style be preserved, or can someone make a version compatible with Stylus browser extension?

Sphinx search behavior changed. Now search "1080 (hevc|265|av1)" doesn't work properly. Is there a way to get back to old behavior so that search "1080" returns the same results with and without Sphinx enabled?

Any chance for a dark mode

Negative-only searches aren't allowed because negative indexes aren't a thing, and I'm not particularly fond of attempts to bypass this. If I add expansion limits, terms like 'a*' probably won't work the way you expect them to any more.
Though I can put in an exception to enable negative-only searches when an aid or eid is supplied. In these cases, there's likely few enough entries such that the performance hit isn't too bad.

If you search using Sphinx syntax, the recent change means that keywords aren't expanded automatically anymore. Previously "1080" could get expanded to "1080p", "1920x1080" etc, but now you'll need to manually specify expansions that you want.

yeah,  'a*' is my users, simply because there's not a way to search with exclusions only so if i want to do eid=1231231&qx="!AC3" i need to add  'a*' because you cant have only exclusions, if you manage to change/fix that I'll remove it from all the queries

I also noticed filtering by "1080" just breaks the feed, you need to do "1080p" now, which is weirdge...

Searching this way with uploader name in brackets always works for me 100% of the time. [Yameii] 1080p or [ToonsHub]  [1080p] or [SubsPlease]  (1080p). Not using brackets for the uploader has given me mixed result (sometimes works sometimes doesn't). Putting the quality (2160p, 1080p, 720p, 480p) in brackets or parentheses or not at all hasn't made any difference for me (they all work).

I can't evaluate what you're referring to, but there are 'flood control' mechanisms in place (though I'd consider the limits fairly generous).

I'm adjusting how the search works; I've made keyword expansion less aggressive, which seems to have lowered CPU.
I might put limits on terms like 'a*' as the expansion of such is particularly problematic.

In terms of client behaviour, I've noticed a lot of non-anime related searches being made. I suspect this is due to automation software sending search terms to all listed indexers, so AT gets a bunch of theoretically unnecessary traffic that way.
There's also the fact that much of that software is married to TVDB and the like, so can't perform AniDB ID based querying and has to resort to title searches, sometimes once per episode.

I'm not entirely sure what the question is, but there's no explicit "quality filter" here. However, since most titles include the video resolution, you can of course just throw in "1080" and it might find what you're looking for.
If you do a search, then use the feed icon (top-right on listing pages), it'll give you the feed for those search results.

hmmm, i wonder if i'm at fault here, as my users might be.... "flooding" the feed with repeat requests of the same

not sure if this was asked but is there a way to filter search by quality? i.e yameii 1080p, subsplease 1080p. i need to adjust my rss feed?

@admin is that possible?

admin will probably say AT is a mirror site and you should take it up with the torrent host.

We want admin to stop Animotvslash promoting his site here. You see there are 3 same videos with the same video size, the only difference isfile size.

TokyoTosho's submit page says users can submit up to 10 torrents per hour, so three would presumably be within their acceptable range.

Ah here https://animetosho.org/series/gekijoub...-hen.16297

Is this spam? 3 videos uploaded, he can't upload to nyaa, because it violates the rules. I hope "tokyotosho" will also be strict with groups like this.

I'm surprised it's running on a shared instance

Thanks for the suggestion.
Generally hosts don't want one client consuming too much CPU on a shared server, as resources are oversold and would limit CPU for others.
Hard limiting CPU usage does mean one can't take advantage of burst availability, so I'm more inclined to look at lowering average usage at first. But it could come in handy some day.

Feedback Page

This is a somewhat temporary page made especially for the beta version of this website. It's to collect all your death threats, love messages, requests to the code monkeys or just suggestions/issues/questions/criticisms/idle talk or whatever feedback you think is worth our database space.

is this feedback? (rhetorical question)

there are no ads here, so adblocker disabled or enable makes no difference, wish more pages would be like this ;)

wtf, is that a restriction of the provider??, i mean normally if cpu peaks at 100% its get hot, fans go to full rpm and its gets laud and basically everything is fine as long as you got enough power and cool air ;)

try running the process that uses to much cycles with "cpulimit" ( https://limitcpu.sourceforge.net/ https://manpages.debian.org/unstable/c....1.en.html )  to limit its cycles accordingly (its a bit like nice only it dose not prioritize others above but limits the cycles to a specific value (like only use 20% ) )...

Thanks, should be fixed now.

Host has been shutting it down due to high CPU usage. Might need to tighten usage of the feed somewhat.

That's the issue we mentioned two post before, the storage server seems to be down.

I'm getting a "pages never finish loading' issue - Tried different browser, removed ad blockers. Still same. Main page is fine, issue seems to be any page with a photo (Series page, episodes page etc) and photos will not load, at all. Might be my end but I haven't changed anything on my setup, and first time seeing the issue and other websites working ok.

Yup, can't download NZB's either.

I have the same issue, the storage server is timing out.

2admin: check please, screenshots and attachments are not available
Requesting https://storage.animetosho.org ..
Error: Connection timed out (Error #10060)

On the "Show: Downloads" listbox above the listing, you can change it to show attachments.