[02:16:23] We ARE the colorpicker. [02:34:46] i wonder if its possible to vector-2022-ify fandomdesktop [02:34:56] it probably is but i'm not skilled enough [02:41:46] I think it's doable, but it'll take some time. [02:52:55] I made my own search engine and it's open source: https://github.com/Tali64/webscouter; I've been working on it for months at this point, and I'm proud to finally be publishing it and not having it languish on my hard drive for all eternity [02:54:28] 🎉 [11:09:26] what is its shtick [11:09:30] I assume its a metasearch [11:20:37] yes [11:20:47] someone did with previous fandomdesktop iteration [11:21:32] [11:37:17] https://www.theverge.com/tech/942302/google-search-ai-overviews-uk-cma-publisher-opt-out [11:38:05] > Google will also have to ensure that publisher content is “properly attributed, using clear links,” in AI‑generated search results [14:48:57] heck yah [15:33:04] We've been internallu chatting on an area adjacent about this and it's one of the exact areas today where AI output is not consistently license compliant [16:40:22] sources: divine intervention [16:47:03] I got scared that Miraheze went straight for a second [16:47:57] the homobeam will return in the next episode [16:50:22] Gonna revert to pride city for now until we have an updated gif loop. Unfortunate why discord doesn't loop by default but I get it [16:51:16] people's devices would be cosplaying db181 [17:10:54] It actually maintains its own index [19:59:16] that's a very tall order [19:59:20] I'll have to look closer [19:59:58] the closer look raised many more questions [20:01:32] I'd love to hear them [20:07:59] [1/6] crawler spec: emulate a js environment (expensive) or not (high chance of blockage especially under vpn ip and depending on behavior). If the latter and going for residential, that will burn up residential quite quick. protection from search history corroboration: defective if its an IP only you posses, adding a hop is an extra step of muted effectiveness. That's [20:07:59] [2/6] kind of the catch 22 with search, either you trust an entity to be your proxy at scale but it might corroborate, or you trust something on your own system at which point recipients have a trivial time corroborating. Hence for maximum benefit I recommend search scopes and distributing across multiple engines if search privacy including provider is a threat you co [20:07:59] [3/6] nsider. [20:08:00] [4/6] Then, competitive value; what is its niche relative to searxng, what I consider the most mature option in the field, or an option like whoogle, or even a local ai which are all kind of the incumbents for the subject and for which a personal conventional scraper is a tough sell imo. all are projects with privacy respecting aims which I feel would be most valuable t [20:08:00] [5/6] o add to, instead of a new wheel on the market unless it has this sell down pat. Options are good, but I have a quick little mental model of the value. 100 options, vs 10, vs 1. 1's too centralized, 100's too fragmented, so the sweet I would find to be 10 actively independent things (not 10 under 1 trench coat mind you like vpns), and each of those with their vari [20:08:00] [6/6] ous approaches get the contributions they need. That's kind of the goal and my want when using and supporting open projects. Variety is good but distinction is critical and for this one there are foundational questions where it fits as you see above. [20:18:20] [1/3] - The crawler uses Playwright with Chromium to fetch webpages and filter out non-content more easily, which necessitates a JS environment [20:18:20] [2/3] - Searches aren't stored except maybe in access logs, which are trivial to disable - I considered fetching results from DuckDuckGo to corroborate with the results from the stored index, but ended up cutting it out late in development to better focus on Webscouter's goal [20:18:20] [3/3] - Speaking of which, WebScouter's niche is that it's lightweight and stores its own index - SearXNG does have to send searches to Google after all, while WS doesn't [21:20:38] Hello guys [21:26:23] How to get verified wiki user [21:26:32] #verify channel [21:29:35] It says I'm already logged in [21:29:40] Which iam [21:30:01] Does this mean I can't verify [21:30:02] ? [21:31:05] Nvm I'm so dumb [21:34:44] homobeam 😭 [21:47:13] DO NOT stand in front of skye's homobeam, you will not be able to resist its powers [22:00:23] https://tenor.com/view/care-bear-stare-carebearstare-gif-19721794 [22:17:52] https://tenor.com/view/my-little-pony-mlp-rainbow-explode-gif-8957195 [22:19:49] Is this just ponyslaughter [22:20:20] Couldn’t resist the homosexuality [22:20:33] sounds like girlbossing to me [22:22:55] [1/3] The character (rarity) was supposed to be bisexual but Hasbro said no, so they just introduced a character who looked identical to her love interest but msle. It’s kinda funny. [22:22:55] [2/3] https://cdn.discordapp.com/attachments/615786602454581249/1512220131579793579/IMG_4113.png?ex=6a234c3f&is=6a21fabf&hm=d068b8d0f3dab90b7ce9a25f0f71f929c1b688a5bb416ef21b0480a11485057d& [22:22:56] [3/3] https://cdn.discordapp.com/attachments/615786602454581249/1512220131839971438/IMG_4112.jpg?ex=6a234c3f&is=6a21fabf&hm=c110423f3574c0de436cc10215e8bc0d8b2b65210489dfdc97746b573909c1bd& [22:23:01] wait what [22:23:06] humanoid mlp? 😭 [22:23:18] Equestria Girls, a spinoff [22:30:15] they look so wrong as humans [22:36:12] https://tenor.com/view/mlp-unimpressed-equestria-girls-maud-pie-my-little-pony-gif-10185030