| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul-jobs] 984141: Add prepare-repos role https://review.opendev.org/c/zuul/zuul-jobs/+/984141 | 00:53 | |
| @fungicide:matrix.org | wow, git.kernel.org has anubis cranked to difficulty 5 now, i just watched the browser on my netbook calculating their challenge response for a solid 15 seconds | 11:35 |
|---|---|---|
| @tafkamax:matrix.org | I was just thinking, what if the PoW logic in Anubis could be turned into mining bitcoin for example and the generated result goes to the owner of the page . So the linux kernel can fund itself for example :) | 11:53 |
| -@gerrit:opendev.org- Zuul merged on behalf of Michal Nasiadka: [opendev/system-config] 983610: Add mirror04.gra1.ovh https://review.opendev.org/c/opendev/system-config/+/983610 | 12:55 | |
| @mnasiadka:matrix.org | Taavi Ansper: I think there are companies providing „pay-per-crawl” service | 13:20 |
| @sean-k-mooney:matrix.org | there are a few like firecrawl | 13:20 |
| @sean-k-mooney:matrix.org | cloudflare had a more interesting approch they have an offerig where if they are your cdn they can serve your page in markdown automaticly to agents doing the converstion on there end | 13:22 |
| @mordred:waterwanders.com | anubis just did its job blocking my coding agent from reading the roles from the zuul-jobs repo via the web. :) (I told it to just clone the repo and keep a copy for local reference which is a better idea anyway) | 13:49 |
| @fungicide:matrix.org | mordred: but also if you told it to supply a user-agent that didn't claim to be a graphical browser, anubis would just pass the requests straight through | 13:52 |
| @mordred:waterwanders.com | heh. I figured teaching it to use git would be better anyway | 13:53 |
| @fungicide:matrix.org | afaik it only presents its challenge to actual browsers (or things trying to pretend they're browsers when they're not) | 13:53 |
| @fungicide:matrix.org | but yes, teaching it to use git when accessing git repositories is going to be more efficient for it and many orders of magnitude less load on us | 13:55 |
| -@gerrit:opendev.org- cid proposed: [openstack/project-config] 984159: Ends project gating for `virtualpdu`` https://review.opendev.org/c/openstack/project-config/+/984159 | 14:06 | |
| -@gerrit:opendev.org- cid proposed: [openstack/project-config] 984159: Ends project gating for `virtualpdu` https://review.opendev.org/c/openstack/project-config/+/984159 | 14:07 | |
| @clarkb:matrix.org | yup if the crawlers would all learn tto maintain a git cache I think they could have acopy of all the data and keep it up todate and no one would care. Zuul's own merger caches are a good example of this being the case | 15:25 |
| @mordred:waterwanders.com | It's almost like there are more protocols than pure http out there | 15:31 |
| @fungicide:matrix.org | probably the reason that doesn't happen (keeping a git cache) is that git caches are expensive compared to lossy llm datasets | 16:22 |
| @fungicide:matrix.org | from a longer-term storage perspective | 16:23 |
| @mordred:waterwanders.com | yah - and harder to throw into an object-storage based data lake | 20:44 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!