4 ms·
That’s not my experience at all, at least on iOS. I notice essentially no performance improvement when accessing AMP vs when I inevitably choose to load the non
by mrtranscendence 6y ago
That’s not my experience at all, at least on iOS. I notice essentially no performance improvement when accessing AMP vs when I inevitably choose to load the non-AMP site because AMP sucks terribly.
- lern_too_spel 6y agoIt's more likely that mobile Safari sucks terribly. On Android, loading an AMP page from a Google or Bing results page is instant.
- crazygringo 6y agoSo which is your experience? I'm quite curious. Is it that AMP sites also take 5-7 seconds to load? Or that native sites load instantaneously?
- ninkendo 6y agoThere’s a really easy answer here: you’re all talking past each other. AMP isn’t just a standard for “fast webpages”. It’s also a technique google uses to embed other pages in their search results, such that: - Google rehosts the result on google.com (with the same HTTP2 connection you already have open to their server) - Sites are eagerly loaded in the background even if you don’t click them (making them appear instantaneous) There’s a difference between “I was sent a link to a site that uses AMP and I opened it” (which may be slow) and “I came across a site while searching on google” (which is nearly always fast.) AMP-the-standard gives google the guarantees necessary to enable hosting your site inside the google search results. This is why it’s perceived as fast. If you’re connecting to the original site itself, outside the context of a google search (or some equivalent... I know Facebook does similar re-hosting) you’re not going to see as much of a speed gain. (Side note: if AMP-the-standard was all that it was and the speed gains only came from having less bloated JS and requiring all on-screen elements to have known sizes to prevent pop-in, I’d love it. It’s what google does to it in its search results that I hate.)