Hacker News new | past | comments | ask | show | jobs | submit login
Twitter rolls back AMP support, no longer sends users to AMP pages (searchengineland.com)
573 points by twapi 10 months ago | hide | past | favorite | 103 comments



Hopefully other sites that normally have a logged in experience will follow. The Reddit AMP experience is, uh...lousy. The end-user experience on news sites is fine for end users, but not great for the publishers. I suspect, though, they will take longer to move away...not a lot of capital around for them to spend on that.

I thought AMP was a trojan horse from the beginning. Happy to see it dying off.


If you’re on iOS, I highly recommend the Apollo client [1] as well as the Safari extension [2] that redirect all Reddit links into Apollo.

[1] https://apps.apple.com/us/app/apollo-for-reddit/id979274575

[2] https://old.reddit.com/r/apple/comments/pryy44/megathread_ap...


> the Safari extension that redirect all Reddit links into Apollo

oh, I didn't realize that was something that existed. thanks, that will help.


Same dude has an app that redirects all AMP links to their non-AMP versions, so you can eliminate it entirely.

https://amplosion.app


Is there an equivalent extension to open Hacker News links in an app on iOS? I'm currently using Hacker, but would be willing to switch for this feature.


AMP reddit's only useful feature is the button that links to the non-AMP page. It's so useless, you can't read more than the first two comments on it, you get sent to the normal version if you expand. Why does it even exist?


On my phone, the stupid dialogue with the button that is supposed to slide up gets stuck like a quarter of the way so I can’t even click it until I refresh.


The complaint for me is even worse: I regularly read on a kind of old iPad (Safari) and those links don't even work for me. This renders the sight unusable, as basically all conversation is hidden.


I'd guess for the SEO?


> ...not a lot of capital around for them to spend on that.

I feel a little bad for publishers. This is the second[1] time in recent memory they've followed the pied piper of a big tech company into wasting money on ultimately useless ventures. I hope someone in management learns a lesson from this and is more skeptical of FAANG in the future.

[1] The first time being the Facebook video push that was based on faulty metrics (https://www.wsj.com/articles/facebook-overestimated-key-vide...).


They're also the folks that decided to host user-hostile ads (full-screen overlays, autoplay video, showing those ads to paying* customers). At some point you've gotta just say they're getting what they deserve.

* typo. trying out a new keyboard, SwiftKey. so far so bad.


Reddit web experience is horrible by design. They don't want you browsing using your browser.


It was terrible for end users as well.


I was trying to give some grace that a lightweight terrible AMP page might be better than a 100Mb newsorg.com page.


Am not getting AMP Reddit links anymore in search results. In Europe.


Honest question from someone completely oblivious to AMP. What problem was AMP trying to solve? "Make websites load faster" is a buzzword-flag to me, and it's not really explained in the wikipedia article [1]

[1] https://en.wikipedia.org/wiki/Accelerated_Mobile_Pages


From a technical perspective, AMP had two, intertwined goals: 1) allow third parties to safely rehost your page, so that they could provide a seamless preloading experience—fetch & render the page in the background, and display it to the user immediately when they click a link. This massively improved the time it took to load search results on mobile web, since it moves the request out of the critical path and allows Google's CDN to serve your assets. 2) Provide a front end framework that focused on performance and UX, forcing developers to get rid of layout shifts, weird scroll effects, and other things that made mobile web browsing a nightmare. This framework allowed for a limited subset of HTML and CSS, but no custom JS, allowing it to enforce these rules for all rehosted pages.


> This massively improved the time it took to load search results on mobile web, since it moves the request out of the critical path and allows Google's CDN to serve your assets

Kind of, some of the time. In practice, that massive amount of render-blocking JavaScript made it slower - I noticed this on a daily basis using the web on the subway here in DC. The Washington Post loaded consistently faster without AMP, but the big win was the long tail where you could read content on normal pages but AMP was either blank until you reloaded, or took 5-10 seconds longer because once the AMP JS loaded it then had to fetch all of the resources it’d blocked behind a lazy-loader.


> Provide a front end framework that focused on performance and UX

And this is huge.

Non-technical users that don't care about the shady things Google did involving AMP probably think AMP is a godsend, if they even know AMP is even a thing.

AMP pages load incredibly fast, are incredibly responsive, and are far less annoying. Your typical end user will likely prefer AMP.

Thing is...we don't need AMP to get those features. But somewhere along the last 10 years, web developers lost the plot and now seem to think a static blog needs to serve several megabytes of JavaScript. They think they need to implement smooth scrolling in code, when every browser already does it natively.


AMP pages have been far more annoying in my experience. It breaks the standard web controls (ie no tap on the top to scroll to top), breaks copying links, and locks you into the google ecosystem at the user's expense.

Thank goodness its falling out of favor.


IMO the issues with controls are worth the tradeoff. The alternative ends up being annoying sites like arstechnica (non-AMP) which relocates content while you're reading it.


> Thing is...we don't need AMP to get those features.

I addressed this in another comment thread (https://news.ycombinator.com/item?id=29269423), but I disagree. The fundamental problem is one of incentives—individual companies don't have the leverage to fight back against ad companies and advertisers that want to implement bad user experiences and slow loading JS. Google's adoption of AMP forces publishers and ad networks fall into line, by enforcing limits on floating ads, popover, interstitials, custom JS, etc. Is this self-interested on Google's part? Maybe. But it's hard to argue that it's not ultimately better for the consumer.


Core Web Vitals achieves this anyway. Which is why Google is switching to using those signals instead of AMP in its search ranking.


> AMP pages load incredibly fast, are incredibly responsive

Compared to an unfiltered web experience. AMP content is still burdened with mountains of unnecessary JS that hurts performance.


> are incredibly responsive

That's not my experience. Every time I try reading a long article on an AMP page, the experience basically crashes. Text rendering stops and you're just scrolling through a blank white pages.

It tries to be too clever with its rendering. And it fails almost every time.


> What problem was AMP trying to solve?

Preserve Google’s control of the ad exchange monopoly; according to allegations in the latest antitrust case.

https://mobile.twitter.com/PatrickMcGee_/status/145161991699...


To expand on this, Google introduced an artificial delay when loading ads on non-AMP pages: https://mobile.twitter.com/tmcw/status/1451938637982142467


AMP solves a social problem for newspaper owners. They are all evil, venal people incapable of hiring for or motivating technical staff to make the right choice for users. This is why your average local newspaper or TV station site downloads 100MB of crap, fires 10000 timers per second, and needs 10-60s to render on your phone. AMP puts these jerks in a straightjacket while also trying to ensure they still get paid.

Theoretically there's a marketplace where better/faster news sites would be preferred by readers, but in actual fact all the news outlets are owned by the same guy.


> This is why your average local newspaper or TV station site downloads 100MB of crap, fires 10000 timers per second, and needs 10-60s to render on your phone

No, the reason for that, almost all of the time, is ads. It's true that local news rarely hires talented developers (they don't have the money) but even if they did they'd still be at the mercy of shitty ad code written by shitty ad providers. One of which is Google! So they generously solved the problem they created by setting up a new walled garden.


AMP solves a social problem for newspaper owners. They are all evil, venal people incapable of hiring for or motivating technical staff to make the right choice for users.

Don't forget how they sit on top of piles of money bags twirling their mustaches, taking lollypops from children, and cackling "Ha ha ha ha ha!" while looking down on the plebs below. Don't forget that!


They made websites faster by restricting what you could do with them. For the average shitty news website it made an enormous difference.


I'm happy to see AMP fade into irrelevance... What was the main driver of its decline? Why did Google stop aggressively pushing it?


I would speculate that the potential legal costs were determined to outweigh the benefits to them https://wptavern.com/amp-under-fire-in-new-antitrust-lawsuit...


Thanks for sharing that. This comment below the article stood out:

Imagine all those budgets that were spent on implementing AMP on major websites. Yikes.

Reminds me of the bogus "Facebook feed will be all video in 5 years" baloney that Zuck et al started promoting in 2016, which resulted in publishers firing journalists and retooling newsrooms for video ... and getting massively burned later when the farce was revealed.

https://mashable.com/article/facebook-video-five-years#IOpNs...

https://www.vanityfair.com/news/2018/10/was-the-medias-big-p...

https://washingtonmonthly.com/2021/03/13/facebook-is-killing...


Well he was kinda right, except the product was TikTok, and not Facebook.


Google changed their SEO tack, allowing non-AMP pages to benefit in the same way AMP pages do. It’s all their push around “Core Web Vitals” that track page performance no matter what the page is made with.

As to what fuelled the change, those of us outside will never know. Quite plausible this was always the plan and page metric functionality needed to catch up first. Equally plausible that Google backed off once they sensed regulatory action would be coming their way if they didn’t.


I guess, after the success of AMP, the Product Manager got promoted or moved to a new company. https://killedbygoogle.com/


It was originally launched as a defensive strategy against Facebook Instant Articles. Those never really took off, so AMP wasn't necessary for anything.


Faster smart phone processors and 4G (now 5G) together with mobile sites actually having better UX makes the advantage of AMP a lot smaller these days.

It simply is a technology that naturally over time will become less and less relevant.


I really liked AMP as a client side framework, and I think it had a lot going for it in terms of guiding you into a performant experience with good UX for mobile devices. Plus, the restrictions on floating elements and third party JavaScript have done a ton to improve ad quality on mobile web. I don't think we get there without Google or another similarly-powerful browser/aggregator coalition really working together to provide the right incentives to publishers.


I used to work in a newspaper’s digital department. From the outside, I can see why AMP might look good. “Faster loading times, higher engagement from ads, and an SEO boost? What’s not to love?” From the inside though, this adds a lot of overhead to sites that are usually old, patchwork, and very complicated. Essentially, the dev team has to have two entirely separate tracks now, one for regular ads and one for AMP. If you have an advertising department, this gets even more annoying since AMP does not allow as much customization of ads as papers are used to (header vs side vs footer vs text-only vs sponsored posts vs in-article, etc.). Not to mention, the main “incentive” problem in the journalism industry right now is not invasive ads, it’s ads and clickbait in general, which AMP doesn’t help with. Most papers don’t want to have to tailor headlines for Twitter, Facebook, and a million other sites and venues. They don’t want to have to deal with ads, but something needs to pay to run the site and no one likes paying for news. Only extremely large players like NYT or financial papers can pull that. So I don’t like when Google says this “helps publishers.” That’s not their goal, that’s just marketing speak. Their goal is to make browsing Google a better experience, going so far to define a Google-Approved HTML Spec and cache your content on their servers. Just my 2 cents. Maybe some newer players in the game do like AMP.


As someone who has maintained an AMP implementation for a big (albeit non-news) site for years, I think "AMP does not allow as much customization of ads as papers are used to" is a feature, not a bug. There are constant demands from sales and advertising partners to add more and more disruptive and broken web experiences, and having a bedrock of "Google will penalize us if we do this" to fall back on is really, really important.

I don't care whether AMP "helps publishers"—I think AMP holds publishers accountable for their (accidental or careless) UX disasters, and prevents them from happening in the first place.


The way I see it, the constant pressures from ads departments is the issue and having to make an appeal to Google is not a good solution. We should not have to rely on having Google back us up on what constitutes a “good” page. So, I don’t agree that it’s “really, really important.” If anything, it’s a sad picture of digital journalism. I think what’s really, really important is having the freedom to design your site without having to worry about Google’s arbitrary judgment of that design and not needing to rely on pleasing a search engine to show up in its results.

AMP definitely does not hold anyone accountable. Look at Reddit. Absolutely atrocious UX after AMP from what is just a basic forum. Instead, AMP makes everyone - regardless of content, design, or purpose - obey the same Google-approved standard of design, using Google-approved fonts and Google-approved CSS. In my view, AMP is the UX disaster because I hate the way it looks and have to remove the AMP from every news article I read. Oftentimes, publishers have a better idea of how their content should look than Google does. Especially smaller ones which have a strong focus on design.


> Google-approved fonts and Google-approved CSS

This is just completely false. You can use webfonts and custom CSS on AMP. There are some common-sense limitations, like limiting the total size of the CSS, but I've never run into an issue where I had a design that I couldn't achieve on AMP due to CSS constraints.

> the constant pressures from ads departments is the issue

The problem is that sales departments don't work in isolation—they're beholden to the broader pressures of the marketplace. It's a classic coordination problem—the industry is caught in a race-to-the-bottom of ad quality, and as long as you can make a few extra bucks per month by adding more and more intrusives ads, you're going to do so. Google has the ability to set different incentives for publishers, because they control a lot of traffic, which turns into revenue.


I mean relying on a massive corporation's total dominance over the web and semi-arbitrary rules for reward and punishment in order to tell the "business" side of your company no isn't exactly something I would be touting as a positive thing.

All of this crap is because sw engineers don't have a professional association with an ethics board. Can you imagine if you could put the experience of your users first and tell your business to fuck off when they ask for this nonsense secure that they really can't just fire you and find someone who will sell out.


I invite you to find any ethics board in the world that would speak out against "I made my company $10MM more yearly by causing the page to load 3 seconds slower for all of our users, with janky scrolling and popup ads". I mean, frankly, that's a very utilitarian calculation—think about just the simple case of a business saving money by removing their caching layer. Who is to say how many seconds of performance deoptimization is worth how many dollars of saved cost? If it only cost users 0.1 seconds in the median case and 1 seconds in the 95th percentile case, and you saved $1MM yearly, would you make that call?


As far as I know you still need to load 200kb+ JS from a third party (ampproject.org) to use even basic built-in browser features like forms. See for example https://amp.dev/documentation/components/amp-form/?referrer=...

If AMP only was a stricter subset of HTML with some limits on page size I don't think the backlash would be so hard. Instead it's a framework that includes massive overhead, third party dependencies, explicit carve-outs for ads and a unnecessary 7 second load delay if you choose not to load their js.

EDIT: Also built in scroll-jacking, hiding the actual origin of the content, and a bunch of other anti-features.

This is one of those things where google might have started with a good idea, but the reasonable person left the room after a single-sentence pitch and left it up to whatever braintrust of marketers and LOC-counting managers stayed.


I think you are the first person that I have encountered that has a positive opinion of AMP.


AMP made some technical sense to speed up page load but it’s proprietary nature and the available http3 and quic protocols mean that It’s high time to say goodbye to amp pages.


Well maybe but also:

> The speed benefits Google marketed were also at least partly a result of Google’s throttling. Google throttles the load time of non-AMP ads by giving them artificial one-second delays in order to give Google AMP a “nice comparative boost.

Page 91: https://storage.courtlistener.com/recap/gov.uscourts.nysd.56...


The stated goals of AMP made sense. But the technical specifics made things feel much slower. Specifically the requirements around a single draw meant waiting on a white screen for everything to load before showing anything. In contrast, normal pages that render as they go lets you start using the page before everything finishes. It's not great when the page changes drastically as it loads, but you can acheive a page that doesn't jump around (or doesn't jump around much) as it loads without forcing a single render.


Only tangential but the only time I really fly into a rage at tech anymore is when a (mobile) website uses the dark pattern of placing a link somewhere where 0.25 sec after load it will be moved down by an ad. When it happens I attempt to click the link, click the ad instead, hit back, and hit the ad a second time. Usually, I'll back out of whatever site this happens on and never go back.


Yeah... that's super annoying... there's maybe some exceptions, but it's usually not that hard to at least just put a height on the ad, so it's a blank rectangle until it finishes loading, and then it's an annoying rectangle, but it doesn't need to change shape.

Also, on my hate list, articles with an ad under a paragraph that somehow moves to the top when it gets half-way up the screen or something, so it's really hard to scroll down while reading. Blerg.


By the looks of it even the founders of AMP think that too... (Google doesn't promote AMP pages any more.)


Google also has some legal docs with unredacted statements about AMP that went public. So, more reasons to not talk about it.


HTTP3 and QUIC don’t get rid of bloat on the pages, and that was one of the big benefits of AMP. But also happy to see it dying.


I love how one of the top things Google pushes with AMP is its performance when Google sites are quite slow. YouTube flunks Google's own Lighthouse benchmark and it's not like it has sorting, filtering and searching features or a very complicated interface, especialy when the old site had the same features at half the memory cost. How hypocritical.

As a user I've always disliked AMP and how it was never the original page. Websites always have a separate AMP page and a separate real page.


What happened to all the AMP defenders that used to flood the comment section?


It's pretty pointless to argue for it on HN, because arguments always turn to technical things like that the website owner can indeed make the website faster and slimmer if they wanted to - and they don't need AMP.

When in reality what I, as someone who supports AMP, is saying is that I don't care about the technical aspects about it - the truth is that with AMP the pages load faster, and are slimmer - with no exceptions. And that's literally all I care about when I need to look up some information on a shitty connection.


The UX sucks on iPhone though. The page fits on my screen, but can still somehow scroll horizontally into blank space, so I have to go out of my way to keep it well-placed. The address bar doesn't disappear when scrolling because of some Javascript UI weirdness.

I'd rather have a slightly slower page.


> the truth is that with AMP the pages load faster, and are slimmer - with no exceptions

I'm not sure this is as universally true as you think it is


Maybe Google fired them?


Retasked obviously.


Did AMP achieve anything? (Aside from the whole everything in hosted by google thing) Wasn't the promise to make pages more responsive and lighter? I'm personally happy I dodged that bullet along with various other front-end fads.


Thank God AMP is dying...

If you are looking to develop a specification for preloading go through the browser standards process, instead of unloading this proprietary crap on the rest of us.


Twitter has become incredibly hostile to users not signed into an account (e.g. many search visitors), so AMP or not isn't going to make a big difference.


Agreed. They've made it almost unusable if you hit it from an embedded browser page.


amp wasn’t a problem. amp prioritized results in search was a problem

and i just saw here they slowed down ads on non-amp pages, that’s not good for business


IMO AMP was still a problem. It’s a version of HTML de facto controlled by Google. There’s window dressing of a foundation or whatever but Google calls the shots on what it does and does not do. To me that made AMP a five alarm fire right from the start.


And it had to be proxied through Google's servers. To hell with it.


I'm curious if anybody else is noticing how G is tracing the same paths AOL once did.


Yes and Google too shall pass.


Would be nice to have a post-mortem on 'how to recognize the outline of an emerging juggernaut'.


Cloudflare, take note.


This is false.


Why, because there's also a Bing AMP proxy? Come on. AMP pages in Google search results are proxied through Google 100% of the time, by design.


In fairness AMP does not have a requirement to use a CDN. It allows the use of a CDN which was another reason why Google pushed it (to do all kind of same-domain tricks), but to keep on topic, I actually don't think Twitter used that Google cache.


If Twitter was serving valid AMP, they didn't have a choice on using the Google cache or not:

"As a publisher, you don't choose an AMP Cache, it's actually the platform that links to your content that chooses the AMP Cache (if any) to use."

"Caching is a core part of the AMP ecosystem. Publishing a valid AMP document automatically opts it into cache delivery."

https://amp.dev/documentation/guides-and-tutorials/learn/amp...


The original article is talking about traffic from Twitter to publisher sites, not traffic from Google to Twitter. Twitter never used AMP for pages on their own site.

In this case Twitter is the platform, not the publisher, and would absolutely have been able to not use the Google cache.

> Now, when using one of Twitter's mobile clients, users will be sent to the amphtml URL in their browser, instead of the link that was shared in the Tweet. Users will load this link directly, not via a page cache. [0]

[0] https://developer.twitter.com/en/docs/twitter-for-websites/a...


Eh? By your own quote:

> it's actually the platform that links to your content that chooses the AMP Cache (if any) to use

Twitter (the platform that links to AMP content) could choose which AMP cache if any to use.

Oh, I guess maybe the confusion here is that your argument is that as a page publisher you cannot opt out of the Google CDN? I read your original statement as "it is not possible to view AMP content without using the Google CDN"


Right, I'm mainly just talking about Google search results. I probably should have made that more clear up front.


We're on a discussion page for an article about Twitter's use of AMP in their own app, which was never proxied through anything, much less proxied through Google.


Twitter directly linked to AMP pages on the publisher's site. They didn't link to a cache at all.


I wonder why someone who works at google bothers to come here to try to defend the worst practices of the company.

I mean, it's never going to go well.

Is it some sort of deranged catharsis in trying to convince yourself?


Google calls the shots on what the web does and does not do via the majority browser - Chrome. I wouldn't consider AMP any more dangerous than Chrome in this argument. There were other problems with AMP.


I'm certainly not a fan of a Chrome monoculture either but AMP crossed all browsers so I see it as more dangerous.


amp was the modern attempt to replicate the turd that was wireless application protocol and I'm glad it's dying.

I'm also glad the community mood shifted about it, no longer than a year ago negative comments about amp where drowned in downvotes

luckily vote count doesn't make ppl right or wrong.


> no longer than a year ago negative comments about amp where drowned in downvotes

This is the strangest case of revisionist history I've ever seen.


No, AMP was a problem. It was a proprietary, ham-fisted subset of the open web entirely forced by Google’s insane search engine dominance. It actively discouraged the creativity, freedom, and self-determination of the web that makes it so special. The AMP boost in search ranking was just the cherry on top.


This is exactly right. Imagine a hypothetical where Microsoft created AMP - no one would use it! The technology itself sucked. The only reason some publishers adopted AMP at all was to get/maintain favourable placement on Google, the only search engine that matters.

I'm just hoping that the death of AMP on the web will also kill Google's ambitions for rolling it into Gmail[1].

[1] https://developers.google.com/gmail/ampemail/


The idea of a proven-safe subset of HTML with no custom JS that's suitable for adding interactivity to emails is actually really interesting to me, and I hope it continues. It's been really great to be able to reply to Google Docs comments from within my email client, and the same things that make AMP good for rehosting make it a good foundation for interactive email.


And that’s not even discussing the fact that Google penalized non-Google ads on AMP.

https://mobile.twitter.com/simonw/status/1451947624513167363


AMP pages being hosted by Google is a pretty big problem as well, though. In my opinion a step in the wrong direction of the centralization of the internet.


Why should it be a browser standard? It’s a web framework. Browsers already support it.

Google did promote some improvements to web standards but they were never tied to AMP, even though they were things AMP would use.


Because without going through the standards process it is a unilateral de-facto standard that has been pushed on the web without community involvement.


Should React be a web standard? How about Typescript?

AMP being built into browsers wouldn't be an improvement.

These technologies are built on top of web standards. It's a different level of the stack.


AMP was built using existing web standards, which means publishers don't have to wait half a decade before Safari lets them use it.


Good riddance. AMP is terrible.


I'm guessing this is related to Twitter's acquisition of Scroll, and the new Twitter Blue feature of sending Blue users to publisher sites with ads disabled. That was probably a nightmare to throw AMP into the mix with everything else that needs to be done.


Has it be definitively proven that the sole purpose of AMP was to lock in advertising revenue?

The DOJ filing accused Google of this; that's not the same as it being true / there being damning evidence.


Maybe Twitter can stop obfuscating URLs as its next step.


AMP used to completely break the cookied login flow to paywalled sites. Thank god I’ll no longer now have to click to ‘Open in Safari’ to read content I actually pay for.


Let's keep this trend going!




Applications are open for YC Winter 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: