Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The End of AMP

lafoo.com

Post reply on HN
Print view View original

41–50 of 166 posts

Re: The End of AMP

#41

Post by jeltz »

jacquesm wrote: post #27
FriedrichN wrote: post #24

Earlier quoted context omitted.

The key part of the article is this: > I’ve had the pleasure of working with more than twenty thousand publishers in the five years since AMP’s launch, and I don’t believe I’ve ever heard a single reason that a publisher uses AMP other than to obtain this priority placement. If there's no reason to use AMP other than to waste money, it's very likely it will be used less and less. And then it's just a matter of time b…

From Google's perspective that is an excellent reason to continue. What I don't understand is why all these parties instantly sign up to any abomination Google decides to foist on us without thinking through the consequences and realizing that they just handed Google even more control than they already had.

Because if they do no do it they fear some of their competitors will. And that is how Google makes everyone implemnt AMp despite it being bad for everyone who is not Google.

Re: The End of AMP

#42

Post by stephenr »

bryanrasmussen wrote: post #38
stephenr wrote: post #36

Earlier quoted context omitted.

I'm not saying it's all, or even a lot of developers. But they exist.

Ok, plenty made me read it as a lot.

I'd say overall there are a pretty decent number of developers/tech people (not limited to just the web) who view Google as some giant benevolent thing.

I'd hope that web developers in particular would see more of the reality of Google's attacks on the open web. But then I've also seen an alarming number of "web developers" making claims like "I just wish Firefox and Safari would adopt Blink too, then we wouldn't have cross browser issues..." without a hint of irony, nor apparently knowledge of the last 15 years of web history.

But I'm also realistic. There are developers, both web- and other, who absolutely defend AMP, and Google's approach.

Re: The End of AMP

#44

Post by detaro »

ksec wrote: post #32
pjmlp wrote: post #26

A phyrric victory for a Web that is basically ChromeOS. Ever heard of Web Bundles? https://web.dev/web-bundles/ https://github.com/WICG/webpackage

How is Web Bundles different from .mht, which I "think", if I remember correctly is the same as MHTML. And Safari somehow support something else..... This is a genuine question. It is such a simple feature, zipping an HTML file with all of its resources. Why do we need that many standards for it. Is it NIH syndrome or are there technical reasons in doing so?

Zipping an HTML file and the resources is not equivalent, since the browser behavior very much depends on HTTP headers etc, which MHTML doesn't capture. From the existing formats, that leaves pretty much just WARC, but that's also already kind of niche, has a a bunch of interoperability issues in practice, would probably need extending/reinterpreting to fit, and just like MHTML uses base64 for binary files, which is needlessly inefficient for the purpose here.

The more important things to specify are the browser behaviors when working with such files and the signature mechanisms, and those specifications you'd also need if you reused a format.

Re: The End of AMP

#45

Post by dannyw »

numair wrote: post #39

Please don’t let your excitement take your focus off this part of the story: > The good news gets even better; non-AMP pages make considerably more revenue per pageview than AMP pages. Initially, I assumed this was due to the nature of how ads load on AMP, however, recent Antitrust lawsuits have proposed that hindering ad competition was a feature and that all non-amp ad tags, such as my company, Ezoic’s, were delaye…

There are active lawsuits going on by something like 49 out of 50 states. That's how damning evidence like this -- as well as Facebook offering to share unencrypted WhatsApp messages with Google (through the "backup to Drive" feature) -- came out to light.

Unfortunately a lot of HN readers are dismissing this or calling it partisan grandstanding, just because it's Texas that filed it.

Re: The End of AMP

#46

Post by chalst »

instagraham wrote: post #15

Whenver i am trying to share a news article with a friend, and Google throws up the AMP page at me, I take the time to select the original article link and share that. Websites deserve their own traffic, not Google's.

Not having to do this is a good reason to use another search engine than Google's as your default.

Re: The End of AMP

#47

Post by kevingadd »

ksec wrote: post #32
pjmlp wrote: post #26

A phyrric victory for a Web that is basically ChromeOS. Ever heard of Web Bundles? https://web.dev/web-bundles/ https://github.com/WICG/webpackage

How is Web Bundles different from .mht, which I "think", if I remember correctly is the same as MHTML. And Safari somehow support something else..... This is a genuine question. It is such a simple feature, zipping an HTML file with all of its resources. Why do we need that many standards for it. Is it NIH syndrome or are there technical reasons in doing so?

AFAIK bundles are designed to enable Google's CDN to serve up signed bundles from their servers and masquerade the bundles as coming from your site's domain. That's not something MHT files were designed for. (This actual feature is separate.)

Re: The End of AMP

#48

Post by chalst »

jacquesm wrote: post #28

I'm flagging this article because of a very misleading title. This is definitely not the end of AMP.

The title is somewhat misleading, but if Google do not come up with a compelling reason to use the widely hated AMP, its future sure doesn't look rosy.

Re: The End of AMP

#49

Post by AgentME »

instagraham wrote: post #15

Whenver i am trying to share a news article with a friend, and Google throws up the AMP page at me, I take the time to select the original article link and share that. Websites deserve their own traffic, not Google's.

Is that very different from "Whenever I find a site hosted through CloudFlare, I look up the origin server's IP address and link to that. Webservers deserve traffic, not CDNs."?
Post reply on HN