So here’s an interesting (to me anyway :) question - is there a way to do this that’s “open”?
I mean, sure, you could do it as a paid service that’s not Google controlled, but it seems like the problem is that you basically need to cache EVERYTHING to make it faster for mobile.
How does one accomplish that in the ideal world we’re pulling for?
I’d go further: is it even a thing that needs to happen?
So, I can see a real use case for this. Here in Boston, AT&T has utterly craptastic coverage. So, I’ll be attempting to browse the web while I’m in a car/bus/whatever, and VOOM my coverage drops to zero. Having a cached page would help a lot here.
Is AMP equivalent to caching, though? I was under the impression that it’s more stripped-down than true caching. Something closer to Safari’s “Reader View” or Instapaper’s rendering – an idealized view of a page that displays with a more default style, with no ads, default colors and fonts, mailing list signups stripped out, etc.
Oh, of course, I see the benefit; but how can this be done without caching, and thus, centralization? Or rather, can the caching be done without centralization?
What’d be really cool is if you could install an “amp” nginx or Apache plugin that would automatically make your page fast. Maybe this already exists, I haven’t looked…
Minify any JS/CSS/HTML, apply barebones styling, use some heuristics to replace social-share-buttons and email-signup-forms with static versions, etc. Seems like it’d be tricky to make it bulletproof but any amount of that would be helpful.
Server-side rendering. Or the oft-mocked idea of a “browser in the cloud”.
Google AMP indeed seems like a bad thing. Less control over our own content generally seems like a bad thing.
What can be done, though? The only thing I know about right now, is “don’t opt in to AMP,” which is great, up and until Google decides that AMP is For Our Own Good and imposes it on everyone. Then what?
I’m fearful about changes that are ostensibly better for the user, and presented in such a way that people willingly adopt them. AMP (“join us and your site will be served faster/ranked higher!”), Medium (“the writing experience is great, no hosting required!”), and the like. Once they gain a foothold it’s very tough to break, it seems.
“What can be done, though? The only thing I know about right now, is “don’t opt in to AMP,” which is great, up and until Google decides that AMP is For Our Own Good and imposes it on everyone. Then what?”
Well hopefully by then we can convince everyone to switch to duckduckgo et al. It’s an uphill battle though.
Hugely uphill battle, yeah. I’ve been trying to switch to DDG myself. It’s not as good as Google a lot of the time, especially when I’m trying to search for things I normally search for (JS/frontend/React-related stuff at work, for instance).
Google’s tendency to show me what I want to see is, somewhat unfortunately, useful a lot of the time. “It just knows.” I think that’s a hard thing to get away from, and simultaneously hard for DDG to implement without tracking its users.
I’ve found the same with quality of results. Particularly for regional results. Fortunately bangs are a useful enough feature that I can at least keep it as my default search engine even if I tend to just use !g for general web searches. Hopefully DDG itself will improve eventually (though as you suggest, is that actually possible without tracking?), in the meantime having quick access to !w !t !mdn !openbsdman etc. is super convenient. It’s a compromise for now. Obviously that doesn’t help solve the issue though…