Comment by xg15 | Hacker News Reader

Comment by xg15 | Hacker News Reader

I don't like AMP at all and fully agree with all the lock-in concerns - but I didn't find this post particularly convincing.

This reads in places like a developer or publisher who has grown accustomated to stuffing his pages with gobs and gobs of javascript and questionable UI patterns and is now outraged at the prospect that someone wants to take that privilege away from him. I can't agree with that.

Examples:

> The underlying message is clear: Google wants full equivalency between AMP and canonical URL. Every element that is present on a website’s regular version should also be present on its AMP version: every navigation item, every social media sharing button, every comment box, every image gallery.

So Google's suggestion is that, if you've already chosen to offer an AMP page, that page shouldn't have intentionally worse usability than the original?

> For years Google has been nudging webmasters to create better websites – ‘better’ meaning ‘easier for Google to understand’. Technologies like XML sitemaps and schema.org structured data are strongly supported by Google because they make the search engine’s life easier.

Other initiatives like disavow files and rel=nofollow help Google keep its link graph clean and free from egregious spam. All the articles published on Google’s developer website are intended to ensure the chaotic, messy web becomes more like a clean, easy-to-understand web. In other words, a Google-shaped web. This is a battle Google has been fighting for decades.

Sitemaps and structured data were the non-proprietary attempts to structure the web. Those could have made the web more accessible for everyone, not just Google. They are something fundamentally different than AMP.

Yet, he condemns them too? Why?

He seems to argue that the chaos and messiness of the web infrastructure is itself a quality that should be defended. Why would that be the case?

>So Google's suggestion is that, if you've already chosen to offer an AMP page, that page shouldn't have intentionally worse usability than the original?

Google is using its dominance in search (again) to force sites to offer AMP versions. Like a "reader view", and lacking evidence at the moment, I bet many sites are building AMP versions with faster load times but reduced features to comply. Now Google is saying "Build your entire site according to our spec, or we'll de-rank you on search".

That is a very bold move, and one that would be making front-page news if they said it straight-up in a way that Buzzfeed could write about.

I, personally, hate websites that serve different content based on what device I'm using. It makes it difficult to customize a device to view things the way I want to view them. All of the changes to navigation and features makes it difficult to navigate and use the site.

I do think there could be a case made for Google abusing their power here but frankly I'm on their side for this: one site, one representation.

If you want a lighter load for mobile users, provide a lighter version for every user. If you want to keep features available, allow users to opt-in or out to such features.

> I do think there could be a case made for Google abusing their power here

Which is the concern here. Abuse of power in order to push your own agenda could just as easily turn into abuse of power to push an agenda you despise.

I think you're ignoring the capabilities and UX that desktop can provide vs. mobile. You're essentially asking for what Ubuntu did with Unity and GNOME with... well, GNOME, in that they're shoehorning a mobile/touch UX into a desktop. That isn't always the way to go.

I mean, an even worse example of that would be Windows 8. Metro everywhere was a colossal failure. Hence the dialing back for Windows 10.

There's nothing wrong with having different mobile and desktop experiences. Just make sure they have feature parity and behave in similar, expected ways.

I agree it's not ideal to have different experiences. But this wouldn't be true if Google didn't start imposing AMP through their influence. This has been an elaborate attempt to push everyone into a requirements system so that they can control how websites are built. Like an app store policy for websites. That's bad news, because their influence is real, businesses rely on search traffic to exist, and will have no choice but to comply.

I think there should be more regulation around such a large gatekeeper. It's a bad state of affairs where small businesses are beholden to one opaque company. Even more worrying if the small company offers a competitive product to one of Google/Alphabet's.

Are you saying you're opposed to responsive design?

Well, I am. We got "responsive design" when we need it least: the moment we have mobile devices capable to display desktop pages just fine, large screen, and tap to zoom-in/out to any area of interest...

I feel like we live in different universes; every time I use a desktop site on mobile, the touch targets are awful and it's hard to see content.

Yes. Moving content around when I resize my window is the opposite of functional. I don't get more screen real estate so websites can have bigger and more fancy text or pictures. I get more screen real estate because I keep things small and the screen full of information; high information density. I resize things often as they get moved around on my screen and resizing the window will move elements around nearly every time. It's the same any time screen rotation is enabled: accidentally tilt or partially rotate the device and then everything resizes and moves around. It's very jarring.

There's also the sites that take control of my keys so I can't use them to navigate; or those that disable operating system features such as copy and paste. Both of those are the opposite of useful.

If you want more trust: build and sign a native application, and don't be overly generous on the permissions you request. There's no reason a message app needs audio permission until the moment I intentionally start or join a voice conversation. Same for the camera. Same for saving data to the local disk, too, really. There's definitely no reason you should need administrator permission. And that permission should be revoked when the conversation is done.

> one site, one representation.

I don't think that's a good idea. "Mobile first" doesn't mean "mobile everywhere". You can use umatrix and stylus Firefox extensions to eliminate most page bloat. Firefox's reader view can give the same look to each page if that's what you want.

Destroying the open WWW for the purpose of making all websites look the same on all devices doesn't seem like a good justification.

Not at all. They are using their dominance to surface websites which provide a faster (and therefore better) experience to users. AMP happens to facilitate that, but if you're able to create a website that loads just as fast without AMP, it will probably rank just as high. I have yet to see proof that any AMP itself (beyond the benefits) is used as a factor to rank results.

Sadly, the reality is that the majority don't want to make the effort.

> I have yet to see proof that any AMP itself (beyond the benefits) is used as a factor to rank results.

For news articles at least, the only way to be shown at the top of the search results is to write your page using AMP: https://searchengineland.com/googles-amp-carousel-working-se....

Edit: from Google's own documentation (https://developers.google.com/search/docs/guides/mark-up-con...): "The Top stories carousel requires that your content be published in AMP".

That's from 2016. A lot has changed since. Try doing a search. There are plenty of non-AMP both at the top and in the carousel.

Just searched for "Trump". Every single carousel entry was an AMP page. So then I tried "Isis", "confirmation hearing", and "Angela Merkle". Same thing, 100% AMP in the carousel.

This is on an Android phone, using Chrome, with a US ip.

How does that show AMP is required, and not just that almost every major media site has chosen to serve AMP pages? I legitimately can't find a large media site that doesn't use AMP, so your observation is essentially meaningless. You're looking for a news site that doesn't use AMP (already extremely rare), makes their pages as fast as AMP pages, and is blocked from appearing in the carousel.

It is a response to the claim that "There are plenty of non-AMP both at the top and in the carousel".

News outlets use AMP because they get pushed out of the carousel if they don't. You have the cause/effect backwards.

> News outlets use AMP because they get pushed out of the carousel if they don't. You have the cause/effect backwards.

Looking at my Google News feed, I see articles from The New York Times and NPR listed prominently in the carousel, and yet neither use AMP.

>The New York Times and NPR listed prominently in the carousel, and yet neither use AMP.

Edit: NYT does use AMP: https://www.google.com/amp/s/www.nytimes.com/2018/09/06/us/p...

It's possible that the article I was looking at was a non-AMP link. At least, it didn't have the lightning bolt symbol next to it.

>I have yet to see proof that any AMP itself (beyond the benefits) is used as a factor to rank results

The "benefits" shouldn't be used as a factor to rank results either.

I want the best page content-wise for my search. Not the fastest to load with irrelevant results.

Besides what kind of "proof" do you expect to see? All their rankings are hidden behind several layers of secrecy, and it's not like Google will come out and say it by themselves.

Web search is a public resource at this point, and Google is percentage-wise a monopoly on it, so we should dispense with opaque algorithms and behind the scenes manipulations.

It's bad for society and bad for democracy.

They should be forced to be able to explain any ranking, and give their algorithm at any point in time.

> > The underlying message is clear: Google wants full equivalency between AMP and canonical URL. Every element that is present on a website’s regular version should also be present on its AMP version: every navigation item, every social media sharing button, every comment box, every image gallery.

> So Google's suggestion is that, if you've already chosen to offer an AMP page, that page shouldn't have intentionally worse usability than the original?

Take a moment to read that again. Mobile has constrained screen real-estate, constrained bandwidth, and navigation typically requires fingers covering the screen - and yet it's still supposed to support everything the full site does, and be more performant, and (it looks like) all be done in a particular framework?

Google is talking crosswise, it's no wonder AMP is a pain that only benefits Google.

I'm confused, I thought AMP was only for mobile (accelerated mobile pages). By regular version, don't they mean the regular mobile version?

Some quick searching to double check, it seems like not. Everything I see is referring to the "canonical" page, without really defining it in context.

But based on the way you use it, the canonical version would be whatever the primary version of the site/page is, most likely the non-mobile version (among other things, this is the URL search results would send you to [0]). It all reads like the AMP version is supposed to replace the regular mobile version.

[0] https://support.google.com/webmasters/answer/139066?hl=en#wh...

>This reads in places like a developer or publisher who has grown accustomated to stuffing his pages with gobs and gobs of javascript and questionable UI patterns and is now outraged at the prospect that someone wants to take that privilege away from him. I can't agree with that.

Well, it's not your job to agree with that or not though. He should be free to do as he pleases (so, the freedom to spew gobs of JS etc. should be unarguable) and you should be free to not visit his page.

Should they be free to spew gobs of JS?

The average person can’t tell in advance — or, often, after the event — that this will happen, and when it does happen it is bad for them.

Aside from the security issues, the current status quo shortens battery life, wastes bandwidth (which is still precious and limited on mobile, even if not on landlines), and is generally done for the benefit of everyone except the user.

>Should they be free to spew gobs of JS? The average person can’t tell in advance — or, often, after the event — that this will happen, and when it does happen it is bad for them.

If they can tell in advance, they can tell it after, and punish the page by not visiting it.

If they can't tell it after the event, then it shouldn't matter.

> If they can't tell it after the event, then it shouldn't matter.

Ah yes, the famous strategy for solving all problems: Hiding them, because if no one notices them anymore, they are not problems anymore.

Isn't that the definition of empiricism and/or utilitarianism?

If something doesn't impact someone in a way that they can ever tell, it is not a problem to them.

Analogy time: Which cigarette gives you cancer?

If they didn’t kill you by cancer and heart disease, but instead one in 1.4 million exploded and blew your head off, you’d know. Everyone would. The most explody brand would be punished.

The constant bit-by-bit risk of 1.4 cigarettes per micromort makes it ignorable, even though the risk is identical.

The only website I punish for their JavaScript content at the moment is forbes.com, and that’s because they have a combination of an adblock-blocker and (whenever I tried turning it off to read a story someone linked me to) ads which redirect me to scams.

That’s dynamite.

On my laptop I disable JavaScript everywhere unless it’s important to enable. I would do the same on my phone except the UI makes switching it in and off on a per-site basis much more annoying. The result is my phone battery, which can trivially record GPS traces all day while also managing day-to-day use of augmented reality translation all with negligible battery use, runs out of battery in a few hours of browsing.

That’s cigarettes.

Which site(s) killed the battery?

In that case Google should be free to not index his page, or at least not show it to users who Google know are not interested in downloading it.

>In that case Google should be free to not index his page

No, it shouldn't. Google should reflect an accurate ranking of pages based on content relevance (as best as possible technically), not a curated list.

Why is Google held to a different standard here?

Because, as it constitutes an effective monopoly on search (monopoly doesn't mean "absolutely sole mean". There were other OSes besides Windows in the 90s too, and other telcos when Bell split (e.g. MCI)), it's ranking has enormous influence over news, information gathering, business, politics, and so on.

Too much to be left to some private interests to play with at their own opaque will (and sell to the highest bidder).

If the metrics Google collect indicate that visitors prefer quickly loading pages, that should be factored into the decision as what external resources to link out to.

The evidence that visitors find bloated pages an issue is slowly growing, from Amazon quantising the financial loss of slower loading pages [1], to more recently, Life Hacks taking advantage of GDPR-triggered quicker-loading less-bloated European-versions of news sites [2].

I guess the question worth asking is if a non-AMP page is objectively more-performant than an AMP version, does Google prefer the publisher to keep the slower AMP version?

In a sense AMP is user-focused, validating Jakob Nielsen's prediction of the end of Web Design, because "people spend more time on other sites" [3].

[1] https://www.fastcompany.com/1825005/how-one-second-could-cos... [2] https://lifehacker.com/to-make-websites-load-faster-browse-t... [3] https://www.nngroup.com/articles/end-of-web-design/

Is Google not a private company that can do what it wants? If they wish to offer a curated list, is that not their choice to make?

> ... and you should be free to not visit his page.

You mean restrict myself to the remaining two websites that don't do this?

Also, it's not like the website obesity crisis is a new concept: http://idlewords.com/talks/website_obesity.htm

Both sitemaps and schema.org had strong influence from what google thought worked best. Other open efforts like microformats and RDF died a cold, lonely death.

Reply on news.ycombinator.com