#dev 2025-08-07
2025-08-07 UTC
grufwub, jak2k1, box4649 and trwnh joined the channel
# 
hughrawlinson I’ve been looking at EU based CDN providers, and wondering if anyone here has a recommendation? From what I can see they seem not to talk about firewall features in terms of preventing scraping by AI crawlers, which is a feature I would like.
Saige_Leah joined the channel
# bugsmith [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
# bugsmith [edit] [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
# bugsmith [edit] @huwrawlinson [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
# bugsmith [edit] @huwrawlinson [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
# bugsmith [edit] @huwrawlinson [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
# bugsmith [edit] @huwrawlinson [Bunny CDN](https://bunny.net) could be what you're after. They're based in Slovenia.
# bugsmith Their '[Shield](https://bunny.net/shield/)' product helps fight against bot scraping and DDoS.
doesnm joined the channel
# hughrawlinson Looks good! The main mention of AI is that they apparently use it to configure their WAFs, rather than “they prevent you from getting too much traffic from crawlers”. Maybe that’s covered under “malicious bots”.
# hughrawlinson I should sign up and give it a go anyway
# hughrawlinson $1/m minimum is great, much lower than alternatives
balintm, rubenwardy, barnaby, doesnm, ttybitnik and bugliker02 joined the channel
# 
[KevinMarks] XSLT << WhatWG proposal to remove XSLT from browsers
# 
Loqi ok, I added "WhatWG proposal to remove XSLT from browsers" to the "See Also" section of /XSLT https://indieweb.org/wiki/index.php?diff=103215&oldid=103194
# [KevinMarks] I did, oops
# [KevinMarks] Quite an interesting discussion. The author made a WASM polyfill of the XSLT code as a substitute
# 
[schmarty] getting some real "Chesterton's Fence" vibes from that discussions lol omg
# [schmarty] particularly this comment, haha https://github.com/whatwg/html/issues/11523#issuecomment-3155739198
# Loqi polyfill is a term used to describe providing capabilities to software, that it was not shipped with https://indieweb.org/polyfill
# [schmarty] "let's just/simply come up with a new way to handle RSS/ATOM feeds in the browser and have every feed publisher implement it"
# [schmarty] some voices of reason stepping up in the last day or so in that thread.
# [schmarty] ok, well, that bummed me out! time for coffee.
# perryflynn no, WASM. :-)
# perryflynn feels like good old tomcat application server again.
# Loqi Front End Study Hall #034 in 5 days https://events.indieweb.org/UzkXwuTEzyuG
# perryflynn only in browser
# [snarfed] answering those kinds of questions for this community was one big motivator for me for https://indiemap.org/
# [KevinMarks] or go back to having browsers render feeds natively, like they used to?
barnaby joined the channel
# 
[artlung] _I've always thought more telemetry in browsers on which features are used, how often, etc would be useful_ ~ [snarfed] there's some talk about that in this podcast from May about "unshipping" features. https://www.igalia.com/chats/unshipping
# [KevinMarks] I suppose it's because I've always seen XSLT as a "now you have two problems" kind of tech
# [KevinMarks] browsers rendering feeds solved the "I clicked this link and got a page of code" UX problem
sebsel joined the channel
# 
[aciccarello] Yeah, I've only used it on ATOM feed files so I'd be okay with that tradeoff
gRegor joined the channel
# [KevinMarks] Firefox does render JSON natively well now, so maybe everyone should switch to JSONFeed (I am joking)
barnaby, cobbland and [pfefferle] joined the channel
# Loqi Feedly is a feed reader service https://indieweb.org/Feedly
# 
[mattl] Feedly << https://www.jwz.org/blog/2025/08/feedly/
# Loqi ok, I added "https://www.jwz.org/blog/2025/08/feedly/" to the "See Also" section of /Feedly https://indieweb.org/wiki/index.php?diff=103217&oldid=101802
# [KevinMarks] well, I have some handy code from 20 years ago to turn JSON into HTML and vice versa https://microformats.org/wiki/xoxo-sample-code
# [KevinMarks] as embodied in unmung
# [KevinMarks] Maybe I should write a version in js
# [KevinMarks] a version in XSLT I leave as an exercise for someone else
ttybitnik joined the channel
# [social] So much of internal enterprise documentation is still in XML formats, but so are architectural structural mappings that are automatically generated are still in XML. Everything is rendered in the browsers with XSLT. So much would break far far beyond RSS feed styling.
# [social] I loathe XSLT, but lived through the SGML days of styling documents in WordPerfect in DOS (no need to invoke Loqi for definitions - other than I deeply miss WordPerfect to this day, but the more modern non-DOS version).
# [social] Many large enterprise have shifted to try and move their XML structured documents and run it through pandoc, but it is a process that isn’t instant and often needs a little hand holding and their XML and XSLT old school methods still function (never solid excuse, but it is one that can have weight).
Kolev joined the channel
# [social] The web browser is the swiss army knife of information readers / renderers. Not everything is a webpage that renders through it, which is a good thing.
bread and jak2k joined the channel; cobbland left the channel
# [KevinMarks] My broad principle is that deprecating things that make files legible should be a last resort. If browsers keep XSLT 1.0 support via them adopting a WASM version and not trying to maintain 20+ year old C code that seems like a decent compromise. I'm sure they can lazy load the XSLT parser.
# mathium05 WASM is a weird thing
# mathium05 It's neat
# mathium05 But it's best for compute heavy things
# [KevinMarks] It's a good way of encapsulating old code. For example:
# [KevinMarks] GitHub - feross/timidity: Play MIDI files in the browser w/ Web Audio, WebAssembly, and libtimidity https://share.google/wfGO6VKsFtM3Bvofw
# mathium05 Kinda, WASM has a lot of limitations
# mathium05 Like it can't dirrectly do nearly anything
rrix, balintm and rodrigo joined the channel
# mathium05 Kinda, nearly everything has to go through JS
jak2k1 and Virtual joined the channel
# [mattl] I added a chipotle page to my website. https://mat.tl/chipotle/
# Loqi Starbucks is a popular international coffee chain and also a named page (typically all lowercase /starbucks) on a few IndieWeb sites describing their preferred drink order(s) from the chain, and sometimes other opinions or aspects of their experiences there https://indieweb.org/Starbucks
duanin2 joined the channel
ben_ joined the channel
# [mattl] Okay, how does this look? A way to make a single, simple web page in your browser. No CLI, no terminal, no nodeJS… https://mattl.github.io/simple-html-file-generator/
lanodan joined the channel
# hughrawlinson I believe you have fixed this! Thanks!