6 ms·

I think HN crowd doesn't fully appreciate the benefits of AMP to people having slow connections. In third world countries still on 2G networks AMP has been an a

I think HN crowd doesn't fully appreciate the benefits of AMP to people having slow connections. In third world countries still on 2G networks AMP has been an absolute boon. Additionally, lot of these users have limited data packs so they're better served by pages which are light in data consumption. This is true even for a lot of low income neighborhoods in the US. Making AMP opt-in would defeat the purpose as a lot of the target users would be unaware/unable to opt-in.

dfischer 10y ago
It should be opt-in then. The UX is lackluster.
milankragujevic 10y ago
They could just use Opera Mini, or have Google fix their Web Compression Proxy so it actually compresses HTTPS, and if you opt in then it's all good, and have it prompt users on slow connections to enable it like how Opera 12 prompted you to enable Opera Turbo.
flukus 10y ago
I think HN and much of the tech crowd also forget how unreliable internet connections can be. I don't know what it's like in silicon valley, but here my 4G connection will drop out several times on my short commute to work and webapps generally don't cope with it well, or at all.
geofft 10y ago
It's a huge benefit with long-term negative consequences. It's the same reason the HN/EFF crowd is opposed to things like T-Mobile's "Binge On" program, where video traffic from certain providers doesn't count against your bandwidth: it's unquestionably great for an individual end user wanting to access that site, but it sets up a system where certain incumbent providers have structural advantages against newcomers—i.e., it violates net neutrality. In the future, that same end user is likely to consider getting content from another site, and be dissuaded because Binge On (or AMP) doesn't apply. A provider-neutral solution wouldn't get in the way of competition and growth. This is the same logic behind monopolies being bad. Nobody thought AT&T was bad for providing phone service to everyone in America; that was unquestionably great. What was bad was the way that nobody else could provide potentially-better phone services. Nobody thought Microsoft was bad for giving people a web browser in the IE 4 days; what was bad was abusing their OS monopoly to gain a web browser monopoly, because IE 5 and 6 started implementing MS-specific technology. There are lots of ways to implement the benefits of AMP without the vendor lock-in Google is pushing. It's certainly technically much easier for Google to implement it in the way they're currently doing so, but that's a short-term and short-sighted gain.
tdkl 10y ago
That's the same argument as giving up privacy for security. Fuck Google and their eagerness to govern the World Wide Web.
fragmede 10y ago
I'd settle for being able to opt-out, but as it stands, no such global setting exists, or is far too hidden.
liquidise 10y ago
I think you are missing the disagreements with AMP entirely. Google, with less effort, could have released a series of requirements for a page to be considered "AMP"ed: (examples) 0 blocking script requests, less than 500KB in total page load, page loads in under 1s, etc. Once your pages meet these requirements, they gain the AMP moniker. Instead, google built an entirely new presentation method itself. This engineering effort can not be ignored and it was not on accident that this happened. They are quite literally stealing the traffic of users who implement AMP.
- fixermark 10y ago
  What you're describing has already been done. https://developers.google.com/speed/pagespeed/ https://developers.google.com/speed/pagespeed/ https://developers.google.com/web/fundamentals/performance/ https://developers.google.com/web/fundamentals/performance/ It didn't work. And given that the solution to Google "stealing" a site's AMP traffic is that site just stops supporting AMP, the "stealing" argument doesn't hold a lot of water. If I'm a small publisher, Google caches taking page hits on popular topics instead of my server is a boon, not a hindrance.
  - angry-hacker 10y ago
    It didn't work because Google didn't create it as a strong signal. Make it matter and the majority of sites will change the next time. One Tweet from Matt Cutts and everyone's already re-doing sites for SEO purposes. Set clear guidelines and rest will follow. But Google wants to build its own news portal to compete with FB.
  - liquidise 10y ago
    A few disagreements with your analysis: AMP was given a wide spread announcement that generated an unusual amount of press. It was impossible to ignore. If page speed was such an important (and viable) solution, google could simply have announced they would more heavily favor sites that performed well. Also, not supporting AMP is not a casual decision for some companies. My blog doesn't earn me revenue so i can happily tell google to pound sand before i support AMP. Can revenue-driven publishers due the same and sacrifice their SEO like that? I doubt it. > Google caches taking page hits on popular topics instead of my server is a boon, not a hindrance I simply cannot disagree with this more. Your most popular posts/topics are what generally drive the viewership and discovery of your site. I want those viewers. With AMP, they are looking at google, with a google URL, and a nice big "X" to close the window, never to see my site again. I don't see how this can be defensed at all. It is the exact opposite of what you are aiming for as a publisher, big or small.

dfischer 10y ago

It should be opt-in then. The UX is lackluster.

milankragujevic 10y ago

They could just use Opera Mini, or have Google fix their Web Compression Proxy so it actually compresses HTTPS, and if you opt in then it's all good, and have it prompt users on slow connections to enable it like how Opera 12 prompted you to enable Opera Turbo.

flukus 10y ago

I think HN and much of the tech crowd also forget how unreliable internet connections can be. I don't know what it's like in silicon valley, but here my 4G connection will drop out several times on my short commute to work and webapps generally don't cope with it well, or at all.

geofft 10y ago

It's a huge benefit with long-term negative consequences. It's the same reason the HN/EFF crowd is opposed to things like T-Mobile's "Binge On" program, where video traffic from certain providers doesn't count against your bandwidth: it's unquestionably great for an individual end user wanting to access that site, but it sets up a system where certain incumbent providers have structural advantages against newcomers—i.e., it violates net neutrality. In the future, that same end user is likely to consider getting content from another site, and be dissuaded because Binge On (or AMP) doesn't apply. A provider-neutral solution wouldn't get in the way of competition and growth. This is the same logic behind monopolies being bad. Nobody thought AT&T was bad for providing phone service to everyone in America; that was unquestionably great. What was bad was the way that nobody else could provide potentially-better phone services. Nobody thought Microsoft was bad for giving people a web browser in the IE 4 days; what was bad was abusing their OS monopoly to gain a web browser monopoly, because IE 5 and 6 started implementing MS-specific technology. There are lots of ways to implement the benefits of AMP without the vendor lock-in Google is pushing. It's certainly technically much easier for Google to implement it in the way they're currently doing so, but that's a short-term and short-sighted gain.

tdkl 10y ago

That's the same argument as giving up privacy for security. Fuck Google and their eagerness to govern the World Wide Web.

fragmede 10y ago

I'd settle for being able to opt-out, but as it stands, no such global setting exists, or is far too hidden.

liquidise 10y ago

I think you are missing the disagreements with AMP entirely. Google, with less effort, could have released a series of requirements for a page to be considered "AMP"ed: (examples) 0 blocking script requests, less than 500KB in total page load, page loads in under 1s, etc. Once your pages meet these requirements, they gain the AMP moniker. Instead, google built an entirely new presentation method itself. This engineering effort can not be ignored and it was not on accident that this happened. They are quite literally stealing the traffic of users who implement AMP.

fixermark 10y ago
What you're describing has already been done. https://developers.google.com/speed/pagespeed/ https://developers.google.com/speed/pagespeed/ https://developers.google.com/web/fundamentals/performance/ https://developers.google.com/web/fundamentals/performance/ It didn't work. And given that the solution to Google "stealing" a site's AMP traffic is that site just stops supporting AMP, the "stealing" argument doesn't hold a lot of water. If I'm a small publisher, Google caches taking page hits on popular topics instead of my server is a boon, not a hindrance.
- angry-hacker 10y ago
  It didn't work because Google didn't create it as a strong signal. Make it matter and the majority of sites will change the next time. One Tweet from Matt Cutts and everyone's already re-doing sites for SEO purposes. Set clear guidelines and rest will follow. But Google wants to build its own news portal to compete with FB.
- liquidise 10y ago
  A few disagreements with your analysis: AMP was given a wide spread announcement that generated an unusual amount of press. It was impossible to ignore. If page speed was such an important (and viable) solution, google could simply have announced they would more heavily favor sites that performed well. Also, not supporting AMP is not a casual decision for some companies. My blog doesn't earn me revenue so i can happily tell google to pound sand before i support AMP. Can revenue-driven publishers due the same and sacrifice their SEO like that? I doubt it. > Google caches taking page hits on popular topics instead of my server is a boon, not a hindrance I simply cannot disagree with this more. Your most popular posts/topics are what generally drive the viewership and discovery of your site. I want those viewers. With AMP, they are looking at google, with a google URL, and a nice big "X" to close the window, never to see my site again. I don't see how this can be defensed at all. It is the exact opposite of what you are aiming for as a publisher, big or small.