π“Ÿ host1.bitchan.one

𓀐 @d+Qyfj*3

⏱ 6h seenβ›“ 0πŸ‘ 0β§– 1

filter [-]

languages:onother

>supports no-js environments Doesn't make much of a difference. Generating an HTML, rather than returning a JSON, just means to concentrate some strings. And scrapers do execute JS, so a modern SPA would have the same issue. That 50ms query looks horrid. Whoever put this together already capitulated and accepted his fate of a full table scan. Why LIKE pat? Is it just adding wildcards to this? If i search for the tag nigger, will it then do LIKE '%nigger%'? Or even worse, is it doing LIKE nigger and commits to a full table scan to something that could be a covering index? This is what you have to look for. That is your issue. Choose your fights. If you can make your query ten times faster, do that >some user input could launch super complex queries But not the LLM scraper, that one is just clicking on buttons. >Depends on what you mean by application level. Just banning them in your application would not really solve an issue of flood like this. Firewall is what you need. what you mean is a basic bitch fail2ban⏱ 41m↩