>>105938450 (OP)
Only if dumb scrapers, who don't even care about whether or not they can access your website, are your largest issue.
So if you are a leftoid who never experienced any struggle, it works.
But if you are anyone else, and you get targeted by random DDoS attacks or even just a single script kiddy who doesn't like you, it does not work.
Anusbis is an enbyware (retardware) piece of shit created by a troon from Canada, filled with filthy pajeets, and it's used by subretards.
Xhe will ACK soon anyway, so I hope this abomination will die with him, or there will be a drama that xhe is a furry pedo with disgusting fetishes and it will be canceled.
So ive been reading the docs and this shit shouldnt work? as far as I understand the PoW is to give it more time to do actual fingerprinting (like cloudflare) so all you have to do to bypass it is give your scrapper the ability to calculate sha 256 right? the pow only serves to speed up real browsers but that can be exploited no?
>>105941035 >all you have to do to bypass it is give your scrapper the ability to calculate sha 256 right?
Yes. And the other end isn't doing any fingerprinting, you burning CPU time hashing is the entire point. The idea is that it slows you down enough that you can't overwhelm the server. It's not to stop you, it's to keep you from being a complete retard and breaking things.
>>105941035 >>105941149 >>105941190
The majority of scrapers do not use javascript if they can't solve the hash challenge then they don't see web content, if they can't see the content then they can't scrape the page and walk the trie of hyperlinks.
>>105941305
It's braindead easy to write a chromium based scraper, almost easier than wrapping wget these days. Javascript doesn't do shit for bots unless they're written by 12 year olds.
>>105941475
Bots still have to go through Anubis, there is a tarpit challenge where bots cannot actually solve the challenge, they'll peg the CPU at 100 % and do not get to scrape anything. The best thing is that you can configure how, so you can also implement a custom way of detecting so you will trap the majority of bots via your config.
>>105941528
No one cares about bots scraping. It's on the web and the data is supposed to be public anyway. The problem is the server load from poorly written ones.
>>105941528
If you use a heuristic to randomly peg your user's machines at 100% CPU as long as they have your page open you deserve to have zero human users visiting your site. No one competent would deploy that on anything important.
>>105941592
You know what proof-of-work mean right? The challenge is about 3 to 7 seconds long for regular users. For bots trapped in the tarpit it's impossible to solve and thus their CPU will be at 100 % until it's manually stopped.
>>105941647
But what's stopping a bot from just solving the challenge like a normal browser? Then it can use the token to scrape the rest of the site. Alternatively, it could just not include "Mozilla" in the user agent, and then it goes through entirely unimpeded anyway.
>>105941956 >But what's stopping a bot from just solving the challenge like a normal browser?
The tarpit. You'll have a hidden page that regular users will never visit. Once the scraper has loaded that Anubis will come up again and destroy it's CPU.
Alternatively you can set up a load based check so that when the server load is high then Anubis fires up.
> it could just not include "Mozilla" in the user agent, and then it goes through entirely unimpeded anyway.
That would mean that the scraper knows about this and actively changes it, that takes manual effort. Scrapers just want to look like real users, that's why they use a regular user-agent.
>>105942243 >Once the scraper has loaded that Anubis will come up again and destroy it's CPU.
If you can already tell bots from humans then you wouldn't need human users to run the POW to begin with you illiterate retard.
