aevai want to make a little art portfolio site, and I'm stuck between wanting to fuck with scrapists and wanting to not fuck with screen readers. I'm considering chopping up images into tiles which would be assembled into their correct relative locations via css or something. i could even put alt text on it with a transparent image layered on top spanning the full area. downside is real people wouldn't be able to download the images though :/
of course that's all assuming they don't just dump the frame buffer somehow. same problem as anything else, it's unclear to what extent anti scraping measures actually work
another idea i had was use webgl instead of images and have the images gradually assemble, so you'd have to wait a short amount of time to see the full thing. that would allow downloading but scraping would need to be more elaborate
the worst idea I've had so far which is also my favorite is make it a 3D game like those old edutainment cdrom games where you walk and in a museum etc. idk how i'd make it screen reader compatible though, but i guess it's a page of visual art anyway so idk
wait wait wait what if i make my website a fmv adventure game lol https://mastodon.gamedev.place/@ramon_wilhelm/116290525951324852
Ramón Wilhelm (@[email protected])
Attached: 1 image Who remembers the classic adventure game The Lost Files of Sherlock Holmes - The Case of the Rose Tattoo (1996)? 🔍 #adventure #game #sherlockholmes
Celes 🌙
Ratsnake Games 🔞
Janek @ IndieDev.site
Erin
Evie 🏳️⚧️
Ronflaix
Clyde (GameTank Guy)@aeva yes you should make a Sonic'sSchoolhouselike
anco@aeva
Someone did something like this!
https://jclahoot.com
(I don’t know how accesible it is though)
Aurora 🏳️🌈 :Dahlia-Unicode-Pink:@aeva Okay but I've actually been working on the exact same idea. I'm planning to implement it as rendered game engine frames with the interactivity driven by HTML buttons overlaid onto the images.
Like yeah, it's probably a really stupid and inefficient way to build a website but I wanna.
Like yeah, it's probably a really stupid and inefficient way to build a website but I wanna.
Video Game King@aeva I had the same idea for a blog - "idea" in this connoting something I have no intention of actually committing myself to.
ferunando@aeva 🤔 if you are going to require javascript anyway, give yourself way more work and hook up that WAI-ARIA thingy? or the web-speech one. well, at least, those are the options *i* am considering for remaking my stuff....
* https://developer.mozilla.org/en-US/docs/Learn_web_development/Core/Accessibility/WAI-ARIA_basics
* https://developer.mozilla.org/en-US/docs/Web/API/Web_Speech_API
WAI-ARIA basics - Learn web development | MDN
Following on from the previous article, sometimes making complex UI controls that involve unsemantic HTML and dynamic JavaScript-updated content can be difficult. WAI-ARIA is a technology that can help with such problems by adding in further semantics that browsers and assistive technologies can recognize and use to let users know what is going on. Here we'll show how to use it at a basic level to improve accessibility.
Simon Richter@aeva I once built a nice interactive "website" that was a full-page Java applet.
Heals 
Samael 
Ela - TechnicalOtter@aeva Are you concerned about these GenAI scrapers or actual humans?
I hear that Techaro's Anubis is quite good at blocking scraping and I'm looking to deploy it on a personal project I'm hoping to spin up soon...
LR@aeva i would only make my site easy to delta and download so the scraper doesn't keep returning all the time. that means no procedurally generated content, no "house of leaves", just pages with correct caching information.
coincidentally that's exactly how our website is set up r/n.
@aeva i would strive for a solution that others benefit also from; for instance, they would not only not download my stuff, but also not download your stuff.
@lritter strictly speaking we all benefit from nvidianthropicrosoft going under, so all they have to do is wait patiently and then we can all have nice things again
@lritter slop coding tools can only "generate" beaten path solutions, and anything that is under represented performs poorly. games being one of the major holdouts for proprietary software development is probably a major inhibiting factor, so it stands to reason that if i don't want to spend the rest of my career cleaning up ai generated diarrhea then i should not release the source code for side project games until this all blows over. same goes for art and writing
@aeva i just pretend they do not exist and go about my merry way. best fuck you i can think of.
Kraftwerk-Das Model Collapse@GyrosGeier @aeva @lritter
in my dream scenario the boilerplate isn't even necessary anymore because it's been abstracted away
in my dream scenario the boilerplate isn't even necessary anymore because it's been abstracted away
@dngrs @aeva @lritter any abstraction layer has an escape hatch to handle the cases the abstraction cannot handle without growing its own ecosystem that will need its complexity abstracted away *looks at CMake*.
That said, it's good to have a bunch of boilerplate in an early iteration to identify desire paths, so you know where to pave.
@GyrosGeier @dngrs @aeva tbh i was more thinking of something that prevents them from downloading anything ever again but, sure.
youen@aeva don't know if the more classical "checking you are human" things actually work? (maybe a custom one with a silly question?)
Otherwise there's the idea of banning visitors that have scrapper behavior, but this comes after they have already scraped a few pages, and scrapers are becoming good at masquerading as real users 😑
It's going to be a never ending fight I'm afraid (well, at least until the bubble bursts)
@aeva could nightshade the pics then add alt text
dbat 