#dev 2025-08-21

2025-08-21 UTC
grufwub, Zen, ggarcia24, CRISPR, alephalpha0, GuestZero_ and geoffo joined the channel
#
[0x3b0b] My Matrix homeserver and microblogpub operate fairly comfortably within the Oracle Cloud free tier.
[KevinMarks]1, revi, angelo, just_jimmy, james1, james2, [benji], kleb601 and Maxpm joined the channel
#
[tantek] wow, haven't heard "Oracle Cloud" mentioned since Iron Man 3 https://www.oracle.com/us/ironman3/omag-mj13-ironman-1936895.pdf
[snarfed], [artlung] and Saige_Leah[d] joined the channel
#
doesnm can't register oracle vm even before 2022
_123nick[d], _secca_[d] and jonnybarnes joined the channel
#
capjamesg TIL about the output HTML element: https://developer.mozilla.org/en-US/docs/Web/HTML/Reference/Elements/output
#
capjamesg [edit] TIL about the output HTML element: https://developer.mozilla.org/en-US/docs/Web/HTML/Reference/Elements/output
benji joined the channel
#
doesnm someone already wrote about this iirc
#
capjamesg Has anyone done any thinking on what a crowd-sourced solution to bot identification would look like?
#
capjamesg My reasoning is that my site saw a massive influx in requests around the time the IndieWeb wiki did.
#
capjamesg It would have been cool if the wiki could have sent a flag to say that a bot is actively crawling and IPs x, y, and z are suspected.
#
capjamesg When I looked in my logs I saw that the bots were trying to crawl broken URLs thousands of times.
#
capjamesg The main concern I would have is that any system to identify bots would have to have a challenge mechanism for suspected bots so that actual traffic isn't blocked, and then it would be a cat-and-mouse game.
#
capjamesg My thinking is along the lines of Cloudflare has a bot solution but why can't there be an open one 😅
CRISPR, mahboubine, rjomara5853, Loqi and sarna joined the channel
#
[mattl] Yeah this is what http://1800www.com is doing
#
zachary.kai I worked out how to add extensions to my FreshRSS instance! Now I have custom date displays.
gRegor and CRISPR joined the channel
#
.pi.r2. honestly detecting bots is a bit of a cat and mouse game. I worked on both CF self-identification through signed request and escaping detection, there is a lot that loses value once it's understood by the escapists
#
.pi.r2. eg headless browsers tend to run at 30 fps, almost nobody else does
#
.pi.r2. capjamesg[d] CF bot detection is a black box that appears quite non deterministic from the outside, so you don't know what affects your reputation for sure and that makes it a lot harder to work around.
Blaps, Cloud and [Sophie_Young] joined the channel
#
[artlung] I used the `<samp>` tag today for my Tilde site. I learned (sorta already knew) no `<label>` tag in HTML 2.0, and `input` elements must be descendents of `<form>` tags (which I already knew from HTML 3.2 and 4.01). But HTML 2 can't keep me from using `:has()` in CSS! 🙂
fluffy1 and Cloud joined the channel
#
[tantek] am a fan of semantic samp and kbd
CRISPR, barnaby and CrimeWave joined the channel
#
[artlung] 😂 My first semantically correct usage, and probably my first usage ever. Though I've done experiments with "every tag" before and likely made one inadvertently.
marc_in_space, extremelymeow and [KevinMarks] joined the channel
#
[KevinMarks] this is an interesting project https://cdm.link/faircamp-intro/
#
[mattl] nice. I can’t find any examples of how it handles the “pay us and we’ll give you a special code” feature but for previewing new stuff this seems great
#
IWDiscord <a​2adam>
#
a2adam Hi there, I'm currently trying to get brid.gy to auto update posts from my website to bluesky.
#
a2adam I get this error on brid.gy:
#
a2adam ``Bluesky doesn't support type(s) h-card, or no content was found.``
#
IWDiscord <a​2adam>
#
IWDiscord <a​2adam>
#
a2adam I don't know how to implement an h-card onto my website (it uses Ghost), is there anything I can do to get this working?
cotsuka joined the channel