@beatricejess @Natasha_Jay I have to use yt-dlp to have a listen because YouTube in its greatness decided to automatically synchronize it so the narration is in German, for me. 🤨

Well, somebody is werkin werking werking … nameline the feckin scraper bots! Remember, to benefit from AI is to benefit from this kind of breaking the Internet first before asking. If you use these large language models, I hope your skills atrophy, your job value goes down, your careers ends as you are replaced by the tool you use, and before you quit, you are trapped in a downward spiral where the AI bros leach all the value out of you as you pay and pay in order to keep up in the AI-bro race, and when you're poor and broke and at the end of the line, you'll find you're still an AI-bro, still a taker, not a giver, still a looter, not a lover, and I hope your family has left you and your kids have disowned you because you made so many mistakes. So many mistakes.

I really have to sharpen this knive of the AI-bro curse.

In any case, the scraper bots are back and as you can see in the graph, as soon as the gate opens, they break in and wreck the place.

My second line of defence is starting to pick up the repeated offenders, the Internet service providers lending compute and bandwidth to the scraper bots and the mobile service providers with the most gullible phone owners who installed the malware participating in this pillaging of the commons.

The autonomous systems in the one-week jail and their IP address ranges are available here, for your perusal:
https://alexschroeder.ch/share/

#ButlerianJihad

This added layer of defense seems to be doing OK, not inconveniencing regular users too much. When the 1 min load average climbs over 10, the website requires basic auth for reading (!) until the 5 min load average drops below 2.
#ButlerianJihad

I guess the line did not hold. I have implemented yet another layer of defence against the scraper bot army: if system load climbs over 10, the major wikis like Emacs Wiki and Campaign Wiki start asking for a username and password. These are shown in the 401 error page, so if you read it, you’ll find it. And with that, my sites take another step into the Darknet, invisible to the exploitative forces ruining the open web – and running the corporate web.

I blogged about it.

http://alexschroeder.ch/view/2026-01-30-lock-the-gate

I'm looking at the scraper armies testing the field. The line is holding, for now. #ButlerianJihad